Seasonality asks whether a calendar window has behaved differently from ordinary time. The answer needs more than an average return.
A monthly study can mean the full calendar month, the next 21 trading sessions, or the remaining sessions from today's date. Those are different samples. State the anchor, horizon, asset history, and treatment of partial years before comparing results.
Average and median returns can hide a typical mid-month drawdown or a result driven by a few exceptional years. Individual paths show when dispersion opens, whether gains arrive early or late, and whether the current year is tracking a familiar route. The median intramonth dip is often as useful as the final return.
An ETF with thirty years of daily prices supports a different claim from a recently launched fund or token. For short histories, the correct output is a boundary: fewer observations, weaker confidence, and no invented extension from an unrelated proxy unless that substitution is explicit.
Treat seasonality as a prior, then combine it with current trend, volatility, and event risk. Compare the selected month with all months for the same ticker and check whether the result survives the removal of one or two extreme years.