NBS-Optiver

Inspiration

Our team wanted to explore how real-time sentiment and fast decision-making could actually improve trading strategies. The idea of combining social media signals with high-frequency arbitrage felt exciting and challenging enough for a hackathon.

What it does

NBS-Optiver monitors dual-listed stocks, checks for price gaps, and uses sentiment from online sources to decide how aggressively to trade. It reacts instantly, adjusts its thresholds based on market mood, and uses built-in safety checks to keep positions balanced.

How we built it

We split the system into multi-threaded modules so each stock could be handled independently. We built a sentiment polling component, a price-tracking and spread-detection engine, and an automated hedging mechanism to fix imbalances.

Challenges we ran into

The hardest parts were syncing sentiment data with real-time price movements, correctly gauging the current market state, and keeping the bot from taking on risky positions during volatile moments. Tuning thresholds without overfitting was also tricky.

Accomplishments that we're proud of

We’re proud that the bot actually reacted to sentiment in meaningful ways, stayed stable under rapid updates, and executed spread trades reliably. Seeing our team’s ideas turn into a working high-frequency system felt like a big win.

What we learned

We learned how important clean architecture is for real-time systems, how sensitive arbitrage strategies can be to delays, and how sentiment can shift market behavior more than we expected. We also improved our teamwork and debugging under pressure.

What's next for NBS-Optiver

We want to refine the sentiment model, test more asset pairs, and add smarter risk controls. Longer-term, we’re interested in exploring stronger ML-based prediction and running the bot under more realistic market simulations.