What is the difference between the probabilistic and the deflated Sharpe ratio?

The PSR asks whether one strategy's true Sharpe ratio beats a benchmark you choose, often zero, judged on its own. The deflated Sharpe ratio (Bailey & López de Prado, 2014) is the same calculation with the benchmark raised to the Sharpe ratio that the best of all the trials run to find the strategy would be expected to reach by luck alone. The more variations were tried, the higher that benchmark, so the DSR corrects for selection as well as for sample length and non-normal returns.

How long a track record is needed to trust a Sharpe ratio?

The minimum track record length answers this. With independent, normally distributed daily returns, an annualised Sharpe ratio of 1.0 needs about 684 trading days, roughly 2.7 years, to be above zero at 95% confidence. An annualised 2.0 needs about 173 days; an annualised 0.5 needs about 2,730 days, nearly 11 years. Negative skewness, fat tails, a higher benchmark or a wide search all lengthen the wait.

Why does skewness change the PSR?

Because the uncertainty of a Sharpe ratio depends on the shape of the returns, not only on their mean and volatility. For a positive Sharpe ratio, negative skewness and fat tails widen that uncertainty. A strategy that earns steady small gains and occasionally loses heavily, such as selling options, therefore needs a longer record to reach the same PSR as one with symmetric returns and the same Sharpe ratio.

Does Opulence Alpha's Promotion Gate decide on the PSR?

No. Each validation report records the PSR against zero, but no gate check reads it, because a PSR ignores the search that found the strategy. The gate's Sharpe check is the deflated Sharpe ratio measured out of sample, and the number of trials behind that deflation is counted from the search's own trial registry, with no default. If the observed Sharpe ratio does not beat the deflated threshold, the report states that no amount of further data would make the result significant.

FAQ · Probabilistic Sharpe ratio

What is the probabilistic Sharpe ratio (PSR)?

Reviewed against the platform's code on Sep 26, 2026

The probabilistic Sharpe ratio (PSR) is the probability that a strategy's true Sharpe ratio is above a chosen benchmark, often zero, given the Sharpe ratio it has shown. It accounts for how many returns were observed and for their skewness and fat tails, so a short or lopsided track record earns less confidence than a long, well-behaved one.

Why it matters

A Sharpe ratio is an estimate, and two equal estimates can carry very different evidence. With independent, normally distributed daily returns, an annualised Sharpe ratio of 1.0 earned over six months gives a PSR against zero of about 0.76; earned over three years, about 0.96. For a positive Sharpe ratio, negative skewness and fat tails widen the uncertainty further, because a few extreme days can dominate the estimate. The Sharpe ratio alone hides all of this. The PSR puts it on one scale, a probability, that can be compared across strategies and track lengths.

How it works

Bailey and López de Prado (2012) start from the fact that an estimated Sharpe ratio is approximately normally distributed around the true one (Lo, 2002). PSR = Φ[(SR̂ − SR*)·√(T − 1) / √(1 − γ₃·SR̂ + ((γ₄ − 1)/4)·SR̂²)], where SR̂ is the observed Sharpe ratio at the frequency of the returns (not annualised), T the number of returns, γ₃ their skewness, γ₄ their kurtosis (3 for normal returns), SR* the benchmark and Φ the standard normal cumulative distribution function. Solving the same expression for T gives the minimum track record length (MinTRL): the number of returns needed for the PSR to reach a chosen confidence, such as 95%.

How Opulence Alpha applies it

In Opulence Alpha's Promotion Gate, each validation report computes the PSR against zero, the deflated threshold, the deflated Sharpe ratio and the minimum track record length from per-period returns, their measured skewness and kurtosis, and the search's trial count. The PSR is recorded, but no gate check reads it. The gate's Sharpe check is the deflated Sharpe ratio: the PSR with its benchmark raised to the Sharpe ratio the best of the search's trials would be expected to reach by luck alone. It must be measured out of sample; an in-sample value cannot stand in. The default bar is 0.95. The MinTRL is measured against the deflated threshold at 95% confidence and is descriptive, not a bar to clear.

PSR and DSR in the Promotion Gate →

Related concepts

Deflated Sharpe ratioThe deflated Sharpe ratio adjusts a strategy's Sharpe ratio for the number of strategies tried and for non-normal returns, giving the probability that its true Sharpe ratio is above zero.Promotion GateA Promotion Gate is the validation step a strategy must pass before any portfolio may use it: overfitting, data-snooping and robustness checks, where a check that could not run counts as a failure.Multiple testingMultiple testing (data snooping) is what happens when many strategies are tested on the same data: some will look significant by chance. How reality checks and step-down tests correct for it.Backtest overfittingBacktest overfitting is when a strategy is tuned to the noise of past data, so its backtest looks strong but its live performance decays. How to detect and avoid it.

All concepts →Sector behaviour in every market regime →

Questions

What is the difference between the probabilistic and the deflated Sharpe ratio?: The PSR asks whether one strategy's true Sharpe ratio beats a benchmark you choose, often zero, judged on its own. The deflated Sharpe ratio (Bailey & López de Prado, 2014) is the same calculation with the benchmark raised to the Sharpe ratio that the best of all the trials run to find the strategy would be expected to reach by luck alone. The more variations were tried, the higher that benchmark, so the DSR corrects for selection as well as for sample length and non-normal returns.
How long a track record is needed to trust a Sharpe ratio?: The minimum track record length answers this. With independent, normally distributed daily returns, an annualised Sharpe ratio of 1.0 needs about 684 trading days, roughly 2.7 years, to be above zero at 95% confidence. An annualised 2.0 needs about 173 days; an annualised 0.5 needs about 2,730 days, nearly 11 years. Negative skewness, fat tails, a higher benchmark or a wide search all lengthen the wait.
Why does skewness change the PSR?: Because the uncertainty of a Sharpe ratio depends on the shape of the returns, not only on their mean and volatility. For a positive Sharpe ratio, negative skewness and fat tails widen that uncertainty. A strategy that earns steady small gains and occasionally loses heavily, such as selling options, therefore needs a longer record to reach the same PSR as one with symmetric returns and the same Sharpe ratio.
Does Opulence Alpha's Promotion Gate decide on the PSR?: No. Each validation report records the PSR against zero, but no gate check reads it, because a PSR ignores the search that found the strategy. The gate's Sharpe check is the deflated Sharpe ratio measured out of sample, and the number of trials behind that deflation is counted from the search's own trial registry, with no default. If the observed Sharpe ratio does not beat the deflated threshold, the report states that no amount of further data would make the result significant.

References

Bailey, D. H. & López de Prado, M. (2012). The Sharpe Ratio Efficient Frontier. Journal of Risk, 15(2), 3–44.
Bailey, D. H. & López de Prado, M. (2014). The Deflated Sharpe Ratio: Correcting for Selection Bias, Backtest Overfitting and Non-Normality. Journal of Portfolio Management, 40(5), 94–107.
Lo, A. W. (2002). The Statistics of Sharpe Ratios. Financial Analysts Journal, 58(4), 36–52.

Educational content about research methods. Not investment advice.