Twelve-month return cycle
SeasonalityAdvanced
A twelve-month cycle in the cross-section says a stock's return in a given calendar month is related to its own return in that same month in prior years. It is not the January effect, which is a market-wide small-cap pattern, and it is not twelve-month momentum, which ranks recent performance. The sort is on the same calendar month across years.
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
- Each month, rank stocks on their average return in that calendar month over prior years, skipping the most recent year if the specification says so. Buy the historically strong names for that month and short the historically weak ones. Re-sort next month. Require a long history before a stock is eligible so one lucky February does not define it.
- Where AI fits
- AI can build the same-month average without using future years and can keep this book distinct from a trailing-twelve-month momentum book. It should show the overlap of the two portfolios. Eligibility rules, such as years of history, have to be fixed first.
- Markets
- Individual stocks with long histories. Survivorship bias is the main data risk.
- Timeframes
- Monthly, using a seasonal lookback of several years.
- Risks
- The same-month average is noisy. Delisted stocks disappear from naive databases and take their bad months with them. The pattern can be a restatement of other seasonal effects. Turnover and borrow costs are easy to leave out of a rank table.