September Effect in Stock Returns
The September effect is a calendar anomaly in which stock markets historically post their weakest returns in September compared to any other month. Academic research traces this to tax-loss harvesting by institutional investors, post-summer portfolio rebalancing, and reduced investor risk appetite after the summer break. While less pronounced in modern ETF-dominated markets, the pattern persists.
The Historical Pattern
Data from the S&P 500 Index since 1950 show September to be the only month with a negative average return. While other months average positive returns of roughly 1.0% to 2.0% annually, September has returned approximately −1.0% on average, with wide swings: some years it is down 3–5%; others it is positive. But the long-run median is decisively negative.
This is not a tiny effect. Over 70 years, a −1% average September return compounds to a meaningful drag on long-term wealth. If every other month returns +1.5%, but September returns −1%, the annual average drops below 13%, a material headwind compared to a hypothetical market that returned +1.5% every month.
The September effect is not the same as the broader “sell in May and go away” adage, which refers to underperformance during the May-to-October period overall. Rather, September stands out as the worst single month, even compared to the rest of summer and early autumn.
Why September Stands Out: Tax-Loss Harvesting
The primary culprit is tax-loss harvesting. In the United States, investors can deduct capital losses against capital gains (and up to $3,000 of other income) on their annual tax return. This creates an incentive to realize losses by year-end.
Consider a taxable portfolio manager who holds positions purchased in January through August. Some have appreciated; others have depreciated. By late August, she sees potential tax losses available. If she realizes those losses in September (ahead of the December 31 year-end cutoff), she can offset gains elsewhere or carry the loss forward. But under the IRS’s “wash-sale rule,” she cannot repurchase the same security for 30 days without disqualifying the loss.
This creates September selling pressure: institutions and high-net-worth individuals selling losers to harvest tax benefits. If a stock is down 15% year-to-date, both the fund manager and a taxable investor have incentive to sell in late August or early September, realizing the loss and swapping into a similar (but not identical) security to maintain exposure. The selling is orderly but persistent.
The effect is magnified because September is the last full month before the year-end tax deadline. October and November see some late harvest selling, but September is when the bulk of decision-making occurs—accounts are reviewed, rebalancing schedules align with tax year-ends, and advisors model tax scenarios.
Post-Summer Repositioning and Risk Appetite
A secondary explanation involves portfolio rebalancing and the psychological reset that comes after the summer break.
Rebalancing windows: Many institutional funds rebalance at fixed intervals: quarterly, semi-annually, or after the summer. In late August and early September, portfolio managers review tactical allocations. If equities have outperformed bonds during the summer (as they often have), the fund rebalances by selling stocks and buying bonds. This mechanical rebalancing, multiplied across thousands of funds, creates September selling pressure.
Risk appetite and sentiment: The summer break often brings a philosophical reset. After extended bull runs, investors return from vacation to news flow (earnings revisions, macro data, geopolitical events) that may have accumulated. There is also a historical pattern: July and August are “vacation months,” with lower trading volume and a softer news agenda. When the market reopens in full force in early September, stale risk assessments are updated, and risk appetite resets downward.
Additionally, September includes the start of the fiscal year for many corporations and the school year for many households. There is an undercurrent of year-end deadline pressure—budget cycles tighten, earnings guidance for the final quarter becomes concrete, and macro uncertainty (inflation, interest rates) looms larger over the final quarter.
Institutional Calendar Effects
Pension funds, university endowments, and large family offices often rebalance in September. Their fiscal years may not align with the calendar year (many endowments run July–June), and their trustees meet in the fall to set annual allocations. This creates a demand-side reduction: less inflow of new capital, more rebalancing-driven selling.
Conversely, retail investors—who dominate trading in late December and early January for new-year savings accounts and resolutions—are less active in September. This imbalance tilts the flow from buyers to sellers.
Data and Magnitude
Academic research quantifies the September effect:
Bouman and Jacobsen (2002) analyzed 109 years of data across 19 countries and found September to be the weakest month in 15 of them, with the effect strongest in Anglo-American markets.
University of Pennsylvania studies estimated the average September underperformance at 1–2% per annum, with September also showing elevated volatility.
In the S&P 500, September daily returns have been negative 55% of the time since 1960, compared to roughly 48% across other months.
The effect is largest in the first half of September (the 11th–20th is often the weakest week) and fades as the month progresses. This timing aligns with the tax-harvesting window: late August through early September for decisions, early-to-mid-September for execution.
Modern Weakening and Passive Investing
The September effect has attenuated in recent decades, particularly since 2000. Several factors explain this:
Passive indexing: As investors shift to index funds and ETFs, mechanical tax-loss harvesting plays a smaller role. Index funds hold all constituents, so they harvest losses more systematically throughout the year rather than in a September rush.
Year-round tax planning: Brokers and robo-advisors now offer tax-loss harvesting throughout the year, not just in September, spreading the effect.
Shorter holding periods: Algorithmic and high-frequency trading firms don’t care about tax consequences—they trade in and out of positions on scales of seconds to days, unaffected by the calendar.
Globalization: U.S. tax rules apply only to U.S. taxable accounts. International investors (who are a growing fraction of U.S. stock ownership) don’t follow the U.S. tax calendar.
Despite these headwinds, the September effect remains detectable in aggregate data, though at a lower magnitude (perhaps 0.5–1.0% rather than 1.5–2.0%).
What It Doesn’t Explain
The September effect should not be confused with broader seasonality or calendar anomalies. It is not the same as:
The “January effect”: a tendency for stocks to rally in January, partly due to new-year cash inflows and the “Santa Claus rally” at year-end. January is typically strong, September weak.
The “Halloween indicator” (“Sell in May”): a broader pattern of summer underperformance. September is the nadir of that pattern, but the effect spans May–October.
Earnings seasonality: Certain industries are sensitive to seasonal demand (retail peaks before Christmas, agriculture after harvest). The September effect is market-wide and persists even when earnings are controlled.
Practical Implications
For passive investors, the September effect is a warning to maintain discipline. The temptation to panic-sell in a weak September is strong, but doing so locks in losses and often leads to buying back in after the rally resumes. A dollar-cost averaging approach or a rebalancing schedule removes the temptation to time September weakness.
For active managers, September is a known hunting ground. Contrarian investors sometimes buy dips in early September, betting that the tax-driven selling is overdone. However, trying to catch the exact bottom is treacherous—September weakness can be severe and prolonged.
For tax planning, the September effect underscores the importance of systematic tax-loss harvesting year-round, not a panicked September sprint. Spreading realized losses across months smooths execution and reduces the risk of triggering wash-sale penalties.
See also
Closely related
- Tax-loss harvesting — the mechanism driving September selling
- Wash-sale rule — the IRS constraint on tax-loss harvesting
- Capital gains tax (investor) — why tax effects matter to markets
- Asset allocation — rebalancing decisions that concentrate in September
- Historical volatility — September shows elevated realized and implied volatility
Wider context
- Index fund — vehicles that have dampened the September effect over time
- ETF — modern alternatives to active management that mitigate seasonal tax effects
- Behavioral finance — the psychology behind calendar anomalies
- Market timing — the failed strategy of trying to exploit seasonal patterns
- S&P 500 Index — the primary market where the September effect is observed