Testing Whether a Trading Strategy’s Edge Is Decaying

Summary

The document asks how to determine whether a strategy’s performance edge has weakened across time periods. It mentions comparing period means with a t-test as an approach encountered elsewhere, while expressing concern that this test may be unreliable for many lower-frequency strategies.

No alternative method, data, or empirical result is supplied. The central learning point is the evaluation problem itself: evidence of changing returns must be assessed with attention to limited observations and the sampling properties of the strategy. The question explicitly asks whether methods differ in robustness to sample size, but it does not define the strategy, return distribution, period lengths, or dependence structure. Those omissions prevent a specific test from being recommended on the basis of this document alone.

Key ideas

• The document frames strategy alpha decay as a comparison of performance across periods.
• It identifies a t-test of mean returns as one possible approach.
• It questions whether mean comparisons are reliable for lower-frequency strategies.
• It asks how sample size affects the robustness of tests but gives no answer or evidence.

Tags

# Calculating performance decay of a strategy


# Calculating performance decay of a strategy












Came across this thread which basically advises t-test for difference in means across periods to calculate whether our edge is deteriorating. The catch being that this test will probably not be accurate for most lower frequency strategies.

How to compute the alpha decay of a strategy?

What other methods can we use to test for loss of edge btwn periods? Are some more robust to sample size than others?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.