Drift Score
Compare recent returns to an already-analyzed report's confidence interval — is your edge already dying?
Retrieved from a previous analysis response on /en/tools/backtest-analyzer.
Financial Methodology: How the Drift Score Measures Strategy Drift
The Drift Score compares the annualized Sharpe ratio of your recent (“live”) returns with the Monte Carlo confidence interval (5th and 95th percentiles) computed during the original audit of your report. It is not a second verdict: it is an early-warning signal to cross-check against sample size.
Data thresholds
With fewer than 5 observations the tool returns “insufficient data”. If the original report was too short to produce an interval, the status is “no baseline available”.
Reports tied to your account
The calculation only runs on a report generated by your own account (Pro plan feature): it reflects your real performance and is never exposed publicly.
YMYL notice: educational, deterministic content. It is not personalized financial advice, a buy/sell recommendation, or a promise of returns. Past performance does not guarantee future results.
Frequently Asked Questions About the Drift Score
The annualized Sharpe of your recent returns is below the 5th percentile of the original backtest's Monte Carlo interval. It can signal strategy drift, but a short recent sample is noisy, so the result must be read alongside the number of observations.
At least 5 numeric values (maximum 20,000). The longer the sample, the more stable the live Sharpe; with very few points the result is statistically fragile.
No. It only means the recent Sharpe is consistent with the report's historical interval. It does not predict future performance and is not investment advice.
It compares real returns with a report you analyzed using your account. That data reflects personal performance and therefore stays private, unlike a verdict's public badge.
Complementary Tools & Guides
Deepen your statistical strategy audit with our other Quant Lab engines and quantitative validation research guides.
More calculators in this hub
Backtest Analyzer
Sharpe, Sortino, and Max Drawdown from QuantConnect/MT5 CSVs.
Multi-Strategy Portfolio
Cross-strategy correlation and combined drawdown risk.
QuantEducator AI
Jupyter notebook generation and Python financial models.