Abstract:
Research by de Prado, Lipton and Zoonekynd that reviews the pitfalls of naive Sharpe ratio analysis and provides a comprehensive framework for its proper use:
Five major pitfalls are identified:
(a) Assumption of normality in return distributions.
(b) Neglect of statistical significance and minimum sample length requirements.
(c) Insufficient statistical power of standard tests.
(d) Misinterpretation of classical p-values as the probability of the null hypothesis given the data.
(e) Failure to adjust for multiple testing.
To address these issues, the paper surveys and extends advanced methods, including:
Probabilistic Sharpe Ratio (PSR)
Minimum Track Record Length (MinTRL)
Observed Bayesian Tail-Area False Discovery Rate of the Sharpe ratio (oFDR)
Deflated Sharpe Ratio (DSR)
Monte Carlo simulations demonstrate that these corrections outperform traditional t-statistics and standard multiple-testing procedures.
The study distinguishes between Familywise Error Rate (FWER), False Discovery Rate (FDR), and hybrid FWER-FDR approaches, highlighting their respective relevance for academic versus industry settings.
Resources:
CFA Society of the UK3rd floor, Boston House,63-64 New Broad Street, London EC2M 1JJ
About Us
Terms of Use
Cookie Policy
Privacy Policy