HackZurich - Optiver Challenge

Inspiration The HackZurich - Optiver Challenge draws its inspiration from the world of financial markets and trading. It's inspired by the continuous quest for innovation and optimization in the field of algorithmic trading. The challenge seeks to harness the power of automated text processing tools and natural language understanding to enhance existing trading strategies, adapting them to the dynamic and information-rich environment of modern financial markets.

What it does The challenge invites participants to modify an existing stock quoting algorithm, typically used in market making, to incorporate automated text processing capabilities. The modified algorithm should be able to analyze incoming news messages and social media feeds in real-time, extracting valuable insights to inform trading decisions. Essentially, it aims to bridge the gap between traditional quantitative trading strategies and the world of textual data, exploring the potential of natural language processing in financial markets.

How we built it The challenge was built through collaboration between HackZurich and Optiver, a leading global market maker. It leverages real-world trading scenarios and data to create a practical and engaging competition. The challenge is designed to be accessible to both novice and experienced data scientists and developers, providing them with the necessary tools, datasets, and support to tackle complex problems in algorithmic trading.

Challenges we ran into While designing the challenge, several challenges were encountered, including:

Data Complexity: Financial market data can be extremely complex and high-dimensional. Ensuring that participants have access to relevant and realistic data posed a challenge.

Algorithm Integration: Integrating automated text processing tools into an existing trading algorithm required careful consideration of technical compatibility and efficiency.

Evaluation Metrics: Defining fair and effective evaluation metrics to assess participants' solutions and strategies was a non-trivial task.

Scalability: Ensuring that the challenge could accommodate participants with varying levels of expertise and computational resources while maintaining fairness was a challenge.

Accomplishments that we're proud of The challenge organizers are proud of creating an innovative and intellectually stimulating competition that brings together the worlds of finance, data science, and natural language processing. They are also proud of fostering a collaborative environment where participants can learn from each other, share insights, and potentially make meaningful contributions to the field of algorithmic trading.

Built With

• amazon-web-services
• jupyter
• optibook
• python