The Signal in the Noise Most hackathons ask you to build an app. We ask you to predict the future. The Martingale is the first dedicated Quantitative Finance competition for students only. We are looking for the top 0.1% of student-engineers capable of applying Natural Language Processing (NLP) and Stochastic Modeling to financial markets. Wall Street no longer runs on gut feeling; it runs on unstructured data. Your challenge is to build an algorithm that digests thousands of financial news headlines to predict market volatility and direction. The Challenge: Sentiment Alpha You will be provided with a dataset containing financial news headlines paired with high-frequency market data.
The Input: Unstructured text (Headlines, Tweets, Earnings Call Transcripts).
The Output: A trading signal (Buy/Sell/Hold) for the S&P 500 (SPY).
The Metric: Your model will be evaluated on its Sharpe Ratio (Risk-Adjusted Return) and RMSE against a hidden Private Test Set.
Why Compete?
The Dossier: The resumes of the Top 10 finalists will be compiled into an "Investment Memo" and forwarded to our partners (University Labs and Proprietary Trading Firms).
The Compute: Top teams gain access to institutional-grade APIs from Polygon.io and Wolfram Alpha.
The Glory: Prove your code can survive a Bear Market.
Start Now: https://www.kaggle.com/competitions/martingale-hacks-winter-2025/data!
Online
Martingale League
$0