Inspiration
To extract new information out of current situations with very low risk. The thing which interested me the most was foreign exchange arbitrage. Since it is low risk and committing arbitrage leads to smoother markets better market liquidity and overall better experience for trades. However the problem with arbitrage is that someone with faster information will beat you in any trade since he has better access to information. However with twitter sentiment analysis I am able to predict changes in prices of foreign exchange to the price of Bitcoin.
What it does
First I get twitter tweets from 4 different countries in 4 different languages. Then I use machine learning to obtain sentiment analysis for the four different districts. Then apply the black scholes equation to the volatile prices of Bitcoin and the time equation of stock to predict the upcoming second changes in the price of bitcoin then apply my own home made arbitrage algorithm to compare the currency value of bitcoin to find discrepancies. Then I execute the most profitable trade.
How I built it
I used a ton of github repos and python libraries. Also I have Azure for the machine learning analysis and a fun point of the project was the comparison of price of flights to the destination of the highest money difference using Amadeus API.
Challenges I ran into
Implementing APIS to get together. Get clean data FAST. Minimizing overhead and enhancing performance of the algorithm.
Accomplishments that I'm proud of
I AM MAKING MONEY!!! COMMERCIALLY VIABLE!! Around 1 cent for $100 for roughly a minute. With 10k it is possible to get $600 assuming a volatile market with fitting to black scholes equation. In perfect market indecisiveness it is possible to obtain return of 8% however the market always recovers.
What I learned
It is incredibly difficult to extract alpha out of the market.
What's next for Money Well
Have access to faster API to lead to faster extraction of Alpha and also better trading platforms which have stop options and hedging options to lead to more choices in trade. We can also increase the number of currencies to trade in and improve twitter sentiment analysis to more location based.
Log in or sign up for Devpost to join the conversation.