In 2015 or thereabouts, I decided that I wanted to work in finance. I had a degree in philosophy and little in the way of financial training, but I had managed an equity portfolio (quite haphazardly I may add) for a number of years, and had long maintained an interest in security analysis. I was bored with journalism and publishing and figured finance would be more exciting (and frankly, more lucrative). I also wanted to determine how I might end up working “for Bitcoin” and I believed that some more training in finance might be my port of entry to the industry. Knowing relatively little about how to “get into finance” I decided to pursue a one-year MSc in Finance and Investment at the University of Edinburgh. I picked Edinburgh because it was just down the road from St Andrews, where I had spent four years as an undergraduate. I liked the city, I liked Scotland, and I wanted to be close to the site of the Scottish enlightenment. Hume, Ferguson, Darwin (who studied medicine at the University of Edinburgh), JS Mill (well, he was half Scottish), Adam Smith — these were some of my favorite thinkers. I felt that perhaps I could absorb some of their auras by physically locating myself in the city. Edinburgh is sometimes dubbed the ‘Athens of the north’. It even has a half-finished Parthenon-like structure atop Calton hill.

The course was intensive, effectively compressing most of the key topics you would learn in an MBA into a single year. Valuably, I picked up a fair amount of stats and econometrics skills, which served me really well in the early days of the project that would become Coin Metrics. In fact, it was a statistics assignment which catalyzed my desire to build a large, denoised repository of data for a variety of blockchains. The contrast between my modest needs at inception and the sheer scale of Coin Metrics today boggles my mind.

I had my hands full learning STATA, playing around with on-chain data, studying Bitcoin (this was midway through my serious period of learning), and reading endless papers about corporate governance and the correlation between the time CEOs spend golfing and subsequent equity returns (yes — there are papers about this). My friends at the time can attest to just how unfun and studious I was. I mean, I still am that way. But I was like that back then, too.

When I had free time, I loved to hike or run up a rock formation bordering the city called Arthur’s Seat. There are many incredible things about Edinburgh — the medieval layout of the city, the castle hunched atop a gigantic rock in the center of town, the cultural explosion that is the Fringe festival every summer, the absolutely gorgeous Princes Street gardens running through the center of it all, the hulking carcasses of the former banks on George street, the yellow gorse that illuminates the city and the adjoining countryside in the springtime — but my very favorite is the presence of Arthur’s Seat not more than a 10 minute walk from the downtown. Arthur’s Seat is actually an extinct volcano.

You generally climb it in two phases — first, up a ridge named Salsbury’s crags that provides stunning views of the town, then up a steep dome with a rounded top. My more demanding runs would take me up the ridge. On certain pink wintry mornings after a brisk scramble up the ridge, I’d watch the sun rise over Leith on the one side, and downtown Edinburgh on the other. (The photos included here are my own.)

Edinburgh is known for its sizeable asset management industry. Baillie Gifford, where my dear friend Allen Farrington works, is based there, as is Ruffer Asset Management. BG recently led Blockstream’s Series C round, and Ruffer took a well-publicized Bitcoin position in 2020. At that time, you could still tally up all the firms digging into Bitcoin and taking a position. Throughout Edinburgh, you can find the legacy of Scotland’s financial sector. George Street is full of grand old buildings which were formerly bank headquarters — dramatically overbuilt and excessively ornate to convey trust and credibility — and have now been converted to bars and restaurants.

The largest banks in Scotland’s feted free banking era towards the end of the 18th century were all based there — The Bank of Scotland, the Royal Bank of Scotland, and the Linen bank. The Scottish free or laissez-faire banking system is of critical interest to economic historians, because it was inarguably successful, and because it offers such a good counterfactual compared to England. During Scotland’s free banking era (generally understood to have lasted from 1714 to 1844), England had a centrally-controlled banking system. The two economies were linked and generally subject to the same pressures and exogenous shocks. Therefore, England’s experience during the period offers a fantastic control case relative to Scotland’s experiment with free and laissez faire banking.

At the time England’s banking system was heavily restricted and largely controlled by the Bank of England. By contrast, in Scotland, no central bank existed, and commercial banks competitively issued banknotes. Banknotes were liabilities of commercial banks, redeemable for specie. The banks did not pay interest on the notes (but did pay interest on deposits) and thus benefited when clients deposited specie and took non-interest bearing notes in return. The longer the notes went unredeemed, the lower bank specie reserves could be, and the more profitable the banks became.

This was not a full-reserve system (nor is it clear that free market systems ever equilibrated at a Rothbardian full reserve). During the period, banks generally held 2–3% in specie reserves corresponding to their demandable liabilities. Equally astonishing, given this narrow margin, was the lack of bank failures. (There was one infamous failure, the Ayr bank, which we shall discuss later.) How could this be possible? Well there were some features of the system that kept banks in check and ensured that they were strongly incentivized not to indulge in the over-issuance of notes. These are well-documented in Krozsner’s 1995 ‘Free Banking: The Scottish Experience as a Model for Emerging Economies’, which is actually a World Bank publication, although I’m not sure they’d publish something like this today. In short, Scotland’s banks had the following features which kept it stable:

• Competitive ‘note dueling’
• A private clearinghouse
• Full liability partnership models
• Until 1765, clauses permitting the temporary suspension of convertibility
• Branching and diversification

Note dueling was one of the direct market mechanisms that prevented banks from over-issuing. This was effectively a form of institutionalized, competitive bank run, designed to push competitor banks into illiquidity and disrepute should they issue too aggressively. Krozsner describes it as follows:

The Royal Bank of Scotland, for example, would attempt to gather up as much of the outstanding note issue as possible from its rival, the Bank of Scotland. The bank hired people called “note pickers” to collect the rival’s notes, some of whom might even offer a little reward to individuals who would exchange their Bank of Scotland notes for the Royal Bank’s notes. The note pickers would then simultaneously converge upon the Bank of Scotland and demand: “redeem these notes as you promised, give us the gold now.”

This was a completely spontaneous market mechanism which kept reserve ratios in check, but it worked startlingly well over the period.

Bank failures were very rare. Aside from the note dueling, another reason for this was a symmetry of risk borne by shareholders. Banks at the time were full liability partnerships; that is, shareholders faced personal liability if the bank were to fail. Due to the significant cost of allowing a bank to fail, such failures seldom occurred. One infamous example, which Adam Smith wrote about in the Wealth of Nations, was the Ayr Bank. They lent too freely and collapsed in 1772; the partners (who were wealthy lairds and landowners) had to make the depositors whole. Over a 20-year process (things happened more slowly back then) creditors were eventually made whole, but a number of shareholding lairds were ruined in the process. A huge fraction of the land in Ayrshire was sold to resolve the Ayr Bank debts, and the shareholders who could not pay went to prison. How quaint now, to have bankers bear personal costs for poor decisionmaking!

Another structural, and spontaneous, feature of the market, was the private clearinghouse that naturally emerged between a subset of banks. Recall that the banks wanted to promote the usage of their notes, but they also had an incentive to accept the notes of other banks from clients, as clients would grumble and complain if they couldn’t deposit a Royal Bank note at a Bank of Scotland. This meant that banks would end up with their rival’s notes, and they began bilaterally exchanging them in the 1760s. The inefficiency of major banks engaging in bilateral settlement became evident and a clearinghouse emerged in 1771, starting with the major Edinburgh banks.

In a clearinghouse, the banks would net their balances against each other. If a bank dramatically over-issued, they’d suffer ‘adverse clearings’ — that is, rival banks would end up with a lot of their notes, and would start to demand redemption in the form of specie. So the clearinghouse would provide an early warning system for a bank engaging in reckless lending. The clearinghouse was private and not overseen by the government, but it worked wonderfully. As Krozsner says,

The clearing system thus turned into a system of prudential private regulation because being in the note exchange system was sending a clear signal to the public that if these other banks, who best knew what was going on within the bank, are willing to accept these notes, then an ordinary person might be willing to accept them also. It thus performed an important information function.

There was no regulatory body, no nasty government entity concern trolling about ‘systemic risks’ or market stability. There was simply a legal structure that discouraged excessive lending, market mechanisms through which banks could competitively keep each other in check, and a vibrant information environment the public could benefit from. The Scottish banking system during the period was remarkably stable; financial crises and panics were rare, contrasting favorably with neighboring England. The Scottish experience of lightly regulated banking shows clearly that such a model can work, and that regulation does not necessarily make a financial system more stable, and in fact is more likely to throw it into disarray.

Now what lessons should stablecoin issuers take from free banking? You don’t have to be particularly astute to see that there are some key similarities there, and indeed our central bank high priests have begun to take note of the resemblance too. I pointed out the similarities in a whitepaper in June 2020. (It’s worth noting that George Selgin has pushed back at the comparisons.) I’ll be brief because I’ve already gone on long enough. First, as exchanges (oftentimes, it’s exchanges issuing stablecoins) continue mutually accepting each others notes, they might consider a private clearinghouse. That way they can achieve efficiency in settlement — moving from real time gross settlement to a net settlement model, saving on fees and on-chain headaches. If they do this, they will be fully incentivized to surface information regarding the solvency of their counterparties. This would solve the coordination problem inherent in entities like Tether being untransparent; their clients don’t have a sufficient economic motive to diligence them. A clearinghouse might in its charter insist that stablecoin issuers disclose their collateral to the group.

Second, one tool that Scottish banks developed in 1750, as an alternative to deposit insurance, was an ‘option clause’. This allowed the bank to suspend redeemability of their notes for specie for a given period of time, effectively allowing solvent but illiquid banks to honor client withdrawals (albeit on a slower schedule). For the privilege, they would pay note holders interest on the normally non-interest bearing notes. This massively reduced the risk of a bank run and it was popular until it was outlawed in 1765. Now for stablecoins to eliminate run risk, they could be structured more like Money Market Mutual Funds, in which you can only withdraw a proportional share of the underlying assets, rather than a fixed claim redeemable for $1. So as a depositor you have no incentive to be the first out the door, as you do with a bank. Or they could implement something similar to the option clause, suspending redeemability if they were faced with a liquidity crunch. Larry White has suggested this, and I believe Tether may have a similar option clause in their ToS but I’d have to double check that.

But the key lesson is simply that unregulated or lightly regulated financial markets can work, and they have worked. There are 60 examples of laissez faire banking over history, and they were only done away with because the modern state demanded control over the economy, and banking was a key part of that. But the track record of central banks is dismal; there’s really no arguing that true laissez faire banking (the American episode was not ‘true’ free banking, as I write here) is superior to the centrally planned and managed alternative. Central bankers need to lie and misrepresent the track record of free or lightly regulated banking, because they want to eliminate stablecoins and install a CBDC. But it’s profoundly embarrassing to them that stablecoins — a fully market-based mechanism — work great, while their CBDCs are still in utero. Stablecoins restored cash to the internet, with its true assurances — privacy and autonomy — plus additional programmability and openness. CBDCs will do no such thing.

For our central banker high priests, afflicted with dizzying ambitions of a fully controlled CBDC system, I prescribe a full dose of the Scottish sunshine. They should take a walk up Arthur’s Seat and gaze upon those old elegant bank headquarters in Edinburgh and reflect on why and how the Scottish system was so remarkably stable. They are either ignorant of the history of free banking, as Selgin relates, or cynically misrepresenting it for political gain; the rest of us don’t have to be. On this, the facts are firmly on our side.

This article was originally published in my Mumurations newletter on the now-defunct Revue platform in September 2021. I had entirely forgotten about it until I dredged it up doing research for a new forthcoming piece on free banking. I liked it so much that I decided to free it from the sands of time and republish it for posterity here.

This week, Venture Capitalist Marc Andreessen appeared on the Joe Rogan podcast and made some explosive claims about the systematic “debanking” of politically disfavored firms and individuals, specifically the crypto industry. At the start of the clip, he fingers the Consumer Financial Protection Bureau (CFPB), an agency created largely by Elizabeth Warren, as the culprit behind the debanking of crypto startups. A number of critics pushed back, saying that not only is such debanking not happening, but the CFPB is actually focused on ending debanking.

The problem is that there’s several different questions being litigated here. The first: what is Marc Andreessen complaining about, and are his concerns valid? The second: what role, if any, does the CFPB play in the debanking of politically disfavored entities — are they a culprit or an inhibitor?

Many on the left are not familiar with the concerns that the crypto industry, and the right in general, have leveled around debanking. Hence the general sense of bafflement or disbelief following Marc’s statements and Elon boosting on X. To start off, I think it’s worth reading Marc and Joe’s conversation in full, since many on X are responding to mere snippets of it, and it was an in-depth piece of commentary that includes many distinct claims. See the appendix for the full transcript. Let’s dig in.

What is Marc Andreessen complaining about?

Marc makes a few distinct and inter-related claims in the episode. He first criticizes the CFPB for being a largely unaccountable “independent” federal agency which can “terrorize financial institutions, prevent new competition, new startups that want to compete with the big banks.”

He then mentions debanking as a specific harm, which he defines as “when you as either a person or your company are literally kicked out of the banking system.” Marc points out that it happens via the banks as a proxy (in a manner reminiscent of Big Tech censorship carried out on behalf of the government) but at arms length, to shield the government from direct responsibility.

Debanking according to Marc “has been happening to all the crypto entrepreneurs in the last four years. This has been happening to a lot of the fintech entrepreneurs, anybody trying to start any kind of new banking service because they’re trying to protect the big banks.” On top of this, Marc mentions certain politically disfavored businesses — under Obama, legal marijuana, escort businesses, and gun shops and manufacturing. This was called Operation Choke Point (by the Obama DoJ itself). This was followed by “Choke Point 2.0” (as dubbed by the crypto industry), which according to Marc went “primarily against their political enemies and then to their disfavored tech startups. And it’s hit the tech world hard. We’ve had like 30 founders debanked in the last four years.”

The victims are “basically every crypto founder, every crypto startup, either got debanked personally and forced out of the industry, or their company got debanked and so it couldn’t keep operating or they got prosecuted [by the SEC], or they got threatened with being charged.”

Marc also mentions that he’s aware of individuals that have been debanked “For having the wrong politics. For saying unacceptable things.”

So to summarize, Marc is making the following claims:

• Debanking means being deprived of bank access as an individual or a firm, either because your industry is politically disfavored, or you hold dissident political views
• The CFPB is at least partially culpable, as are various other unnamed federal agencies
• This works by having the regulators outsource financial repression to the banks so that the government can’t be held directly responsible
• The primary victims of debanking under Obama were legal but disfavored industries — weed companies, adult businesses, and gun shops and firearms manufacturers
• Crypto firms and entrepreneurs, alongside fintechs, have been the primary victims of debanking under Biden. Conservatives also occasionally get debanked simply for their political views
• 30 tech founders within the a17z portfolio have been debanked

We will assess these at the end of the article.

What are critics saying about Marc Andreessen’s claims?

I’m oversimplifying, but the left-libertarians are upset about Marc’s comments, because they feel that he’s coopting the debanking narrative to suit his own ends (supporting crypto and fintech), while ignoring the more “valid” victims of debanking — Palestinians getting kicked off Gofundme for sending money to Gaza, for instance. As for the mainstream left they tend to straightforwardly support the debanking of their political enemies, so they would just rather ignore the whole thing.

But there is a section of the left that is at least somewhat ideologically consistent, and is skeptical of corporate and state power over speech and finances. (A growing cohort, perhaps, now that the right has reclaimed control over a few tech platforms, and regained State power). They have been speaking out against debanking for a while. They recognize that while dissident voices on the right have been the primary victims of debanking (think Kanye, Alex Jones, Nick Fuentes and so on), this could just as easily happen to the left if the tables were turned. They more narrowly define debanking in the following way: “Debanking, or, as some financial institutions prefer, derisking, refers to a bank cutting ties with a customer deemed politically incorrect, extreme, dangerous, or otherwise out of bounds.” (from this TFP article). Rupa Subramanya in that article talks about how banks have the power to completely ruin someone’s financial life if they consider them too reputationally risky to do business with. And individuals across the political spectrum have been affected — Melania Trump, Mike Lindell, Trump himself, Christian charities, the Jan 6ers, and Muslim crowdfunding groups and charities.

Yet many on the left are nevertheless critical of or nonplussed by Andreessen’s comments, especially regarding the CFPB. Here are a few examples:

Lee Fang: The CFPB has loudly opposed debanking, why is Andreesen saying this? Where’s the evidence? Left unmentioned in this screed is the fact the CFPB has investigated Andreesen backed startups for scamming people, not political speech. Debanking comes from the FBI and DHS, not CFPB

Lee Fang: There’s a huge issue re debanking. We saw anti-Covid mandate truckers losing access to bank accounts over activism, pro-Palestinian orgs losing access to Venmo, etc. But now predatory lenders & scammers are conflating consumer protection w/ “debanking” to call for deregulation.

Jarod Facundo: I have no clue what @pmarca is referring to bc literally just a few months ago at a FedSoc event, CFPB director Chopra warned the audience of Wall Street debanking conservatives without explanation.

Jon Schweppe: I’m with @dorajfacundo. No idea what @pmarca is actually referring to here. The CFPB has been leading the charge against discriminatory debanking. What is this actually about?

Ryan Grim: The CFPB put out a legitimately good rule that went after banks over debanking users based on political views. Yes, a populist left wing CFPB head stood up for the rights of conservatives. Now VCs and Musk who don’t like the CFPB for other reasons are straight up lying to whip people into a frenzy and defang the CFPB.

Overall, this group resents crypto and fintech and does not consider firms in those industries valid debanking victims, certainly not on a moral par with the crowdfunding platform sending funds to Gaza. Crypto people, according to the left-libertarians, made their bed. Crypto folks to issue tokens, scam and commit fraud, the thinking goes, and they deserve to be treated with contempt by the banks. “If crypto founders were debanked, that’s some bank regulation business. Not our fight.”

Moreover, Marc erred, according to these critics, by casting blame at the feet of the CFPB. The CFPB, we are told, is an agency fighting debanking. Marc is simply upset at the CFPB because he invests in fintech platforms and the CFPB is in charge of making sure they don’t abuse their customers.

Since Marc appeared on Rogan, dozens of tech and crypto founders have opened up about their stories of unilateral deprivation of bank access. Many in crypto see a light at the end of the tunnel and believe that bank regulators’ unconstitutional attacks on the crypto space via banks are at an end. Cries to investigate “Operation Choke Point 2.0” have reached a fever pitch. So who is right? Andreessen or his critics? Is the CFPB really culpable? Is debanking really as big a problem as Marc maintains? Let’s investigate, starting with the CFPB.

What is the CFPB?

The Consumer Financial Protection Bureau (CFPB) is an “independent” agency established in 2011 by the Dodd Frank act in the wake of the financial crisis. It has a vast mandate, being authorized to supervise banks, credit card companies, fintechs, payday lenders, debt collectors, and student loan companies. As an independent agency, it is funded outside of Congress (and so immune to Congressional funding review). The director can not be easily removed by the President, it can make rules directly, and it can bring enforcements and legal cases under its own name. It has a considerable degree of power. The CFPB was basically set up at the exclusive behest of Senator Elizabeth Warren.

The CFPB is a frequent target of conservatives and libertarians as it is yet another federal agency, and a largely unaccountable one at that. It was set up by Elizabeth Warren, a frequent target of the right, to effectively harass fintechs and banks. And of course, most of these firms were heavily regulated already. Banks have to submit to state or Federal (OCC) oversight, as well as answering to the FDIC, the Fed, and the SEC if they are publicly traded. Credit Unions, mortgage lenders, and so on all have their own regulators. It’s not like there was a gaping lack of federal financial regulation prior to the establishment of the CFPB. The US has more financial regulators than any country on the planet. So you can forgive the right for wondering why Elizabeth Warren, who seems to operate purely on spite, was granted an entire pet agency that she could use to harass her political enemies at will.

Now in terms of what the CFPB is meant to do.

There are specific regulations in the CFPB’s mandate that generally oppose discrimination in bank access. Notably, these include the Equal Credit Opportunity Act (ECOA) and the Unfair, Deceptive, or Abusive Acts and Practices (UDAAP) portions of Dodd-Frank. The ECOA prohibits discrimination in credit transactions based on certain protected classes: race, color, religion, national origin, sex, marital status, age, or receipt of public assistance.

As far as the broader Choke Point conversation that Marc Andreessen brought up is concerned, this doesn’t really apply. “Crypto entrepreneurs” or “Conservatives” are not a protected class under law. Thus, this portion of the CFPB’s mandate does not address, even in theory, the issue at hand: the political targeting of specific disfavored industries. Moreover, this regulation applies to credit access, rather than banking generally.

The UDAAP portions of Dodd-Frank is another regulation that could plausibly pertain to debanking. It gives the CFPB broad authority to go after practices they view as unfair, deceptive or abusive. Their monster settlement with Wells Fargo fell under UDAAP. In theory, if the CFPB were to go after debanking, they would do it under UDAAP. But they haven’t done anything just yet, aside from make noise.

What the CFPB is saying

Rohit Chopra, the director of the CFPB, specifically came out against politically-motivated deplatforming from payments platforms in a June speech at the Federalist Society. In his speech he expresses concerns around big tech payments platforms unaccountably censoring users with no recourse. He worries specifically about platforms like PayPal or Venmo shutting out users for (politically unpopular) views expressed elsewhere. This is a very real phenomenon and it’s encouraging to hear him verbalizing these issues.

There’s two issues here. The first is that Chopra seems primarily concerned with the unaccountable (undemocratic) exercise of private power, especially if these networks have monopoly-like characteristics. He doesn’t really address the risk of the state, which controls much larger payments networks through the banks, using regulatory tools to compel banks to redline whole industries (which is what Marc Andreessen is complaining about).

Second, as welcome as Chopra’s comments are, the CFBP hasn’t followed up this talk with much action. As we will explore, they are moving in the direction of potentially regulating large nonbank payment networks. However, the Choke Point 2.0 question pertains to state power over the banks wielded via financial regulators. The CFPB isn’t concerned with this. This is the ambit of the Fed, the FDIC, the OCC, and the Executive which oversees them (or Congress in the case of an investigation). The CFPB doesn’t really have the authority to regulate other financial regulators. So their ability to tackle Choke Point-style behavior is limited. (I will say though, to add a feather to Andreessen’s cap, Chopra sits on the board of the FDIC, so he’s at least partially aware of or responsible for the FDIC’s generally lawless behavior).

Notably, the CFPB did in an August court filing describe the debanking of Christians as a discriminatory practice which the agency has statutory authority to tackle. This was cited by Lee Fang as a positive (and somewhat surprising) development, as the CFPB is not known to be an agency which is particularly sympathetic to the plight of conservative groups. As I mentioned in the prior section, religious groups are a protected class, so there is little ambiguity around the CFPB’s ability in law to protest their financial exclusion. We have yet to see the CFPB stick up for a non protected class (such as conservatives generally, or an industry like crypto) that has suffered debanking, which I’ll detail in the next section. Regardless, this is a step in the right direction.

What the CFPB is doing

Most recently, the CFPB finalized a rule that would subject digital wallet and payment apps to their supervision, treating them more like banks. The rule would require large digital payments apps including Cash App, PayPal, Apple Pay, and Google Wallet to provide transparent explanations for account closures. They specifically mention “debanking” in their release on the rule. Keep in mind this rule doesn’t pertain to banks, but rather to “big tech” or p2p payments apps. Either way, no actions have been brought under this rule yet, so we have yet to see how they would implement it in the real world.

Would this rule inhibit something like OCP2.0? Almost certainly not. First of all, it covers behavior by tech companies, not banks. Second, Chokepoint-style behavior is not discretionary at the bank level, but rather the consequence of federal regulators targeting entire industries via the banks. If the CFPB noticed, for instance, that crypto startups were being systematically cut off from banking, they would have to go toe-to-toe with the FDIC, the Fed, and the OCC (and ultimately the White House) to end the practice. Given Elizabeth Warren’s staunch anti-crypto sentiment, one wonders if the CFPB would do this. And more to the point, the issue with Chokepoint has to do with bank regulators stepping outside of the boundaries of the law in order to debank an entire industry. It doesn’t have to do with malfeasance at the individual banks specifically (the banks are just haplessly carrying out the orders of their regulators).

Under UDAAP, in theory, systematic account closures targeting an industry (like crypto, for instance), could be examined by the CFPB. But this recent payment apps rulemaking, cited by some critics of Marc Andreessen to demonstrate the CFPB’s anti-debanking stance, does not apply to banks. And the CFPB in their actual enforcement actions has been silent on debanking so far.

What about the CFPB’s main enforcements?

I couldn’t find any CFPB settlements that pertained to debanking specifically. Here are their largest 30 settlements in order of size:

The closest or most adjacent cases I could find were the Citigroup 2023 case, in which they were revealed to be discriminating against Armenian Americans in the context of credit card applications. (The bank was ostensibly doing this because of heightened fraud rates in the Armenian community in California, due to fraud rings). Citigroup paid $25.9m in penalties.

In 2020, the CFPB found that Townestone Financial had discouraged African American applicants in their marketing from applying for mortgages. They paid a $105k penalty.

National origin and race are both considered protected classes under US law, so neither of these cases pertains to purely political redlining, in the same sense that critics of the crypto debanking are complaining about.

Additionally, I looked at the 50 most recent CFPB settlements dating all the way back to March 2016 and none of them covered the arbitrarily deprivation of access to banking services. Of the 50 most recent cases, 15 pertain to UDAAP violations (like the infamous Wells Fargo case), eight pertain to fair lending violations, five refer to student loan servicing, five to credit reporting inaccuracies, five to mortgage servicing, four to auto lending discrimination, three to illegal overdraft practices. Debanking: not present.

What about the substance of Marc’s critique, regarding crypto/fintech firms and conservatives getting debanked?

On this point, there is no real ambiguity. I have documented in detail the phenomenon known as Operation Choke Point 2.0, whereby the Biden administration resumed the practice of financial redlining begun in the Obama administration, which started the practice around 2013. Back then, Obama’s DoJ launched Operation Choke Point, an official DoJ program to target legal but politically disfavored industries like payday lending, cannabis, adult, and firearms manufacturers via the banking sector. This was well discussed in Iain Murray’s piece Operation Choke Point: What it is and why it matters. Obama’s FDIC, under Marty Gruenberg, used insinuations and threats to persuade banks to “derisk” firms in over a dozen industries. Conservatives cried foul, and the practice was brought to light by members of the House led by Rep Luetkemeyer. Critics felt that the practice was unconstitutional, since it involved secret regulation by persuasion, rather than official rulemaking or legislation.

In 2014, DoJ memos detailing the practice leaked, and the House Committee on Oversight and Government Reform published a critical staff memo on the practice. The FDIC issued new guidance to banks to encourage them to assess risk on a case by case basis rather than redlining entire industries. In August 2017, Trump’s DoJ officially ended the practice. In 2020, Trump’s Comptroller Brian Brooks published the “Fair Access” rule aiming to end debanking on a reputational risk basis.

However, in May 2021, Biden’s acting Comptroller Michael Hsu revoked the rule. In early 2023, following the FTX collapse, folks in the crypto industry including myself noticed that similar choke point tactics were being utilized against crypto founders and firms. In March 2023, I published Operation Choke Point 2.0 Is Underway, And Crypto Is In Its Crosshairs, following it up with another piece in May detailing new revelations. Specifically, I found that the FDIC and other financial regulators had secretly imposed a 15 percent cap on deposits at banks for crypto-related firms. Moreover, I felt that the crypto-focused banks Silvergate and Signature had been unjustly forced to liquidate or shuttered due to this anti-crypto animus in the government. Since then, crypto firms have continued to have an extremely hard time getting banking access — even though there has been no official public regulation stipulating bank restrictions on banking crypto firms, nor any legislation to that effect. Again, Choke Point 2.0 was alleged to be unconstitutional by law firm Cooper and Kirk.

More recently, with the crypto industry still suffering under the yoke of this covert regulation, I took another look at the practice, and found fresh evidence that Silvergate bank had been executed, rather than dying a natural death. Today, the 15 percent cap on crypto-focused banks remains in place, stifling the industry. Every single US-based entrepreneur has suffered from this — I can attest that every single one of our roughly 80 portfolio companies has dealt with it. My firm, Castle Island (an ordinary venture capital fund dealing only with fiat), has also suffered the abrupt closure of its bank accounts.

After Marc’s appearance on Rogan, a number of crypto executives also chimed in. David Marcus explained how the Facebook Libra project was killed off by Janet Yellen. Kraken CEO Jesse Powell, Joey Krug, Gemini CEO Cameron Winklevoss, Visa’s Terry Angelos, and Coinfund’s Jake Brukhman also shared their stories. Caitlin Long has long spoken out against Choke Point 2.0, going as far to launch her own bank, Custodia, which was deprived of a Master Account by the Federal Reserve. Critics may not be sympathetic to the crypto industry, but the fact remains that it is a perfectly legal industry, quashed by secret missives and insinuations whispered from bank regulators. The result has been a broad-based crackdown on crypto banking in the US, not done in a democratic manner through legislation or rulemaking, but entirely via the administrative state.

Outside of crypto, a similar campaign is being more quietly waged against fintechs. Since early 2023, according to research by Klaros Group, a quarter of the FDIC’s enforcement actions have been leveled at banks partnering with fintech firms (the odds for non fintech partner banks were just 1.8%). As an investor in the fintech space, I can attest that finding bank partners for fintech firms has become an almighty challenge, rivaling the difficulty that crypto firms face in getting banking. The WSJ has criticized the constitutionality of the FDIC’s campaign, stating that the agency is “engaged in de facto rulemaking while bypassing the notice and public comment period legally required under the Administrative Procedure Act.”

And regarding Andreessen’s comments on conservatives being debanked, we have ample anecdotal evidence this is happening. Melania Trump mentioned that she was debanked in her recent memoir. Gab.ai, a right-wing speech platform, suffered debanking. General Michael Flynn was debanked by JPM in 2021, citing reputational risk. BoA canceled the accounts in 2020 of Timothy Two Project International, a Christian nonprofit, and froze the accounts of the Christian preacher Lance Wallnau in 2023,In the UK, Nigel Farage was debanked from Coutts/NatWest, causing a minor scandal. These are just a few of scores of examples. Under current law, banks in the US can close accounts for any reason whatsoever, and are not required to give an explanation. So on the substance, Andreessen is correct.

Why are so many commentators trying to gatekeep “debanking”?

A common thread among the critics is that Andreessen is somehow coopting the debanking term to advance his own economic agenda. Writer Lee Fang says “There’s a huge issue re debanking. We saw anti-Covid mandate truckers losing access to bank accounts over activism, pro-Palestinian orgs losing access to Venmo, etc. But now predatory lenders & scammers are conflating consumer protection w/ “debanking” to call for deregulation.” Writers at Axios have darkly implied that Andreesen is concerned about the CFPB because his firm invests in sketchy neobanks like Synapse which collapsed earlier this year. This has been a common thread in commentary on the episode: Andreessen is only concerned about “debanking” because he wants to lighten regulation on the crypto and fintech industry and get away from the CFPB’s attempts to protect customers.

And this sounds true enough, so it’s resonating with a lot of folks on the left that don’t want to believe that the government would lawlessly deprive an entire industry of access to banking. Unfortunately for them, it’s true. The Obama administration did develop tactics to use bank regulation to unconstitutionally clamp down on industries like firearms manufacturing and payday lending. And again, the Biden administration refined these tactics and used them very effectively against the crypto space. They are now going after fintechs by harassing partner banks. These things provably happened, and in both cases they represented a vast (and unconstitutional) overstepping of administrative power, which will now be revealed and rolled back under Trump.

Whether or not writers like Fang think the Biden admin’s tactics to deprive crypto firms of access to banking takes away the moral weight of his own views on the debanking of more sympathetic groups is irrelevant. It happened, it’s debanking, and its lawless. And whether Marc Andreessen has some kind of economic motive to criticize the CFPB is also not that relevant. (I was able to find one enforcement action by the CFPB against an a16z portco, a 2021 lawsuit against LendUp). The bank regulators (and Andreessen does talk about multiple agencies (plural), not just the CFPB) have indeed deputized the highly regulated financial system for political ends. Whether or not the motives of the messenger are pure is irrelevant. What matters is whether federal agencies engaged in a dangerous overreach of executive power and strayed far beyond their mandate to harass a legal industry. And the fact is that they did.

The verdict on Andreessen’s claims

So based on the full analysis, let’s evaluate the claims made by Marc on his Rogan appearance:

• Debanking means being deprived of bank access as an individual or a firm, either because your industry is politically disfavored, or you hold dissident political views

In my view, this is an accurate description. Debanking doesn’t have more or less weight because the victim is unsympathetic from your perspective.

• The CFPB is at least partially culpable, as are various other unnamed federal agencies

The CFPB definitely makes it a habit to harass fintechs and banks, and probably doesn’t need to exist. Based on what we know about Chokepoint 2.0 though, they aren’t the primary culpable agency. More directly involved are the FDIC, the OCC, the Fed, with coordination from the Biden White House. Contrary to what the critics say, the CFBP isn’t really a mitigating force, as they haven’t actually brought any cases pertaining to debanking, despite making some noises about it recently.

• Debanking works by having the regulators outsource financial repression to the banks so that the government can’t be held directly responsible

This is an accurate description. Just like using big tech to censor dissidents, getting the banks or fintech platforms to kick off tech founders is an effective way to financially suppress regime enemies without facing too much scrutiny.

• The primary victims of debanking under Obama were legal but disfavored industries — weed companies, adult businesses, and gun shops and firearms manufacturers

This is an accurate description of how Operation Chokepoint (which was an official Obama-era DoJ program) worked. It actually began with payday lending, which Marc doesn’t mention.

• Crypto firms and entrepreneurs, alongside fintechs, have been the primary victims of debanking under Biden. Conservatives also occasionally get debanked simply for their political views

Certainly both true, although we have more evidence for the coordinated crackdown against crypto than the anti-fintech campaign (although we know the FDIC is going after fintechs by bringing enforcement actions against partner banks). Regarding the debanking of conservatives, we have plenty of anecdotal evidence of it happening, but as of yet no evidence of standing internal policies at banks. It seems to be done on a case by case basis, justified by “reputational risk”. Ultimately, the banks are total black boxes, and they don’t have to give reasons for derisking individuals or firms.

• 30 tech founders within the a17z portfolio have been debanked

Eminently possible and indeed quite likely. a16z is a very active active crypto investor and virtually every domestic crypto startup has faced banking issues at some point.

In what ways did Marc err?

• He overstates the CFPB’s role a bit, as it is their sister regulators, the FDIC, the OCC, the Fed that are more responsible for the recent spate of crackdowns against crypto and fintech firms. However, he does specify that other unnamed ‘agencies’ are behind the debanking (although he doesn’t mention the FDIC, OCC, or Fed). And also, on the CFPB, Elizabeth Warren is the creator of the agency, and she was the individual most responsible for OCP2.0 (it was her appointees, specifically Bharat Ramamurti in the NEC, that led the charge under Biden). So I can forgive Marc for apportioning the CFBP outsize responsibility.
• His discussion of politically exposed persons is a little simplistic. Being classified as a PEP does not result in your automatic exclusion from banking, but it does generally result in you being subjected to more due diligence. Marc may have been thinking of the Nigel Farage/Coutts/Natwest case, in which Nigel was considered a PEP, which was a factor in his debanking from Coutts.

Overall though, Marc is right and the critics are wrong. The CFBP is not (yet) some potent anti-debanking force. Debanking is real, it unequivocally applies to crypto and fintech, and more evidence of it will come out as Republicans take control and the inquiries are launched.

Appendix: Full transcript of Marc Andreessen’s relevant comments on the Joe Rogan podcast

[Extract beginning at 1:33:47]

Marc Andreessen: And then my favorite twist is we have this thing called independent federal agencies. So like for example, we have this thing called the Consumer Finance Protection Bureau, CFPB, which is sort of Elizabeth Warren’s personal agency that she gets to control. And it’s an independent agency that gets to run and do whatever it wants, right? And if you read the Constitution, there is no such thing as independent agency, and yet there it is.

Joe Rogan: What does her agency do?

MA: Whatever she wants. Basically, terrorize financial institutions, prevent new competition, new startups that want to compete with the big banks.

JR: Really?

MA: Oh yeah, 100%.

JR: How so?

MA: Just terrorizing anybody who tries to do anything new in financial services.

JR: Can you give me an example?

MA: Oh, you know, debanking. This is where a lot of the debanking comes from, is these agencies. So debanking is when you as either a person or your company are literally kicked out of the banking system.

JR: Like they did to Kanye.

MA: Exactly, like they did to Kanye, my partner Ben’s father has been debanked.

JR: Really? For what?

MA: For having the wrong politics. For saying unacceptable things. Under current banking regulations, after all the reforms of the last 20 years, there’s now a category called a politically exposed person, PEP. And if you are a PEP, you are required by financial regulators to kick them off, to kick them out of your bank.

JR: What? But what if you’re politically on the left?

MA: That’s fine. Because they’re not politically exposed.

JR: So no one on the left gets debanked?

MA: I have not heard of a single instance of anyone on the left getting debanked.

JR: Can you tell me what the person that you know did, what they said that got them debanked?

MA: Oh, well, I mean, David Horowitz is, you know, he’s pro-Trump. I mean, he’s said all kinds of things. You know, he’s been very anti-Islamic terrorism. He’s been very worried about immigration, all these things.

JR: They debanked him for that.

MA: Yeah, they debanked him. So you get kicked out of your bank account. You can’t do credit card transactions.

JR: How is that legal?

MA: Well, exactly. So this is the thing. This is where the government and the companies get intertwined. Back to your fascism point, which is, there’s a constitutional amendment that says the government can’t restrict your speech, but there’s no constitutional amendment that says the government can’t debank you, right? And so they, if they can’t do the one thing, they do the other thing, and then they don’t have to debank you. They just have to put pressure on the private company banks to do it. And then the private company banks do it because they’re expected to. But the government gets to say we didn’t do it. It was the private company that did it. And of course JP Morgan can decide who they want to have as customers, because they’re a private company. And so it’s this it’s this sleight of hand that happens it — basically it’s a privatized sanctions regime that lets bureaucrats do to American citizens the same thing that we do to Iran. Kick you out of the financial system. And so this has been happening to all the crypto entrepreneurs in the last four years. This has been happening to a lot of the fintech entrepreneurs, anybody trying to start any kind of new banking service because they’re trying to protect the big banks.

And then this has been happening, by the way, also in legal fields of economic activity that they don’t like. And so a lot of this started about 15 years ago with this thing called Operation Choke Point, where they decided to, as marijuana started to become legal, as prostitution started to become legal, and then guns, which there’s always a fight about. Under the Obama administration, they started to debank legal marijuana businesses, escort businesses, and then gun shops, just like your gun manufacturers, and just like you’re done. You’re out of the banking system. And so if you’re running a medical marijuana dispensary in 2012, guess what, you’re doing your business all in cash, because you literally can’t get a bank account. You can’t get a Visa terminal. You can’t process transactions. You can’t do payroll. You can’t do direct deposit. You can’t get insurance. Like none of that stuff — you’ve been sanctioned, right — none of that stuff is available.

And then this administration extended that concept to apply it to tech founders, crypto founders, and then just generally political opponents. So that’s been super pernicious.

JR: I wasn’t aware of that.

MA: Oh, 100%. It’s called Operation Choke 1.0, that was 15 years ago against the pot and the guns. Choke Point 2.0 is primarily against their political enemies and then to their disfavored tech startups. And it’s hit the tech world hard. We’ve had like 30 founders debanked in the last four years. It’s been a big recurring pattern. This is one of the reasons why we ended up supporting Trump. We just can’t live in this world. We can’t live in a world where somebody starts a company that’s a completely legal thing, and then they literally get sanctioned, and embargoed by the United States government through a completely unaccountable — by the way, no due process. None of this is written down. There’s no rules. There’s no court. There’s no decision process. There’s no appeal. Who do you appeal to, right? Like who do you go to, to get your bank account back?

And then there’s also the civil asset forfeiture side of it, which is the other side. That doesn’t happen to us, but that happens to people in a lot of places now who get arrested and all of a sudden the state takes their money.

JR: That happens to people if they get pulled over and they have a large amount of cash in some states.

MA: Right. There have been well-publicized examples — there’ll be some investigation into safe deposit boxes and the next thing you know the feds have seized all the all the contents of the safe deposit and that that stuff never gets returned. And so this is when Trump says the deep state — it’s administrative power. It’s political power being administered, not through legislation. So there’s no defined law that covers this. It’s not through regulation. You can’t go sue a regulator to fix this. It’s not through any kind of court judgment. It’s just raw power. It’s just raw administrative power. It’s the government or politicians just deciding that things are going to be a certain way, and then they just apply pressure until they get it.

JR: So what happens to those 30 tech people that you know?

MA: To go into a different field like try to do something different. Complete upending of your life. Try to get away from the eye of Sauron. Try to get out of whatever zone got you into this and keep applying for new bank accounts at different banks and hope that at some point a bank will say, you know, okay, you know, it’s okay. We’ve checked in. It’s now all right.

JR: So what do they do with their money? Like what happens?

MA: I mean, you go to cash.

JR: So where do you put it?

MA: Under your mattress. you put it. Yes, exactly.

JR: That is so insane. So if someone has 30 million dollars in the bank and they get debanked.

MA: Diamonds, arts, do you, I don’t know, go overseas somewhere? And again, there’s no fingerprints. Like there’s no person who… It just happened. And we can trace it back because we know the politicians involved and we know how the agencies work and we know how the pressures apply and we know that these banks get phone calls and so forth. And so we understand the flow of power as it happens, but when you’re on the receiving end of this, your specific instance of it, you can’t trace it back.

JR: So what are the instances? What is the company? What are they trying to do? And how do they run afoul?

MA: Well, all the crypto startups in the last, basically, four years. So remember the crypto thing, everybody got excited, and like, NFTs, and all that stuff. And then it just stopped. And the reason it stopped is because basically every crypto founder, every crypto startup, they either got debanked personally and forced out of the industry, or their company got debanked and so it couldn’t keep operating or they got prosecuted, charged or they got threatened with being charged.

This is a fun twist. So the SEC sort of has been trying to kill the crypto industry under Biden. And this has been a big issue for us because we’re the biggest crypto startup investor. The SEC can investigate you, they can subpoena you, they can prosecute you, they can do all these things. But they don’t have to do any of those things to really damage you. All they have to do is they issue what’s called a Wells Notice. And the Wells Notice is a notification that you may be charged at some point in the future.

You’re on notice that you might be doing something wrong and they might be coming after you at some point in the future.

JR: Oh my god. Terrifying. That’s the eye.

MA: The eye of Sauron is on you. Now trying to be a company with a Wells Notice doing business with anybody else. Try to work with a big company. Try to get access to a bank.

JR: That’s when they support DEI initiatives.

MA: Well the SEC under Biden became a direct application of DEI. They started they did a lot with that and then the ESG stuff — ESG is a very malleable concept and they piled all kinds of new requirements into that. So through this process the SEC could basically just simply dictate what companies do with no accountability at all. Like there are hearings where they get yelled at, but like nothing changed. Nothing ever happened in a hearing that ever changed anything. It was just the raw application of power, right?

JR: And this is your friends, this has happened to.

MA: Oh yeah, for sure, yeah. And we had, like I said, we had an employee who got debanked because he had crypto in his job title. He was doing crypto policy for us and his bank booted him. Because they did a screen across their customer base. And anyone involved with crypto became a politically exposed person because crypto was politically controversial. You hear this sometimes, it’s these terms, “compliance, reputation management, tone at the top.” They have these lovely sounding terms that make it sound like everybody’s going to be an upstanding citizen, but what they’re all code for is “destroy the enemy.” Bring the hammer of God and the bank and the government or whoever. Bring it down and just crush the individual with no due process.

And look, there’s an argument in the long run that this is all unconstitutional because the constitution gives us all the right to due process and this is government pressure. So there’s probably a Supreme Court case in five years that’s going to find retroactively that this was all illegal. But in the moment when you’re the guy who’s been debanked–

JR: And then also the potential that if you do challenge them in court and lose, the repercussions will be even heavier. Is it really worth your effort? Is it worth the risk? Especially if you’ve already had your life upended. You ready to do it again? Yeah, that’s right. When you barely built yourself back up?

MA: And I think this is important context, where like when Elon and Vivek talk about reducing regulation, there’s two ways of thinking about reducing regulation. It’s like, oh my God, the water in the air is going to get dirty and the food’s going to get poisoned. Now, some of those regulations, I think, are very important. But the other way to think about it is examples like this, which is just raw government power being applied to ordinary people who are just trying to live their lives, are just trying to do something legitimate, and they’re just on the wrong side of something that the people in power have decided.

JR: Well, there’s something that isn’t illegal, but they don’t want to be done like crypto.

MA: Exactly, like crypto, or having the wrong political points of view. The other great example is the trucker strike up in Canada was an even more direct version of this, because here you had truckers physically showing up, and it was something like step one was, they take away your driver’s license, which by the way, is just somebody pressing a button on the keyboard. No more driver’s license. Step two is I take away your insurance. And step three is I take away your kids. That was the threat at the end to the truckers, because the strike was gonna jam up these cities. They were doing a nonviolent protest, but they wanted to stall the cities to be able to exert political pressure back on the government. And the government was like, we’ll tolerate it for a little while. Then we’ll take your trucker license, then we’ll take your insurance, then we’ll take your kids.

JR: How do they say they would take their kids?

MA: Because it’s administrative power. The theory would be, these aren’t good parents if they’re sitting in a truck in the middle of Calgary preventing goods and services from reaching people, putting people’s lives at risk. You know, child seizure. Now I don’t know if they actually seized any kids, but it’s just an example of: there is an agency in the Canadian government, just like in the US government, that if they want to, they can take your kids.

JR: Well, they were doing debanking there with people who donated to the trucker convoy, which is even crazier. Not even people who were there, people who were opposed to the mandates that Trudeau’s administration was imposing on people. And so they donated to these truckers, and then they got their bank accounts taken away, which is really crazy.

MA: Exactly. And I think that I think that right way to think about this is when we think about totalitarianism we think about literally World War Two, you know we think about Nazis and jackboots with tanks and guns, beating people up and killing people. You might call it that hard totalitarianism, right? But there’s this other version you might call soft totalitarianism, which is just rules and power exercised arbitrarily that just simply suppresses everything. And this is speech control and debanking and all these other things that we’ve been talking about. The good news is they’re not coming up and like beating you up in the middle of the night. The bad news is you are under their complete control and they can do whatever they want to you that doesn’t involve physical violence, which basically includes every aspect of how you actually conduct your life and support your family and get an income and everything else.

JR: And most people aren’t even aware of it.

Update 12/01/24: This story has been updated to reflect that the CFPB has indeed brought an enforcement action against an a16z portfolio company, LendUp.

At this point, stablecoins are a demonstrated success — crypto’s first killer app. The data on this front is abundant and increasingly clear. As I have argued, stables have dollarized the crypto market, and they are crypto-dollarizing a whole host of economies, particularly in emerging markets. There is an important discussion to be had about the long term prospects for native tokens like Bitcoin, Solana, or Ether in a world where almost all transactions on chain are settled in stables, but that is a question for the crypto-natives to grapple with.

What I am interested in here is what the post-stablecoin world looks like. It’s apparent to me that stables are becoming a truly dominant global settlement infrastructure, and one that is increasingly integrated into the existing financial system. As I have laid out, I believe stables are the new Eurodollars, and once they reach critical mass (perhaps in the $300-$500b range (versus their $160b float today), the Federal Reserve and other major central banks will be forced to integrate them into their financial toolkit, rather than impolitely ignoring them as they do today. This would mirror the transition that Eurodollars went through in the early ‘70s.

Thus, I would like to graduate the discussion from its current state that mostly centers on the prudence of creating stables, whether they can hold their pegs in the long term (see eg this paper from the BIS), whether they can be accepted as a true money substitute (see eg Gorton and Zhang), and consider a world in which stables continue to thrive and ultimately achieve ubiquity. Already, some former and current policymakers have begun to adopt this posture, considering not the question of “should stablecoins exist?” but rather “assuming stablecoins, then what?” It’s these thinkers — specifically Fed Governor Chris Waller, former CFTC chair Timothy Massad, and former Comptroller Brian Brooks — that I want to focus on. However, in an effort to capture the entire spectrum of debate, I’ll start with more skeptical views espoused by Rohan Grey and academics Gorton and Zhang.

1. Greyism: Stablecoins are bad because they are unregulated bank deposits

I initially wanted to start with Sen. Warren to represent the most dogmatically anti-stablecoin perspective, but since Warren is generally opposed to most things crypto, I don’t think she is the most representative voice. Her anti-stablecoin stance isn’t particularly notable within the broader cluster of her crypto contrarian views, so she isn’t necessarily the best example of a critic.

Instead, I am selecting Rohan Grey to represent what I think is an internally consistent anti-stablecoin perspective. I find Grey’s views on stablecoins interesting because they appear to be principled, but from virtually the opposite side as me. Grey is a legal scholar who notably helped author the STABLE Act introduced by Rep. Tlaib in 2020. The STABLE Act (which didn’t pass) was arguably a reaction to Facebook’s doomed Libra ambitions. It mandated that stablecoin issuers obtain bank charters, asked for them submit to oversight from the Fed and FDIC, and generally behave like banks. Suffice to say, regulating stablecoin issuers in this manner would have effectively destroyed the stablecoin industry.

From what I recall, when crypto people pointed out that this would go beyond stablecoins and effectively condemn fintechs like Paypal to bank regulation (since stablecoins and Paypal dollars are substantively the same), Grey and those in his camp bit the bullet and said “ok”. My understanding of his position is that fintechs and stablecoins and other near-money, nonbank depository instruments should in fact be folded into the more highly regulated bank system. Grey on several occasions described stablecoin issuance as “counterfeiting”.

Interestingly, Grey is supportive of financial privacy for individuals, but feels that private sector issuers are unlikely to provide it; or even if they are providing sufficient privacy, the fact that this is happening outside the guardrails of the regulated banking system makes it unacceptably costly (since stablecoin issuers will ultimately, in his view, require a bailout). If unregistered deposit-taking ends up with a bailout and taxpayers are guaranteed to be on the hook in the end, why not just regulate them as banks in the first place?

I will say, I understand the moral logic here. If unregulated deposit-taking and dollar liability issuance ends with issuers moving up the risk curve and developing asset-liability mismatches (like Terra’s UST), and stablecoins reach systemic size, you could end up with shadow bank crises that require government intervention. The “deal” that banks make with the government as public-private partnerships is that they are backed by the FDIC and other government liquidity facilities and have to submit to regulation and supervision, and in exchange, are allowed to engage in lending with the savings of ordinary citizens. Because household savings cannot politically be allowed to evaporate, the government has to be involved both in terms of supervising banks and in terms of providing a liquidity backstop. Stablecoins and other unregulated issuers, the argument goes, are wagering the savings of individuals and firms that deposit with them, without submitting to the other side of the deal (supervision and depository insurance). So they are essentially getting something for nothing.

My reactions to this are fourfold. First, stablecoin issuers seem to be getting more risk averse with time. Tether used to hold all kinds of unusual assets on their balance sheet, but now they mostly hold short duration treasuries. Circle had a snafu with SVB and drastically reduced their exposure to cash in banks. Newer stablecoins seem to be prioritizing bankruptcy-remoteness and structures that privilege holders in liquidation. Paypal’s PYUSD is a standout, operating under a NY Trust License with a bankruptcy remote model. In fact, PYUSD is so ironclad that it’s even more secure than conventional user funds held with Paypal. Stablecoins like PYUSD are regulated by a sophisticated state regulator; it’s a stretch to consider them opaque shadow banks.

Additionally, there now exist a number of ratings firms — both crypto-native ones like Bluechip, as well as S&P and Moodys — that help the general public understand the risk of stablecoins. Overall, my assessment is that the stablecoin space has reacted really well to the failure of UST and has become much healthier overall in the last two years. It’s unclear to me how much additional regulation is needed. The stablecoin regulations we are seeing globally seem to be acknowledging this, rather than forcing stables into the ill-fitting rubric of bank regulation.

Third, it seems like deposit-taking and lending are being decoupled anyway. Firms and households are increasingly holding treasuries directly, enticed by higher yields that aren’t passed along in savings accounts. Banks are increasingly not lending out customer deposits, but rather parking cash at the Fed (and sometimes outsourcing the capital side of lending to private lenders, as Matt Levine points out). So “narrow banking” is becoming increasingly popular. Stablecoins and money market funds are effectively a form of narrow banking. This does potentially reduce the importance of commercial banks as a financial intermediary, but this isn’t just a stablecoin thing. It’s a systemic change that financial regulators will have to grapple with eventually.

Lastly, on privacy, I feel that stablecoins do offer a fairly good blend of both privacy for individuals and transparency for illicit actors. Because blockchain analysis is fairly hard, in practice most on-chain deanonymization efforts tend to focus on significant financial crimes. Grey’s preferred solution is a government-run digital cash product that offers similar privacy to physical cash (at least in small denominations). But to me it doesn’t seem like the government is at all interested in financial privacy (quite the contrary, if you followed the Samourai or the Tornado cash cases), so I find it highly unlikely the state would be a reliable sponsor of a private digital cash system. Stablecoins in my view are a reasonable middle ground. The big baddies are frequently identified and have their funds seized, while ordinary citizens going about their business can have reasonable (if not perfect) privacy assurances.

2. Gorton and Zhangism: Stables don’t work in theory, so they won’t work in practice

Following Grey’s rather principled rejection of stablecoins, we move on to two economists who simply refuse to incorporate the existence of stablecoins into their model of the world. I speak of Gary Gorton and Jeffery Zhang who wrote an infamous paper in 2021 entitled Taming Wildcat Stablecoins.

Gorton is a professor of economics at Yale and Zhang was formerly an attorney at the Federal Reserve, now a law professor at Michigan. This is an important paper because it typifies an entrenched establishment belief: stablecoins represent a return to the US “free banking” era that was riven by crises and bank failures. Ergo, stables themselves are likely to be an inferior form of money. Paul Krugman recommended the paper, and the same talking points were echoed by St Louis Fed president James Bullard and Senator Elizabeth Warren. Central bankers simply love to invoke the antebellum “free banking” era to attack the stablecoin sector.

The only problems: the US free banking episode in the 1830’s-60s wasn’t “true” free banking (and hence not that useful as an analogy), and stablecoins aren’t really all that similar to free banks.

As I wrote in my Coindesk piece at the time, the US version of “free banking” wasn’t exactly a representative episode. Free banking refers to a setting in which banks operate without central bank oversight and charter issuance is relatively open. Scotland is the archetype, and that system was stable for over 100 years. As I wrote:

The American antebellum episode did not constitute genuine laissez-faire banking. Banks during that period were forced to hold risky state government bonds and were restricted from engaging in “branching” — meaning they couldn’t establish branches nationwide. This inhibited them from geographically diversifying their depositor base and from having free choice in their asset portfolio. It’s no wonder that bank failures were common.

In the US free banking period, bank failures were common, but that’s because banks themselves were fragilized due to regulation which prohibited them from diversifying their deposits, and forced them to hold inferior assets. Other forms of genuine free banking such as that found in Scotland in the period were genuinely unrestricted, and much more successful and stable, as George Selgin has spent a career pointing out. The fact is, proper free or “laissez faire” banking has a long track record of creating stable, crisis-free financial systems. The US episode that American policymakers fixate on simply isn’t a good example of the phenomenon.

Additionally, it’s misleading to compare stablecoins to banks as they aren’t engaged in maturity transformation or risky lending. Mostly, they hold short term US Treasuries or overnight repos, which are highly liquid, short duration assets. And stablecoins, though they do occasionally suffer redemptions, distribute liabilities to global, heterogeneous userbases and thus are less exposed to acute liquidity crises such as those faced by tiny regional banks (which was the problem with the US “free banking” system).

Because they find stables analogous to banknotes issued by banks that sometimes traded below par, Gorton and Zhang assert in their paper that stablecoins do not satisfy the “NQA” (no questions asked) principle. NQA “requires that the money be accepted in a transaction without due diligence on its value.” For them, MoE is downstream of NQA. As they say, “it cannot just be assumed that an object will be used as a medium of exchange. For that to happen, the object must satisfy the NQA principle.” So they conclude that stables are a poor MoE because they believe that users cannot ever have complete faith that a given stablecoin can be exchanged or redeemed at par.

This is an interesting case of reasoning from the armchair. While stablecoins have historically faced crises, if you ask a stablecoin user today whether they review the Tether or Circle balance sheet before each transaction, they would laugh at you. It is empirically observable today that stablecoins are used as a dollar substitute not just for crypto purposes, but for general digital dollar activity by tens or hundreds of millions globally. The vast majority of the time, major stablecoins trade at par on highly liquid markets, both on DeFi and centralized exchanges globally. Deviations from the peg are quickly arbitraged away.

Additionally, the dogmatism of G&Z regarding NQA is questionable. Few doubt the quality of commercial bank money, and it’s generally treated as functionally identical to central bank money (cash). However, during the bank crisis of 2023, the quality of deposits in certain banks above the FDIC limit of $250k was indeed called into question. For Silvergate, Signature, SVB, and others, there certainly were Questions Asked. Does this mean that commercial bank money is forever doomed to be considered unreliable? No, it simply means that users need to incorporate the possibility of bank runs (and possible government reactions) into their risk model. Similarly, UDSC had a depeg as some of it reserve in SVB were called into question in March 2023, but that doesn’t doom it forever. The stablecoin reacted and updated their reserve policy to de-emphasize exposure to commercial bank dollars, becoming more robust to future shocks.

As we often see with crypto critics, G&Z believe that because a certain system doesn’t work in (their) theory, it can’t work in practice. Yet in practice, stablecoins are thriving and going from strength to strength, and getting more embedded into the real economy. It’s clear that they are increasingly treated as a form of money, and it’s time that central bankers acknowledge that — or change their definition of money.

3. Massadism: We should engage with stablecoins, because they threaten sanctions enforcement

Recently, Brookings published a piece by Timothy Massad entitled Stablecoins and national security: Learning the lessons of Eurodollars. I encourage you to read it in its entirety.

Massad’s piece is one of the most remarkable discussions of stablecoins to date, not just because of its substance, but because of its author. Massad, a Democrat, was Obama’s CFTC chair and is no crypto booster. However he betrays an understanding of the stablecoin sector that is deep and realistic. His posture is not to ignore or write off stables as an ersatz form of money invented by crypto bros, but rather to acknowledge their success and to consider how their emergence affects US interests. (Compare him with certain progressives who simply want stables to not exist anymore, or with central bankers who think stables will be trivially replaced with CBDCs.) Massad’s piece focuses on the actual reality of stablecoins today and frames them as the successor to Eurodollars. This is a comparison that has also been made by folks like Izabella Kamiska and myself, more belatedly.

Massad makes a series of important observations. He points out the similarities between the emergence of Eurodollars and stablecoins:

• They represent dollar liabilities issued by entities outside the banking system (largely, but not exclusively in the case of stablecoins)
• They both emerged due to concerns around the use of US banks via entities who wanted to reduce their onshore risk (for eurodollars, Cold War adversaries, and for stablecoins, crypto traders who had been systematically debanked)
• They were both initially ignored by policymakers, and in the case of Eurodollars, ultimately accepted once systemically important
• A big portion of their growth stemmed from the opportunity to offer and earn higher yields than were available onshore
• They both cement the international role of the dollar, giving US policymakers strategic leverage (although this is less determined with stablecoins today)

Massad pinpoints an important difference between the two systems. Eurodollars, he points out, still ultimately require transactions to clear through US banks, which establishes a nexus of control that can be utilized for strategic purposes.

What he is chiefly worried about with stablecoins is their use for illicit finance, and in particular their possible undermining of our established sanctions regime. Massad acknowledges that known illicit stablecoin flows are still relatively minor, but he is concerned with a future in which they are more widely used.

Massad differs from some of his democrat colleagues by spurning the “let crypto burn” viewpoint held by many post FTX. He says that “crypto is not going away” and thinks that stablecoins could “arguably have greater potential to become a widely accepted medium of exchange for international payments” than CBDCs. Accordingly, he urges Congress to seriously consider stablecoin regulation, understanding that the US has more leverage if a larger share of stablecoins are issued by accountable, onshore firms. He understands that US crypto regulation must be situated within a global framework of competitive regulators. As he says: “ignoring the market on the assumption that it is small and can be contained could be risky, especially with other jurisdictions moving to permit wider use of stablecoins.” If the US pushes too hard against crypto, some regulator elsewhere will benefit, and this is already happening with stablecoins.

This is a viewpoint derived not from any fondness for crypto, but a serious acknowledgement of its successes and likely trajectory. I expect that Massad’s views will prove increasingly popular with members of his party as they realize that crypto is not dying off post the debacles of ’22, and that stablecoins are indeed thriving. Today, virtually all stablecoin metrics (aside from supply) are at all time highs. The empirical reality of the continued success of stablecoins, combined with positive regulation elsewhere, means that inaction is not an option for the US.

Massad ends the piece by musing about a few ways that stablecoins could be made to comply with sanctions enforcement, although he doesn’t settle on any single policy or technology. Encouragingly though, Massad doesn’t take the fire and brimstone approach of fellow Democrat Warren. With regards to stablecoin freeze and seize, he suggests “striking a balance between having adequate tools to detect and prevent illicit activity on the one hand and preventing unreasonable searches and seizures and protecting privacy on the other.”

While Massad and I almost certainly disagree about the merits of current US sanctions regime and the Bank Secrecy Act, his paper is a great example of pragmatic engagement with the stablecoin space. I hope he can make the case to more of his colleagues on the Democratic side.

4. Wallerism: Crypto, via stablecoins, is good for the dollar

Flipping over to the Republican side of the aisle, but remaining within the establishment, we have Federal Reserve Governor Chris Waller.

Waller pricked my ears in February with his comments in a speech entitled The Dollar’s International Role. The purpose of the speech is to defray concerns that the dollar is losing its status. Like Massad, he has an interest in maintaining the soft power that comes with sanctions-making ability, but he also stresses other benefits of the dollar’s globalized nature, such as lower borrowing costs for the Government and individuals, and insulating the US economy from macro shocks.

Though his comments on crypto were brief, they caught my attention. In his speech he says:

People often conjecture that cryptocurrencies like Bitcoin may replace the U.S. dollar as the world’s reserve currency. But most trading in decentralized finance involve trades using stablecoins, which link their value one-for-one to the U.S. dollar. About 99 percent of stablecoin market capitalization is linked to the U.S. dollar, meaning that crypto-assets are de facto traded in U.S. dollars. So it is likely that any expansion of trading in the DeFi world will simply strengthen the dominant role of the dollar.

And on this point, Waller is completely right. Around 60–80% of all value settled on blockchains is done with stablecoins, and over 99 percent of stablecoins reference the USD as their unit of account (and these are generally backed with dollar assets). New blockchains launch with native stablecoin integrations and this is a priority for every new L1 and L2 I’m aware of. Even Bitcoiners are focusing on stablecoins, after dismissing their importance for around a decade. From a data perspective, the overwhelming dollarization of the stablecoin sector has been consistent since we first pulled the data in 2020.

Today, the figure is higher, at over 99 percent dollars. To summarize the argument:

1. Crypto activity (whether DeFi, perp trading, or spot trading) generally utilizes stables as the MoE and core collateral type
2. 99 percent of stables reference the dollar, and this figure has actually been growing over time
3. Ergo, crypto is good for the dollar.

As with Massad’s commentary, the argument itself isn’t necessarily that interesting, it’s who is saying it. Virtually anyone active in the crypto space should be aware of this line of reasoning, which is an inversion of the previously popular talking point that Bitcoin is displacing the dollar. But this is the first time I’ve seen someone at the Federal Reserve making this exact point. (As I was writing this article, I found some more prescient comments on stablecoins from Waller dating back to 2021 — this man knows his stuff).

Waller is a Republican and was nominated by Trump, but he’s no kook. He was confirmed in the Senate at the same time that gold standard-enthusiast Judy Shelton was rejected. He appears to have fairly conventional if dovish attitudes to monetary policy.

Regarding Waller’s argument, it’s worth imagining some scenarios in which it might break down. Starting with the first premise, that crypto is becoming dollarized, I don’t see how this trend would reverse. When I first pulled the data in 2020, stablecoin settlement value had grown to around 40% of all value settled on blockchains, and I thought the trend would continue.

I pulled this data again in 2023 for a series of talks and more recently in 2024 for a refresh. In 2023, stablecoins touched as much as 80 percent of value settled on chain. Keep in mind, this data is subjective and requires a considerable amount of denoising and spam elimination, so these figures aren’t exact. But the overall trend is clear — whereas Bitcoin and Ether were historically media of exchange within crypto, stablecoins are gradually displacing them.

This makes sense. Transacting in a volatile cryptoasset is simply more complicated from an accounting perspective. Everyone who traded prior to 2017 remembers Bitcoin as the unit of account for altcoin trades — but no one does this any more, as a volatile UoA is too mentally taxing to keep track of. If you’re purchasing goods with Bitcoin or Ether, you are subject to tax accounting, and you have to recognize a capital gain if the price appreciated during your holding period. And if you want to use a volatile cryptoasset as a bridge currency for remittances, for instance, you are subject to FX risk for the duration of the transfer. These are all frictions that naturally push people towards using stablecoins instead. As major stables like Tether and USDC have recovered from crises and regained their pegs after turbulence time and again, they have become highly trusted within the crypto space. Newer stables even offer real-time interest provision from the portfolio of underlying treasuries, eliminating the opportunity cost of using stables.

The main factor I can imagine that might undermine confidence in the stablecoin sector and push crypto users back towards BTC or ETH as media of exchange would be if stablecoin settlement assurances were significantly impaired. This could happen if, in response to government edit, the seizure rate of stablecoins went from a few hundred instances a year to thousands or tens of thousands. If there was a 0.5 percent chance that any given stablecoin transaction might be reversed, users might defect from them en masse, preferring instead the settlement assurances of digital bearer assets like Bitcoin. Barring a total wipeout of the stablecoin sector, I don’t see the dollarization of blockchains being reversed any time soon.

Waller’s second premise also seems secure. The dollar has been strengthening relative to most fiat currencies, as the US economy is generally stronger than the rest of the developed world. With China in crisis and the EU shrinking its economic role in the world, real competitors to the dollar seem as remote as ever. Empirically, no one seems to want Euro stablecoins. I do expect that as crypto becomes more entrenched and regulators pass protectionist legislation (such as the EU has done with their MiCA), we could see limited instances of non-USD stablecoins flourishing. But crypto is a global market, and the overwhelming dominance of dollar stablecoins evidences that when the sovereign walls come down, the distribution of outcomes settles in a kind of extreme pareto distribution. And after a decade of stablecoins existing and a $160b float, I think we have enough data to assert that the dollar dominance as UoA isn’t just a weird data artifact, but an actual indication of which currency the world prefers, if given the choice.

In my view, crypto markets are a natural experiment demonstrating that if state-level monetary barriers didn’t exist, there would be many fewer currencies than there currently are. It’s possible that nation states might try and reassert their waning monetary privilege by banning stablecoins or their liquidity nexuses like exchanges, as we are seeing in Nigeria, but this doesn’t seem to be very effective. Crypto-financial infrastructure is simply too ubiquitous, and grey and black markets create p2p crypto liquidity virtually everywhere, even when bans exist. So while I think the dollar’s share of stablecoin UoA may come down to the 90–95% range in the coming years as we see more crypto-protectionism at the state level, I believe that dollars will continue to predominate.

The most questionable part of Waller’s argument is actually the notion that stables are good for the dollar. They’re good for the dollar in the abstract sense that they distribute dollars (and dollar assets like US Treasuries) in a frictionless manner to virtually anyone on the planet with a smartphone. This will likely collapse weaker fiats with episodes of crypto-dollarization. However, they may not be good for the Dollar Establishment, as in the set of entities that benefit from the current configuration of the dollar system. As Massad says, stablecoins (if issued abroad, via less cooperative issuers) may impair the dollar sanctions regime that characterizes the dollar system today. If stables are able to retain their current permissioned pseudonymity privacy model (whereby most on-network transactions are p2p and largely unsurveilled), they may be bad for those in government who seek to express power by creating political conditions on whom can transact. And if stablecoins become truly successful and create a high-tech form of “narrow banks”, stablecoins could be bad for the domestic banking system — by accelerating the disintermediation of commercial banking. So stablecoins could well get the dollar into the hands of many more individuals worldwide, and even turn the dollar into an apex predator which collapses many weaker fiats, but it may not be good for the Washington Consensus — the dollar establishment in DC today. I’ll cover some more of the winners and losers in stablecoin-world in the final section.

5. Brooksism: Stablecoins can help keep the dollar the world reserve

Brian Brooks should be a familiar name to virtually anyone who follows crypto policy. Formerly Coinbase CLO, he was Trump’s Comptroller of the Currency, where he passed a rule banning banks from engaging in Choke Point style behavior and devised a federal charter for crypto and fintech firms. He’s probably the most pro-crypto admin official this country has ever seen.

Brooks penned an op-ed in the WSJ last year entitled Stablecoins Can Keep the Dollar the World’s Reserve Currency. Brooks’ argument is simple, and one I align with (and have echoed in my talks). The dollar is the world reserve, although its status is under threat. Trade is increasingly being invoiced in other currencies (especially after sanctions on Russia and the emergence of the Russia-Iran-China axis), and major holders of dollar assets, like China and Japan, are divesting.

Stablecoins by contrast represent approximately $160 billion of net new dollar exposure — and a substantial fraction of them are held by foreigners. Each of those dollar liabilities is backed, generally speaking, by dollar assets like short term treasuries or overnight repos. Ceteris paribus, more buy pressure for the debt makes it cheaper to service, and less makes it more expensive. The US happens to be engaged in historically elevated deficit spending and has a relatively high debt to GDP ratio, so we need all the buyers of the debt we can afford. Stablecoins are approximately 99 percent dollarized — and this tendency has held for their approximate 10 years of existence. Thus, their existence as the native collateral of crypto, and, long term, a dominant form factor for global digital money, is positive for dollar proliferation and creates a potentially large buyer of the federal debt. And crypto is increasingly synonymous with stablecoins. Historically, assets like Bitcoin or Ethereum served as crypto collateral types (SoV), media of exchange, and even units of account. This is no longer the case. Stables now dominate for margin and collateral at exchanges, quote currencies on these exchanges, and represent 70–80% of all value settled on chain (as I showed in my Token2049 talk last year). Stablecoins have won the MoE race in crypto, even if the most idealistic crypto-natives haven’t realized it yet. Blockchains are all about dollars.

This turns a common crypto-native talking point on its head. Far from eroding the dollar’s use globally, crypto — via stablecoins — seems to be extending it in a new digital terrain. We see incidents of crypto-dollarization taking place, notably in Venezuela, Argentina, Turkey, and Nigeria. (Castle Island is currently undertaking an on-the-ground survey to get a more quantitative sense of what is happening in certain highly-adopted EM markets). Stablecoins, being instruments that can be held directly in a fully sovereign manner with no intermediary, seem to be more credible to individuals in these countries seeking dollar exposure than, for instance, dollarized bank liabilities in their local banks. They are also highly liquid and available on centralized exchanges, via moneychangers, or local OTC networks. For many folks in the global south, stablecoins offer a dollar liability that is far more credible than dollar deposits in banks, far easier to access than physical USD cash, and one that can easily be deployed to earn interest in DeFi or the US T-bill rate (via an emerging cohort of natively interest-bearing stablecoins). Stables are globally liquid and can serve as remittance rails or settlement medium for cross-border commerce without the hassle of banks and other costly intermediaries.

Brooks’ views contrast with Waller in that Waller’s comments reflect a more passive view that even if crypto succeeds, it’s unlikely to threaten the dollar (and the dollar isn’t really under threat, anyway). Brooks by contrast takes the more active position that stablecoins should be encouraged because they could actually help rescue the beleaguered dollar.

Brooksism is the position I align with the most, but there are a few pieces that trouble me.

The first is the embedded assumption that the dollar as the global reserve is a desirable state of affairs. Following Pinketty and Lyn Alden, I’m not actually convinced that the dollar as the global reserve is actually good for most Americans. In fact, I’m persuaded that the dollar reserve is great for coastal elites who work in finance, for folks the government (and recipients of that patronage), and bad for the working class. The structural nature of the system is such that the US as the issuer of the global reserve must maintain a persistent trade deficit in order to supply the world with dollars, causing debt accumulation in the US, and an offshoring of manufacturing. (This concept is too complex for a full treatment here, but you can read Lyn Alden, the BIS, or Greely in the FT for more.) This setup causes the world to accumulate dollars (and dollar assets like US Treasuries, US equities, bonds, real estate, stocks), benefiting those who work in finance or those sectors. Meanwhile, the relatively strong dollar makes US exports uncompetitive relative to our supplier nations, causing our industrial sector to fade into irrelevance, immiserating a large portion of the population.

So, even though I have at times promoted the idea that continued dollar dominance is generally supportive of US interests, the truth is that we need to be specific with whose interests we are specifically interested in. Certainly, dollar dominance is good for US policymakers (they can spend more loosely), for DC-adjacents, for the US’ continued sanctions-making ability, and for individuals (like myself) who work in the US financial sector — globalized Capital, essentially. But it’s not, as far as I can tell, particularly good for the working class, or inequality generally.

One reason I would hypothesize this system persists even if it creates discontent and populism is simply that it’s extremely convenient for the US government to retain oversight over the global nexus of all financial transactions, especially as it gives policymakers the ability to project power (via sanctions) without exerting kinetic force. But this may be less of an incentive, especially as our sanctions-making power has meaningfully diminished in recent years.

So it could be said that if there is a rethinking of the petrodollar system, more commodities are invoiced in other currencies, the world de-globalizes, mercantilism re-emerges, the American industrial base is rebuilt, we become a competitive exporter once again, and we rethink our debt-financed, consumerist economy dependent on China, the prospects for the American middle class might actually improve meaningfully. For sure, stablecoins probably won’t move the needle much here either way. But I am always mindful of this possibility when I think about the “stablecoins promote the dominance of the dollar (and that’s good)” talking point. It may well be the case that the dollar’s dominance isn’t really good for most Americans — and the negatives are no longer offset by the perceived benefits like our sanctions-making ability, which is an increasingly blunt tool.

The second misgiving I have is simply a matter of scale. Although I have made the exact same case as Brooks in some of my talks (see my talk at Messari Mainnet in 2023), the fact is that stablecoins are still relatively minor in the grand scheme. At $160b in float, they are still a minor (yet growing) buyer of the debt. If they were a sovereign in their own right, they would be the 16th largest sovereign holder of Treasuries. If they were a money market mutual fund, they would be the 14th largest. But they don’t — right now — move the needle in terms of making the deficit cheaper to monetize. Stablecoins at ~$160b only represent 88 bps of US M1, and their holdings only 47 bps of total US government debt (at $34T). Large foreign holders of Treasuries like Japan ($1.1T) and China ($0.85T) are still multiples larger than the debt instruments held by all stablecoins. And it’s possible that a new cohort of stables that aren’t based on US treasuries and dollar assets emerges. For instance, Ethena (which is the fifth largest USD token today — they prefer “synthetic dollar” to “stablecoin”) derives its value from Ethereum and Bitcoin collateral, offset by short positions. Its success may do little for US interests.

All of this could change, of course, as crypto balance sheets continue to expand and the stable float grows as they become a major payments system. But for now, the “stables make the National Debt cheaper to monetize” concept remains a relatively speculative talking point.

The ought of stablecoins

It’s important to decouple the normative from the descriptive, i.e., what should happen from what is likely to happen. Where some of these thinkers go wrong, like Gorton and Zhang, is that they focus on ought rather than is. Because stablecoins shouldn’t work (according to G&Z), or because it would be bad if they worked (following Grey), they are dismissed. But the plain truth is that they are working, and in my view have achieved exit velocity.

It’s very clear to me that stablecoins aren’t going anywhere. They are by far the most important application of public blockchains. What started as a hacky solution to Bitfinex’ banking problems became the most important development in financial technology in decades, and our best chance at creating a true form of digital cash and regaining the ground we’ve lost on financial privacy in the last half century.

By my estimate, stables settle approximately 10 trillion dollars a year, a similar level to Visa (you can quibble whether this is an apples to apples comparison). Almost 100 million addresses on-chain hold stablecoins today. They are increasingly being integrated into major payments networks, such as those run by Visa, Stripe, Checkout, Paypal, Worldpay, Nuvei, or Moneygram, to name a few. The mythical convergence between crypto and tradfi is finally, actually happening, with stablecoins as the beachhead.

And even as the US continues to be hostile to stablecoins on virtually all fronts, new jurisdictions are embracing them. In most cases regulators acknowledge that the vast majority of stables are dollar-backed and are permitting dollar stablecoin issuance by local issuers. This growing list includes Singapore, Hong Kong, Dubai, Japan, Bermuda, and others. Certain places, like the EU, take a more cautious view, passing stablecoin regs, while aiming to limit the influx of dollars through that vector. For stablecoin issuers though, all that matters is that there are a few safe havens that are willing to give issuers a reasonable regulatory terrain to operate on. And this seems to be the case.

Objectively speaking, therefore, it seems that stablecoins are here to stay, and will likely continue to grow untrammeled unless policymakers globally launch a coordinated campaign to destroy them.

On the normative front, my synthesis of the views summarized above is simply that each individual policymakers’ views are informed by their own political objectives and underlying views. Gorton and Zhang prefer a state monopoly on money. Many of the critics listed here tend to favor the exploitation of financial infrastructure for political ends to varying degrees. Thus, any cash-like network, especially one that’s particularly unaccountable, is hostile to this agenda. Massad (who appears to be a more moderate, Obama-style Dem) is more nuanced. He acknowledges the importance and historical significance of stablecoins, but is concerned about sanctions evasion. Brooks is a libertarian-leaning Republican, and is therefore strongly favorable towards stablecoins. Waller is a Republican-appointed Fed governor, and is able to see the conditional benefit of stablecoins for the dollar.

Normatively, I have libertarian sympathies, and strongly believe in individual liberty, a reversal of financial surveillance trends that began 50 years ago with the digitization of finance, and the right to financial self-determination. I am also skeptical of the US’ continued ability to deputize financial rails for political ends; this status quo appears to be ending (and I don’t particularly lament its departure). I am generally supportive of dollarization as a policy that promotes restraint in untrustworthy jurisdictions and therefore see the welfare benefits of spontaneous bottom-up crypto dollarization that we see happening today.

Stablecoins disintermediate banks, remitters, and in conjunction with other forms of crypto-financial infrastructure like exchanges, give billions of savers globally direct access to digital dollars that they may not have had before. In each case, disintermediation means cheaper transactions. We see this on the ground directly with remittances. Settling remittances on stables via exchanges shaves meaningful basis points off global remittance rates, which still average 6.2 percent according to the World Bank, although this varies by channel. For billions of individuals in the global south, this makes a meaningful difference.

Thus, for me, stablecoins — especially those that are truly cash-like (i.e., exhibit minimal embedded surveillance) — are an immensely powerful tool, and represent a largely unmitigated force for good globally, especially in countries with immature or unstable financial sectors. The downsides of stablecoins, such as more scalable illicit flows, can be managed as issuers become more aggressive with ‘freeze and seize’ policies, and law enforcement builds sophistication around blockchain analysis.

So who are stablecoins good for?

As with any disruptive technology, there are winners and losers in any transformation. Instead of labeling stablecoins “good” or “bad” broadly, I will segment by stakeholder my views of who they are good or bad for.

Stablecoins create opportunities for:

Individuals living in unstable currency regimes: the existence of this market is well-established. If you look at the IMF or Chainalysis data, it’s evident that crypto adoption correlates meaningfully with inflation, unstable monetary regimes, and past histories of sovereign default. Stablecoins offer a self-custodied, credible USD liability outside the banking system, and this is clearly attractive to individuals in places like Argentina, Nigeria, and Turkey where currencies and banking systems are not reliable.

Sophisticated ex-US financial hubs: just like the UK thrived as a hub due to the emergence of Eurodollars, certain jurisdictions have begun to see stablecoins (and crypto more generally) as an opportunity, especially when contrasted with the US’ reluctance to productively regulate the sector. Already, we see a flight of founders and capital from the US to these emerging hubs. The default place to launch a stablecoin today is outside the US.

Digital nomads: stablecoins renders savings far more portable on a cross-border basis. They also simply payroll, especially for businesses with globally distributed employees. Already, it’s common to see crypto-native employees asking for payment in USD stablecoins, both due to lower transaction costs and because their local currencies may be unreliable.

The US government (from a fiscal perspective): as mentioned, stablecoins largely back themselves with short term US government debt. Collectively they rank 16th among sovereign holders of the debt, and 14th among US money market mutual funds. While the scale of this activity is relatively small today, it’s not inconceivable that stablecoins collectively could become a top three holder of T Bills. This would meaningfully improve the US’ fiscal prospects.

Stablecoins create challenges for:

Foreign governments with unstable currencies: stablecoins make spontaneous dollarization (the likes of which we saw historically in Ecuador in 2000) much more efficient. I expect to see multiple episodes of crypto-dollarization over the next decade, whereby savers engage in currency substitution in a bottom-up manner, accelerating the devaluation of weak local currencies. Already, we see crypto exchanges and stablecoins blamed for a wave of dollarization and Naira inflation in Nigeria.

Banks: stablecoins accelerate the rise of narrow banking, or the transition from banks as the default savings devices for individuals and firms. They are part of the broader trend of a rotation towards fintechs, money market funds, direct Treasury ownership, and crypto instruments. If interest rates stay high (and as stablecoins increasingly pass along interest), a low-yielding savings account looks less attractive relative to a money market fund or an interest-bearing stablecoin.

Legacy financial gatekeepers: any financial business that benefits from regulatory barriers to entry (banks are the epitome of this) will likely face margin compression due to the increasing popularity of stablecoins. For instance, legacy remitters relying on the correspondent bank system will see their business model challenged, as they are forced to compete with digitally native remitters who can offer transfers more cheaply with crypto-financial infrastructure.

Law enforcement: stablecoins create a paradox for law enforcement. They are highly legible, and indeed records of illicit transactions are present on the blockchain forever, once on-chain addresses are linked to illicit entities. However, they also facilitate the (relatively) final settlement of value on a cross-border basis in a peer to peer manner at scale. Physical cash faces the constraint of being costly to move and store, and as such exhibits physical constraints on its use in illicit finance. There’s only so many banknotes you can stuff in a duffel bag and take on a commercial flight, for instance.

Stablecoins face no such scale constraints. But stablecoins are also subject to intervention by issuers who maintain “freeze and seize” capabilities. And these issuers do frequently interrupt illicit activity when they are made aware of it by law enforcement, at an increasing scale. (See for instance the seizure last year of $225m USDT linked to a crime syndicate.)

Unlike cash, illicit stablecoin flows can be frozen at a distance. It’s also, at present, easier for law enforcement or courts to reach out to a single stablecoin issuer rather than trying to wrangle a network of banks through which illicit funds are flowing. There’s only a handful of major stablecoin issuers (and they maintain active dialogues with law enforcement); there are thousands of banks.

It’s important to understand stablecoins as an evolving tool in the cat and mouse game between illicit actors and law enforcement. In the short term, illicit actors may feel that stables offer them more flexibility and convenience in terms of terrorist finance, scams, or money laundering, but as the sophistication of global law enforcement ramps up, I expect that they will be increasingly perceived as a relatively worse vector for illicit flows. However, as we are in the transitional period, there is rightly a lot of concern on the part of governments as to the possible use of stables for crime. Long term though, the fundamental higher legibility of stables plus the freeze-and-seize quality of these networks should make them overall less good for crime than cash — and possibly legacy digital rails too. This will of course require law enforcement to continue to grow their toolkit as it pertains to these networks.

The US national security apparatus: similarly, stablecoins will force a rethinking of the standard sanctions toolkit employed for foreign policy objectives. Personally, I’m very skeptical of the status quo in US sanctions policy. It appears to have manifestly failed in the case of Russia. As far as I can tell, the US seizing Russia’s reserves, and attempting to interrupt the flow of commodities to Russia via dollar networks has only accelerated de-dollarization. In 2023, 20 percent of oil trade volume was settled in currencies other than dollars according to JPM. This is mainly catalyzed by the emerging Russia-China-India-Iran commodity axis.

At the same time, sanctions don’t appear to have dissuaded Russia in its imperial ambitions or meaningfully impaired its ability to wage war. Nevertheless, Washington hasn’t changed course with its approach to sanctions, and you can see this thinking reflected in the comments made by Massad, who is chiefly concerned with stables interfering with US sanctions-making ability.

In my view, Massad is shutting the barn doors long after the horse has bolted. Nonetheless, if US policymakers want to deputize stablecoins for their decaying sanctions regime, they still can. They would have to encourage onshore issuance via sensible stablecoin legislation and encourage accountable, regulated financial institutions to become issuers (rather than keeping banks firewalled from the sector and making stablecoins toxic via edicts like SAB121). Their current posture, which pushes stablecoin issuers abroad, will only make enforcing national security directives via stablecoins more challenging. Already, it’s an open question as to how the US would exert pressure on Tether, the world’s dominant stablecoin.

The Treasury seems to be increasingly taking the perspective that any dollar liability, no matter the domicile of its issuer, end users, or backing assets, falls under the aegis of the United States. But this would call into the question the nature of the entire offshore market, and in my view, constitute a significant change in policy, and would likely accelerate the dedollarization which is already underway.

Reading list

• Rohan Grey in The Nation, Facebook Wants Its Own Currency. That Should Scare Us All.
• Gorton and Zhang, Taming Wildcat Stablecoins
• Timothy Massad in Brookings, Stablecoins and national security: Learning the lessons of Eurodollars
• Chris Waller, The Dollar’s International Role (speech), Reflections on Stablecoins and Payment Innovations (speech)
• Brian Brooks in the WSJ, Stablecoins Can Keep the Dollar the World’s Reserve Currency

Thanks to Austin Campbell for his feedback.

Update 05/07/24: I have edited the piece to reflect a correction from Rohan Grey regarding his position on the State monopoly on money.

John Griffin, he of the infamous Griffin and Shams paper which gave academic grist to the Tether Truther crowd, has poked his head up once again. Unsatisfied with spinning vague conspiracies about Tether to undermine Bitcoin, he has now moved on to claiming that the crypto market finances slavery at a mass scale via romance scams. His new paper is just as flawed as his effort with Untethered, and if unchecked, could do a similar amount of damage. This man is a dangerous sophist who hides his weak arguments behind the veil of academia — but leaves a trail of real fallout that affects everyone in crypto. It’s time we scrutinize his record.

As a new industry which is largely derided by the establishment, crypto suffers from the problem of academic truth-laundering. What happens is that some anti-crypto academic out to make a name for himself publishes a questionable critical piece of research about crypto — maybe it even passes peer review — and this gets reported as a fact in the press, and then used as justification for anti-crypto policymaking. This happens time and time again.

The reason it’s hard to deal with is that crypto doesn’t have a particularly active academic presence, and the academics that engage in peer review often have no experience with crypto, so they have a hard time spotting a fake. Also, most elite academia is basically an anti-meritocratic clown show and most journals are essentially frauds, but that’s a story for another day.

The reason this is so successful, and you keep seeing these awful, fake papers like Mora et al (2018), Griffin and Shams (2020), or anything by De Vries get referenced, is because they provide a valuable resource to the crypto-critical press and policy establishment: truthy sounding research with an academic sheen.

It doesn’t matter that these articles are completely fabricated nonsense with no relationship to the truth. They provide academic air cover to cynics who hate crypto (the Liz Warrens of the world) or to journalists that are often simply too dumb to understand what they’re reading and simply need talking points.

You know how this works. The direction of travel isn’t truth -> analysis -> policy. It’s desired policy outcome -> half-baked analysis -> fake academia.

One great example is the infamous WSJ claim about Hamas using crypto in apparently large amounts, which was immediately weaponized by Elizabeth Warren as the moral and factive justification for a sweeping anti-crypto bill, which came very close to passing. Never mind the fact that all the major chain analysis companies refuted the WSJ analysis, and that the Treasury Undersecretary later confirmed that the WSJ numbers were false.

We only got the truth in that case because we kicked off a gigantic fuss about it. Although it didn’t undo the fact that Sen Warren almost passed extremely punitive anti-crypto rules based on the analysis. And the problem is that it takes a huge amount of work to actually debunk fake academia and fake news; and it’s far easier to produce plausible-sounding falsities that play to the biases of the critics than it is to refute them.

Another great, if even more egregious example, is the cottage industry of fake academics that create anti-Bitcoin mining content. De Vries is a good example. Virtually none of his work is peer reviewed, and he works for the Dutch Central Bank. But because he knows how to place his work in the Commentary (blog) section of academic journals like Joule, he is able to trick journalists into listening to him.

The journalists that he syndicates his takes out to don’t know this of course. They don’t really understand that academic journals are a dime a dozen, and that 99% of academia is basically fake, p-hacked nonsense. They don’t realize that journals have commentary sections with no or little peer review (that’s where virtually all of De Vries work has been “published”). There’s no meaningful difference between me publishing this article on Medium, and De Vries putting his idle thoughts in Joule’s Commentary section.

Yet of course when he sends his “peer reviewed papers” to journalists, they print his claims without a second thought. And then these claims make it into the policy discussion. This is extremely damaging. Right now, the Biden Admin is trying to use fake academia like De Vries to justify virtually eliminating the Bitcoin mining industry in the US, despite abundant evidence that Bitcoin mining is the most benign (if not salubrious) presence on power grids in the U.S.

Griffin and Shams (2020) is debunked garbage

Griffin (and his coauthor Shams) have a simple approach to their crypto-critical “academia”. They make outrageous, extremely negative claims, and then justify them with an obscure, largely proprietary analysis of on-chain data that is extremely hard to parse. Since none of their peers, nor most journalists, actually understand on-chain data, they aren’t able to spot the fact that the empirical support for these claims is basically gobbledygook. And because the anti-crypto claims are generally popular things to believe (“the price of Bitcoin is fake and manipulated upwards by Tether”; “Bitcoin is rampantly used for crime”), their papers tend to be well received, at least in academia and the press.

It’s worth spending some time on Griffin and Shams because they use the exact same strategy there as Griffin does in his new Pig Butchering paper. Outrageous, ludicrously anti crypto claim, followed by a soup of on-chain data which purportedly supports it. Of course, despite his claimed expertise, Griffin is not an expert in on-chain data analysis. He is a tourist who looks at large datasets and finds illusory relationships to justify claims he’s already decided on. What he does is a semi-sophisticated form of p-hacking. With blockchain data, you can make it say virtually anything you want, especially if your analysis boils down to “these flows are BAD, and there are LOT of them.” John abuses this feature of blockchain data more than anyone I’ve ever encountered.

So we will spend some time of Griffin’s Untethered paper to show just how bad his analysis is, especially as we have around six years of data since his sample period to evaluate his claims with.

In the infamous Griffin and Shams paper, Is Bitcoin Really Untethered, the authors consider whether Tether really is just a useful means of accessing crypto markets, or whether Tether is used to systematically inflate the price of bitcoin via unbacked issuance. They call this the ‘push’ hypothesis.

[U]nder the “pushed” hypothesis, Bitfinex prints Tether regardless of the demand from cash investors, and additional supply of Tether can create inflation in the price of Bitcoin that is not due to a genuine capital flow.

They also speculate that under the ‘push’ hypothesis, the Tether team could choose to wind down Tether if Bitcoin crashes and they can no longer handle redemptions.

[I]f cryptocurrency prices crash, the founders essentially have a put option to default on redeeming Tether, or to potentially experience a “hack” or insufficient reserves where by Tether-related dollars disappear

Unsurprisingly, the authors decide that, yes, indeed, Tether is ‘pushed’ rather than ‘pulled’, effectively alleging Tether of running a ponzi and using it to inflate Bitcoin prices.

They say this plainly:

Our results are generally consistent with Tether being printed unbacked and pushed out onto the market, which can have an inflationary effect on asset prices

The bulk of their “quantitative” study refers to the time period from march 2017 to march 2018, a time period when Bitcoin went from around $1000 to $10k (touching $19k on the way) and Tether supply went from a couple hundred million to $2.3b. As you might expect, if you evaluate the hypothesis that Tether inflates Bitcoin prices on the data from the period where everything in crypto went up, you will probably achieve that result. But of course, that’s a preposterously short window, and all of the subsequent data since then throws their result into question.

The actual evidence Griffin and Shams present for their view is:

• During the sample period (2017–18) Bitcoin prices generally increased after Tether prints
• During the sample period, Bitcoin prices cumulatively rose a lot during hours in which Tether had the highest flows
• Tether flows and Bitcoin returns were high at times when Bitcoin was trading at round numbers (i.e. $1000, $2000)
• Bitcoin had negative price performance at the end of the month, indicating that Tether was selling Bitcoin to raise cash to make attestations each month (yes… they actually posited this)

The academics conclude that:

By mapping the blockchains of Bitcoin and Tether, we are able to establish that one large player on Bitfinex uses Tether to purchase large amounts of Bitcoin when prices are falling and following the printing of Tether

And they decide that the ‘push hypothesis’ — aka, unbacked Tether issuance to support the price of Bitcoin, probably by a single large entity — is the right one.

And the cherry on top of this paper, is that it was published in the Journal of Finance, aka the literal most prestigious finance journal in the world.

These outrageous claims were of course repeated uncritically in the press, with Bloomberg reporting breathlessly that a Lone Bitcoin Whale Likely Fueled 2017 Price Surge, among many others. Griffin’s claims also served as the basis for a frivolous $1.4T lawsuit against Tether and Bitfinex. His paper bears more in common with the fabricated Steele Dossier used to undermine Trump’s 2016 election with claims of Russian collusion. This form of academia isn’t just the benign pursuit of truth: it serves as potent fuel for lawfare against Tether, Bitfinex and other market participants; it helps the press delegitimize the crypto space by allowing pundits to claim that the valid appreciation in Bitcoin was due to pure manipulation; and it gives policymakers air cover to harass vital pieces of crypto infrastructure. This isn’t academia: it’s a naked attack on the industry.

The only problem with Griffin and Shams’ analysis is that… it’s completely bunk.

Starting with Griffin and Sham’s evidence. The sample is very small (again, about a 12 month period during a crazy bull market), so of course Bitcoin returns and Tether issuance were correlated. Tether represents the balance sheet of crypto market participants, and balance sheets were obviously climbing during the rally.

The flow hypothesis is explained by virtue of the fact that obviously there’s more on chain settlement when price is volatile. And if you accumulate the hours of the most flows, you will obviously isolate the hours with high returns, and get a high cumulative return figure, especially during 2017–18 when the crypto industry 10x’ed in size.

The round number thesis is just gibberish and can be immediately dismissed.

The end of the month thing is cute, but we know that Tether actually has the cash, so it’s also irrelevant. And at best it’s extremely circumstantial.

In terms of the more conceptual arguments made by the paper, we know they are false, for a few reasons.

1. Tether has processed major redemptions

Since the Griffin and Shams paper, Tether has processed tons of redemptions. The Tether supply drew down by 36% in Oct/Nov 2018 without incident, so it clearly wasn’t unbacked. In May-July 2022, Tether processed $17b in redemptions without issue. If it was unbacked, it wouldn’t have been able to redeem. Large Wall St firms engaged in massive Tether shorts over the years to no avail — they would have had all the incentive in the world to discover that Tether was unbacked and leak that information. They weren’t able to, because it wasn’t unbacked, and always processed redemptions smoothly.

2. Tether has the cash

Tether had an accounting issue where Bitfinex was hacked, and they gave a loan to Bitfinex to cover some of the $850m losses, which was then repaid. This didn’t actually cause any issues with Tether, and they settled with NYAG for the paltry sum of $18.5m. This was a misstep, to be sure, and depending on whom you ask it was either a valid legal transaction, or accounting fraud, but it certainly wasn’t the kind of malfeasance Griffin and Shams were alleging. They weren’t even accidentally right.

After the settlement, Tether agreed to supervision from NYAG for two years with no issues — clear proof that they had the funds (there’s no auditor more compelling than the NYAG itself). More recently, Cantor Fitzgerald CEO Howard Lutnick has taken to loudly proclaiming on TV that Cantor (service provider to Tether) has seen the financials, and they are sound. Tether also continues to release quarterly attestations via audit firm BDO. Tether has the cash, and that’s not disputed by anyone serious.

3. The analysis has never been validated or replicated

At no point since 2018 has the mysterious “lone actor” that Griffin and Shams allege influenced Bitcoin prices with fake Tethers been identified — because there is no such actor. Griffin and Shams, being rank amateurs, simply misread the on-chain data. Tether did not collapse despite two 80% selloffs in the crypto market. Tether did not collapse despite decabillion redemption events or redemptions that took down over a third of supply (let’s see a bank do that). At no point has any actual manipulation by Tether regarding Bitcoin markets been proven. Tether has only become more transparent, more reputable, more surveilled, and better understood.

Griffin and Shams’ research has not been replicated, validated, or proven in any way. It has only been disproven by subsequent events.

4. There is a very simple explanation for this behavior

The core claim by Griffin and Shams is that it’s somehow unusual or evidence of ‘manipulation’ that Tether prints after Bitcoin sells off. However, this is explainable with recourse to simple dip-buying activity — namely, people have cash in fiat, and they want to buy Bitcoin, and go via Tether when the opportunity looks good. So they create Tether with cash, and buy Bitcoin. The other explanation could simply be that participants find a safe haven in stablecoins when they see Bitcoin selling off — or even more directly, the outflow from Bitcoin to Tether is what causes Bitcoin selloffs. So Tether rallying on Bitcoin dips and Bitcoin rallying after Tether creations isn’t particularly surprising or nefarious. Tether represents the balance sheet of market participants and a waystation between fiat and Bitcoin.

5. Tether clearly does not support Bitcoin prices

Since the short sample period in the Griffin and Shams paper (when everything was going up anyway), there have been plenty of occasions in which Tether supply and the price of Bitcoin have gone in opposite directions. This undermines the core G&S claim — that printing unbacked Tethers was sufficient to buoy the price of Bitcoin. If it was so easy to support BTCUSD by printing Tether, why did Bitcoin sell off from $19k to $3k in 2018–19? Why did Bitcoin sell off ever? Why have the price of Bitcoin and the supply of Tether gone in different directions on many occasions? Why didn’t the lone actor just print to infinity? If the lone actor was printing with unbacked Tethers, why didn’t Tether collapse when Bitcoin did? G&S don’t have answers to these questions.

Let’s look at a few periods in Bitcoin’s history to show how asinine the view is that Tether has any role in supporting Bitcoin prices.

From Jan 2018 to Oct 2018, Tether supply grew from $2b to $2.8b. In that period, Bitcoin fell from $13.5k to $6.5k. Where was the mysterious Tether whale supporting the Bitcoin price?

Tether began November 2021 at $73b and was at that exact same level in March 2023. Over that period, Bitcoin fell from $59k to $24k. Where was the mysterious price-supporting Tether whale?

Obviously, these are just price observations rather than fancy statistical regressions, but you don’t actually need advanced econometrics to understand the facts here. (Given that I have a Masters in econometrics, I could put this in regression format and publish a fake academic paper. But any idiot with STATA can ‘prove’ anything at all with a regression. The important thing is that the data matches intuition and financial theory.) The fact is, Bitcoin has sold off plenty of times when Tether is rallying. Griffin and Shams did their analysis when both Bitcoin and Tether were going up together (as everything in crypto was parabolic at that time), and so it was easy for them to “prove” that their movements were correlated. However, we have six additional years of evidence since them showing that they were simply wrong.

6. Other academics have convincingly pushed back

Most notably, Viswanath-Natraj and Lyons evaluated Griffin and Shams and found that their analysis was, basically unsound. First of all, they find no actual evidence that Tether issuance was followed in the near or medium term by Bitcoin or ETH appreciation, so there’s that.

They explain stablecoin issuance not with recourse to a mystery whale printing Tethers to buoy the Bitcoin market, but with a much more straightforward explanation: Tethers are printed when Tether trades at a premium, which can appear during bullish phases, or when the market is in safe haven mode. Everyone who understands stablecoins knows this. Of course, Griffin and Shams make scarce mention of market makers, arbitrageurs, authorized participants, or these more precise mechanics of Tether creation and redemption in their paper. Because they are ignorant and unfit to have an opinion on the topic. For an actual paper explaining the fundamentals of stablecoin issuance and the factors that drive it, consider Viswanath-Natraj and Lyons.

G&S’ peers haven’t exactly come to their defense. For such a blockbuster paper in the Journal of Finance, you might expect it to have been corroborated at some point. But this didn’t happen. No one has proven Tether-Bitcoin manipulation — despite billions of dollars of Tether shorts from sophisticated hedge funds, who would have loved to prove such a thing. No one has identified the mysterious “lone actor” that Griffin and Shams think manipulated Bitcoin with Tether prints. It appears to have been a chimera that they dreamed up, or an misunderstanding how OTC desks work. No one, not the NYAG or any other court, has proven that Tether “printed unbacked Tethers” for any reason whatsoever. The NYAG found that Tether wasn’t truthful about their loan to Bitfinex, but they didn’t find the kind of wrongdoing that Griffin and Shams allege — and they found nothing thereafter, despite an inordinate amount of political pressure of Tether. Tether was emblematic of the crypto industry, and it has been the primary target of attacks from hedge funds short crypto, the press, academia, and the policy and law enforcement establishment. The fact that all of these entities looked as hard as they could, for the last decade or so, and came up empty, is more evidence that Griffin and Shams were chasing a boogieman.

The simple explanation here is that Griffin and Shams just have an anti-crypto animus, and decided to make up a fantasy which they veiled in fancy math to explain why Bitcoin was going up so much, with Tether the culprit. In the mind of the crypto critic, the success of Bitcoin can never be due to real factors, like actual, real-world demand. It has to be due to fakery and manipulation. This is why the Tether truther brainworm is so prevalent. It convinced an entire generation of investors that Bitcoin was only going up because of “fake Tethers”, and not because Bitcoiners were actually right. The critic and the hater can never accept that they are wrong. If the market proves them wrong it is because there is a crime being committed, a conspiracy, some kind of trickery. They can’t admit they were just plain wrong, and millions of people worldwide simply want what Bitcoin is selling. Untethered is the most successful piece of literature in this genre of Bitcoin Cope. Unfortunately for G&S, hindsight has showed us that their analysis was unsound. This hasn’t stopped them though.

John Griffin uses his academic platform for self- promotion and profiteering

John Griffin’s academic poverty is matched only by his own arrogance. Just read this simpering story in Fortune about him from 2023 — if you can get through it. The article describes him as a “6-foot-2 former high school football star [who] views himself as a crusader for good, a moral sleuth who, as he tells Fortune, “looks to expose financial evil, to shed light on the world and expose dark things in the markets.’ ” Definitely a dispassionate and truth-seeking academic right there.

And the very fact that he reached out to Fortune to get a sycophantic profile written (that’s how these things work, journalists don’t just randomly profile mid-tier academics) is remarkable. He probably knows that his 2020 Untethered paper is bunk given all the evidence we have received since then, but he actually dug his heels in and had the nerve to take a victory lap. In the Fortune piece he claims that the utterly benign and unremarkable price action in 2023 is “very suspicious. The same mechanism we saw in 2017 could be at play now in the still unreal Bitcoin market.” Like a Sasquatch chaser who got a writeup in local news once for allegedly spotting the beast decades ago, Griffin has kept at it and now sees his mythical manipulator around every corner.

He continues to be deliberately blind to the reality of Tether, maintaining over and over again that Tether is used only as a lubricant for crypto speculation. In the Fortune article he claims that “Bitcoin and Tether aren’t used for buying things like cars and pizzas, they’re used for buying other coins. So in that closed system, a relatively small amount of manipulated buying, spurred by creating new coins from nothing, can cause an outsize increase in the Bitcoin price.”

Of course, this is false. Tether is widely used in emerging markets like Argentina, Nigeria, Turkey and many others as a digital dollar equivalent. Right now, Nigeria is experiencing a panic of sorts as they grapple with crypto-dollarization driven mainly by citizens engaging in Tether-powered currency substitution. These individuals use Tether as a savings device, a means to access crypto yields, a tool for remittances, payroll, b2b transfers, cross border trade invoicing, and other conventional uses. Griffin refuses to acknowledge these facts, because they undermine his story that Tether is created only to jack up crypto prices.

In fact, the data clearly shows that Tether (and other stablecoin) usage has increasingly become de-correlated with liquidity and trading dynamics in crypto. To be sure, stablecoins still enjoy an important role as crypto collateral, margin, and as a settlement rails between traders and exchanges. But in the data, we see increasing use of stables for non-crypto activities, too.

Griffin would be surprised to learn that stablecoin usage, particularly Tether, grew throughout the bear market both in terms of the dollar value of transactions, and monthly active wallets. This was while every other metric in crypto declined, like price, volume, and so on. If stablecoins had solely been a speculative instrument used to post collateral on exchanges or move funds back and forth, their use would have declined with the broader crypto selloff. But that didn’t happen. They found real product market fit as digital dollar equivalents, particularly in EM. Much of this usage has nothing to do with the crypto economy at all. But Griffin refuses to engage with this reality, either out of his sheer incapability, or because he’s too stubborn to acknowledge that Tether actually has a huge amount of real-world traction.

Because Griffin refuses to engage with today’s stablecoins reality — which is that they are actually quite untethered from crypto market cycles, and have achieved meaningful usage in the real world — he remains completely ignorant about his own subject, and unfit to write about it.

What’s more, Griffin uses his papers as ad space for a company he owns, Integra. I don’t know what the University of Texas code of ethics is regarding this approach, but it’s sketchy at best. On the very first page of his Pig Butchering paper he loudly advertises Integra:

We further thank Integra FEC for use of their bulk tracing tools and for substantial crypto-research support. Griffin is an owner of Integra FEC and Integra Research Group, which engage in financial consulting, research, and recovery on a variety of issues related to the investigation of financial fraud.

Yes, thank you Integra, company that I happen to own! We couldn’t have done it without you!

As to what he’s using Integra for, its not very inspiring. In the paper he says: “We use a variant of the Ether, ERC-20 token, and Bitcoin bulk-tracing tool developed by Integra FEC to simultaneously follow multiple paths.”

This is amusing to me because there are a dozen different software tools and analytics providers that can do this. “Simultaneously following multiple paths” is not exactly a challenging problem. But of course, only Griffin’s Integra was fit for the job. Can’t have him shouting out Alchemy or any of the sophisticated chain analysis software products that have benefited from hundreds of millions in R&D. No, it was Integra that was conveniently uniquely positioned to solve this problem for Griffin. Thanks, Integra.

And lastly, Griffin writes in an emotional and grandiose manner, which is completely unbecoming for what should be sober academia. He clearly thinks he’s on a divine mission to root out evil in crypto, and he uses these papers as a pulpit to preach as much as to inform. It’s honestly embarrassing. Here’s a few choice passages:

From the Pig Butchering paper:

Our findings highlight how the “reputable” crypto industry provides the common gateways and exit points for massive amounts of criminal capital flows. We hope these findings will help shed light on and ultimately stop these heinous crimes.

He self-importantly dedicates his paper to “all pig butchering victims, those defrauded and those enslaved, and especially the victim who gave us the impetus to write this paper.” I’m sure those enslaved are grateful for John Griffin.

As we will cover in the next section, he’s extremely aggressive towards the crypto industry, using his unsupported assertions to harass major players in the space. For instance, even though his $75b figure is extremely dubious, he straightforwardly asserts the following with no caveats:

[L]arge crypto exchanges like Binance, Huobi, OKX act as exit points for $75.3 billion in criminal proceeds.

And he ends the paper with this absolute clanger:

[O]ur analysis shows that the “legitimate” crypto space commonly serves as the entry and exit point to the illegitimate space and in so is facilitating the cheap and easy flow of funds that is the lifeblood enabling both pig butchering and modern-day slavery.

These are absolutely wild, possibly slanderous allegations, especially when he singles out specific firms like Binance, Huobi, OKX, and Tether. They are especially startling when juxtaposed against the utter poverty of his analysis itself, which we will cover in the next section.

On Pig Butchering, Griffin doesn’t know better than Chainalysis

Griffin’s latest contribution to the academic literature is a paper written with Kevin Mei entitled “How Do Crypto Flows Finance Slavery? The Economics of Pig Butchering”. The title is indicative: Griffin thinks that crypto flows facilitate a huge number ($75b worth of these in his estimate) of romance scams (what he calls “pig butchering”), and what’s more, they are carried out by a vast network of slaves, mainly in South East Asia.

As is always the case with this academia truth laundering, the press has already reported uncritically on his numbers, even though the paper is an un-peer reviewed preprint. It’s only a matter of time between Senator Warren uses this talking point to further harass the crypto space.

Griffin holds crypto culpable for both romance scams and this apparent slave network. He doesn’t contemplate in the paper anything like other means of payment. Reading the paper, you’d think that crypto invented both slavery and scams.

Before we even start on his numbers, it’s absurd that he appears to hold crypto responsible for these alleged slave networks, implying that they only exist because of crypto. He says the following:

Though varied in nature, the origin of these scams is often even darker, as the manpower powering the communications is often enslaved in compounds thought to hold 220,000 victims in Southeast Asia. This paper examines how these criminal organizations are financed through cryptocurrencies.

In support of that 220k figure, he cites a UN report which says “Credible sources indicate that at least 120,000 people across Myanmar may be held in situations where they are forced to carry out online scams, while credible estimates in Cambodia have similarly indicated at least 100,000 people forcibly involved in online scams.” The UN is considerably softer in their language than Griffin, and they don’t actually cite any third party for their estimate, just saying that the 220k figure comes from “information on file with OHCHR”. I suppose you can just blindly trust the UN’s assertion. But it remains the case that the source for this eye-popping figure is “just trust me bro”.

It’s also worth noting that, unlike Griffin’s sly attempt to place the blame for all of these 220k potentially indentured laborers on crypto (even the UN report is coy in terms of calling it “slavery”, unlike Griffin), the UN document mentions several different financing mechanisms for such scams. They clearly do not place the entire blame on crypto — and no reasonable person would, because romance scams long predate the existence of crypto. Digital fiat transmission methods have existed for well over 100 years.

One last amusing note is that the very UN report that Griffin cites, itself cites an estimate of crypto scams (in the aggregate) that’s far, far lower than Griffin’s own estimate of flows deriving from romance scams (by a factor of ten). So Griffin’s own estimate is contradicted by a paper he is relying on for a key claim he is making.

Griffin and Mei’s major quantitative approach is the following: they sum up flows from addresses that they attribute to romance scammers, and determine that they represent “$75.3 billion in criminal proceeds” in the 2020 to 2024 time period. They go as far as to say that if you apply a looser clustering heuristic, the figure could be as high as $237 billion. This should be a clue that something is wrong. They’re twiddling a dial that represents the looseness of the heuristic they’re using to determine which wallets they think belong to scammers, and getting results that vary by hundreds of billions of dollars.

The base of their analysis is wallets collected from “online message boards, dedicated crypto-scam reporting websites, and first-hand personal accounts” reported to be associated with romance scammers. They “apply screens through information collected on Etherscan to remove any addresses that are likely unrelated to pig butchering.” Ah, that’s settled then. They looked at Etherscan, so that’s all the false positives removed then. Good job.

They then take those addresses and use “proprietary criteria” developed by Griffin’s company Integra (thanks, Integra!) to determine if flows are attributable to romance scammers. Some of the criteria they mention are:

• A completely arbitrary 5-hop rule
• The coins reach an exchange
• The coins reach a wallet that they think aren’t related to romance scams

Notwithstanding the obvious issue in both advertising his own company in the paper, and relying on an unreplicable black box methodology invented by said company, their disclosed criteria are ludicrous. Coins can go anywhere in five hops. It’s very likely in this case that they are being comingled with flows to brokers.

Taylor Monahan, one of the premier experts on crypto fraud and scams, says that Griffin’s numbers are “a few orders of magnitude off,” noting that the authors have “clearly double counted certain flows and have misidentified OTC / money broker addresses as scammer-controlled.” This second mistake is exactly the one the WSJ made in their erroneously high estimates of Hamas crypto wallets.

In terms of actual estimates, Chainalysis helpfully published in their 2024 Crypto Crime Report estimates of scam volume. Chainalysis is a multi-billion-dollar company with hundreds of employees dedicated to accurately characterizing entities and illicit flows. Their objective is not to minimize illicit flows (nor to exaggerate them) but rather to be as faithful as possible. Needless to say, they have massively more resources than two lone academics, and they face stiff competition from other chain analysis providers and hence have an economic reason to get these things right (unlike the academics, who are clearly motivated by an anti-crypto animus and whose motive appears to be to embellish the data to paint crypto in a negative light).

Chainalysis report that “scamming revenues globally have been trending down since 2021”. They do admit that romance scam activity is trending up, and that tracking these scams on the blockchain can be difficult. As with all on-chain tracking of real-world flows, there’s a significant amount of ambiguity, and anyone that doesn’t acknowledge this should be treated with suspicion.

Chainalysis doesn’t specifically break out romance scam flow volume in their report, but even for the aggregate scam category (which includes other categories like charity scams, giveaway scams, impersonation scams, investment scams, NFT scams, phishing and extortion, and rug pulls), the total aggregate figure since 2020 is $26b.

Again, romance scams are a subset of this, so the $75b to $237b range Griffin and Mei land on is wildly overestimated, likely by a full order of magnitude. Chainalysis in a tweet refer to only $1b in flows attributable to approval phishing and romance scams since 2021 (although this may only be a subset of the romance scam category). I would invite the blockchain analysis firms to publish their own estimates of romance scams, as they did when the WSJ published their excessively high estimates regarding Hamas-related flows.

More fundamentally, blaming a financial settlement network for romance scams, or for slavery, is just asinine. If crypto didn’t exist, do you think scammers would just stop? Clearly, they simply use other digital payment networks. It’s not like the world is going to lose the ability to transact at a distance. Griffin might say that there’s something unique about crypto that makes it a great conduit for fraud or crime, but the very fact that these transactions are so traceable and leave a permanent record (and in the case of stablecoins, can be frozen by the issuer) is why you don’t want to use crypto for crime. Indeed, law enforcement officials with crypto expertise frequently say that they would prefer that illicit activities be conducted on crypto rails as opposed to using wire transfers through a network of banks. Just recently, OKX and Tether were able to freeze $225m USDT associated with romance scams. This speaks to the usefulness of stablecoins for law enforcement purposes. Try that with cash, or try sending a subpoena to a cluster of international banks, and see how fast you can interdict funds.

For sure, there are some genres of fraud and harm that are specific to crypto itself, like phishing scams that target funds held in browser wallets. But romance scams are a kind of fraud that can be perpetrated via many types of payment networks, whether it’s p2p apps like Venmo and Zelle, or wire transfers or the physical delivery of cash. If Griffin wants to hold crypto particularly culpable for romance scams, he needs to establish that there’s something unique about crypto that makes it particularly bad here. But beyond his erroneous flow estimate, there’s just nothing there.

John Griffin abuses blockchain analysis to support a conclusion he has already decided on

In both Untethered and Pig Butchering, Griffin does the exact same thing: he looks at masses of blockchain data, declares that certain nodes are bad (in Untethered, it’s an evil manipulator printing Tethers and buying Bitcoin, in Pig Butchering it’s the scammers themselves). And then he uses his super proprietary Integra-assisted advanced blockchain analysis to connected those clusters of addresses to many, many more, and he derives eye-popping flow data from that. Then out of this toxic stew he is able to come up with outrageous figures and conclusions; in the case of Untethered, it’s that a single actor was responsible for printing Tether and manipulating upwards the price of Bitcoin; in Pig Butchering, it’s that these scammers have apparently harvested $75.3b from their victims.

The problem with this is of course that the blockchain is a massive network of interconnected nodes, and if you get even one tiny detail wrong, such as mistaking a market maker, OTC desk, or exchange for an individual user, you will get a massively overestimated figure. And this is what he’s doing. He’s being extremely careless with false positives and badly polluting his dataset. This is why we generally leave this stuff to the professionals, because blockchain analysis is very hard, and without an extremely high level of entity characterization, these blind estimates are completely worthless. And Griffin definitely makes this mistake!

By including tons of false positives in his dataset, Griffin erroneously decides that Bitcoin is manipulated by a single entity (which no one has ever confirmed), and he finds an estimate for pig butchering flows that are many times greater than the state of the art estimate from Chainalysis. But this isn’t a problem for Griffin, because he feels that he’s on a God-given mission to root out the evil of crypto. So landing on an extremely inflated figure that makes crypto look extra bad is just fine to him! It’s all in service of a just cause, after all.

This is interestingly the exact same problem that plagued the misleading WSJ coverage of the Hamas crypto flows. The WSJ was relying on a naïve analysis of addresses published by Israeli intelligence and simply added up all the flows they could find relating to them, without ever stopping to wonder who those addresses belonged to. As Chainalysis and Elliptic later confirmed in their corrections, some of these addresses were brokers that had many types of flows, both licit and illicit. So naively adding up all those flows and ascribing them all to Hamas was totally wrong. This was later confirmed by the Treasury Undersecretary Nelson in sworn testimony in Congress.

At the end of the day, as I tried to point out during the WSJ/Hamas affair, there is considerable ambiguity when you try and ascribe real-world flows to blockchain data. It’s extremely easy to overestimate flows from clusters of addresses, especially when you are using naive heuristics. And yet time and again we see academics pollute the popular and policy discourse by tying [bad thing] to these inflated flows. A toxic brew of poor or lacking peer review, gullible journalists, and hostile policymakers means that these exaggerated or false claims tend to have a huge amount of staying power, and are often used to justify bad and hostile policy versus the industry.

The curse of legibility

If you consider the WSJ/Hamas affair, Griffin and Shams, or Pig Butchering, the problem is the same. Blockchains are the most legible financial databases in history. Any member of the public can read the entire history of transactions of anyone that has ever used a blockchain — although they don’t necessarily have the overlay between entities and addresses. No other financial database is as transparent, and this is crypto’s curse. Any third party can simply look at on-chain data and tease out sinister conspiracies based on their interpretation of the data. If JPM’s correspondent bank database was public, I’m sure plenty of academics and internet sleuths would be poring over that too.

So crypto is held to a far higher standard than any other transactional medium, especially when illicit or questionable activity is concerned. Because the moment wrongdoers are identified on chain, all of their flows can be derived (and over-extrapolated, as is the case here). Even though the scammers have used many different transactional media over the years (and this stuff dates back decades), the crypto transactions get the most attention, quite simply because bank records aren’t available to third parties. So crypto takes all of the blame, because it’s legible, and other transactional tools are not.

But what’s interesting is that the fairly sophisticated reporting apparatus that has emerged in crypto, and its intersection with law enforcement, has meaningfully impaired the ability of these scammers to operate. Hamas stopped soliciting donations after their donors were targeted. The Tether/OKX seizure of $225m in USDT associated with romance scams was quite material and would have caused the scammers to think twice about using stablecoins. Rooting out these scams and freezing and seizing them is getting more practical by the day, as law enforcement gets more sophisticated about blockchain analysis.

Thanks to Taylor Monahan for her feedback on this article.

Last week, Bloomberg posted an article entitled “Crypto Fans Lured by 20% Stablecoin Yields Even After 2022 Bust” [archive link], which was followed by Matt Levines’ column [archive link] which riffed on the article. Having talked to Hannah Miller for the piece, I was excited to read the article, but I was immediately disappointed at the editorial tilt. I know journalists don’t necessarily write their own headlines, and this sometimes causes the tenor of otherwise-good coverage to end up skewed. Regardless, the coverage, both from Bloomberg’s crypto desk and subsequently Matt Levine sorely misrepresented the issues, and I feel compelled to clarify them.

The Bloomberg article opens with “The same type of investment product that led to widespread disaster in the cryptocurrency market in 2022 is proliferating once again.” Right off the bat, this is false. The authors are referring to the infamous Terra/Luna protocol, whose UST “stablecoin” went from a nominal value of $18b to virtually nothing in spring 2022. The article in question covers tokenized US treasuries /interest bearing stablecoins like Mountain’s USDM or Ondo’s USDY, and Ethena’s USDe. None of those systems bear much resemblance to Terra’s UST.

No one to my knowledge is trying to recreate UST today (nor should anyone in their right mind do it). The uncollateralized UST stablecoin that was backed mainly by Luna — which itself was a pseudoequity derivative of the system — was doomed, and some, including myself, warned about this. For instance, I posted this tweet thread taking aim at Terra on April 21. On May 8th, Terra/UST collapsed.

nic 🌠 carter on Twitter: "of all the bad ideas the crypto industry has come up with algo stables are among the worst / Twitter"

of all the bad ideas the crypto industry has come up with algo stables are among the worst

I also warned about it on my podcast repeatedly prior to the collapse, and directly warned many of my colleagues and friends in the industry. I say this not to pound my chest — I’ve done enough of that — but to be extremely clear that I actually did see Terra coming, I understood the flaws in the system, and I identified the risk before it collapsed. (For a detailed postmortem of Terra/Luna, see my piece All Falls Down coauthored with Allen Farrington.)

At the time, few VCs were willing to speak up about Terra, because many of them were in the trade, or didn’t want to jeopardize the possibility of investing in the Terra ecosystem if it did succeed. (As an aside, this is why virtually no VCs were willing to publicly criticize SBF (as I did), because he controlled the largest pool of capital in crypto at the time). I’ll note that Bloomberg was mainly a cheerleader and stenographer for SBF. And I don’t think they foresaw the failure of Terra either for that matter.

So as we’ll cover here, when Bloomberg implies I’m investing in the next Terra/Luna, they are being unfair in a few ways:

• Most egregiously, implying that all interest-bearing stablecoins have the same risk profile as UST, which was a completely bonkers outlier
• Implying that I lack the ability to evaluate the risk of an interest-bearing stablecoin, when I was one of the few voices that accurately identified the Terra collapse before it happened
• Implying that I am being duped by founders, when it’s in fact me who has been championing the notion of a (responsible) interest-bearing stablecoin for months

If this sounds personal, it’s because it is. Here’s a few passages from the Bloomberg article:

The spectacular implosion of the TerraUSD stablecoin — and the subsequent string of bankruptcies of firms dependent on the nearly 20% yields once offered by a related project — laid bare how risky such investments can be.

Yet memories can be short in the crypto world. […]

As crypto players aim to profit off of the renewed interest in stablecoins, they’ve turned to yield-bearing products that some worry may cause the tokens to be much-less stable than their name implies. These projects, some of which offer interest rates of more than 20%, have raised concern that holders will be stranded with worthless tokens if the mechanics behind them collapse — which was the case with TerraUSD.

The article goes on to quote myself and Martin Carrica, the founder of Mountain Protocol, a company where Castle Island led the seed.

Nowhere in the article do they mention an un- or under-collateralized stablecoin, which UST was. They mention Mountain, Ondo, and Ethena’ USDe. Mountain’s USDM and Ondo’s USDY are fully collateralized stablecoins that simply pass along the interest from the underlying treasuries held in reserve (more detail on Ethena later). This is exactly how the biggest stablecoins work, with the difference being that the issuer passes along most of the interest to holders, rather than keeping it for themselves. This is, obviously, an improvement from an end user perspective. Would you rather keep all your cash in a checking account that pays 0%, or in a money market mutual fund that pays 550 bps? This is the exact same thing, except the issuer liabilities circulate on a public blockchain, as opposed to a bank ledger.

That stablecoins haven’t done this historically is because stablecoin holders could tolerate the lack of yield when rates were low, and there was a high “convenience yield” for cash instruments on chain. Basically, the opportunity cost of forsaken yield was bearable since there were abundant ways to make money with stablecoins with yield farming and so on. However, as crypto rates fell and the treasury rate climbed, capital was sucked out of crypto and back into tradfi (as evidenced by USDC’s declining market cap). This more than anything convinced me that stable issuers would be compelled to externalize yield from their underlying asset portfolio.

Today, tokenized treasuries total $850m, up from almost nothing at the state of 2023, although not all of them are issued in the stablecoin form factor. But these are all virtually the same kind of instrument.

These are straightforward products. An issuer holds treasuries and they distribute the interest to holders. They differ in terms of regulatory approaches and compliance/permissioning.

It goes without saying that USDM or USDY have nothing in common with UST, aside from the ‘stablecoin’ title (which I’m against anyway, as evidenced by my ill-fated campaign to rename these instruments cryptodollars). If you had asked me, I would have said that Terra’s UST was not a stablecoin at all, but an unbacked liability of the Terra Foundation. Bloomberg is clearly going for gasps when they conflate the two. But the similarities are merely cosmetic.

Mountain passes along 500 bps of interest derived from an underlying Treasury portfolio. UST paid 2000 bps of interest based on Do Kwon’s ability to raise capital.

The risks to something like a USDM are the conventional risks that all stablecoin issuers run: namely, smart contract risks, the (slim) possibility of a liquidity crisis if there’s a run or a problem with a service provider, and the standard regulatory risks. The actual “yield” is derived from holding short-dated Treasury securities. (And by the way, in Mountain’s case, these are held in a FBO Trust which is bankruptcy remote, which is the optimal structure for a stablecoin). If Bloomberg thinks 3-month treasuries are as risky as Terra’s ability to honor an arbitrary 20% rate in perpetuity, they have a very poor opinion of the US government indeed. It’s possible the US government could default. But if it does, we have much bigger problems than the <$1b tokenized treasury sector.

On the other risks, I firmly believe that the Mountain team has appropriately developed processes to manage all of these, and they have my full confidence, which I why I led the round in the first place. The system is working. Today, they are the largest interest-bearing stablecoin, after only a couple months of existence.

Beyond the incorrect conflation of USDM with Terra’s UST, the article also quotes a purported expert who completely misrepresents the regulatory strategy of Mountain at a minimum. Bloomberg quotes Michael Selig, a biglaw partner, who says:

[Stablecoin issuers] attempt to avoid the registration and licensing requirements by only issuing and redeeming directly offshore, but allowing the stablecoins to flow into the US through secondary sales.

This is just blatantly incorrect. Mountain is a Bermuda company. In fact, they were the first stablecoin licensed with the Bermuda Monetary Authority as a Digital Assets Business. They don’t do business with US clients (based on the Reg S exception, which they judiciously adhere to). So there is no notion of “avoiding registration or licensing requirements”. Last I checked, foreign companies that aren’t based in the US and don’t serve US clients aren’t beholden to the SEC or any other US regulator. Contrary to Selig’s assertion, in no way is Mountain “unregulated”.

The BMA is no fly by night regulator. They are a highly sophisticated, top-tier regulatory body which oversees one third of the global reinsurance market, and has issued digital asset licenses to some of the largest crypto firms on the planet like Circle and Coinbase. According to KnowYourCountry, Bermuda is the 16th ranking jurisdiction globally for AML and sanctions compliance, ahead of the US or the UK.

Contrary to what Selig suggests, Mountain maintains strict protocols that are designed to stop the flow of tokens on the secondary market to US individuals. These include on-chain address monitoring for addresses connected to US persons, off-chain monitoring (including social media, blogs and news), as well as deterrence measures by which holders breaching terms and conditions can have their assets frozen. If a major US exchange were to list USDM, Mountain would explain to them that they are in breach of the T&Cs and, if necessary, freeze the tokens (although I don’t expect it would ever come to that).

And what’s more, an interest-bearing stablecoin isn’t particularly relevant for US individuals, since it’s trivial for an American to get access to a money market fund or a high yielding savings account. This is a product that, definitionally, is useful for someone that doesn’t have direct access to the US financial system.

If Selig or anyone else thinks that Mountain’s registration in Bermuda is tenuous, they just need to listen to Bermuda leadership. Because the US maintains a hostile stance towards stablecoins, dollar stablecoin issuers are going overseas and being welcomed by high-quality regulators like Bermuda, Singapore, or Hong Kong. This evidences the blatant failure of US leadership, but that’s a story for another day.

When I went to Bermuda for their fintech summit last year, the only crypto project Premier Burt mentioned in his remarks was Mountain. The Bermudan regulators are proud that they’re attracting the best fintech/crypto talent with their progressive regulatory regime, and they should be. The US’ utterly shambolic treatment of crypto means that startups and capital will continue to flow abroad, and I’m not at all ashamed of the fact that I’m participating in this exodus.

Today, most stablecoins (roughly 75%) are crypto-eurodollars, which is to say, they are dollar stablecoins issued offshore. If Selig thinks a dollar stablecoin that doesn’t serve US clients has registration requirements vis a vis the SEC or others, he should identify the law that says that.

On Ethena, it could be the case that Bloomberg was referring to them with the Terra mention, but that’s still an unfair comparison. Ethena is a synthetic USD token which works by combining a long staked ETH position with a short futures position. Staked ETH pays a yield, and short futures normally (but not always) pays a hefty rate (because in crypto markets there’s normally a structurally positive demand for long-biased leverage). The two combined create a delta-neutral USD position which maintains a positive carry. Granted, this is a different flavor of risk than a mere tokenized US treasury, but USDe holders also get a higher yield in exchange for this. And it has vanishingly little in common with Terra’s UST, which offered a completely arbitrary 20% yield based on … nothing. It’s like comparing a high yield credit fund with Madoff’s scheme. And, for what it’s worth, I’ve spent months diligencing the Ethena system and believe it to be extremely well thought-through and very sound — albeit more complex than something like USDM.

Adding fuel to the fire, Bloomberg columnist Matt Levine piled on as well. He quoted the section from his colleagues’ article where I was quoted, and added the following:

Even beyond stablecoins, there is a general category of financial story that is like “somebody convinces venture capitalists that reinventing banking is a good idea,” and those stories often end up in, uh, interesting places. (A lot of fintech stories are that sort of story, but the story of FTX is really that sort of story.) Reinventing banking is a great idea! You can make a lot of money, pretty consistently, by taking risks that your venture capitalists don’t notice. “We’ll take deposits, pay interest, invest the deposits, earn more interest than we pay, keep the spread and not be subject to bank capital or prudential regulation” is a great trade some large portion of the time, surely large enough to raise some venture capital.

This again is quite unfair, and unbecoming for Levine, who is normally excellent. We’re not reinventing banking. Mountain is not “not subject to prudential regulation” (they are regulated as a Digital Asset Business by Bermuda which has a specific expertise and body of work on digital assets dating back years). The “venture capitalists” (aka, me) aren’t ignorant of the risks.

Martin Carrica, the founder of Mountain, didn’t convince me that interest bearing stables are a good idea. I’ve had conviction on the idea for well over a year now (and have looked at many startups in the category). I went on a speaking tour last fall making the case for interest-bearing stables, keynoting Token2049, Messari Mainnet, and Boston Blockchain week, where I gave talks explaining why I think they’re a good idea. I haven’t been duped in any way, and I’m fully apprised of the risks here.

And I don’t think we’re reinventing banking for that matter. We’re merely moving the database tracking the liabilities away from a bank or an asset manager’s back office to the blockchain. There’s particularly nothing new or unusual — at least from a risk perspective — about what tokenized treasury issuers are doing. What’s different is that the liabilities freely circulate on a public global ledger, opening up access to virtually anyone worldwide. I believe to be an unambiguously good thing, and have placed my bets accordingly (in the arena trying stuff, etc.) I believe both Mountain and Ethena have been extremely transparent about the risks, and are going about them in a responsible way — while providing a superior experience to users.

USDe or USDM bear no similarities to UST besides the stablecoin title. They aren’t the same thing at all, and it’s completely irresponsible for a financial journalist to make the conflation.

I’ve talked to Bloomberg journalists, no exaggeration, hundreds of times over the last six years. They have been kind enough to invite me on TV more than a few times. I consider some of their journalists and analysts good friends. And this is why I’m extra disappointed at their recent, very strong, anti-crypto tilt. I expect sophisticated coverage from them, and I know they can do much better than this.

As I have made clear, as discussed in my Messari Mainnet presentation, supporting dollar stablecoins is in the national interest. Stables export the dollar globally, creating strong benefits for the large fraction of the world’s population that struggles with unstable banking systems, weak currencies, and inferior property rights. Interest bearing stables are the next evolution and further enhance the value proposition of these instruments. The journalists that cover these systems would do well to treat them fairly, rather than writing histrionic missives chasing the hobgoblins of yesteryear. We don’t ask for preference or a rose tint: just the truth, and a fair assessment of the risks and benefits of these systems.

Nothing in this piece should be construed as investment advice. Mountain’s USDM is not available to US individuals. CIV disclosures. Personal disclosures.

<< click here for this article in PDF format >>

On October 10, following the attacks by Hamas on Israeli civilians, Angus Berwick and Ian Talley of the WSJ published the article “Hamas Militants Behind Israel Attack Raised Millions in Crypto”. The WSJ followed up with “Why Hamas Uses Crypto to Raise Money” and then published an op-ed from Sen. Elizabeth Warren entitled “Cryptocurrency Feeds Hamas’s Terrorism”. Sen. Warren then wrote a letter to the White House and Treasury signed by over 100 members of Congress, reliant entirely on the WSJ claim that Hamas and affiliates had raised “over $130m in crypto” to fund their operations since 2021. Warren is now leading a campaign to aggressively marginalize the crypto industry in the US based on her provocative claim. Yet subsequent evidence has thrown her entire empirical basis into question.

The WSJ initially refused to update or retract their story, but subsequently issued a correction on October 27, after Elliptic released a blog post saying that the WSJ had misinterpreted their data.

On October 26, fearful that the WSJ would not do their jobs and would refuse to investigate the actual number regarding Hamas’ crypto fundraising, I created a bounty program designed to solicit blockchain analysis from the crypto community. I had initially earmarked $10k for the bounty, but I received third party contributions and the pot swelled to $50k (with much of it still outstanding). The objective wasn’t simply to rebut the WSJ’s reporting; but to dig into the data and get to the bottom of the story — whatever the answer was. If my analysts discovered that Hamas had genuinely raised $130m in crypto, as Sen. Warren claimed, I wouldn’t have concealed that.

As it has now been a week since the bounty was created, and I have reviewed over a hundred submissions, I wanted to take stock, summarize the state of our knowledge around this question, and showcase some of the better answers I’ve received.

Unfortunately, I don’t have a concrete anwer to give you as to the extent of Hamas fundraising with crypto. I’m not sure any entity aside from Hamas itself knows. I doubt even that the Israeli intelligence have a perfect estimate. (Although I would encourage firms like Chainalysis and TRM to further weigh in with their own specific estimates.) What we can say confidently however is that the $130m figure so confidently declared by Senator Warren is not supported by the evidence we have at our disposal. In the remainder of this article, I’ll explain why I think this is the case.

Timeline of events:

• July 2022: Israel’s National Bureau for Counter Terror Financing (henceforth NBCTF) issues ASO 15/22 (link), targeting 47 addresses associated with “Dubai co for exchange” (datasource used by BitOK)
• April 2023: NBCTF issues ASO 19/23 (link), targeting 5 addresses and 77 Binance client accounts associated with “Dubai co for exchange, Al Matahadun for exchange, and Al Wefaq co for exchange” (datasource used by BitOK)
• April 2023: Al Qassam (Hamas affiliate) stops soliciting donations in Bitcoin, citing safety risks to their donors
• July 2023: NBCTF issues ASO 34/23 (link), targeting 26 addresses and 67 Binance client accounts associated with Palestine Islamic Jihad (PIJ) (datasource used by Elliptic)
• July 2023: Elliptic claims that PIJ had received $93m based on addresses disclosed in ASO 34/23
• August 2023: BitOK releases an analysis of addresses found in ASO 29/23 and 34/23
• Oct 7, 2023: Hamas attack on Israel
• Oct 9, 2023: Hamas-affiliated Gaza Now starts a crypto crowdfund, raising only $21k, of which 2k is frozen at Binance and 9k is frozen by Tether
• Oct 10, 2023: TRM releases an analysis finding <$1m raised by Hamas in crypto-based crowdfunding. They do not make an estimate based on the Israeli orders
• Oct 10, 2023: WSJ releases their first article on the topic citing both Elliptic ($91m) and BitOK ($41m)
• Oct 17, 2023: Senator Warren writes a letter which is signed by >100 members of congress, citing the WSJ article exclusively
• Oct 18, 2023: Chainalysis released a blog questioning the methodology without referencing either provider directly
• Oct 25, 2023: Elliptic releases a follow up disputing the WSJ’s approach and the Warren letter
• Oct 25, 2023: BitOK releases clarification of their own methodology
• Oct 27, 2023: Initially defiant, the WSJ ultimately corrects their article, softening their claims, while still refusing to issue a full retraction
• In the Senate hearing on Oct 26, numerous Senators and witnesses continued to question the WSJ/Warren claim that crypto had funded “over $130m” in donations to Hamas or affiliates
• Oct 31, 2023: in a Twitter space, Elliptic CEO Tom Robinson says that he “wants to see the WSJ go further with their retractions” and that he “doesn’t believe that the WSJ has provided evidence to support the title of the article”
• Oct 31, 2023: Sen. Warren doubles down on her claim, aware that the WSJ issued a correct, stating “it’s not about one report” (even though her letter cites only one report)

On the ambiguity of blockchain data

Ever since I cofounded Coin Metrics in 2016, I’ve been looking at on-chain data almost every day. My number one challenge with this type of data is its ambiguity. The essential premise of on-chain data analysis, whether it’s used for risk assessment (as with Chainalysis or Elliptic), or for capital markets (as with Coin Metrics, Nansen, or Dune), is to try and connect real-world identities to on-chain transactions. I understand how difficult this can be. Ultimately, blockchain data is ambiguous, and concrete numbers are few and far between. Many mistakes are made when people attach excessive levels of certainty to on-chain numbers that are inherently vague and complex to interpret.

Because of idiosyncrasies in how blockchain wallets work, and the ease of gaming or inflating on-chain data, it’s very common for third party observers who aren’t deeply familiar with on-chain data to make critical mistakes when looking at it. Occasionally, these reverberate in the press and create persistent myths. For instance, take the example of the infamous Griffin and Shams paper from 2020 published in the Journal of Finance (the most prestigious financial journal) that claimed that a “single large entity” was manipulating Bitcoin markets with fake Tethers, by looking at a lot of blockchain data and finding perceived relationships there. As it turns out, this was not the case, and they were probably looking at the activity of a market maker arbitraging the Tether peg. Subsequent developments showed that their claimed Bitcoin-Tether relationship did not hold.

But people believed the academics, because they backed their faulty analysis with a large amount of inscrutable blockchain data — and a story that many wanted to believe (“the price of Bitcoin is fake, and is propped up by unbacked manipulation relating to Tether”).

In the last few weeks, the same thing has happened with Hamas and alleged crypto funding. The idea of crypto powering the terrorist attacks on Israel was a convenient one for crypto-critical members of Congress, and a narrative blaming crypto for the attacks was quickly contrived. Israeli intelligence did identify clusters of addresses that they felt had some relationship to terrorist financing, but third parties over-interpreted that data and drew conclusions that weren’t supported by the evidence.

Key issues

Having contemplated the totality of the story and the data for a couple weeks now, I feel that there’s a few ways the epistemic pipeline starting with “Israeli NBCTF issues seizure order for specific wallet addresses” and ending with “terrorist groups raised over $130m in crypto” could have been polluted. Specifically, the issues are as follows:

1. Israeli intel is not focused on precision, but on freezing addresses that are terror-affiliated. This is a dragnet that can catch unrelated, or only semi-related third party addresses. Analysts need to be mindful of the difference
2. It is very likely that the addresses in the Israeli intel dataset included third-party brokers that had plenty of non-terror related flows. Elliptic (initially), the WSJ (before their correction), and Sen. Warren in her letter conflate these numbers
3. A “gross flow” methodology on a per-address basis artificially inflates flow data. My analysts have shown that the Elliptic numbers are inflated in this manner (although only by about 12–18%)
4. Wallet flows are not equivalent to “funds raised”

We will dig into each in turn. To clarify the setting, I’m attaching a diagram of how action by Israeli intelligence was ultimately warped into a (now known to be exaggerated) claim by Sen. Warren in her letter that was signed by over 100 members of Congress.

To briefly summarize, a game of telephone took place, starting with Israeli intelligence issuing seizure orders for Tron and Binance addresses, and ending in the halls of Congress. At each step the data became less concrete and the claims more extravagant. Blockchain analytics firms took addresses listed by Israeli intelligence and made aggressive inferences from the data, which were then repeated and extended by journalists, particularly at the WSJ. The press coverage was then once again reinterpreted by Sen. Warren in her letter. By the time it reached Congress, buried under three layers of interpretation, the claims bore little resemblance to the original data, and were presented without ambiguity, hedging, or an acknowledgement of the inherent uncertainty. Elliptic later significantly revised their level of confidence in the data, and the WSJ also issued a substantial correction, thus undermining the major claim made in the Warren letter.

The NBCTF’s Precision versus Comprehensiveness Tradeoff

First, and most importantly, the NBCTF is not an academic organization. They are not focused on specifically singling out wallet addresses which belong to Hamas or affiliates, but rather stopping terror in its tracks. I am speculating here, but I would posit that to them, a false negative is much worse than a false positive. As in, if they determine that terrorists have some on-chain identity, and they fail to designate the entire set of addresses, that’s a much worse outcome than if they tag the terrorist wallets as well as a few unrelated ones. They are presumably focused on minimizing type II error (false negatives), not type I error (false positives).

Their objective is to eliminate terror financing. This means that they are more willing to accept false positives rather than risking missing a few while trying to be surgical. This is really the crux of the matter: NBCTF may not care that they are tagging an affiliate or payment processor or brokerage/exchange that has a history of transacting with Hamas and others, rather than Hamas itself. From their perspective, they’re still intercepting funds earmarked for terror, even if there is some collateral damage. And from a national security perspective, this makes total sense. It’s just that when you strip this information out of that context, and declare that all activity associated with those wallets is specifically dedicated for funding terrorism, you start to introduce errors. It’s Elliptic and BitOK that are making the key mistake, by blindly looking at a list of addresses published by Israeli Intel and deciding that 100% of those flows specifically pertain to funds raised by Hamas.

The NBCTF address listing likely includes data from third party brokers

The crux of the matter is that some of the addresses claimed to be owned by Hamas or affiliates may not have belonged not to PIJ, Hamas or their affiliates directly, but rather third parties, with only a fraction of their volume attributable to terrorist financing. For instance, the two orders the BitOK number of $41m was based on activity not by Hamas or PIJ, but rather “AL Mutahadun for Exchange, Dubai Company for Exchange, and AL Wefaq Co. for Exchange”, which Binance describes as “currency exchange outfits.” This can also be inferred from their names. In Reuters reporting, Al Mutahadun describes itself as “a money exchange company”. This is not to say that they’re not involved in funding Hamas — of course, the reason Israel sought to freeze their funds is because of that specific allegation. But because they operate currency exchange businesses, they may well maintain flows that are non-Hamas related.

So what we’re dealing with here is an artifact of the nature of chain analysis itself. It’s entirely valid for Chainalysis, Elliptic, or Israeli law enforcement to label addresses as terror-finance-affiliated, but this doesn’t mean that all funds flowing through these addresses are specifically funds being raised directly for Hamas. I’m attaching commentary from each of the major chain analysis providers on this specific question. Chainalysis’s entire blog post is focused on cautioning against this specific conflation. They are very careful to point out that wallets that are vaguely terror-affiliated may well be payment processors, OTC brokers, hawalas, or other types of service providers. They actually list 20 such service providers that have interacted with known-terrorist financing wallets, knowingly or unknowingly. As Chainalysis states, “it is often not productive to continue following funds once they’ve been deposited at a service [like one of these money transmitters]”. Here are the key extracts from blog posts by Chainalysis, TRM, and Elliptic (in my opinion, the three most “blue chip” blockchain analysis firms) on this issue:

Chainalysis: We have seen recent estimates related to the attacks on Israel that appear to include all flows to certain service providers that received some funds associated with terrorism financing. In other words, those totals include funds not explicitly related to terrorism financing. Of course, these service providers are supporting terrorism by acting as facilitators, and cutting off terrorist access to them through sanctions or other offensive operations is an important component to disrupting terrorist finance. But it would be incorrect to assume all of the transaction activity conducted by those service providers is related to terrorism.

Translation: Chainalysis is taking aim at Elliptic, and pointing out that they are wrongly bundling in service providers with specifically PIJ-related wallets, and assuming that the entirety of those flows are terror-related.

TRM: According to Israeli authorities, the key address [in ASO 34/23] was controlled by Tawfiq Muhammad al-Law, a Syria-based hawala operator who worked with key Hezbollah and IRGC financiers. TRM identified on-chain links between the seized addresses and entities and exchanges located in Iran, Syria, Iraq and the Gaza strip, all of which have ties to the IRGC and Hezbollah.

Translation: TRM is pointing out that one of the main entities tagged by Israel in the order ended up composing the BitOK estimate was a hawala operator. Hawala refers to a kind of remittance process that works outside of the formal banking system. It’s reasonable to assume that some of their flows were licit or non-terror related. At the very least, we specifically have an example of the kind of service provider Chainalysis is referring to that might be cofounding the analysis.

Elliptic: In July this year, the NBCTF issued a seizure order for crypto wallets linked to Palestinian Islamic Jihad […]. Elliptic analysis of the wallets seized by the NBCTF shows that these wallets received transactions totalling just over $93 million between 2020 and 2023. As we made clear in our research, in no way does this mean that PIJ had “raised” all of these funds or that they even all belonged to PIJ. It is not known what proportion of the funds received by those wallets are directly attributable to PIJ or other terrorist groups. It is likely that some of the wallets listed by the NBCTF belonged to small service providers such as brokers that were used by PIJ.

Translation: Elliptic is pointing out here a difference between tagged addresses showing flows of $93m in the aggregate and PIJ “raising” that amount of money. They align with the view of Chainalysis and TRM that some of the wallets were likely related to service providers. Elliptic was extremely direct in their language and the CEO has stated (on a subsequent Twitter space) that he “does not believe that the WSJ correction went far enough” and that he “doesn’t believe that the WSJ has provided evidence to support the title of the article.”

On-chain fund flows are net, not gross

Another mistake that researchers have made (update: Elliptic disputes these numbers / this section. See correction below) is grossing up incoming funds to wallets and taking their nominal value at face value. This can lead to inflated estimates of how much a cluster of wallets really has. One respondent to my challenge, Mrfti_plus, explained it well in this thread.

If I receive $100 and send it to Bob, and he sends $50 to Alice, and you naively add up all of the inflows at the wallet level, you get a gross inflow figure of $250 (my first $100, Bob’s $100 inbound transaction, and Alice’s $50), even though the total amount of money between us three is only $100. This is an easy way to inflate figures when aggregating on-chain data at the wallet level without considering the relationships between those wallets. If you look at the Elliptic chart (note: the heading to the chart was edited without comment after the WSJ article was published), you can see that it refers to “incoming USD” in the legend.

Elliptic’s methodology involves adding up the flows received by all of the addresses tagged by Israeli intelligence as being Hamas affiliated. But some of these addresses also transacted between each other, driving up the numbers. Once you net out the double-counting, you get a lower figure.

The participants in my Bounty Program contributed a lot of different datapoints to my study, but the most important finding is the following: Elliptic over-counted the flows, even assuming that all NBCTF-tagged wallets were 100% terror-related, by ~$11m.

Here’s a summary of findings by a few participants in my bounty program, compiled by @NFTherder:

Importantly, I asked that the data and analysis be posted publicly. Here’s four Flipside dashboards demonstrating this same finding that anyone can replicate and evaluate:

• MoDeFi
• Sam
• Dayvidjosh
• Deeofweb3

Each of these analyses finds roughly the same thing: Elliptic over-counted in their initial estimate. Moreover, some analysts, like Lamoka, also felt that BitOK overstated their findings (by 10% in their case). Notwithstanding the points made in the prior section about the addresses themselves not being fully attributable, we still find a consistent overstatement of the data. It is therefore clear to us at this point that the “over $130m” claim made by Sen. Warren in her letter is unsupported by the data.

UPDATE 11/09/23: Elliptic disputes these analyses — see below thread:

Tom Robinson on Twitter: "Following on from the bounty program issued by @nic__carter, a number of people have posted analyses of the value of crypto received by the accounts listed in ASO 34/23, many of which appear to be erroneous. So let's take a look at the data... / Twitter"

Following on from the bounty program issued by @nic__carter, a number of people have posted analyses of the value of crypto received by the accounts listed in ASO 34/23, many of which appear to be erroneous. So let's take a look at the data...

Flows are not equivalent to “funds raised”

Ultimately, flows between wallets on the blockchain are not necessarily indicative of actual “funds raised”. Lacking evidence of actual crowdfunding campaigns, it’s not enough to look at flows in a cluster of wallets tagged by Israeli intelligence and determine that those transaction volumes represent new funds raised. The few crowdfunding campaigns we have actually witnessed were relatively meager and have been interrupted. GazaNow, a Hamas affiliate, has raised only $800,000 in crypto in the last two years, and some of these funds have been seized via Binance or Tether.

The WSJ coverage and the Warren letter nevertheless conflates on chain flows with funds raised. Here’s some example of language used by the WSJ and Sen. Warren in her letter:

WSJ headline: “Hamas Militants Behind Israel Attack Raised Millions in Crypto”

My view: this headline doesn’t comport with the reality. We do not know that Hamas “raised money” in crypto. Indeed, both Elliptic and TRM point out that public crowdfunding campaigns by Hamas and affiliates raised less than $1m. Elliptic specifically states in their blog post: “there is no evidence to suggest that crypto fundraising has raised anything close to this amount [$91m], and data provided by Elliptic and others has been misinterpreted.” My verdict: it’s wrong to use the term “raised” here. A more accurate claim would have been “wallets targeted by Israeli authorities as potentially being linked to Hamas showed flows of ~$83m.”

(Title of chart in WSJ article): “Crypto funds received by PIJ”

My view: we know now that Elliptic overstated fund flows via double-counting (see the prior section). We also know that not all of the funds were attributable to PIJ directly. My verdict: this language is not supported by the evidence.

WSJ body text: “Digital-currency wallets that Israeli authorities linked to the PIJ received as much as $93 million in crypto between August 2021 and June this year”

My view: the WSJ moderates their language slightly by noting that the wallets were “linked to the PIJ” by Israel authorities. They do not caveat or hedge sufficiently. My verdict: this claim is not supported by evidence.

Warren letter: “In the months leading up to their brutal and horrific October 7th attack on Israel, Hamas and Palestinian Islamic Jihad raised millions of dollars in crypto — evading U.S. sanctions and funding their operations […] between August 2021 and this past June, the two groups raised over $130 million in crypto”

My view: we don’t have the evidence to support this claim. The $130m claim is based on the gross flows to Israeli-tagged wallets, but as we have discussed, that figure was overestimated by methodological errors on the part of Elliptic, and likely further exaggerated by the presence of third-party brokers in the sample. While we don’t currently have a better estimate, it’s clear at this point that Warren’s claim is not reliable. It’s also a misrepresentation to characterize wallet flows as “funds raised”. All we know is that funds are moving between addresses tagged by Israeli law enforcement as potentially having a Hamas connection; we don’t know whether these were funds were “raised.”

Further context

It’s true that Hamas has been active in trying to raise money in crypto; in fact, they’ve been one of the most active terror groups in trying to utilize the technology.

Nevertheless, crypto appears to be a relatively minor tool in the Hamas financing toolkit. Shlomit Wagman, formerly chair of the Israeli Money Laundering and Terrorism Financing Prohibition Authority lists in her congressional testimony the main sources of funding for Hamas as being state funding (transmitted via cash, hawala, and banks), business portfolios, fundraising (of which a portion is crypto), and stolen humanitarian aid.

Indeed, Hamas’ enthusiasm for crypto seems to be trending downwards, given the effectiveness of law enforcement in interrupting their crypto flows. Hamas’ journey with crypto began around 2019 when they started soliciting crypto donations, and seems to have trailed off in April 2023 when they announced a halt to their crypto fundraising activities (right after an Israeli seizure order went out). The problem with raising funds in crypto is that exchanges very actively seek to freeze suspected accounts; stablecoin issuers themselves routinely freeze funds suspected to be held by terror organizations; and donors can be deanonymized and identified (and perhaps even prosecuted). Transacting with crypto leaves a permanent paper trail, and even if a donor believes that either they or their recipient are pseudonymous at the time, if those addresses are ever subsequently connected to real-world identities, the pseudonymity is broken and the participants can be identified and prosecuted. This is an improvement over the banking system or conventional remittance or hawala channels, where records and databases are not strictly public, as they are with crypto.

Of course, it’s worth noting that physical cash cannot be seized remotely like a stablecoin can, and this is a significant improvement over cash from a law enforcement perspective. Crypto exchanges can be duped into servicing terror organizations (for instance, if they lie on their KYC, which Binance frequently points out), but this is the same for banks. It’s not just crypto exchanges which are exploited for this purpose but banks as well.

If I were a member of law enforcement, it would be my preference that terror organizations solely utilize crypto, rather than cash or banks. This leaves an indelible paper trail, and creates abundant opportunities to seize or interrupt financial flows, which law enforcement have become quite adept at doing. Yes, there is a period of latency while illicit actors achieve a level of sophistication before governments do and are able to transact relatively unencumbered. However, governments today are quite sophisticated at blockchain analysis, and have closed the skill gap. It’s for this reason that we see organizations like Hamas apparently moving away from crypto financing, as opposed to growing their on-chain footprint. This reality cannot be ignored when we consider illicit uses of crypto.

I am republishing this newsletter entry, first published two years ago in my short-lived newsletter, Murmurations. Murmurations was published on Twitter’s newsletter platform, Revue, which has since been deprecated. Because I like this piece, I am posting it here for posterity. Keep in mind this was intended for a smaller, more intimate audience.

***

I grew up in the swamp. Well, more precisely, I spent most of my formative years in the suburbs of Washington, D.C. The question of where I’m “from” is a bit more complicated. I touched on it a bit in my blog On Writing (Ok, I have to come clean — to my embarrassment, I learned later that Stephen King had written an entire book with the same title and some of the same lessons, but of course from the position of a much accomplished writer. I later listened to this on audiobook with my dad on a 1,500 mile drive from Boston to Miami and thoroughly enjoyed it. I can’t recommend it enough.)

Did any of you get the joke in On Writing (the above blog post post referenced, not the book)? I’m not sure anyone did. In the post, I’m scolding myself, as well as my readers, for derailing the process of communicating information from your brain to another — telepathy, as Steven King describes it — with overly flowery language. My point was that vanity impairs clarity. But the post itself is full of unnecessary rhetorical flourishes. So it’s a classic case of myself not being able to take my own advice.

“Vanity of vanities,” says the Preacher;
“Vanity of vanities, all is vanity.”

Ecclesiastes 1:2

George Orwell is actually guilty of the same in his famous essay Politics and the English Language. He makes many of the errors of writing he insists a good writer should not commit. But we still love his essay. So I’m developing a new thesis: perhaps the flourishes are why readers stick around. No one wants to read robotic, syntactically optimal prose. Fair warning: due to a lack of an editor on this new project, I plan to fully indulge in digressions, flourishes, parentheticals, footnotes, asides, and needless rhetoric. Consider this my natural undiluted self.

Back to the swamp. Well — one point of clarification. Washington D.C. was never actually a swamp. George Washington picked it quite deliberately for the nation’s capital as he knew the area well given its proximity to his estate at Mt. Vernon. While D.C. is built around the Potomac, and as any resident knows, the soil in the region is virtually all clay, inherited from the river’s journey over the centuries, the city itself isn’t particularly swamplike. The swampiest era in the history of the nation’s capital came after the Civil War, when poor farming practices caused mudflats to emerge on the outskirts of the city. These were later transformed into the reflecting pool, the Tidal Basin (the place with the cherry blossoms), and much of the national mall where you can find the Lincoln and the Jefferson Memorials (my favorite). If you’ve ever visited, you will know that it can get extremely hot and humid, especially in the late summer months. But it’s no swamp.

By the way, there’s an incredible 18-mile bike trail from Alexandria, VA to George Washington’s estate at Mt Vernon. One of the best things about D.C. is the abundance of fantastic bike trails, many of them running along old canal towpaths or converted from trolley trails. I spent a couple years cycling to work from Bethesda (12 miles north) to downtown D.C. on a fantastic trail called the Capital Crescent. Downhill on the way in, following the curve of the Potomac down to Georgetown. Then uphill on the way back. There’s nothing like cycling uphill in the dark, sometimes in the rain or snow, for 12 miles after a long day’s work. At the time I worked for a magazine covering corporate law of all things. I think I earned around $40k/year at the time but I was quite content.

As any DC resident knows, there’s a spectacular natural landmark not far from the city. I speak of Great Falls, a series of enormous rapids 14 miles upstream of DC. The falls drop 80 feet in under a mile, including a 50 foot drop over 1/10 of a mile. The flow rate increases dramatically as the Potomac is crammed into a narrow gorge. It’s quite breathtaking, especially when the river is engorged after a rainstorm. These are Class V+ rapids, with only the most experienced kayakers able to traverse them. Fatalities are not uncommon. Swimmers in the area also frequently drown. Gigantic signs adorn the banks of the river grimly warning of the number of drownings in the area. The rapids are so dangerous that kayakers are cautioned to attempt them only during off-peak hours so as not to attract attention from park visitors who get dangerously close to the river in their enthusiasm to watch.

There’s a great trail with fantastic views of the falls on the Maryland side called the Billy Goat Trail (Section A is the one you want). It has plenty of scrambling over rocks for those who enjoy that sort of thing. Growing up our family must have done it a dozen times at least, maybe more. My dad had a habit of doing the relatively demanding trail in flip flops. Invariably one would break and he would march on, stoically, with one foot bare, insisting that he was just fine. If you ever visit D.C., take a break from the monuments and give the trail a shot. The first half winds right along the river on the sheer rocky banks, and the latter half takes you back along the canal and towpath.

This is where George Washington features again in our story. Washington, aside from being a general, nation-builder, statesman, and one of the largest landowners in the U.S., was also a huge aficionado of canal-building. Long before he commanded troops in the revolutionary war or assumed the Presidency, he pursued canals with a passion. In 1748, at the mere age of 16, the youthful Washington was hired by his mentor Lord Fairfax to join a surveying expedition in western Virginia. (Washington’s surveying career spanned his entire life; he continued to undertake surveys up until his death in 1799.) This first trip convinced Washington that a series of canals around the Potomac’s impassable rapids was the key to linking the coast to the fertile Ohio valley. He would spend the rest of his life, political career permitting, attempting to realize this vision.

In the wake of the Revolutionary war, Washington formed the Patowmack Joint Stock Company in 1785 in pursuit of this dream. The Company through great adversity eventually completed five canals, including a canal skirting Great Falls, which opened in 1802.

To build the locks descending alongside Great Falls, Washington’s canal required black powder to blast out the rock, one of the first engineering projects in the U.S. to use this technique. Given the general lack of modern engineers in the country at the time, this work was slow and arduous. In the end, the Patowmack Company made a 218-mile stretch of the Potomac somewhat passable, with boats taking the river most of the way, and using the canals and locks to descend impassable portions. That’s why it was called a ‘skirting’ canal — it didn’t replace the river but rather supplemented it during the rockiest bits. The system was nevertheless somewhat rudimentary. As the national park service relates:

Some years there were only about 45 days when the water reached a sufficient level for the locks to operate. The Potomac itself was unpredictable and often tore up boats in rapids and whirlpools. Because no one could pole against the strong current, boats had to be broken up in Georgetown and sold along with the other cargo. A more effective way was needed to navigate the Potomac.

In the pre-steamboat era, traveling upstream on the Potomac was basically impossible, so the direction of travel was one-way. The vestiges of some of these canals and locks, including the ones at Great Falls, still exist and can be visited today.

Economically speaking, the Patowmack Company was not a success and folded in 1832. Its operations (and some of its locks) were incorporated into the more substantive Chesapeake & Ohio (C&O) Canal, which ran parallel to the Potomac. The C&O canal, completed in 1850, linked the coal mines of Cumberland, MD with Washington D.C., operating until 1924. The C&O Canal was passable in both directions, with mules towing boats along the towpath. Today the towpath running along the 184-mile length of the canal (or its desiccated remnants) remains intact, thanks to Richard Nixon’s 1971 designation of the Canal as a national park.

This is a picture of the ‘Little Falls’ section of the original Patowmack canal, later folded into the C&O canal. The present day C&O canal and towpath looks like this along much of its length. In the summer it’s chock full of turtles sunbathing on logs, fish, and birds, which are attracted to the warm, placid waters.

For his part, Washington had to abandon his canal-building plans in order to preside over the Constitutional Convention in 1787, which would end with him assuming the role of the presidency in 1789.

A man, a plan, a canal

But Washington’s work on canals was not wholly distinct from his nation-building ambitions. Indeed, it could even be said that his desire to unite the interior frontier of the country with its coasts led to the constitutional convention itself. The canal was a multi-lateral endeavor; the canals along the Potomac were bordered by Maryland and Virginia and thus both states had an economic interest in the project. Postwar relations between the two states were anything but harmonious, though. Washington took on the unenviable task of persuading the assemblies of both states to support the canal. With the help of a convention hosted at his estate in Mt Vernon, he was able to reach an accord. Future ideological opponents James Madison and George Mason also supported the canal and associated convention. There, delegates from Maryland and Virginia agreed on the Mt Vernon Compact, which governed river navigation rights, fishing rights, and the sharing of toll duties from canal passage.

After the success of the Mt Vernon Convention, the future-framers decided to invite other states to the discussions, broadening their scope from simple questions of river tolls to commerce more generally. In 1785 letter, Madison urged Washington to take advantage of the traction from the Mt Vernon conference:

It seems naturally to grow out of the proposed appointment of commissioners to Virginia and Maryland, concerted at Mount Vernon, for keeping up harmony in the commercial regulations of the two States. Maryland has ratified the report, but has invited into the plan Delaware and Pennsylvania, who will naturally pay the same compliment to their neighbors.

Thus in 1786, Madison organized a second conference, this time in Annapolis. The objective was to “take into consideration the trade of the United States” and “consider how far a uniform system in their commercial regulations may be necessary to their common interest and permanent harmony.” This time, delegates from Virginia, New York, Pennsylvania, Delaware and New Jersey attended. Among them were Madison, Alexander Hamilton, and Edmund Randolph, future Secretary of State under Washington.

At the time, the nation was in deep recession, and the incipient Confederation was unable to finance itself through taxation nor able to devise a consistent trade policy. Shay’s rebellion, a populist uprising in Massachusetts triggered by a postwar debt crisis, was also underway at the time. The political atmosphere was grim. Thus at the Annapolis convention, Hamilton became convinced of the need for a wholesale reimagining of American governance. He proposed a subsequent convention in Philadelphia the following year, “to take into consideration the condition of the United States, and to devise such further provisions as shall appear to them necessary to render the constitution of the Federal government adequate to the exigencies of the Union.”

Thus it could be said that Washington’s canal ambitions set off a causal chain that led directly to the Constitutional Convention in 1787. National Geographic in their June 1987 issue described the Patowmack Canal as the ‘waterway that led to the constitution’.

Constitution or a Coup?

There’s a strain of revisionist history, which I am partial to, that treats the constitutional convention, and the ensuing establishment of the United States, as a kind of bloodless coup. The Confederation was certainly ineffective, but it had not failed entirely at the time of the convention. And if you compare the Constitution with its predecessor, it was considerably less democratic. Curtis Yarvin puts it plainly:

The Constitution is an elected monarchy because the Constitution was installed as a right-wing coup. Its goal was to replace the chaotic and ineffective Congress of the Confederation with an effective government that was monarchical and national, while retaining the cosmetic appearance of a federation of sovereign states. Later this fatal, unresolvable ambiguity produced a war; it worked quite well for most of a century.

adding,

In America, something like pure, decentralized democracy lasted roughly from 1776 to 1789. The Congress of the Confederation worked so poorly that Americans airbrushed our first national government out of our own history books.

Indeed, the colonies did organize a government between the end of the revolutionary war in 1783, and Washington’s first term, which began in 1789. The Continental Congress had no fewer than ten separate presidents (although these presidents had little executive power, being part of the congressional body). This government didn’t function particularly well, but it existed.

Now to Yarvin’s question, if you went to school in the U.S., how much time was devoted in history class to this era in history? I would venture that you recall an unbroken narrative from the Declaration of Independence straight to the framers, with no discussion of the 13 year interregnum. Was the name of John Hanson, the first individual known as ‘President’ in the United States, mentioned at all?

There’s a very detailed book on this topic called The Framer’s Coup, by Harvard Law professor Michael Klarman, that casts the Constitutional Convention as a kind of reprisal of the elites against the incipient democracies that existed at the State level. Indeed, it’s quite clear that the U.S. Constitution was much less democratic than the Articles of Confederation. The Constitution established the Senate (at the time, Senators were chosen by the States, rather than being directly elected), the Electoral College, clearly intended to insert elite discretion into the presidential selection process, and firmly entrenched the Federal government as superior to the States.

So why did the Constitution include such apparently anti-democratic measures? Klarman argues that, post revolutionary war, these State-driven democracies had embraced populist policies imposing a kind of economic class warfare designed to disempower large landholders and benefit poorer farmers and the working class. He lays it out in a 2010 talk:

The Framers’ constitution was mostly a conservative, aristocratic response to what they perceived as the excesses of democracy that were overrunning the states during the 1780s.

The Framers were trying to create a powerful national government that was as distant from popular control as possible: very long terms in office, large constituencies, indirect elections. They thought of democracy as rule by the mob. They didn’t think poor people could be trusted with the suffrage. They didn’t think women should vote.

A lot of what the original Constitution was about was constraining the power of the states to pass laws beneficial to debtor farmers in a time of economic distress and expanding the power of the national government to that it could efficiently raise taxes in order to pay off government bond holders, who often were merely speculators in such debt rather than initial suppliers of credit.

According to this reading of history, it may well have been the case that the Founding Fathers were motivated by a desire to reign in loose monetary policy that was benefiting debtors (by inflating away the real value of their debt) and harming the creditor class (which they, as a kind of American landed gentry, by and large consisted of).

This is a monetary theory of the Constitution not often advanced, but one that seems to align with many of the Framers’ stated views on monetary policy. Jefferson for instance once claimed that “paper is poverty,” describing it as “the ghost of money, and not money itself.” He additionally sought to advance an amendment abolishing borrowing at the federal level. Washington was no fan of fiat either, writing in a letter: “paper money has had the effect in your state that it ever will have, to ruin commerce, oppress the honest, and open the door to every species of fraud and injustice.” James Madison was equally aggrieved, claiming that “paper money is unjust,” adding: “it is unconstitutional, for it affects the rights of property as much as taking away equal value in land.” (And if you read Madison’s Federalist 10, it seems he isn’t a huge fan of direct democracy either.) Virginia’s rampant currency printing and devaluation during the revolutionary war likely affected Madison, Jefferson, and Washington, as they were all Virginia landowners. They would all have likely dealt with losses stemming from the inflation, with Jefferson hit particularly hard.

In this context, the ‘elite economic reprisal’ theory promoted by Klarman seems rather persuasive. The New Republic further elaborates on the theory:

Not surprisingly, common people began to use their new political influence to create economic policies that were favorable to themselves (and disadvantageous to creditors and wealthy citizens), such as inflationary monetary policy and progressive taxation. The Constitution, according to the economic interpretation, was the 1 percent’s revenge, a countermeasure designed to undermine the democratic governments in the states, thereby returning power to wealthy elites and insulating them from popular opinion.

The notion of the founding of the nation as an elite-led reprisal against democracy, motivated at least in part by concerns regarding excessively loose monetary conditions, is an idea that tickles me no small amount.

I’m generalizing, but Bitcoiners by and large aren’t noted fans of democracy. We tend to vote with our wallets. Granted, it’s hard to be a fan when democracy created for us the worst debt overhang in peacetime in the history of the nation. We are at wartime debt to GDP levels, without having fought a war, and without any clear way out. The government commands the largest role it has ever had in society when measured by expenditures as a share of GDP. It’s not hard to see why we might be disenchanted with the electoral process.

I have long considered Bitcoin a bit of an inversion of prior populist monetary movements. Historically, populism generally entails a demand for loose monetary policy and the easing of credit. Jubilees and the cancellation of debt as Graeber relates are among the oldest political traditions in existence. So when I think of populist monetary movements I think of William Jennings Brian and his ‘cross of gold’. His ‘free silver’ movement aimed to expand the money supply, thus softening the currency and bailing out indebted farmers. Bitcoin, representing a hard money, low time preference, high interest rate ideology is the opposite. But how often do you see populist monetary movements in favor of hard money? I’m not sure I can think of one.

Some of the critics are right, I think. Well, partially. Any political description of Bitcoin will be both overly broad and incomplete; Bitcoin is a gigantic tent and is used by likely north of 100m individuals worldwide; no single idea unites them. The protocol is cold and unfeeling and manifests no ideology. It’s the collision of this neutral system with the established political world that is revealing. And certainly, groups of bitcoiners exist with definite political ideas.

For instance, the popularity of Bitcoin in the developed world could well be interpreted as an elite reprisal against the current populist desire to loosen monetary conditions, collapse asset prices, and eliminate general indebtedness, especially among young people. If you wanted to be dramatic, you could say that we currently face the choice between becoming a permanently indebted nanny state like Argentina or undertaking a hard money reset, such as the founding fathers engineered. In other words, default or austerity. I have my preferences, but I don’t think they’re widely shared.

Back to the canals. Something Washington’s economically unsuccessful canal project makes me think about is the importance of timing in tech investing. His travails are a stark reminder that being a visionary with regards to the progress of technology is not sufficient; you have to execute at the precise right time too. The challenging thing is that knowing when a technology is “due” is impossible a priori and without experimentation; thus tech innovation is necessarily built on the back of many prior failures that were directionally right in their thesis. Gwern puts it well in his stellar essay, Timing Technology: Lessons from the Media Lab:

Technological forecasts are often surprisingly prescient in terms of predicting that something was possible & desirable and what they predict eventually happens; but they are far less successful at predicting the timing, and almost always fail, with the success (and riches) going to another.

Gwern points out that accurate technological predictions are abundant, but what is unknown are the precise conditions and timing that will give rise to economically viable technological outcomes. In the case of canals, it may have been the most obvious thing in the world that transporting goods via waterway was more efficient than by cart and ox; but the the skills and technology required to build the ambitious canal projects of the late 18th century U.S. were virtually absent in the burgeoning republic. Thus it took Washington’s company 17 years to build the 1,800-yard canal at Great Falls (one of five in the Patowmac plan). He would not live to see it completed.

As someone who is paid to predict the future, or at least to try to anticipate trends before they are obvious, I think about Gwern’s essay quite often. It has a vaguely fatalistic tint to it. Founders may know the future, but the only way to learn whether it’s time for the future to arrive and whether conditions are economically suitable is to throw themselves at the problem. In practice, ‘transformative’ technologies involve numerous failed attempts before one finally hits at the right time. But the tragedy is that you have to be irrationally optimistic to have a chance at hitting. Almost definitionally, winners will be attempting to pull off a similar idea which has failed many times before.

This is why Gwern says that the future is built “by individuals acting suboptimally on the personal level, but optimally on societal level by serving as random exploration.”

In Washington’s case, he couldn’t achieve sufficient political buy-in to get the multilateral canal project off the ground as the returns from a canal were not sufficiently obvious in the pre-industrial era and serious canal-building technologies were not yet available. The subsequent C&O canal did much better, as transporting coal to the coast made a lot of sense — although it’s worth noting that the C&O canal was nearly obsolete upon completion in 1850, as the B&O railroad had already reached Cumberland by that time. So canals were a kind of cursed technology, at least in the agrarian and industrializing USA. When they were really needed, they were impractical and costly to build; and once they could be built economically, railroads quickly supplanted them.

All of that said, the towpaths along the contemporary remnants of the canals, especially the C&O Canal, make for a really nice crushed gravel bike path today. I like to think that Washington, an itinerant rambler at heart, would have felt that it worked out quite nicely after all.

Your faithful correspondent,

Nic

Proof of Reserves (PoR) isn’t just a niche procedure debated by crypto bros and accountants any longer. It’s increasingly making its way into legislation and policy. The Texas legislature recently saw a bill introduced asking for segregated custody at exchanges alongside quarterly PoR attestations. Wyoming also mentions PoR in their 2021 Digital Asset Custody Framework. I expect others will follow, both at the state and federal level. At the same time, PoR is widely misunderstood and sometimes derided, both in the crypto space, and among policymakers.

Just recently, Sens. Warren and Wyden wrote a letter to the PCAOB calling PoRs “sham audits”. The senators asked the accounting regulators to dissuade CPA firms from engaging with PoRs. Bloomberg Tax calls PoRs “misleading reserve estimates”. Paul Vigna in the WSJ says PoR “isn’t all it appears to be”. Amidst the departure of PoR stalwarts Armanino and Mazars from the crypto space, conspiracies swirled about the potential falsity of the PoR procedures they oversaw. Critics saw their exit as confirmation that PoR was indeed a sham. Why else would audit firms overseeing the procedures quit abruptly?

While some of these criticisms are made in bad faith (as is surely the case with Sen. Warren), I also tend to think that a better understanding of PoR would help allay some of these concerns. Given the noise around the procedure today, a refresher is warranted.

Proof of Reserves is widely deployed in the crypto industry today

PoR isn’t hypothetical or a mere thought experiment promoted by crypto bros. It’s a widely deployed procedure that many exchanges follow today. By my count, eleven major crypto exchanges (major defined as holding a minimum of $500m in client assets) engaged in at least one PoR attestation since November of last year. In the aggregate, these most recent PoRs covered $33b worth of assets. On average, these PoRs covered 73% of total assets held by each exchange. Five of these exchanges are doing PoRs on a monthly frequency or better, with two producing attestations on a daily or bi-weekly basis. (All these figures available in my recent article.)

The procedure has been done by exchanges dating back to 2014 (I briefly cover the history here). It is fairly mature from a technical perspective, with lots of academic and practitioner debate, and has undergone significant refinement in its decade of life. Newer innovations such as ZK liability proofs have considerably enhanced the privacy of PoR (which was historically the biggest reason exchanges didn’t undertake it). Policymakers owe it to the industry to engage with the reality of this proactive, self-regulatory phenomenon.

Proof of Reserves does cover both sides of the equation

Probably the biggest misconception around PoR is that it represents “only half of the equation”. It’s commonly said that Proof of Reserves only pertains to assets, not liabilities. This confuses the issue. In the wake of FTX, some exchanges did informal asset attestations as a short term stopgap measure in which they shared coldwallets without corresponding liabilities. This indeed, was “only half the equation,” and hence not very useful. But this isn’t what Proof of Reserve is. PoR has always meant proving ownership of client assets and demonstrating outstanding liabilities owed to clients. That has been the meaning of PoR since it was first discussed in 2013 and that’s what it means today. If we were just talking about demonstrating some quantity of assets held, we would say Proof of Assets. “Reserves” implies something held in, well, reserve — something held on behalf of someone else. So a Proof of Reserve necessarily includes the liability side too.

Proof of Reserves is uniquely possible with digital assets

Proof of Reserves is only possible with cryptoassets, as these are the only type of digitally-stored value whose ownership can be proved to a third party on a purely peer to peer basis. In other words, I can prove unequivocally that I own some quantity of Bitcoin to you, without any third party or intermediary needing to be present or participate. This isn’t true of equities, dollars in a bank account, or any other financial asset. Even a commodity like gold fails this test. Verifying the integrity of gold requires physically assaying it — or relying on a third party like LBMA to attest to its authenticity. Digital assets are genuinely unique in this context, and policymakers should appreciate that by virtue of this quality, exchanges can be reformed and made more accountable than any other type of custodial institution before. This introduces a new model of custody: one in which the assets held on behalf of depositors can be proven and demonstrated to clients or any type of third party at any time. Traditional modes of oversight are upended by this. This is a novel and truly remarkable quality of the asset class.

Proof of Reserves is not a substitute for an audit, but a different thing entirely

I’ll admit, some exchanges haven’t covered themselves in glory here. Several exchanges have called PoR procedures “audits”, which has given conniptions to more than a few accountants. It’s important to understand the difference between the two. A PoR — done well — evidences that an exchange possesses assets under their control equivalent to client liabilities. It’s not exhaustive, and it’s not equivalent to an audit. For instance, it does not prove the following:

• The exchange exclusively controls those assets
• The exchange is practicing good key management
• The exchange maintains strong controls around key man risk, privileged information, etc
• The exchange is immune to hacks or key loss
• The exchange doesn’t have any large hidden liabilities
• Customers will be senior claimants in the case of insolvency
• The exchange cannot become insolvent in the future

And so on. PoR demonstrates something pretty narrow, but it’s nevertheless quite compelling. It’s a non-audit, non-institutional, technical way of demonstrating credibility around the most important part of the exchange business: the integrity of user assets. The other issues above must be addressed through legal and contractual structuring, proper controls, a sound custody setup, good governance, and yes, actual auditor oversight. PoR does one thing, and does it well, but it doesn’t give you universal assurances about exchange quality. I could see PoR eventually being incorporated into traditional audit approaches, but not replacing them. And if you consider what traditional audits actually cover, a financial statement audit covers information included in financials, which in some cases does not cover user assets.

In the case of a publicly traded company with quarterly audited financials and strong controls, depositors may not seek the additional assurances of a PoR. There’s already a strong institutional architecture which makes it hard for the exchange to lie about its financials. However, the set of exchanges that this caveat applies to is very few — in the U.S., only one, as far as I can tell. Generally, the status quo is for crypto exchanges and platforms to not be publicly traded. They also have trouble securing financial statement audits from Big Four firms. Additionally, a PCAOB financial statement audit doesn’t really focus on reserves. It may involve taking a sample of customer assets to test. But if digital assets are held off balance sheet (as is common), they wouldn’t necessarily be within the scope of a FS audit. A controls audit would focus on things like key management and internal controls. In many cases, management simply signs off on the effectiveness of those controls, rather than the auditor investigating them. A SOC II covers controls, but that doesn’t prove the money is actually there, just that the controls are reasonable and active over the coverage period. PoR is simply a different (and complementary) part of the stack, and a vital one, in my view.

As for the CPA firms that oversaw PoRs — as far as I am aware, a list that consists entirely of Armanino and Mazars — they did not represent PoRs as “audits”. CPA firms are notoriously careful with their language. In all cases, the CPA-supervised PoRs were described as “Agreed Upon Procedures” engagements (see here for an example). An AUP is nothing like a financial statement audit. The AICPA defines it as “an attestation engagement in which a practitioner performs specific procedures on subject matter and reports the findings without providing an opinion or conclusion. The subject matter may be financial or nonfinancial information. Because the needs of an engaging party may vary widely, the nature, timing, and extent of the procedures may vary, as well.” Effectively, it’s an engagement whereby a client describes some procedure and the audit firm reports on it. Typically, but not always, it’s done for the private benefit of some third party, rather than the general public. Generally, it’s a way of surfacing facts in a controlled manner and sharing them officially with a third party, who then draws their own conclusions. The auditor provides no assurance or opinion.

Both Mazars and Armanino in their AUPs were careful not to characterize them as audits. So the allegation that these PoRs were “sham audits,” at least from the perspective of the audit firms who oversaw them, is unfair. They were very clear about what they were doing. If there was misrepresentation, it was on the side of the exchanges — and admittedly, some of them used “audit” in a more colloquial sense to refer to their PoRs. When I talk about PoR, I describe it as a “procedure” or an “attestation.” For now, the major standards-setting bodies in the US have yet to cover exchange PoR, so we don’t have a rigorous way to talk about them in an accounting context. Calling them audits probably implies an unearned rigor, so I suggest moving away from that.

Proof of Reserves attestations are heterogeneous

There is as of yet no standard Proof of Reserve attestation. AICPA nor FASB have not issued any guidance as to what a PoR might consist of or how a CPA firm might go about doing on. And most audit firms simply have no experience overseeing a PoR. Those factors don’t help the issue. PoRs vary widely. Here’s a few of the axes along which they vary:

• Some PoRs are overseen by auditors (always AUPs), many are not. There are currently no active CPA firms overseeing PoRs
• Exchanges prove ownership of assets in different ways. Some do self-send txns, some sign messages with their keys, some publish partial transaction data
• Exchanges disclose user liabilities in different ways: some don’t share it at all, some share liability data with third parties, some use Merkle trees to let clients verify liability inclusion, some use Merkle sum trees, some disclose the entire liability set, and others use ZK proofs to disclose the presence (but not the contents) of the liability set
• Some PoRs are done on a daily basis, others every six months
• The number of assets covered varies widely, as does the share of client assets attested to (recent PoRs vary between 15% and 100% of client assets demonstrated)

In short, ‘PoR’ is not a singular concept, but the general idea of demonstrating the equivalence of assets held on deposit alongside outstanding liabilities. Today, due to this heterogeneity, PoR attestations must be evaluated individually. This is why I created a simple rubric to help users reason about their credibility.

Proof of Reserves would have helped prevent Mt Gox, Quadriga, or FTX

Imagine a parallel Earth identical to ours in every way, except for the fact that crypto exchanges had embraced PoR far earlier. In this world, only exchanges unable to carry out the procedure (on account of malfeasance poor controls or accounting) would have failed to produce one. These exchanges would have stuck out like sore thumbs (and indeed, each of Gox, Quadriga, and FTX was likely insolvent for years before collapse). Clients could have acted accordingly and steered clear, and regulators could have intensified their investigations into these platforms. On-chain analysts would have started to interrogate why these exchanges weren’t producing PoRs while all of their peers were. In the case of FTX for instance, sleuths looking at the blockchain after the Alameda balance sheet revelations were some of the first to confirm that something was amiss.

Proof of Reserves has drawbacks, but they aren’t dealbreakers

PoR skeptics like to say things like “PoR is worthless” or “PoR is easily gamed.” This isn’t really the case if you run through the objections. In fact, none of the objections appear to be dealbreakers. PoR works well and provides strong and distinct assurances which can’t be obtained through other methods. Here are the main issues people have with PoR and my responses:

“PoR leaks data, especially if exchanges are disclosing the full liability set”

If you consider one of the more advanced PoR attestations, like BitMEX’s implementation — which allows any third party to compare the entire liability set to the assets on chain — you will see that there are controls which preserve user privacy. Of course, no user PII is leaked, as everything is anonymized: users are assigned a string in their own client dashboard which uniquely identifies their account. User balances are split randomly into parts, so third parties cannot triangulate user behavior over time. What is leaked is the aggregate asset base on the exchange, and the distribution between assets. However, this information is published already by numerous chain analysis companies (including Coin Metrics, a data firm I cofounded), and there’s no way for exchanges to prevent these inferences. Additionally, newer zero-knowledge tools are coming to market which allow exchanges to perform their liability attestations while keeping all balance and distribution data private. As far as I’m concerned, the fact that these ZK-PoR tools exist now obviates the privacy objection, which was historically one of the biggest issues people had with PoR.

“PoR is unnecessary because you could simply regulate exchanges through some other mechanism”

Right now, exchanges in the U.S. are lightly regulated. Mainly they are regulated on a patchwork, state-by-state basis as money transmitters. This approach isn’t really fit for custodial institutions holding billions of dollars of client assets. In this context, PoR legislation (and all the accompanying features, such as requirements that client assets to be segregated from operating capital, or held in a separate trust which is insulated from bankruptcy) definitely ameliorates this. Some may think that a unifying federal framework for exchanges might simply require better custodial practices, making something like a PoR irrelevant. However, we aren’t there yet, and such legislation could take years.

Additionally, PoR measures are being actively undertaken by numerous exchanges throughout the industry, so PoR legislation simply codifies an existing process that exchanges have embraced. PoR is a crypto-native solution which, in my view, surpasses the level of assurance you get from traditional audits in a reserve context. Other types of custodial oversight are purely regulatory. If you were reinventing bank oversight from scratch, but this time it was possible to prove to depositors (rather than just state or federal supervisors) that banks literally had sufficient liquidity, wouldn’t you prioritize that? After all, all of this regulation is meant to be for the benefit of end users and depositors.

Lastly, many exchanges are offshore and completely unregulated. We can debate the morality of this, but if you take a harm reduction approach, supporting PoR is an unalloyed positive. While no proof of reserve legislation could compel offshore exchanges to undertake the procedure, if all onshore exchanges were doing it, that would put pressure on their offshore peers to do the same. Additionally, a regulatory compulsion to use PoR domestically would create a market for more and better technical tools and CPA firms to oversee the attestations, making it more convenient for offshore exchanges to engage in PoR. PoR is most useful for exchanges where traditional assurances don’t exist. There are many of these, so standardizing PoR and encouraging CPA firms to cover them would improve the overall credibility of these exchanges, even if offshore.

“PoR can be cheated by borrowing funds (“window dressing”)”

This is a very common objection by folks in the industry who believe (mostly erroneously) that exchanges have already cheated PoR by borrowing funds. In fact, the recent Gate/Crypto.com transfer was not actually a case of window dressing to cheat a PoR (although it was weird and puzzling). The fact is that proving asset ownership means disclosing wallet addresses. If you wanted to cheat a quarterly PoR by borrowing huge amounts, this would be obviously visible on-chain. Would you really risk committing fraud if anyone with an internet connection could trivially expose you? Additionally, higher frequency PoR is immune to this issue. Deribit does PoR on a daily basis, and BitMEX on a biweekly basis. In this case, window dressing is also out of the question, because it would make no sense to borrow and return funds daily. Frequency solves the window dressing issue — which is another way in which PoR vastly outperforms traditional audits (which are done yearly, or at best, quarterly).

“PoR can be cheated by hiding liabilities”

First of all, for modern PoR like the ones done by Derebit of BitMEX, the entire liability set is released, so there’s no real uncertainty around the completeness of liabilities. Any standard PoR is also user-verifiable, so presumably any user could blow the whistle if they found that their liability entry was understated. Today, most PoRs are done with the Merkle proof method where liabilities are only disclosed on a per-client basis, which creates more possibilities for liability hiding. But this is solved with next generation PoRs which rely on ZK proofs, making disclosure of the full liability set possible without privacy drawbacks. Newer cryptographic technologies have largely made this objection obsolete.

“Exchanges could have massive out of scope liabilities”

Yes, PoR doesn’t fix this kind of issue (no one has ever claimed it is a panacea). Exchanges could have some massive hidden liability. This is just a general problem though, not a PoR problem. This is more the domain of legal and contractual structuring. The way to fix this is to ask exchanges to hold assets in a segregated trust held for the benefit of clients, which is insulated from other liabilities. In the case of liquidation or insolvency, depositors are whole. Any PoR legislation should include this stipulation. To their credit, NYDFS have already laid out how this would work. Problem solved.

“PoR doesn’t fix poor key management, key loss, or fraud”

Yes, but it makes it impossible to run at a fractional reserve for any sustained period of time. In the case of prior exchange collapses like FTX, Quadriga, or Gox, these exchanges were insolvent for months and years. They never had sufficient reserves to honor all possible client withdrawals. The moment an exchange was even under-reserved, the PoR would have been impossible to pass. So PoR makes it virtually impossible to behave badly for any meaningful period of time. If any exchange did PoR and became insolvent, they would stop doing PoRs. This would be a massive red flag.

“Forget PoR, just do an audit”

Imagine if there was a type of audit that allowed a custodial institution to prove with no uncertainty, on a daily or weekly basis, to their customers, the government, or the public, that they had all the assets they said they had. This simply doesn’t exist in traditional audit land. Financial statement (FS) audits are slow, expensive, infrequent, and very broad in scope, covering far more than just reserve management. To the extent they cover client reserves, they generally involve sampling — rather than investigating all client assets. Certain major exchanges that did have FS audits did not include customer assets in their scope prior to 2022.

Practically speaking, audits are expensive and cumbersome, and CPA firms are very averse to working with crypto companies. This isn’t helped by folks like Sen Warren trying to bully audit firms into further spurning crypto. Given this reality, Proof of Reserve is a highly complementary solution. It is frequent, narrow in scope (but covers the specific thing that clients care about), and relatively cheap. It doesn’t even strictly require an audit firm — BitMEX’s and Deribit’s PoRs, two of the best in my opinion, don’t have audit oversight. Think about PoR as a targeted tool to give depositors confidence over one domain of an exchange’s practice — their custody — which can, and probably should, be supplemented with traditional assurances such as audit and contractual depository assurances. A PoR isn’t sufficient on its own, but it’s a vital piece of the puzzle.

Proof of Reserves introduces a whole new design space for assurance

Ultimately, PoR is far broader than exchanges proving the existence of coins to their clients or to regulators. The key technologies that make PoR possible are generalizable to all sorts of financial trust-provision. I could see interlocking or recursive Proofs of Reserve allowing an ecology of custodians, exchanges, prime brokers, trading firms, and lenders to transact with each other with confidence. These proofs could allow counterparties to demonstrate the existence and nature of assets on their balance sheet (or facts about the assets, without actually revealing sensitive info — a perfect use case for ZK proofs). Imagine lenders able to demonstrate the solvency of their portfolio by pulling through balance sheet data provided by their borrowers. Borrowers could also demonstrate the exclusivity of pledged collateral to their lenders (eliminating Archegos or Three Arrows type problems). The design space is enormously large, and has barely been explored yet. ZK-proofs are particularly well suited here, as they allow for the provision of metadata without revealing the underlying data. Exchange solvency is just the most pressing need, so that’s where this tech is being applied first.

Takeaways for policymakers

As you start to investigate PoR and consider incorporating it into policy, keep a few things in mind. Post FTX, exchanges are doing their best to demonstrate their credibility and solvency to clients, regulators, and the general public. If they have gaps in their PoR implementations, that’s most likely because the procedure is still fairly new and still not standardized. Ultimately, a fraudulent or insolvent exchange will likely not go to the trouble of creating a PoR and hiding liabilities or engaging in other high-risk fakery. They would simply not do a PoR and hope they could avoid the pressure to do one. If an exchange is voluntarily incurring the costs associated with doing a PoR, they are demonstrating their willingness to come to the table and prove their credibility.

PoR is still evolving technologically, so don’t fixate on one specific implementation over another. Any guidance should be technology-neutral. Today, most PoRs rely on the Merkle tree method to prove inclusion in the liability set, but in the future, I expect ZK proofs to largely take over, because of their superior privacy guarantees.

If you do consider creating legislation on the topic, the objective should be to create a framework which actually helps enhance exchange credibility, while still being mindful of the cost of compliance (so as to not create gigantic fixed costs which erode the ability of small exchanges to compete).

There’s a couple things which could make PoR requirements prohibitively expensive:

• Insisting that an exchange cover 100% of assets by AUM with a PoR. Because many exchanges have a long tail of smaller assets, and a PoR requires a new treatment (requiring a fixed cost) for each blockchain (or L2), implementing a PoR universally would be very expensive. For this reason, most exchanges that have done PoRs have not covered 100% of assets (in fact, there’s only one exchange that was able to achieve universal coverage in the most recent round of PoRs). I would suggest a high, but doable threshold, like 75% or 90%.
• Insisting that PoRs be supervised by audit firms is also potentially exclusionary. Today, there are no active CPA firms that supervise PoRs. I do expect this to change in the future, but this will take time. So asking an exchange to get coverage from a CPA firm would require them to pay a significant amount of money and coax an audit company to enter the business. This disfavors smaller exchanges. Also, getting an audit is at best a quarterly undertaking. However, PoRs provide the best assurances when they are done weekly or monthly (to avoid the potential issue of borrowing funds on a short term basis to match the liabilities). Thus, insisting on a slower cadence, as would be required by audit firm oversight, would actually stand to inhibit some of the guarantees that PoR provides. Thus, I would ask for something like a frequent PoR (maybe at a monthly cadence) with more periodic audit oversight. Additionally, a sunrise period might be warranted, to give audit firms time to reenter the market.

Audit firms will be leery of Proof of Reserve as long as accounting standards don’t exist. I would consider asking their regulators and trade organizations to start to build a body of knowledge around PoR and eventually release standards around them, so CPA firms can take these engagements without incurring excessive risk.

To get a sense of the state of the art in PoR, I would investigate various PoRs. There is no need to reason from the armchair. I maintain a dossier of all major PoR procedures on my website, alongside related attestations like stablecoins. BitMEX’s and Deribit’s are of particular interest, because any third party can verify them without being a client of the exchange in the first place. There is also a vibrant academic and practitioner literature on PoR. An accessible and thorough introduction is the Digital Chamber’s Proof of Reserves: The Practitioner’s Guide.

And lastly, because PoR is only one piece of the puzzle, it should be paired with other requirements that give end users strong guarantees around custody. These include segregating client and operating capital and protecting depositor capital from bankruptcy. Proof of Reserve is an increasingly important part of the toolkit, but not sufficient on its own.

Thanks to Jeremy Nau for his feedback and contributions.

It’s undeniable that momentum around cryptographic Proofs of Reserve is gathering. This is a genuine silver lining from the FTX debacle. We may well yet emerge from this crisis with a major step forward in exchange credibility. But there remains a huge amount of confusion around Proofs of Reserve, and some of the exchanges taking their first steps towards the procedure have plenty of work to do.

In the last two months, major exchanges Kraken, BitMEX, OKX, and Binance, among others, have all published attestations they call Proof of Reserve, although these don’t all grant the same assurances. Coinbase also published a blog post explaining their status as a public company with audited financials and quarterly disclosures.

Here I’ll answer some lingering questions around PoR and dig into some of these recent attestations, so users can better understand what these recent procedures actually mean. I also summarize PoR efforts to date and introduce a simple framework through which to judge their effectiveness.

Who has done a PoR recently?

I’ve compiled data on all the PoRs I could find in 2022 above. I have mainly included procedures which give clients the ability to verify their inclusion in the liability set (with the exception of Luno, which just published an attestation and no corresponding cryptographic proof). For the purposes of this exercise, I’m tallying up custodial, centralized exchanges, with liabilities outstanding to their end users. Other things commonly referred to as ‘proof of reserve’ are not included. (A brief sidenote for my Chainlink readers: when I talk about Proof of Reserve I’m mostly referring to the process whereby exchanges or custodial institutions compare user liabilities to assets held in reserve. I know there’s a spectrum of similar attestations that cover all sorts of scenarios, like comparing WBTC to BTC held in reserve. I’m not deliberately snubbing Chainlink, just that their PoR procedures generally do not pertain to exchanges proving reserves to clients.) For a broader look at historical PoR beyond the table see my website. A few notes on the table:

• Impressively, I found that these exchanges in the aggregate underwent Proofs of Reserve covering $33b worth of assets (or 4% of total crypto market cap). Of course, these attestations are of varying quality, but it’s still a good and growing level of coverage. This is not to say that the other tens or hundreds of billions of custodial cryptoassets are not safe — many are with credible institutions like Coinbase, Fidelity Digital Assets, Gemini, etc.
• Only BitMEX and Deribit are undertaking these procedures with a high frequency. This is doable because they don’t rely on auditors to oversee their process. Kraken has a slower cadence due to their usage of an auditor for oversight. For auditor-overseen PoRs, I don’t expect much faster than a monthly policy.
• Only BitMEX and Deribit allow third parties to verify liabilities for themselves. Most of the exchanges covered allow their clients to individual verify whether their liabilities are included in the set (thus creating a kind of ‘herd immunity’ assuming that some clients actually did the verification and found it proper), but mostly these exchanges didn’t publish the full liability set. However, as a third party, I much prefer the model where anyone, whether a client or not, can undertake the verification for themselves.
• Armanino and Mazars oversaw all of the PoR attestations where there was an auditor present, and they’re both out of the PoR market now. This is problematic for the sector. I’m hoping a few audit firms dip their toes back into the water. In the short term I expect these exchanges will mostly be unable to persuade audit firms to oversee their PoRs, as the public backlash against these audit firms has been pretty fierce.
• There’s still a lot of room for improvement. My six point test (elaborated below) lays out very simple ways to improve. For a few exchanges, simply committing to running a PoR on an ongoing basis would help their score. For others, covering a larger share of assets or introducing an auditor would help. Generally, it is not too difficult for exchanges to improve their scores, should they want to.

I’ll get into my assessments of some of these PoRs a bit later. First I want to address the lingering question of nomenclature, which has dominated the PoR narrative recently, and made it hard to discuss the substance of the issue.

Does ‘Proof of Reserve’ not count liabilities?

It’s frequently said that PoR refers to only half of the equation. This mostly stems from a terminology issue. When I, and most others that have been tracking PoR for a while, refer to PoR, we are talking about the procedure whereby both the assets and the liabilities are attested to. Proof of Reserve dates back to at least 2014 (arguably, Mt Gox’ infamous 424,242 BTC transaction in 2011 was the first attempt at a PoR, but it included only the asset side), and even back then, PoR was used to refer to both the assets and liabilities.

The earliest reference to the procedure I can find is a 2013 chat in which Greg Maxwell discusses the possibility of proving ‘reserves’ with an on-chain tx for assets, and a merkle proof of liabilities. So PoR has always referred to both sides of the equation. I moved away from calling it a ‘Proof of Solvency’ because solvency is a broader question which requires a fuller analysis of all of the claims on the exchange. This strays into more typical audit territory and can’t be cryptographically resolved.

Some people have suggested calling it a Proof of Reserve & Liabilities. I don’t love this, because ‘Reserve’ already includes Liabilities. You could also call it a Proof of Assets and Liabilities (PoAL) but that’s unnecessarily pedantic. Reserve is more concise — and it’s the existing term of art.

At the end of the day, the thing initially dubbed ‘Proof of Reserve’ as described by Maxwell is all about the liabilities. Saying “Proof of Reserve doesn’t count the liabilities” doesn’t make sense, because the original and enduring usage of the term refers to a process in which both assets and liabilities are attested to.

A reserve is something that’s being held … in reserve. Assets held on behalf of someone else. So the word ‘reserve’ implies a custodial relationship. The ‘Proof’ of ‘Reserve’ describes the orderly functioning of that relationship — in which assets equal liabilities. If it was called “Proof of Assets” I’d understand the claim that it’s incomplete. But it’s not. So I’m going to stick with ‘Proof of Reserve’ for now.

What do you make of the recent Proofs of Reserve?

I am very encouraged exchanges are taking up the practice. Not all of the recent PoRs are the same though. On my PoR tracking site, I have split up recent exchange attestations into ‘gold standard’, ‘good quality’, and ‘other’ to reflect these distinctions. To get the gold standard label, the exchange should do the following:

• Satisfy the basic qualities of a PoR: cryptographic attestation to assets held, and a disclosure of liabilities
• Optionally, but optimally, incorporate a third-party auditor in the process, to ensure that the attested-to liabilities match the internal database
• If no auditor is involved, demonstrate a high level of credibility by undertaking a PoR for substantially all of the assets on the platform, and allowing third party verifiers to check the completeness of the liability set, including the non-negativity of liabilities
• Commit to an ongoing procedure. There is a wide spectrum of frequency, and this is due to the different types of PoRs (supervised and unsupervised). I would like to see quarterly for PoRs supervised by auditors. For unsupervised ones, a more frequent cadence is possible

Kraken, BitMEX, Deribit, and Kucoin meet the above criteria, and get the ‘gold standard’ moniker on my tracker.

I gave the ‘good quality’ moniker to exchanges scoring at least a 4/6 on my rubric, and ‘other’ to exchanges scoring less than 4. The test assigns one point for each feature present, and a half point for partial credit:

Asset side:

• Did the entity perform cryptographic verification of assets held? (1 pt)
• Does the PoR cover the vast majority of assets on the platform? (1 pt)
• Is the procedure undertaken on a recurring basis with reasonable frequency? (1 pt)

Liability side:

• Can users verify their inclusion in the liability set? (1 pt)
• If the exchange has complexities around margin and lending, are these fully accounted for in the PoL? (1 pt)
• Did a credible auditor provide oversight over the liability attestation? (1 pt)

You can see my findings below. Note that some of these are subjective, and I am relying on my own judgment for these assessments. For exchanges that want to provide feedback, please email nic@niccarter.info.

Additional ‘extra credit’ line items I’m not insisting on, but I’d like to see, are:

• Is the entity able to prove exclusive ownership of assets held?
• Is the entity able to demonstrate that there is no window dressing to shore up cash positions prior to PoR attestations?
• Does the entity clearly stipulate the segregation of client and operating capital?
• Does the entity clearly ensure the seniority of client deposits in a bankruptcy or liquidation scenario?
• (Longer term) Is the entity part of a consortium of provers collectively attesting to the non-duplication of client deposits?

I’ll dive into a couple PoRs here, starting with Kraken. Kraken employed Armanino LLP in their attestation, which gives clients a good level of confidence that they aren’t hiding liabilities, publishing negative or undercounted balances in the merkle leaves, or engaging in window-dressing (aka borrowing funds on a short term basis to pass the attestation).

Kraken also did PoRs for BTC, ETH, USDT, USDC, XRP, ADA, and DOT, representing the majority of platform funds. They even covered staked funds for ETH, ADA, and DOT. Right now they are doing PoRs every six months, although I hope that becomes more frequent with time. In this post, they are realistic about PoR shortcomings, and do not represent it as a panacea for exchange issues.

BitMEX’s approach also deserves praise. They are not relying on an auditor, choosing instead a highly transparent model. On the asset side, they list all BTC balances held by the exchange and the execution scripts for these UTXOs which prove that they are spendable by the BitMEX multisig. On the liabilities side, they publish the Merkle tree of user balances in full. This is different from the standard Maxwell approach whereby users are only exposed to their leaf in the merkle tree (and path to the stem) in the interests of preserving privacy. This means that there are no issues with excluded or negative balances since anyone can vet the liability set in full. To deal with the privacy leakage, they randomly split user balances into two, so specific balances can’t be tracked over time. And impressively, BitMEX now publishes PoR attestations twice a week, a more frequent cadence than most other exchanges.

Deribit also appears to be following a model similar to BitMEX, which is why they score highly on my rubric. They release the full set of liabilities, meaning anyone — client or not — can evaluate the PoR for themselves. They actually split up the accounts into many pieces (rather than four as BitMEX does) to make forensic analysis difficult. And they release liability snapshots on a daily basis! Something I think is unique to Deribit as well is their publication of cumulative margin locked so third parties can evaluate overall client leverage relative to exchange assets.

And uniquely among exchanges I’ve evaluated is that Deribit’s PoR covers 100% of client assets (granted, BitMEX’s PoR coverage isn’t far away). This is because Deribit is a derivatives exchange with relatively few collateral types.

As you can tell, BitMEX/Deribit and Kraken trade off along the cryptographic versus institutional trust continuum. Kraken does not publish the full liability set to the general public but we can still trust that they extracted it faithfully given that they used a credible auditor with experience overseeing the procedure. Additionally, Kraken users can verify their inclusion in the liability set, so that gives us another layer of assurance, provided they do so (and have the wherewithal to spot an issue if there is one). BitMEX on the other hand did not use an auditor but publishes frequent, complete attestations which allow anyone to validate the PoR.

Binance’s PoR on the other hand scores poorly, as it is incomplete. CZ has extolled the virtues of PoR ever since FTX collapsed, but hasn’t yet risen to his own challenge. As an aside, it’s worth noting that CZ’s own history with PoR is mired in controversy. Whether the OkCoin PoR that CZ oversaw in 2014 was genuine or not remains an outstanding question. Whatever the truth of the matter, CZ has known about PoR for eight years and is quite familiar with its possible shortcomings. So he should naturally be unsurprised if some are skeptical of Binance’s PoR if it fails to include controls against similar understatements.

And indeed, Binance’s first PoR doesn’t grant strong assurances. It only covers Bitcoin, which only represents 16.5% of their client assets. It does allow individual users to verify their inclusion in the liability set but does not contain the entire liability list, making it hard for a third party to verify the procedure. Audit firm Mazars supervised the attestation with an ‘Agreed Upon Procedures’ and then deleted their AUP. Shortly thereafter they curtailed their entire PoR practice. Given the relatively small PoR relative to Binance’s depository base and balance sheet, questions arose regarding whether Binance used assets exclusively attributable to clients or marshalled other assets they controlled. Technically, security firm Mysten Labs found a number of issues in the Binance PoR, including four possible ways Binance could have understated liabilities.

Now Binance does mention that they intend to involve third party auditors in subsequent attestations, expand the set of assets covered, and explore ZK proofs to ensure that negative balances attributed to margin are fully matched by user collateral. Given that this appears to be a somewhat rough first draft, I’m inclined to be lenient for now, but I would want to see significant improvements before I am satisfied. I do think that some of the criticism around Binance has been overstated lately, as people appear to be hoping that Binance goes the way of FTX. But the situations are extremely dissimilar. Regardless, Binance could quell many rumors with a more complete PoR.

Of course, evaluating exchange credibility is broader than this simple rubric. For instance, we might aspire to privacy-preserving liability attestations that don’t leak anything material — not the distribution of user balances, or the number of users, or anything like that. Then there’s complexities around margin accounts, which some exchanges deal with by inserting negative user balances in the merkle tree, offset by positive collateral balances. The more the exchange permits margin and lending, the less straightforward the PoR becomes, and the more auditor oversight is warranted.

Proving the exclusivity of assets held is another tricky question. Exchanges can of course borrow funds and engage in window dressing. This isn’t a PoR problem exclusively, it’s a general well-known accounting issue. Incorporating auditors into the process could help provide assurances that exchanges are just borrowing funds on a short term basis. And lastly there’s legal and contractual questions which can obviate the need for a PoR or at the very least give clients very strong assurances that their assets are safe and that they are entitled to them in a variety of scenarios. Here we are getting into more legal and contractual territory, where cryptography is less addressable.

I’ll cover some of these additional considerations in the final section.

Should exchanges release only the merkle tree or a full list of liabilities?

There’s a degree of controversy over this. Exchanges aren’t keen on releasing a full dump of liabilities like BitMEX or Deribit do. If they have 100m+ users, like Binance does, such a file would also be insanely large and unwieldy. And even if you split up balances randomly, you’re still releasing a lot of data — and there’s always ways to extract some signal from such a large dataset.

So there’s a tradeoff between the desire for privacy on the part of the exchange and the auditability of the PoR. If the full liability set isn’t released, third parties can’t meaningfully evaluate the quality of the PoR: they simply have to trust that sufficient exchange clients are diligently performing the verification. And that doesn’t feel great.

Here we run into contradictions and paradoxes. On the one hand, I would say auditor oversight would likely be sufficient if an exchange doesn’t want to publish the full liability set — but in that case, why even bother with user verification and the merkle tree? Why not just do a simple AUP and ask everyone to trust the auditor?

Ultimately, the current state of PoR doesn’t lend itself to good answers in this case. I’m sympathetic to exchanges who don’t want to list a full liability set, even if they have obfuscated the distribution of user balances. But in my view PoR should be public facing, not just something that clients can verify for themselves. So I generally support the publication of as much liability information as possible.

I think the concern regarding privacy leaks from the merkle approach is not unfounded. While auditor oversight can naturally help here, I prefer crypto-native solutions if possible. So I am hoping that ZK approaches, such as the ones Vitalik mentions, can fill in the gap here. ZK liability proofs could potentially grant third parties — not just exchange clients — strong assurances, while maintaining privacy regarding the distribution of client balances. Such schemes date back to at least 2015 with Provisions so the idea is certainly not new. However, ZK tech has come a long way and is now generally trusted throughout the industry in the case of rollups which are deployed in production. The time is ripe for ZK liability schemes, in my view.

What’s next for Proof of Reserve?

Now that PoR appears to be catching on, there are many possible refinements. The core procedure hasn’t changed much since it was proposed by Maxwell in 2013 — improvements have been largely incremental. It was 2015 when the Provisions paper was first published proposing ZK proofs for a more privacy-preserving liabilities side, and yet we still find ourselves with no deployed implementation of the idea. Newer schemes like Ji and Chalkias’ Generalized Proof of Liabilities rely on Zk proofs (specifically, bulletproofs) and look promising, though.

Historically, attempts at PoR have suffered from a variety of technical issues, as a gulf between the academic literature and actual practice of PoR has persisted. Chalkias, Chatzigiannis, and Ji identify a number of vulnerabilities in historical liability assessments, including privacy constraints like leaking the number of users and leaking individual liabilities. There’s clearly scope for more academics to engage in the PoR space. I’m hoping the gulf between academia and industry in this sector is bridged fairly soon.

What I’d like to see in the future regarding PoR is a combination of a few things:

• ZK proofs of liabilities. These don’t leak client data but still provide credible attestations. The merkle approach, even if privacy is sought by splitting up accounts into random pieces, still leaks all sorts of data regarding client behavior. I think exchanges should be comfortable sharing aggregate deposits, but they may not wish to share the distribution of ownership on a weekly or daily basis. Eliminating these privacy concerns makes exchanges more likely to pursue PoR and on a more frequent basis.
• Legal and contractual assurances on top of PoR. PoR is not a substitute for clear terms that establish the seniority of depositors in a liquidation situation and the segregation of client and operating capital.
• Audit firms reentering the space. Right now, the major CPA firms that did AUPs for PoR have deprecated their practices. I would like to see some audit firms step up and start supervising PoR attestations again, as the non-supervised PoRs just don’t provide the same assurances. This is especially the case for more complex liabilities relating to margin and lending.
• Standardization of PoR. One issue we haven’t addressed is the possibility to engage in window dressing by borrowing prior to a PoR and returning the funds after. More frequent attestations mostly fixes this (hard to engage in window dressing if you are doing daily PoR attestations), but another way to address it would be getting a number of exchanges on a shared PoR standard. If they were in some kind of PoR consortium, they could attest to the respective uniqueness of capital and it would be relatively easy to verify that.
• Dedicated custodians building out their own PoR practice. Exchanges are unbundling and some are outsourcing custody. This is because we have good, high quality custodians, that now support all the requisite assets. So we can expect that crypto might end up more like tradfi, with order matching, clearing/settlement, and custody being distinct functions. In this world, a handful of custodians might end up being very important. Clients of the exchanges relying on these custodians deserve to know that their funds are accounted for. For this reason I want to see these custodians start to build their own PoR practice, so that they can cater to these requests as they emerge. This was the case with the Bitcoin held by Coinbase on behalf of GBTC — but Coinbase hadn’t built the proving infrastructure yet, so we were left with an unsatisfying answer.
• A larger set of ‘PoR watchers’. Right now most PoRs are being treated as equally valid or equally stupid, depending on your perspective (there’s a few BTC maxi cynics that hate PoR because they think it normalizes third party custody at the expense of the pure and holy self-custody). I would like to see more critical eyes affixed on PoRs so that exchanges were encouraged to provide better and more complete attestations. I would be much happier if there were dozens of people like me that took the time to evaluate these PoRs.
• DEXes that obsolete CEXes. Of course, functional DEXes are the equivalent of a continuous PoR, because clients generally retain their own assets until it is time to conduct a swap. Something a few folks have noted is that exchanges like StarkEx are kind of a middle ground between a pure on-chain DEX and a centralized exchange that does a PoR. You can think of a proof of reserve as an attempt to bring off chain functions on chain. DEXes are the end state there. If we can get performant and trustworthy DEXes, then we won’t have to worry about CEXes (and PoRs) as much.

To sum up my own feelings on PoR, I am cautiously excited and feeling somewhat validated. I’ve been pushing PoR/ Proof of Solvency for years now, and for a long time it felt like we were making no progress as an industry. Sadly, the catalytic event that caused this latest round of PoRs (with the exception of BitMEX and Kraken who were working on PoR beforehand) was the collapse of FTX. Unfortunately, reform only follows a crisis.

I do reflect with irony on my prior Coindesk column from 2020 — How to Stop the Next Quadriga: Make Exchanges Prove their Reserves. We did not sufficiently ask exchanges to prove reserves, and therefore suffered something much, much worse than Quadriga. Indeed, it appears that the FTX shortfall, if it ends up being $8b, would be 37 times worse in fiat terms that Quadriga’s loss. Incredibly, it seems that no one ever scrutinized the FTX hot or cold wallets carefully. A mere ‘proof of assets’ would probably have exposed them, let alone a full Proof of Reserve.

Each of Quadriga, Gox, and FTX would have been avoidable had PoR been entrenched in the industry, as none of those exchanges would have ever been able to ‘pass’ a PoR. Each was insolvent for some time before the collapse — in the case of Gox, for years. The way PoR works is, if enough exchanges do it, the few exchanges that don’t do it end up sticking out like a sore thumb.

This is why it’s so important that even the most credible exchanges actually commit to the procedure; otherwise, the signal is lost and the sketchy exchanges can hide in the crowd. So I would encourage every exchange that has undertaken the PoR process to continue to publish these attestations and grow their frequency and asset coverage. For the exchanges that don’t, take a careful look at the process. It will get more convenient, safer, and more secure as better tools emerge. And regulators and lawmakers may end up asking for exchanges to do PoRs anyway, so it would be best to get ahead of that and start doing them voluntarily now.

I have annotated the study in full. You can find my annotated version by clicking the image below or the link here [pdf]:

Note: I’ve highlighted in yellow text that I am responding to in the margins. I have highlighted in purple claims that I consider to be untrue or misleading. I have also implemented a Pinocchio system (🤥🤥🤥) to qualify the egregiousness of these claims with five being the worst.

I wasn’t able to fully elaborate on my points in the margins, so I’m including some companion notes here. Note: I am not laying out a positive case for Bitcoin mining here. By this I mean a discussion of sustainable power models used by miners, or the propensity of miners to induce new renewables, stabilize the grid, and support location-agnostic energy infrastructure. I have done that in plenty of other places. See an index of my prior work on Bitcoin mining on my website, or the On The Brink mining miniseries. I am just responding to the most egregious parts of the OSTP report.

In the study, there are a few bright spots:

• The OSTP acknowledges that PoW and PoS may not grant identical assurances, and there remains uncertainty as to whether PoS might be a perfect substitute for PoW
• The OSTP acknowledges the interesting developments in mining with otherwise-flared or stranded natural gas (often released as an unsaleable byproduct of oil extraction)
• The report meaningfully acknowledges the contributions to grid flexibility miners can offer via participation in demand response programs
• The report notes the potential to mine with stranded renewables
• And importantly, the report offers only lukewarm recommendations

However, the overall effort is weak. The paper is effectively an extended literature review with very little in the way of new data or analysis presented. It relies on a mélange of academia, anti-PoW consultancies like the Crypto Carbon Ratings Institute, and quasi-academic bloggers like the infamous De Vries. Many of the academics cited are conflicted. Virtually no sources from the mining space itself are included. Citing the non-peer reviewed work of (highly conflicted) amateur hobbyists just doesn’t cut it for a purportedly scientific paper. Remember that this report was nominally authored by the “White House Office of Science and Technology Policy.” The report should therefore adhere to a high scientific and academic standard. Relying on non-peer reviewed, non-scientific estimates, especially by individuals with known conflicts of interest, should be an absolute dealbreaker. Unfortunately, this paper does so with gusto.

Virtually no case studies are presented in the report, even though there are many examples of miners operating sustainably, adding power to the grid, taking zero-carbon approaches, and engaging in grid stabilization. There are a number of outright falsehoods in the report, which I’ve highlighted in purple. The report takes an extremely harsh stance on miners utilizing renewable sources, chastising miners for using renewables and laying out an extremely narrow set of conditions where their use would be acceptable. The section on demand response concludes that it’s mostly irrelevant, because miners, in the eyes of the government, are raising aggregate power demand without inducing any additional supply. The section on flared gas mining is a brief olive branch, but the report then concludes that because all oil and gas extraction must be shuttered to meet climate goals, no gas flaring can exist under Net Zero™.

Generally, the subtext is that because mining is a “bad” usage of power, it doesn’t matter if you promote grid stability, or use renewables, or even mine off-grid with stranded resources — you’re not welcome here, and you should buzz off.

Some of the chief issues are as follows:

• Presenting virtually no new data
• Ignoring contributions of industry subject matter experts
• Relying on the partisan De Vries / Digiconomist
• Relying on the conflicted Gallersdorfer, Klaassen, and Stoll
• Citing the absurd Mora et al 2018
• Urging caution on data while using it recklessly
• Pushing a “can’t win” approach to miners using renewables
• Refusing to project Bitcoin’s energy consumption trajectory
• Making stupid and counter-productive recommendations

I’ll tackle these main issues in turn.

The White House presents virtually no novel data

This report is mainly regurgitation of data presented (and in some cases dreamed up) by academia and bloggers. I can tell that the authors have very limited experience with the debates around PoW, or are being lazy in their approach and citing folks willy-nilly, because they cite Mora et al 2018 (genuinely unforgivable) and have an extremely heavy reliance on De Vries/Digiconomist, as well as Stoll, Klaassen, and Gallersdorfer. The Mora reference is shocking. It’s a bit like reading a scientific government report on the history of the moon landing and finding a reference to a conspiracy website claiming that the entire thing was faked.

To get to the bottom of Bitcoin’s (we can ignore everything else, since post-merge, only Bitcoin will remain as a meaningful PoW blockchain) energy consumption, likely future trajectory, and emissions profile, you will need to do some original work. This report contains very little.

The battle lines are already drawn. We have data from hostile academics that are highly partisan and conflicted in many cases, and then data from industry groups like BMC. The report glowingly cites the former as if the data (which includes massive uncertainties, which I’ll delve into later) is settled science. It’s not. If OSTP wanted to make a mark on the debate instead of just repeating De Vries, they should propose some new data of their own.

The White House totally ignores any contributions from industry subject matter experts

So this is to be expected. The OSTP doesn’t cite any of the great work on mining done by Coinshares. They don’t cite Bitcoin Net Zero by Ross Stevens/NYDIG and myself — even though they ask for help modeling Bitcoin’s future energy and emissions profile. They don’t cite Arcane Research, even though the OSTP covers many of the topics where Arcane demonstrates helpful expertise. They don’t cite any of the data aggregated by the Bitcoin Mining Council, which aggregates sustainability data for half of the Bitcoin mining network.

Now you might say “well the OSTP can’t just rely on partisan data and commentary written by industry participants. They adhere to a higher standard of academia and rigor.”

That is where you’d be wrong.

As I will further demonstrate below, the OSTP is relying on the following in this report:

• The personal blogs of a Dutch Central Bank employee with a widely documented antipathy to the subject matter and a clear anti-crypto motive
• Non-peer-reviewed content published in the “commentary” sections of journals like Joule and Nature Climate Change
• Non-peer-reviewed journal content published by authors with a serious, commercial conflicts of interest
• Non-peer-reviewed journal content written by undergraduates, thrice-debunked in its own journal, widely considered junk science
• Reports published by a for-profit ‘carbon ratings institute’ listing numerous Proof of Stake cryptocurrencies as clients — indicating a clear anti-PoW bias

These aren’t just one or two stray citations. These questionable sources (which I will elaborate on below) constitute dozens and dozens of citations, and in some cases constitute sole support for claims made in the paper. The OSTP is perfectly willing to rely on estimates published by conflicted individuals with clear commercial interests that incentivize them not to pursue the truth. They are also willing to publish non-academic and non-peer reviewed, vaguely academic content. So if their reasoning to the total stonewalling of crypto industry-derived content is “it’s not sufficiently academic,” or “they have competing, commercial motives,” this applies to the crypto-critical sources they’ve relied on in this report too. Ignoring the industry leaves the report starved of current data and analysis from subject matter experts. It is ultimately individuals and firms that are direct participants in or adjacent to mining that best understand the reality on the ground. Including these perspectives would have saved the authors many blunders (enumerated in the annotated version).

Additionally, the total exclusion of industry sources reinforces the fact that this report is almost completely one sided; that is, it is deeply hostile to Proof of Work.

Reliance on De Vries / Digiconomist

The De Vries body of work is neither academic, scientific, nor unbiased. He is a blogger who works for the Dutch Central Bank (an anti-crypto institution), an affiliation he routinely fails to disclose on his papers. He maintains an extremely antagonistic outwards stance towards his subject, Proof of Work cryptocurrencies, evident in his tweets. His Digiconomist website is just that — a personal website. It’s a hobbyist project run by an obviously conflicted individual. It’s not academia tier, not even close. His motive and mission is to exaggerate Bitcoin’s climate impact, whether it’s energy consumption, emissions, or e-waste (the White House paper cites him heavily on all three topics), which he does with gusto.

Quite simply, citing him at all is completely inappropriate for a purportedly scientific study. He is not a climate expert; he is not an authority on mining; he is a data scientist who works for the Dutch Central Bank. Most of his papers are published as “Commentary,” often in the journal Joule. These “Commentary” articles are not subject to peer review. They are basically glorified blog posts that happen to be hosted on the journal website. The Joule editorial team freely admits that Commentary articles are Opinion. Science, they are not. Let’s investigate a few of De Vries’ citations in the White House document:

• Bitcoin Energy Consumption Index: personal blog, not peer reviewed (cited over a dozen times)
• Revisiting Bitcoin’s Carbon Footprint, with Gallersdorfer, Klaassen and Stoll (2022) — commentary in Joule, no required peer review
• Renewable Energy Will Not Solve Bitcoin’s Sustainability Problem (2019) — commentary in Joule, no required peer review

These look to the uninitiated observer like peer reviewed scientific papers. In fact, they are non-reviewed blog posts in scientific garb. You’d think the White House Office of Science Policy might be exercising more caution and perform a cursory review of the resources they are citing. Maybe they don’t care.

This “commentary” trick is also done by De Vries collaborators Gallersdorfer, Klaassen, and Stoll, for instance in their 2020 article (of course cited in the OSTP report). It’s a common way to launder sciency claims into the scientific literature and press without actually facing any scrutiny for those claims. Journalists almost never verify that the articles aren’t actual scientific journal articles.

As for Digiconomist, it’s even more straightforward. It’s just not science. It was even rejected as a valid citation by Wikipedia for the energy section of the Bitcoin page (thanks to Level39 for pointing this out). As a hobby website/blog it is not rigorous. When fact checked by industry experts it is frequently shown to be erroneous. Its estimates are consistently far, far above those issued by more rigorous academics. But the White House Office of Science is much more permissive and is willing to accept De Vries’ amateur assertions as fact.

De Vries is cited 16 times in the document. Digiconomist is cited 23 times. Collectively, this makes De Vries the #1 source for the White House report. The far more authoritative, less exaggerated, and genuinely academic and neutral (taking no industry financing) Cambridge Center for Alternative Finance is cited 10 times.

Take table A4 in the appendix, which purports to compile GHG emissions for a number of cryptocurrencies from an array of academics. Of the 24 datapoints cited, fully 58% are De Vries/Digiconomist, one is Mora et al 2018 (basically tinfoil hat climate trutherism — more on that later), and one derives from the questionable Stoll, Klaassen, and Gallesdorfer. So that’s 16/24 cited datapoints in the table that are extremely questionable or completely meritless (I can’t overstate the absurdity of Mora’s paper in particular).

To address the substance of De Vries would take an entire dissertation, because his strategy is to generate a huge volume of material and claims, all of which take far more energy to rebut than they do to dream up. There’s three broad claims the White House relies on for De Vries/Digiconomist. Energy consumption estimates, emissions estimates, and e-waste estimates. Each has been dealt with elsewhere capably, but I’ll address them briefly.

1) Energy consumption

De Vries’s guesses come in way, way higher than his more credible peers. Look at the table A1 in the appendix: De Vries comes in at 144 TWh/y for Bitcoin versus Cambridge’s 88 TWh/y. For ETH, De Vries guesses 93.9 TWh/y versus Kyle McDonald’s 22.9 for the same time period. His estimates are consistently far, far higher than competing ones by other researchers. (He biases the estimate upwards for Bitcoin by assuming that the average ASIC hardware on the network is very old and hence inefficient, requiring more electricity per unit of hash than a newer fleet would.)

More generally, the energy consumption model De Vries relies on (that the White House is citing here) has been heavily, and in my opinion fatally, criticized by Ben Gagnon, the Chief Mining Officer at Bitfarms. The De Vries model doesn’t actually follow its own stated methodology. It didn’t reflect changes in the actual Bitcoin network in 2021, indicating that De Vries was manually tuning the model to make it show a higher energy consumption figure than even the model would have outputted. There’s additional issues with De Vries’ assumptions around hardware too: he contradicts himself, claiming at once the hardware turns over extremely quickly with his e-waste estimates, yet claiming Bitcoin has a very old fleet with his energy estimates.

2) E-waste

The e-waste claim, which relies entirely on a De Vries paper, is just absurd. De Vries flips from thinking the Bitcoin ASIC fleet is extremely old with his energy consumption estimates, to thinking that it turns over very quickly — falsely assuming in the e-waste paper that Bitcoin ASICs are fully depreciated and junked every 1.3 years. This 1.3 year estimate comes from a misapplication of Koomey’s law, a completely irrelevant observation about the historical growth in the efficiency of computers. This obviously has nothing to do with ASICs specifically, and cannot be used as the basis for a top-down estimate of e-waste. We have plenty of real, actual data on how long ASICs last. Certain vintages, like the s9, were still chugging away 5 years into their lives in 2021. These are machines that you turn on and spit out money. Someone will always run them, if they are producing above breakeven given electricity costs. This is why — contra de Vries — there is a vibrant secondary market for ASICs. Older models get “retired” from initial buyers to operators who have low electricity costs and who can still run them profitably.

This 1.3 full depreciation — and trashing of ASICs! — assumed by De Vries is a premise completely out of step with reality, easily determined if you to talk to miners or read their public disclosures or look at the secondary market. ASICs aren’t thrown away when they are old, they are sold on in the secondary market. They contain virtually no toxic parts and can be recycled (they are mostly aluminum by weight). I address the e-waste paper from De Vries and Stoll more fully in this letter to the EPA — signed by the CEOs of many large financial institutions active in the Bitcoin space. (Skip to section 7 in the letter)

3) Emissions

De Vries makes many estimates regarding cryptocurrency emissions, rather than just energy use. This is extremely questionable, and there’s a reason few academics do this: because we don’t have energy mix data which would be necessary to derive an emissions estimate. The Bitcoin Mining Council publishes some data on miner sustainability via voluntary disclosures covering about half the network — this is ignored in the White House report — but this alone isn’t sufficient. To estimate emissions, you’d need to know what power source miners are actually using. Many miners are foreign and run by private entities. Unlike the subset of the network that is run by publicly traded companies in the US, these private firms don’t disclose much about their energy inputs. So emissions estimates are extremely vague, to the point where presenting that data without extreme caveats is irresponsible. Generally, they are based on average energy mixes in regions where miners are located. But this is questionable, because miners often use highly idiosyncratic local energy sources (eg, collocating with a hydro dam), and they disguise their locations, so IP-based tracing is not reliable.

Cambridge, an authority on Bitcoin energy consumption, has so far held off on making any kind of emissions estimate for Bitcoin mining because that would require distinct bottom-up knowledge regarding miners’ energy mixes. So emissions estimates that the White House cites are from the most aggressive and least rigorous sources, like De Vries, who are happy to interpolate (always in favor of higher, scarier numbers) and make wild estimates. Most actual academics mostly shy away from making these guesses. You can’t “debunk” this with real data, because the data doesn’t exist. The correct attitude would be to admit you don’t know and wait for more data, rather than relying on some wild, inappropriate energy mix assumption.

So what should the White House have done? The better approach would have been to rely on credible, non-conflicted actual scientists and academics. For instance, Cambridge’s estimates are academic, nonpartisan, and widely cited. Of course, they don’t make fantastical claims about GHG emissions associated with Bitcoin, because it really isn’t possible to make any estimates like that with any confidence. Naturally, the OSTP sought to amplify the most exaggerated, sensational claims about Bitcoin’s climate impact. Merely academic ones are of little interest to them.

The OSTP may not care that the #1 source they are relying on is the non-academic, largely non-peer reviewed work of a Dutch Central Bank employee (who appears to have no practical knowledge of Bitcoin mining). But they should, because this fact alone discredits their paper. Many sections of the report rely entirely on De Vries’ wild guesses.

Reliance on Gallersdorfer, Klaassen, and Stoll

These three academics (who also collaborate at times with De Vries), also rely on this same trick of getting blog posts published in journals under the “comment” section. Their tactic is actually more insidious, because rather being motivated by an anti-PoW ideological crusade like De Vries, they actually do cash in on their academic efforts with a consultancy called the Crypto Carbon Ratings Institute (CCRI). The basic model appears to be the following: obtain credibility by publishing quasi-academic work lambasting the emissions associated with PoW, and use that to sell ESG-focused reports to help Proof of Stake blockchains [pdf] launder their reputations. These blockchains can then proudly proclaim themselves green (even if a financial consortium happily claiming to be green because it’s NOT Bitcoin is an absurdity), armed with a report from verified academics™ with a track record of publishing on PoW (never mind that many of the articles aren’t peer reviewed).

GKS are cited in the White House report as well as the CCRI directly. Shockingly, the CCRI Report [pdf] referenced in the OSTP paper, Energy Efficiency and Carbon Footprint of Proof of Stake Blockchain Protocols, was actually commissioned by the PoS blockchain Avalanche. Avalanche was founded by longtime PoW antagonist Emin Gun Sirer, and the protocol aggressively markets itself on the basis of its supposed sustainability. Commissioning reports from the likes of GKS to greenwash Avalanche is part of the normal playbook for PoS blockchains. It is remarkable that the White House would cite such a clearly conflicted report though.

I am not ultimately that concerned about GKS’ rather transparent grift. What does trouble me is the OSTP ignoring all data originating with the crypto industry, as if it’s tainted by capitalism, yet embracing data that is clearly biased in the opposite direction — in this case, by consultants selling anti-PoW, pro-PoS reports to PoS blockchain clients.

As for the rigor of their work, it’s questionable. It goes without saying that the CCRI reports cited in the OSTP report (footnotes 40 and 67, as well as repeatedly in the Appendix table A1) are completely non-academic. Stoll collaborates with De Vries on the infamous e-waste article (addressed earlier). The report cites the 2022 Revisiting Bitcoin’s Carbon Footprint, featuring De Vries and each of GKS, which is published as Commentary in Joule (non-peer reviewed). Energy Consumption of Cryptocurrencies Beyond Bitcoin, cited in the report, also appears in Joule as Commentary.

Citing Mora et al

There’s not a lot of things that can shock me here, but the inclusion of Mora et al as a citation really did surprise me. It’s well-known as probably the worst paper ever written on the topic of Bitcoin’s emissions impact. Entitled “Bitcoin emissions alone could push global warming above 2°C,” Mora et al is a running joke in the mining space.

Mora et al 2018 is part academic fraud, part performance art. It was written by undergraduates (Mora himself reportedly didn’t actually contribute to the paper, just lent his name to it so it would get published) as a class exercise in “how to get a paper published”. The paper itself, if you care to read it, is completely ludicrous. It supposes a model of Bitcoin that bears no relation to Bitcoin at all, and gets an obviously erroneous result (Bitcoin will increase the earth’s temperature by 2 degrees). I discuss it in this video and in greater detail in this blog post.

It’s worth noting that the Mora et al paper also appears as a probably-not-peer-reviewed ‘Comment’ in the journal ‘Nature Climate Change’. And if you actually read the paper, it’s hard to miss the three rebuttals included right there on the journal page. I’ve helpfully pointed them out so the OSTP can find them.

Easy to miss, I know.

If you read them, the rebuttals completely discredit the paper. The fact that this paper is still up, unredacted, is quite remarkable. Maybe because Nature Climate Change doesn’t really care that a blog post written by some undergraduates turned out to be unscientific garbage. Again, the problem is really that the OSTP figured this would be worthy of inclusion in their study. I’m sure they didn’t read the paper, because any sane person would realize how completely ludicrous it is. The problem is that they are further validating it by citing it and using it to bolster their arguments. I hope this is a case of ignorance rather than malice on the part of OSTP, but either way, it isn’t encouraging.

Deeply conflicted approach to data

The report stresses in many places the uncertainties involved in these estimates. That’s a step forward from the old model of governments simply rebroadcasting lunatic estimates or projections around Bitcoin’s climate impact and using that to set policy.

But the report fluctuates wildly between characterizing wild guesses from the likes of De Vries as fact, and then saying the estimates are uncertain and data is lacking. In places where we could have reasonably good models, like estimate Bitcoin’s future energy consumption (see the next section), they refuse to make an estimate.

Even though the report does emphasize data gaps and stresses the epistemic limitations of this topic, the authors are generally undeterred and plow ahead with naked assertions. Consider their assertion on page 6 around the emissions of major cryptoassets. There’s no footnote. It’s just a statement of apparent fact. Turns out it’s a Digiconomist estimate (which you have to do a fair amount of detective work to figure out, given the odd lack of footnote). So on the one hand, the OSTP is saying that the data is complicated and nuanced. And on the other, they’re making naked assertions regarding the most complex and nuanced piece of the entire debate (emissions estimates), and ones that are based on the Digiconomist, no less!

Another example of this is the report’s coverage of Texas. Texas represents a narrative violation when it comes to crypto mining; Bitcoin miners in Texas are aggressively going after cheap energy in West Texas particularly, which is a consequence of a massive renewable buildout (driven by federal subsidies) paired with a lack of local demand and insufficient long-distance transmission to load centers. Bitcoin miners in Texas are scooping up negatively or 0-priced energy, and opting into demand response programs such that they are offline during grid scarcity events. In short, they pay for cheap power that no one else will pay for (thus improving the economic profile of new renewables), and they don’t compete with households when energy is scarce. There isn’t a better energy buyer you could hope for.

The report however opts to cite erroneous figures in order to imply that miners are imminently taking over the Texas grid. They do this in multiple places. Here’s one such passage:

Texas, and has a peak summer electricity demand of about 76 gigawatts (GW), and current crypto-asset mining activity of about 2 GW. ERCOT has about 17 GW of crypto-asset facilities that are in the process of connecting to the grid, with an expected 5 to 6 GW of new demand in the next 12 to 15 months (equivalent to the power demand of the city of Houston). ERCOT may also see an additional 25 GW over the next decade. While many of these projects may not be completed, the prospect of up to 25 GW of new electricity demand from crypto-asset mining equivalent to a third of existing peak electricity demand in Texas raises potential challenges for maintaining electricity reliability, especially with rising power demands and extreme temperatures over recent years.

These numbers are huge. Seventeen gigawatts of power is an enormous quantity. As they helpfully point out, 25 GW would be equivalent to five Houstons. The only problem with this section is that the data is completely false.

ERCOT does not have 17 GW of crypto data centers that are in the process of connecting to the grid. There will most likely not be 5–6 GW of Bitcoin mining data centers connected in ERCOT in the next 12–15 months. ERCOT will not see 25GW (1.7 current Bitcoin’s worth) connected over the next decade. This 17 GW number in particular refers to non-binding, non-secured requests made speculatively to ERCOT. Nowhere near that amount of power actually stands to come online.

Every TX-based miner I connected with told me that the OSTP report numbers as described are erroneous, and the quoted 17 GW figure from ERCOT is deeply misleading. To confirm, I reached out to Lee Bratcher at the Texas Blockchain Council, arguably the organization with the most complete birds-eye view of mining in Texas. Lee related me the following:

Yes, we’ve asked the ERCOT Comms team to stop giving out that number without any context. There is currently between 1.5 GW and 2 GW of bitcoin mining in Texas. Another 2- 3 GWs could come on in the next few years. The 17 GW and 25 GW numbers are simply interconnection applications. Those do not require security to be posted. We are working with the utilities like Oncor and AEP to see if they would be willing to give an aggregated and anonymized number of who has posted security for their site. My guess is it would be around 1 GW of that 17–25 GWs that has actually posted any kind of collateral. Unfortunately the White House and the media are not interested in that nuance though.

Shaun Connell, EVP power at Lancium, a Texas-based data center company focusing on renewables, echoed Bratcher’s points. He told me that there was no financial cost for large putative loads to make requests to ERCOT to review new projects. Only once ERCOT greenlights the project does the applicant have to decide whether to move ahead and make the investment. So you had miners applying for the same massive project several times over with different utilities, hoping ERCOT would approve one.

Getting a request in the interconnection queue is a matter of submitting an Excel spreadsheet. There’s no meaningful filter or cost to doing this. Of the 8 GW in the territory of the utility Texas-New Mexico Power (these requests are evaluated jointly by ERCOT and the local utilities), 1 GW has been evaluated and only 50 MW have been approved so far. So that 2% yield should be understood as indicative of what you might expect in terms of energization.

So basically anyone — a miner, or broker trying to tout these contracts or relationships, could make a connection request to ERCOT at virtually no cost. During the bull market froth of 2021, there was a huge incentive to demonstrate confirmed capacity with ERCOT, even if it was just a trivial request for which no collateral had been posted.

From a more conceptual perspective, believing that 17 GW of new Bitcoin hashrate is imminent doesn’t make sense for other reasons. All of Bitcoin uses around 15 GW globally. If you more than double that, with new hardware, hashrate would increase by 3–4x. That massively compresses miner margins, and would lower the electricity price at which miners are breaking even. Adding this 17 GW would push miner breakevens to levels below average power prices in Texas. So the OSTP is basically positing that miners would expand their business well into the territory where they are guaranteed to lose money. This simply makes no sense.

The OSTP forgets that bitcoin mining is a business, and miners need to make money, or they will close up shop. The market size is bounded by the value of the Bitcoin rewards available to miners. This is a very real constraint on the amount of investment miners are willing to make.

The reality on the ground couldn’t be more different from the apparent apocalypse the OSTP is forecasting. The current status quo in ERCOT is a much slowed deployment pace, given worsened miner economics, a weaker capital markets environment and higher energy prices that further reduce the appetite of miners to build.

It’s also worth noting that the report assumes that energy is geographically fungible, as if a miner active in West Texas is depriving a household in Dallas of electricity. This isn’t the case; there is limited transmission from renewable-rich West Texas to the DFW triangle, explaining why power prices so frequently diverge between the two regions. Electricity decays with distance. Absent further high voltage transmission (and ERCOT’s CREZ, capable of carrying 15 GW, is already at full capacity ), you will inevitably develop local pockets of energy that simply go unused. Texas is a perfect case study showing how Bitcoin miners can surgically target these low-priced energy pockets. This is a category of energy I have dubbed ‘nonrival,’ because it doesn’t compete with other load centers at all; it’s only accretive, increasing the economic incentive to build more.

The debate really does suffer from a shortage of data, that is true. But the government isn’t exactly taking a welcome stance and inviting miners to collaborate with them in data sharing initiatives. Instead, they’re lambasting the sector, using junk data from fake academics or data that is simply erroneous, and threatening to ban the entire industry. If they had bothered to engage with actual miners with a knowledge of the Texas grid, they wouldn’t be making such mistakes.

“Can’t win” approach to miners using renewables

Probably the most frustration portion of the report concerns miners utilizing stranded natural gas or mining with renewables. Basically, the report dismisses all the efforts of miners to decarbonize their operations, laying out extremely narrow conditions in which mining with renewables might be considered acceptable. Suffice to say, I’ve never encountered the government insisting on conditions this stringent to any other buyer of grid electricity. Given that the report considers a full ban on mining in the US, the dismissal of miners’ genuine efforts to decarbonize should be deeply alarming. The OSTP sets them up to fail with a “can’t win” approach.

1) Dismisses flare gas mitigation

Despite a token admission of its usefulness, the report generally dismisses the merits of mining with otherwise flared or vented natural gas. First, the authors betray ignorance of how flared gas mitigation actually works. On page 24 they claim that mining with otherwise-flared gas doesn’t affect emissions one way or another. This is false: miners incorporating flared gas are able to combust the methane with a near perfect combustion efficiency, whereas generic flaring is low efficiency, especially in windy conditions. This is well documented in the literature. Thus, with flaring methane (a far more potent greenhouse gas) ends up vented anyway. Bitcoin miners are able fully combust the methane and convert it to energy and CO2. This is a direct improvement from an emissions perspective; it’s not merely emissions neutral like the report maintains.

The report also claims that the methane ought to be used for other uses, like hydrogen production, or that it should be exported via pipeline. This is a completely markets-blind claim. If the methane were market relevant, it would have been piped out. Many of these installations — the ones most suitable for flared gas mitigation — are off grid and have no pipeline infrastructure. Methane is a byproduct of oil extraction and it is not always the case that there are pipelines ready to go near a wellpad, nor is it necessarily economical to construct them. “If I were you,” the government is saying, “I would simply pipe out and sell the gas.” I am sure miners reading the report will appreciate the helpful tip.

Regarding grey/blue hydrogen (or other location-agnostic use cases) the OSTP should rejoice. Several Bitcoin miners have already announced their plans to build repurposable energy infrastructure that could be used for hydrogen product, if that market further develops. The fact is, the hydrogen market is generally not cost competitive with Bitcoin mining, especially as it requires additional physical infrastructure (you can just mine the Bitcoin into the cloud and sell it immediately. Getting the extremely flammable hydrogen to market is a bit more challenging). Bitcoin is just more location agnostic. OSTP again misses the mark by ignoring the insights of actual practitioners on this topic.

Third, the report goes on to dismiss the flared gas mitigation use case entirely by flatly asserting that “climate policy aligned with achieving net-zero emissions would have zero methane venting and zero methane flaring.” How convenient! According to the Net Zero Trajectory that we are all apparently on, the world will magically have no flaring, so mitigating the emissions associated with flaring or venting isn’t commendable.

Notwithstanding the Malthusian lunacy of a climate policy that purports to eliminate all oil extraction (of which waste natural gas is a byproduct), this statement is both myopic and utopian.

First, there’s no way to enforce a global ban on O&G, no matter how much the OSTP bleats about net zero. If the US bans all oil domestic and gas extraction, then competitor groups like OPEC would simply reap a massive windfall. (This would also put the US into the unenviable position of losing all energy sovereignty, a truly disastrous state to be in as our European allies are just learning.) In that scenario, mining with otherwise-flared gas would still be relevant, just overseas.

If the US bans flaring entirely, and, say, insists that all waste natgas is piped out from oil wells and brought to market, it will effectively be a subsidy to foreign oil producers that don’t share America’s flaring qualms. Foreigners will be able to mine more cheaply, because they will not be burdened with this additional step of building gas pipelines whenever they want to drill a new oil well. So the US will just be effectively imposing an artificial tariff on its on oil production and obliquely funding Russia, the Saudis, etc.

Maybe this is what the administration wants — the US disempowered, dependent on foreigners for essential petroleum (we aren’t deprecating this any time soon, all heavy industry depends on diesel, and planes still require jet fuel), and uncompetitive on the global markets. So if that’s their objective, by all means, illegalize flaring. But flaring will occur as long as oil is extracted from the earth, and miners will still be there to improve the emissions profile of that waste gas, and put it to use economically. And if the government thinks that miners should be producing grey hydrogen with that flared gas instead of Bitcoin, they are basically saying that the market is wrong.

2) Claims that using stranded renewables inhibits transmission

This is a wild one. The report claims that, basically, miners using stranded energy (as they famously do in Texas and other places with abundant renewables, which is well-documented at this point) means that there won’t be enough price signals for grid operators to know where to build transmission.

Using curtailed electricity can provide additional revenue to renewables developers and incentivize the construction of additional renewable energy capacity. However, it can also reduce the financial incentives to construct transmission from these renewables to existing users, or reduce the incentives to store excess renewable electricity to use when demand is higher. In addition, crypto-asset miners would not be likely to operate only during periods of curtailment, requiring consumption of grid electricity at all other times.

This is quite the claim. Basically, the government is saying that miners buying renewable energy — even if curtailed or undermonetized — isn’t praiseworthy, because that inhibits the signal to build more transmission.

Right. If only we had some kind of a sign that areas with lots of renewables (typically far from load centers) need to be linked to areas with lots of demand for electricity.

This is a chart of electricity prices in Texas on April fifth of this year. Notice how the west has instances of negative pricing (This is due to tax credits which compensate renewables for power generated regardless of local prices. This requires prices to be negative before its uneconomic for them to generate power), and the south and east is paying massively high prices? This is because electricity can’t just be teleported. If it needs to move long distances, it needs to be transported through long range, high voltage transmission, of which there is only a finite amount. So the market cannot clear at a single price.

As renewables further penetrate US grids, the cost of generation will go down (as renewables do not have marginal costs like electricity) but the need for transmission will go up. Bitcoin miners help monetize some of these renewables, but they don’t fix the core need for transmission. And I can assure you, grid developers can figure out where to put them. The signal is perfectly clear. The inhibitors are really just a lack of political will, NIMBYism (no one wants transmission lines in their back yard), and of course financial costs.

3) Gives miners no credit for subsidizing a renewable buildout

Lastly, the report dismisses Bitcoin miners buying renewables, adopting a weirdly rigid and unrealistic attitude to power:

When a crypto-asset mine purchases electricity from existing renewable sources, it displaces the GHG emissions in the near-term, shifting users of renewable sources to fossil fuel sources. This is because coal and natural gas often supply electricity generation for each additional unit of electricity demanded in the United States. As the amount of renewable sources is held constant, but electricity demand increases, additional fossil power will likely be dispatched. This displacement results in no net change or in increases in total global emissions through a process called leakage.

Given the report’s earlier insistence on the ‘electrify everything’ school of thought, it’s truly odd that they take such a zero sum approach here. Basically, the OSTP claims that buying renewable power just forces other grid consumers to buy thermal power, so it’s a wash. This assumes, wrongly, that:

• Bitcoin miners aren’t subsidizing the addition of new renewables in any way. This isn’t true. For instance, Aspen Creek is just one example of a miner that focuses on mining with renewables with additionality, meaning that they are bringing new power to bear, and only consuming part of it.
• Electricity is geographically fungible; that is, it can be transmitted anywhere at no cost instantly. In reality, power needs transmission and ends up trapped in pockets throughout the grid. Miners buying this power when it is negatively or cheaply priced are directly improving the economics of these renewable installations, and making it easier to justify building more.
• There are no ancillary benefits to having a flexible load on grid. This isn’t true either. Miners are a uniquely responsive load that can enhance grid flexibilization. Ever more renewable grids need massive flexibility from both the supply AND demand side. Miners can do this better than any other load resource, period. Here’s a great paper explaining this. More flexibility = more renewable penetration.

Indeed, a heavily renewable grid must be overbuilt to several times its nameplate capacity, because wind and solar are so intermittent. So in the glorious Net Zero future, there will necessarily be a massive overbuild of renewables. Having day-1 buyers for these renewables, especially location agnostic ones which can travel to the generation source, fundamentally improves the economics of these new buildouts. The report implies a scarce, zero-sum world, where the available power is fixed. This isn’t the case, nor is it even consistent with statements made earlier in the very same report! To achieve a real emissions reduction, tons of renewables will need to be built and grid flexibility will need to increase. Bitcoin Miners directly and indirectly help achieve both of these objectives.

Refuses to even take a guess on future energy trajectories

While the OSTP House is willing to repeat De Vries’ unscientific fever dreams about Bitcoin’s emissions with no issues whatsoever, they oddly draw the line at projecting future energy usage. This is a really odd move, because doing a back of the envelope projection for Bitcoin’s future energy demands is really very simple, and should be within the reach of the Scientists at the OSTP.

Refusing to propose any model at all helps the OSTP’s anti-crypto case, because it leaves future crypto energy demands wide open to the imagination. Most people, if not guided by reasonable models around future usage, tend to panic, relying on linear extrapolations of prior energy usage growth into the future. Most people, however, are ignorant of the effect of the halving, the declining growth rate of price, the effect of rising energy prices on consumption, and the real-world constraints to price growth and fees. These models — provided generously by industry, which the government has chosen to disregard — show that even under the most optimistic price models, it’s very unlikely Bitcoin becomes a ravenous energy basilisk.

In fact, in my own Bitcoin Net Zero report (coauthored with Ross Stevens of NYDIG), we find that even in the “high price” scenario of gold parity (which would price Bitcoin around ~$500k per coin!) Bitcoin mining doesn’t even reach 0.5% of global primary energy consumption.

None of this is complicated. Proof of Work is really just Bitcoin, so we can ignore everything else. For price, you can generate a few scenarios over a certain timeframe — say, you think Bitcoin might match gold’s capitalization in 15 year’s time. You already know the supply ahead of time. You then determine what you think fees might be, given historical trends. Right now, they are de minimis, but you can assume generously that fees will pick up again. Then you simply determine what share of revenue miners are likely to spend on pure electricity. This data is also attainable from historical trends. It varies, but if you think ASICs are likely to be more commoditized over time and miners will focus more on opex rather than capex, electricity spend might head to 50–70% of their revenue. Those are all the tools you need to construct a simple energy consumption estimate. I am not including an excel file — I trust the Office of Science and Technology Policy can do that part?

Generating energy mix projections and hence emissions trajectories is much more difficult, but you can always just use the generic US carbon intensity, which you expect to become more renewable anyway. (Even though many miners are low or zero carbon, especially newer ones). There you have an easy back of the envelope estimate. No need to fearmonger and pretend that Bitcoin miners in west Texas will be depriving hospitals in LA of electricity in the hyperbitcoinized future.

There are many ways to skin this cat. The OSTP, with all its abundant resources, doesn’t even try.

The OSTP recommendations are stupid and counter productive

The report recommends the following:

The Environmental Protection Agency (EPA), the Department of Energy (DOE), and other federal agencies should provide technical assistance and initiate a collaborative process with states, communities, the crypto-asset industry, and others to develop effective, evidence-based environmental performance standards for the responsible design, development, and use of environmentally responsible crypto-asset technologies. These should include standards for very low energy intensities, low water usage, low noise generation, clean energy usage by operators, and standards that strengthen over time for additional carbon-free generation to match or exceed the additional electricity load of these facilities.

Should these measures prove ineffective at reducing impacts, the Administration should explore executive actions, and Congress might consider legislation, to limit or eliminate the use of high energy intensity consensus mechanisms for crypto-asset mining. DOE and EPA should provide technical assistance to state public utility commissions, environmental protection agencies, and the crypto- asset industry to build capacity to minimize emissions, noise, water impacts, and negative economic impacts of crypto-asset mining; and to mitigate environmental injustices to overburdened communities.

To simplify, they want to require that bitcoin miners are bringing net new renewable generation online in order to be eligible to mine (“additional carbon-free generation to match or exceed the additional electricity load of these facilities”). No such requirement exists for any other industry in the US, period.

If miners can’t do this, “Congress might consider legislation, to limit or eliminate the use of high energy intensity consensus mechanisms for crypto-asset mining.” Of course, no such rule exists for any other extractive industry in the US, like gold mining (even though gold mining uses a comparable amount of (much dirtier) energy).

Effectively, the OSTP is asking that miners be forced to meet a completely impossible standard — again, one that isn’t asked of any other industry — and if they can’t meet that, they are asking Congress to regulate mining out of existence.

Overall, the report is laced with a profound neo-malthusian attitude. Even though it gives lip service to Net Zero goals like “electrify everything” which would require a massive buildout of power and transmission (implying that we will have an energy abundant future), the fact that the US government is so intent on marginalizing an industrial sector that accounts for around 0.5% of electrical generation — much of it otherwise going unused in places like West Texas — should give anyone pause. If the government was really that confident in the energy-abundant, everything-electrified green transition, why would they be worried about 0.5% of current generation? If their Net Zero trajectories are met, the electrical sector will be greening at a rapid clip anyway, and with it, purchasers of grid power.

Unlike virtually any other industry, Bitcoin is already fully electrified, so it can benefit from the (envisioned) greening of the grid. So why is a large buyer of energy, that is fully electrical, location agnostic, interruptible, and portable, a threat to the green transition? Could it be that the government doesn’t believe their own incantations around Net Zero? Could it be that their motives have more to do with using ESG to politicize the electricity sector and determine politically acceptable uses of energy, like an electrical version of Operation Choke Point?

Reading the report, and in particular the section on miners using renewables, it is very clear that the current administration views electricity as a good that only politically favored firms should have access to. If they are able to “choke point” electricity, they won’t stop with Bitcoin mining. They will move on to demanding that utilities shutter electricity used by politically disfavored entities like firearms manufacturers, religious institutions, and right-wing educational facilities.

The government already politicizes access to finance, and in a ghastly partnership with big tech, politically determines who has access to internet infrastructure. Why should electricity be any different? The totalizing state doesn’t know any restraint. Obviously their political enemies should be deprived of any resources, whether financial, communications, or literal energy. Bitcoin just so happens to straddle all three of those sectors.

From a practical point of view, a domestic ban on PoW would be counter-productive. I’ve already pointed out that policymakers in Western countries concerned about Bitcoin’s emissions should convince miners to stay domestic. A ban would be massively counterproductive to the US’ ability to influence both Bitcoin and the network’s emissions trajectory. If the US were to successfully ban industrial Bitcoin mining, miners elsewhere — almost all of them dirtier than US miners — would immediately receive a massive dividend. Naively assuming the 30–40% of BTC mining (the current US total) comes to a complete halt, non-US miners would almost immediately be producing 42–66% more units of Bitcoin with their same level of exertion. US miners experienced this when China banned the vast majority of domestic mining and ex-China miners were rewarded a bumper crop.

Regulating Bitcoin mining is like squishing the water inside a water balloon and expecting the balloon’s volume to decrease. Squeeze hard enough, and it just goes to the other side of the balloon. Xi Jinping, the great dictator of our time, wasn’t able to stamp Bitcoin out when he forced miners out of China. They just packed up and left — Bitcoin was completely unaffected, and western miners profited greatly. (Sidenote: if even the most powerful man in the world, Xi Jinping, was unable to eliminate mining — not even within his own borders — how does the generally hapless Biden admin expect to achieve anything different?) It’s also worth noting that there is still ‘black market’ mining in China.

If you hate the ecological impact of gold extraction, you don’t ban gold mining domestically — that simply advantages competing producers (in case you’re wondering, that’s China, Russia, and Australia). Gold will still be mined and make it onto global markets, so a ban does nothing to affect the gold mining industry overall.

Instead, you regulate the industry and ask participants to please not dump mercury in rivers. The same could be done with Bitcoin miners. Ask them to disclose their energy mix and be transparent with regards to their emissions and participation in grid stabilization. You could easily request that they stay away from populated areas given the noise pollution. You’ll find that miners are very willing to play ball.

In a sense, the naïve approach the Biden admin is taking around mining echoes carbon accounting generally. American pundits like to celebrate our declining emissions while ignoring that this is a consequence of America’s deindustrialization and offshoring of almost all heavy industry to places like China. Not to mention that the green revolution built on wind turbines, photovoltaics, and batteries is almost entirely dependent on emissions-intensive Chinese heavy industry.

Yes, you can “decarbonize” by decommissioning nuclear and coal plants and substituting them with wind and solar (albeit at the cost of a more unstable grid). But you are now importing new emissions from China via those PV panels and batteries. You can eliminate energy-intensive aluminum smelting or industrial manufacturing and import the finished goods from coal-heavy China. You haven’t reduced the CO2 molecules in the atmosphere. The sky doesn’t care about the borders between countries and your fancy country-based carbon accounting.

Equally, pushing mining outside of the US doesn’t help the climate (most likely, it makes things slightly worse by putting mining in the hands of Russia, Iran, Venezuela, and North Korea). It might make western policymakers feel better, but it raises the carbon intensity of the network. The U.S. is the most capacious energy grid in the world, is adding renewables at a rapid clip, and has many pools of stranded energy, which are only growing, as generation (especially of new renewables, generally not located near population centers) outpaces transmission. This is empirically observable.

On the more insidious side, the report is definitely infected with this sentiment that perhaps PoS could be a plug and play replacement for PoW, and maybe this problem will solve itself. (Note again that the government is citing non-academic lobbyist work funded by PoS protocols in this paper.) Partly I ascribe this to naivete, but you could also point to more sinister motives, given how much more controllable PoS networks are than PoW. It’s very much in the interest of the government (that seeks to politicize all of finance) to advocate for PoS over PoW. I’ve pointed this out before: PoS is simply more capturable, because coins tend to accumulate in large financial institutions, which are trivially influenced by the State. Miners are far more resistant to capture and control. Empirically, hashrate is far, far more distributed among a diverse, global set of miners than capital held by stakers is.

The envisioned substitution of PoS for PoW the Biden admin hints at is simply a move to outlaw an asset the US government can’t control and replace it with one it can.

Ultimately, what the administration thinks of Bitcoin mining is a reflection of whether it’s willing to embrace a high-energy, pro-American dynamism, energy-independent future in which we reshore our industrial capacity, or if it would rather take a neo-malthusian scarcity mindset and play political games about who is entitled to which resources.

Think of it this way. If a new industry came along tomorrow with the following properties:

• It consumes energy, but it doesn’t matter where that energy is produced
• It is fully digital, so requires no physical infrastructure aside from electricity, transformers, and a datacenter
• It is completely interruptible, and can be curtailed fully at any time, so it could be fully offline whenever the grid needs it (a feature that renewable grids increasingly require)
• Because it’s fully digital, it can be rendered as sustainable as its electricity inputs, and indeed, by most measures, it appears to be more sustainable than any single other heavy industry in the US
• It renders a service giving property rights to the entire world
• It buys up low or negatively-priced renewable energy, vastly improving the economics of new and existing renewable projects that would otherwise not be able to monetize
• If the industry goes away, the infrastructure built could be repurposed for other location-agnostic uses, like the generation of green hydrogen, a key part of the renewable transition
• It mitigates emissions associated with unavoidable gas flaring, a byproduct of oil extraction
• It rebuilds a high-energy infrastructure in the heartland of America, laying the groundwork for a desperately needed reshoring of industrial capacity
• If you ban it, its aggregate emissions will go up. If you embrace it, its aggregate emissions will go down
• If you ban it, you empower your enemies, like Russia, Iran, Venezuela, and North Korea. If you embrace it, you directly hurt them, and give their citizens tools to free themselves from those oppressive regimes.
• It represents tens of billions of equity value and offers a massive tax windfall to states that can sell their otherwise unsold energy to the producers of this product

… would you ban this industry? Or would you embrace it?

I know my answer.

I only hope that the White House can eventually come to their senses and reach the same conclusion.

Thanks to Level39, Shaun Connell & Lancium, and Lee Bratcher for their assistance with this article.