After countless late-night coding sessions fueled by caffeine and determination, our team proudly presents a market-making solution that is both very simple and effective.
In a world where risk-taking is the norm, we decided to play it smart by operating in the background, orchestrating the market to churn out profits. Our main strategy is as straightforward as it gets—we aim to be the Usain Bolt of trading, the fastest in the game.
Our bread and butter? Arbitrage from dual-listed stocks. We've fine-tuned a simple approach that cuts through the noise, allowing us to ride the waves of market discrepancies with precision. No fancy language models here—those take up too much precious time and performance. Instead, we opted for well-tuned classic machine learning algorithms, quick and efficient like a well-oiled machine.
How did we pull it off? Picture a motivated, sleep-deprived team of four, each laser-focused on their part of the puzzle. We dived into a sea of experiments, optimizing and tuning our simple algorithm until it was as sharp as a razor's edge. Along the way, we explored various ML approaches, discarding the complex ones for the sweet spot of performance and speed.
Challenges? Oh, we had our fair share. Perfecting the ML model was like walking on a tightrope—little training data, big expectations. Testing our market simulation theories in a constantly changing landscape added another layer of complexity. But hey, challenges are just opportunities in disguise, right?
Our proudest accomplishment? We're almost always holding no stock at all. We're not risking it all; we're making the market itself.
What did we learn? Market making and quantitative trading became our second language. We also learned the art of persistence in a cutthroat environment. Never giving up became our mantra, and it paid off.
Log in or sign up for Devpost to join the conversation.