The September effect is the observation that September has historically been the weakest calendar month for U.S. equities on average. It's one of the more persistent seasonal anomalies — real in the long-run data — but, as always, it's background context rather than a signal you can trade mechanically.
What the data shows
Across many decades, September's average return has been the lowest of the twelve months, and more often negative than most. The effect is a statistical average: plenty of individual Septembers have been positive, and the “weakness” is a modest tilt over a long history, not a reliable annual decline. It shows up robustly enough across markets and eras to be more than coincidence, which is why it gets attention every fall.
The theories
Proposed causes include post-summer repositioning as institutions return and de-risk, tax-related selling in some fund structures, and the tail end of the weaker summer half-year. None is definitive, and some of the effect may simply be the kind of pattern that appears in any long dataset. The honest stance: it's a documented tendency with plausible but unproven drivers.
September's weakness is a decades-long average, not a forecast. The market doesn't owe you a down month because the calendar flipped — and it often doesn't deliver one.
What it means for scalping
For an intraday trader, September's seasonal tilt barely touches your decisions — you're trading the live structure, not holding for the month. At most it's a reminder to respect that autumn can bring elevated volatility as participants return from summer. Don't carry a standing bearish bias into every September session; read what the tape actually does. Seasonality is context you weigh, never a thesis that overrides the map in front of you.