Stats › Backtesting and evidence › Strategy decay

Published predictors earn 58% less after publication

In sample: 0.582%+0.582%In sampleOut of sample, before publication: 0.402%+0.402%Out of sample, before publicationAfter publication: 0.264%+0.264%After publication

Source: McLean and Pontiff, Does Academic Research Destroy Stock Return Predictability, Journal of Finance 71(1). These are the raw period means from their Table I, which imply falls of -30.9% and -54.6%. Their headline figures of 26% and 58% come from regression coefficients of -0.150% and -0.337% measured against the in sample mean, which is a different and slightly gentler calculation.

Once a return predictor is published, anyone can read the paper and trade it. Across 97 predictors, returns fell 26% out of sample and 58% after publication. About a third of that loss is other people acting on the paper.

97Predictors studied
26%Out of sample decline
58%After publication
median 73%Live Sharpe deterioration

Members only

The rest of this page is for members

Below this point there are 6 sections, 1 chart, 1 table and 5 named sources, roughly 1300 words of it. Every figure carries the source it came from and the date the data is from.

You can keep browsing every statistic in the library for free. The intro and the summary are always open.