SmartStonks

Inspiration

The financial markets are driven by complex dynamics, where inefficiencies create opportunities for those who can adapt quickly. Inspired by the rise of AI in trading and the semiconductor industry’s importance in the global economy, we wanted to develop a solution that thrives in volatile environments while pushing the limits of algorithmic trading. The HackaTUM challenge provided the perfect opportunity to combine our passion for finance, technology, and innovation.

What it does

SmartStonks is an algorithmic trading platform that identifies and capitalizes on arbitrage opportunities between ETFs and their underlying stocks in real-time. It:

Detects price discrepancies between ETFs and their correlated stock baskets.
Executes hedging strategies to reduce risk.
Functions as a market maker, providing liquidity while optimizing exposure limits.
Adapts dynamically to market conditions, ensuring profitability even in non-overlapping trading hours.

How we built it

We built SmartStonks using:

Python: The primary programming language for algorithm logic and API integration.
Optibook API: To interact with the simulated market environment.
Pandas and NumPy: For analyzing large datasets and performing real-time computations.
Custom Risk Management Module: To ensure compliance with risk limits and competition rules.
An iterative process of backtesting strategies and optimizing performance in live simulations.

Challenges we ran into

Non-overlapping Markets: Trading effectively when one market (e.g., US or EU) was closed required precise estimation of theoretical prices.
Real-Time Decision Making: Handling high-frequency data streams and making split-second decisions was computationally intensive.
Exposure Management: Balancing risk and reward while adhering to strict exposure limits required dynamic adjustments to our strategy.
Algorithm Calibration: Finding the optimal parameters for price thresholds, order sizes, and timing in a volatile environment.

Accomplishments that we're proud of

Successfully developed a working algorithm that consistently captured arbitrage opportunities in the simulation.
Improved market liquidity by functioning as an effective market maker.
Adapted to challenges posed by non-overlapping trading hours with minimal errors in theoretical pricing.
Demonstrated a robust hedging mechanism to manage risks while maintaining profitability.

What we learned

Financial Dynamics: Gained a deeper understanding of how ETFs and their underlying stocks interact in global markets.
Risk Management: The importance of balancing aggressiveness with caution in trading strategies.
Collaboration: Working as a team to iterate on ideas, solve complex problems, and adapt under pressure.

With SmartStonks, the future of algorithmic trading is here! 🚀