Bitcoin overnight seasonality

Bitcoin trades all week, but returns have not been uniform across the hour of the day in historical samples. Research write-ups of overnight or intraday seasonality test whether a fixed clock-time window carries the return. A rule that exploits it is long only during the stated hours and flat otherwise. The clock is the signal. There is no chart pattern.

Not investment advice. Nothing here is a recommendation to buy or sell anything. Backtests and historical studies are not guarantees, and a rule that worked in one sample can fail in the next. Trading can lose money. Links are for education, not endorsements. Do your own research.

Rules

Convert timestamps to one time zone and one exchange calendar. Hold bitcoin only during the precommitted window, such as a defined overnight segment, and stay flat outside it. Use the same hours every day, including weekends if the study includes them. Subtract fees on every round trip. Do not shift the window after seeing which hours were green.

Where AI fits

AI can aggregate trades into the stated timezone, handle daylight-saving mistakes, and net fees. It can test the window on a later sample that was not used to pick the hours. If the hours were chosen by scanning the whole history, the model should label the result as in-sample, not as a strategy.

Markets

Bitcoin on one liquid venue, or a futures contract with an explicit roll. Mixing venues mixes different clocks and fees.

Timeframes

Intraday clock time. The position is flat most of the day.

Risks

A clock effect is easy to mine because there are many possible windows. Fees and funding can exceed it. Liquidity at the chosen hour may be the reason the return existed and the reason you cannot have it. Exchange downtime clusters in ugly ways. Past hour-of-day returns are not a schedule you can trust.

← Back to all strategies