Inspiration
Wall Street Bets GME 2021. We also have no experience in this field, therefore we would like to try a new challenge.
What it does
It is an algorithm that does pair trading, which buys ETFs and sells its components stocks, and vice versa when an opportunity comes where the bot can sell for more than it can buy.
How we built it
Following the guidelines, it is separated into two parts. For the first one, we managed to build the pair-trading bot. The second part of the project was a market maker bot. However, due to time constraints and lethargy, there were serious bugs in the second part of the project. In the middle of bug fixing, we managed to optimise the first part really well, so we ended up abandoning the 2nd part of the project.
Challenges we ran into
Had no finance background and struggle to understand concepts introduced in a challenge, therefore took a long time to get started. Online resources did not provide direct help so we discussed it together
and eventually understood the assignment. Besides, the provided Cloud9 IDE only allows one person to edit files, therefore this causes inconvenience and disturbs the workflow.
Accomplishments that we're proud of
Initially, we are in second last place and have significantly lower PnL values, but after 4 hours of work, we managed to see improvements and were second before the first submission. As a result, we developed a working product with significant improvements to non-functional versions. We ended up in 2nd place out of 15 teams when the hackathon ended, and was 1 out of 3 teams whose bots actually managed to make money.
What we learned
We developed a deeper understanding of how markets work, in particular about liquidity, and how bid and ask orders are used under the hood. We also learnt how to properly work as a team, work under pressure and proper time management skills.
Built With
- cloud9
- optiver
- python
Log in or sign up for Devpost to join the conversation.