Seasonality: Month-End and Quarter-End Effects

Calendar-driven flows create repeatable, measurable moves because institutional behavior is scheduled. Month-end and quarter-end effects are among the most documented — and the most over-traded. The edge exists, but only inside narrow windows and with realistic expectations.

The mechanics

• Window dressing: funds buy strong performers and sell weak ones in the final days of a period so month-end statements look better. This pushes winners up and losers down into the close.
• Rebalancing: target-date and balanced funds realign to fixed weights at month/quarter end, generating predictable buy/sell pressure in large caps and bonds.
• Quarter-end is stronger than month-end because more funds report quarterly and rebalance on that cadence.

The tradeable edge

• Last trading day: SPY has historically shown a modest positive bias into the close, driven by rebalancing inflows. The effect is small — roughly 0.1–0.3% average — and not tradeable as a standalone bet after costs.
• First 2–3 days of the month: positive seasonality ("turn of the month") across equity indices. The combined last-day + first-days window is where most of the documented edge sits.
• Bond yields: tend to drift lower into month-end as funds buy bonds for duration rebalancing, then stabilize.

Rules for trading it

• Trade the window, not the day. Buy index exposure 1 day before month-end, hold 3 days into the new month. Single-day bets are noise.
• Size small. The edge is 0.2–0.5% per window. Risk 0.25–0.5% — this is a frequency strategy, not a conviction trade.
• Filter by trend. The turn-of-month effect is stronger when the index is above its 200-day SMA. In downtrends, the effect weakens or inverts.
• Avoid September quarter-end specifically if you are trading long equity — it historically has the weakest seasonal tailwind due to mutual-fund fiscal-year-end tax-loss selling.

What not to do

• Do not generalize "stocks go up at month-end" into an unlimited rule. The effect is statistical, not deterministic, and has decayed as more participants trade it. Test on the last 10 years, not the last 30, because the older data overstates current edge.
• Do not pair this with leverage. A 0.3% average edge with high variance is untradeable at 3x leverage after financing costs.

Seasonality is a supporting overlay, not a standalone system. Use it to time entries within a broader trend strategy, not as the strategy itself.