Join waitlist

Stats › Currencies and commodities › Commodity seasonality

Commodity month-of-year seasonality does not survive an out-of-sample test

Best seasonal variant: 0.245+0.245Best seasonalvariantEqual weight benchmark: 0.183+0.183Equal weightbenchmarkClassical SSA signal: 0.014+0.014Classical SSAsignalRobust L1 SSA, vol normalised: -0.365-0.365Robust L1 SSA,vol normalised

Out-of-sample Sharpe ratios, fifteen commodity futures, January 2016 to December 2024, 108 months, costs at 8.6 basis points one way. Across 500 bootstrap paths with within-family Holm correction, none of the eighteen paired Sharpe tests rejects, with adjusted p values from 0.192 to 1.000.

Seasonal tables for corn, gas and gold circulate every year. The one peer-reviewed positive result is marginal by its own authors' description, and a 2026 test on 2016 to 2024 across fifteen futures rejects nothing after multiple-testing correction.

t = 1.93The one peer-reviewed positive result, called marginally significant by its own authors
NoneOf 18 paired Sharpe tests that reject after Holm correction, 2016 to 2024
0.5 ppBank of England estimate of seasonality's share of expected heating oil returns
0.245Best seasonal Sharpe out of sample, against 0.183 for equal weight

Members only

The rest of this page is for members

Below this point there are 5 sections, 1 table and 5 named sources, roughly 500 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.