Active Competitions
Research Library
Strong OOS survival
Mixed OOS evidence
Strong OOS survival
Why most factors fail
Curriculum
Introduction to Quantitative Finance
Foundations
Machine Learning for Markets
Intermediate
Market Microstructure & HFT
Advanced
Alpha Decay & Portfolio Construction
Advanced
The Problem
Most platforms grade the wrong thing.
In-sample Sharpe is easy to manufacture. Fit the noise, overoptimize the parameters, and your backtest looks excellent. Then you trade live and lose money. ConvexPi grades you the way the market does — on data you have never seen.
What most platforms grade
IS Sharpe
In-sample performance on data you trained on. Meaningless in live trading. Easy to overfit.
What ConvexPi grades
OOS Sharpe
Performance on a holdout market with a secret seed. You cannot overfit what you cannot see.
Evidence
Most published anomalies decay out-of-sample.
We track the Fama-French factor zoo against live markets. Some effects survive. Many do not. This is what you are competing to find.
| Anomaly | IS Sharpe | OOS Sharpe | Status |
|---|---|---|---|
| Market | 0.47 | 0.44 | weakened |
| Size (SMB) | 0.20 | -0.02 | dead |
| Value (HML) | 0.47 | 0.25 | weakened |
| Momentum (Mom) | 0.63 | 0.38 | weakened |
| Profitability (RMW) | 0.56 | 0.29 | weakened |
| Investment (CMA) | 0.69 | -0.16 | dead |
Platform
Three environments. One question.
Can your strategy survive data it has never seen?
Lab
Alpha Discovery Lab
Write strategies against a synthetic equity panel with embedded factor signals. Six structured missions take you from naive overfitting to disciplined out-of-sample evaluation.
- Python-native (pandas, numpy, scikit-learn)
- IS / OOS breakdown on every submission
- Six missions: intro through ML methods
Arena
Live Market Arena
Deploy agents that submit live orders to a limit-order-book simulation running in real time. Compete against classmates, market makers, and noise traders.
- WebSocket agent API — connect in Python
- Inventory risk, spread capture, adverse selection
- Survival score replaces Sharpe ratio
Courses
Instructor Cohorts
Private competitions for classroom settings. Assign missions as homework, grade via OOS Sharpe, export gradebooks. FERPA-compliant by default.
- Private leaderboards with deadlines
- Instructor dashboard and gradebook export
- University of Cincinnati — Fall 2026 cohort active
Community
Build in the open. Follow the research.
Every submission is graded and published to a public leaderboard. Follow other researchers to see their strategy results in your feed. Share your GitHub to let others learn from your code.
Follow researchers
See their submissions and OOS Sharpe scores in your activity feed.
Link your GitHub
Share your strategy code publicly so others can learn from your approach.
Research library
8 factor deep-dives with key papers, OOS survival evidence, and platform missions.
Anomaly tracker
Pre- and post-publication Sharpe ratios for canonical equity anomalies.