HFT-bot

Inspiration

Money Money Money

What it does

We implemented a trading bot to trade Stocks and ETFs in a completely random market. The market consisted of 3 stocks and 2 ETFs, which were distributed in 2 different time-zones. When both markets were open, our bot used both a hedging algorithm and an arbitradge algorithm to profitably trade against bots and other users. We managed risk by balancing the bot's dependency on our implemented algorithms. We leveraged a modified arbitradge algorithm to trade with a smaller volume when only one market was open, to balance risk and profitability.

How we built it

Our bot was implemented in python. Using the given API to place bids and selling offers, we were able to create a continously running algorithm that would adjust to the changing price books, to place the most profitable trades. We performed calculations to check the price of the ETF in its different markets, to find trading opportunities. To debug and follow our transactions, we logged important values to be able to monitor them.

Challenges we ran into

During the development of the individual algorithms, we were successful in becoming profitable in the market. Challenges started to arise when we began trading against a multitude of people, where we had to improve our algorithms to adapt to different behaviours in the market. We also had trouble when we began to combine our algorithms altogether, as we had to manage aspects such as our position and trade volumes. This required a lot of refactoring to ensure none of the set market rules were breached with our bot.

Accomplishments that we're proud of

We became the only team to reach a PnL (Profit and Loss) of 1 Million!

What we learned

Collaboration, teamwork, planning, ... This was an incredible learning experience for our team, who had never participated in such a rigorous event. We had to learn how to deal with time and pressure, and we expanded our understanding in programming as well as project management.