chainmeter
EN | 中文 | ES

Article

Bitcoin Seasonality: Is There Really a 'Best Month' to Buy?

Chainmeter · ~2 min read

TL;DR

Every month of Bitcoin's trading history, colored by return, reveals some months with historically stronger average performance than others — but with roughly a decade and a half of data, each calendar month has only 13–15 independent observations, far too few to statistically distinguish real seasonality from random noise.

What the heatmap shows

This site's Monthly Returns page plots every calendar month of Bitcoin's trading history, colored by that month's percentage return, letting you scan for patterns like "September has historically been weak" or "Q4 has historically been strong" — claims that circulate widely in crypto commentary.

Why the sample size problem is severe here specifically

Bitcoin has meaningful trading history going back to around 2010–2011, meaning any single calendar month (say, every September) has occurred only 13–15 times. That's an extremely small sample for drawing a statistically confident conclusion about a genuine recurring seasonal effect versus random variation that happens to look patterned in hindsight. Traditional financial markets with a century-plus of data face this same seasonality debate with far more observations per month, and the debate remains unsettled there too.

Why check it yourself instead of trusting a claim

The heatmap here is computed live from the full available price history; nobody curated it to make a particular argument. That means you can verify any specific seasonality claim ("Bitcoin always dips in month X") directly against real historical returns instead of taking a repeated internet claim at face value.

The honest conclusion

Some months do show higher average historical returns than others in the available data. Whether that reflects a real, persistent seasonal effect or is largely explainable by a handful of outsized months (a single strong or weak month can dominate an average calculated from only 13–15 data points) is genuinely not resolvable with the amount of data that exists.

Scan every month of Bitcoin's trading history colored by real return, and check any seasonality claim yourself, on the Monthly Returns Heatmap.

FAQ

Has any month never had a negative return?
No calendar month in Bitcoin's history has been positive in every single occurrence — every month has had at least one significantly negative year within the available sample.
Why not just trust the month with the best historical average?
With only 13–15 observations per month, a single unusually large gain or loss can swing that month's average substantially, making the 'best average' month more sensitive to a few outlier years than to a genuine recurring pattern.
Do other assets show more reliable seasonality?
Some traditional markets with over a century of data still debate the reliability of seasonal effects like 'sell in May.' If a century of data doesn't fully settle the question there, a decade and a half for Bitcoin is a meaningfully thinner basis for a confident claim.