SensiTrade

Info

Team Number: 26

trading algorithm file: sentiment_analysis/trader_working.1.py

What it does: Following a traditional trading strategy (dual listing is combined into one stock) at the same time monitoring the news with a language model and reacting to them by bying & selling bigger amounts with predicted prices.

Inspiration:

The inspiration behind SensiTrade stems from the Optiver Challenge, where the objective is to design an efficient trading algorithm to navigate the complexities of financial markets. Drawing inspiration from the challenge, we aimed to create a trading bot that not only maximizes profits but also demonstrates adaptability to the dual listing situation and reacting to news.

What it does:

SensiTrade is a sophisticated trading bot designed to operate seamlessly within the Optiver Challenge framework. Leveraging advanced algorithms and the real-time exchange, SensiTrade executes strategic trades, aiming to optimize returns while managing risk using hedging. The bot is equipped with features to analyze market trends, identify potential opportunities, and execute trades.

How we built it:

SensiTrade was built using a combination of NLP technologies and some financial modeling techniques. We utilized Python for ease of integration with the trading framework. The bot relies on machine learning models to analyze news data and make predictions about market trends. Additionally, we incorporated risk management strategies to ensure the bot's resilience in the face of market volatility.

Challenges we ran into:

Building SensiTrade posed several challenges, including adapting and selecting parameters, model fine-tuning, and the need to optimize for the right trading frequency execution between exploiting knowledge about news and traditional trading.

Accomplishments that we're proud of:

SensiTrade successfully navigates the Optiver Challenge environment, consistently making strategic trades and demonstrating competitive performance. We are proud of achieving a balance between risk and reward.