Verify a strategy

A pretty backtest is cheap: run enough configs and the best one looks brilliant by luck alone. This tool runs the three statistics that catch that — the Deflated Sharpe Ratio (corrects for how many backtests you ran and for fat tails), the probability of backtest overfitting (does your in-sample winner survive out-of-sample?), and the out-of-sample Sharpe haircut. It catches selection bias and overfit if you disclose your full search; a hidden-trials overfit can still slip through, and a backtest stamp is always weaker than a forward track record. Same math as our own certification engine, running entirely in your browser.

This is a research and education tool, not investment advice. A CERTIFIED verdict means the disclosed backtest family clears the anti-overfit statistics — it says nothing about future returns, costs, capacity, or whether the returns were generated honestly (lookahead, survivorship). No verdict here is a recommendation to buy, sell, or allocate to anything. See the disclaimer.