Year-End Window Dressing by Funds

Mutual funds and pension funds disclose their holdings periodically, usually quarterly, and managers know that whatever appears in that snapshot is what clients, consultants, and journalists will judge them on. Window dressing is the practice of shuffling the portfolio right before the reporting date: sell the stocks that fell hard during the period (so the client never sees "we owned the loser"), and buy the stocks that just ran up (so the client sees "we owned the winner"), even if those trades reverse shortly after the report is filed. The disclosed portfolio ends up looking more skillful than the actual portfolio held through the period.

This creates a predictable, if modest, pattern around quarter-ends and especially year-end: recent big winners get a bid from managers piling in for the snapshot, and recent losers get extra selling pressure from managers exiting before the snapshot, then often bounce back once the reporting date passes and the "dressing" trades unwind. The effect is strongest in categories with the most reputational scrutiny, such as pension and mutual funds reporting to retail-facing intermediaries, and weaker among funds not required to disclose holdings publicly.

For a researcher, window dressing matters less as a standalone tradeable signal (the excess buying pressure is often already anticipated and is small relative to trading costs) and more as a confound: if you build a factor using end-of-quarter fund holdings data, some of what looks like "informed fund positioning" is actually cosmetic repositioning that reverses days later, and treating it as a genuine signal about manager skill would be a mistake.