Is My Backtest Bullshit?
Is Your Backtest Real or Overfit?
Most backtests look good. Most backtests are overfit. This tool tells you the difference. Paste your trade returns and get:
- Monte Carlo bootstrap (400 runs) — what's the probability of loss?
- Walk-forward stability — does your edge hold across different time windows?
- Deflated Sharpe ratio — adjusted for multiple testing
- Robustness score (0-100) — above 50 is likely real, below 30 is probably overfit
How to Use
- Paste your trade returns (one per line, in bps or %)
- Click "Analyze & Score"
- Get your robustness score and verdict
Works with trade lists from TradingView, Excel, Python, or any platform. Just paste the returns.
Frequently Asked Questions
- What is backtest overfitting?
- Overfitting is when a strategy looks good in backtest but fails in live trading. It happens when you optimize parameters to fit historical data too closely. This tool detects it.
- What is a good robustness score?
- Above 50 means your strategy likely has real edge. Below 30 means it's probably overfit. Between 30-50 is borderline — proceed with caution.
- How does Monte Carlo bootstrap work?
- It resamples your trade returns 400 times to simulate different orderings. If most simulations are profitable, your edge is robust to trade ordering. If many are losses, your edge depends on lucky sequencing.
- What is walk-forward stability?
- We split your trades into 4 windows and check if the Sharpe ratio is positive in each. If edge only appears in one window, it's not stable — it's likely overfit to that period.
Try dMoERA Studio — build, backtest, and deploy strategies with production-grade validation →