Testing Market Seasonality with Monte Carlo Permutations in Excel

Summary

The article demonstrates a spreadsheet-based permutation test for assessing whether an observed market pattern could arise by chance. Its example examines whether Bitcoin returns are unusually high on Tuesdays: daily returns are randomly shuffled, grouped by weekday, and summarized across repeated realizations. The observed Tuesday average is then compared with the distribution of randomized averages using a percentile rank and an empirical cumulative distribution plot.

The example places the observed Tuesday return above most randomized results, while a t-test provides weaker evidence; the author ultimately treats the finding as inconclusive given its uneven presence across years and lack of a clear market mechanism. The method is accessible without specialized statistical software, but its conclusions depend on sensible hypothesis framing and limited data mining. Shuffling also assumes independent, identically distributed observations, so autocorrelation can undermine the test. Evidence that an effect is unusual does not establish that it will persist or cover trading costs.

Key ideas

• A permutation test compares an observed statistic with statistics from randomized versions of the data.
• Shuffling returns can preserve their aggregate values while removing weekday assignments.
• The Bitcoin Tuesday example shows suggestive but inconclusive evidence of seasonality.
• The test depends on appropriate assumptions, including independence that autocorrelation may violate.
• Statistical unusualness alone does not show that an effect is tradable or durable.

Tags

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