Does this strategy deserve capital, or is it noise dressed up as edge?
TrueVerdikt runs your returns through a statistical engine that never flatters: Deflated Sharpe Ratio, purged cross-validation, regime stress-test. A clear verdict — GO, CAUTION or NO-GO — in two minutes.
No raw data stored • Signed, verifiable verdict • Free to start
A score built to reduce validation bias, not to flatter you
Most backtest platforms prove you right. TrueVerdikt does the opposite: it actively tries to break your strategy before the market does it for you. That's the only way to know if your edge is real.
Academic rigor
Deflated Sharpe Ratio (Bailey & López de Prado), purged CPCV, Lo's autocorrelation adjustment. The real formulas, not marketing approximations.
Translated to plain language
A DSR of 0.62 never shows up alone: always paired with an actionable verdict and concrete recommendations to fix it.
Cautious by design
A catastrophic drawdown blocks the GO, whatever the Sharpe. The score sees risk, not just average performance.
Far beyond the Sharpe ratio
Deflated Sharpe Ratio
Adjusts the Sharpe for how many variants you tested — anti-overfitting.
Purged CPCV
Stability measured across all combinatorial splits, not one lucky split.
Regime stress-test
Does your edge survive high volatility, or only when markets are calm?
Tail risk
Max drawdown, 95% CVaR, Sortino, skew/kurtosis — the risk the Sharpe ignores.
Position sizing
Fractional Kelly weighted by statistical confidence: with how much capital?
Verifiable badge
Every verdict is signed and shareable, without ever exposing your raw returns.
Drift Score
Is your live edge already diverging from the backtest? Automatic alert before it becomes obvious (Pro).
Portfolio analysis
Two GO strategies can be correlated and concentrate risk more than it appears separately.
Analyses, tutorials and tools
New free ebook: Systematic Portfolio Optimization
A free reference guide on the complete institutional quant pipeline: RMT spectral filtering, shrinkage, risk parity, regime detection and Half-Kelly sizing.
Why the Sharpe ratio isn't enough to judge a strategy
A high Sharpe can hide a catastrophic drawdown or pure luck from multiple trials. Here are the blind spots of the most-used ratio in quant finance, and how to fill them.
Data-snooping in trading: the beginner quant's #1 trap
Testing 1000 variants then keeping the best guarantees a nice backtest… and a live letdown. Here's how to spot data-snooping and protect yourself with the Deflated Sharpe Ratio.
Tech & Finance ebooks & guides
Practical guides to go from backtest to real capital, written by practitioners.
From your CSV to a verdict, in three steps
No install, no brokerage account to connect: your data stays yours.
Export your returns
A .csv file with a "return" column: one return per period, as a decimal. Minimum 30 rows. Every other column is ignored, so you can upload the raw export from QuantConnect, MetaTrader or pandas.
Describe the backtest context
How many free parameters, how many variants you tested before keeping this one, and the annualisation convention. Those two numbers are what let us deflate a Sharpe inflated by repeated trials.
Get the verdict
GO, CAUTION or NO-GO, with the Deflated Sharpe Ratio, maximum drawdown, CVaR and a plain explanation of what tipped the balance — in seconds.
Example of an accepted file
The "date" column is optional: only "return" is read.
What the verdict does not say
TrueVerdikt assesses the statistical robustness of a return series, not whether a trading idea makes sense or how it would behave in a market regime it has never seen. A GO verdict lowers the odds of fooling yourself; it guarantees no future result.