is september historically a bad month for stocks

Overview

As of January 15, 2026, according to major market analyses, the question "is september historically a bad month for stocks" remains a frequently asked seasonal-investing query. In plain terms: the "September Effect" refers to the observation that many U.S. equity indices have shown weaker average returns in September than in other months. This article explains the term, summarizes long-run evidence (including S&P 500 figures), surveys proposed explanations, and highlights practical implications for different types of investors.

What you'll learn: whether "is september historically a bad month for stocks" is supported by data, why it might occur, which statistical caveats matter most, and how investors can respond without overreacting to calendar noise.

Definition and terminology

The phrase "September Effect" denotes a calendar-seasonality claim: that returns for equities—typically measured on major U.S. indices such as the S&P 500—tend to be poorer in the month of September than in other months. When readers ask "is september historically a bad month for stocks," they are usually asking about two related measures:

• Average monthly returns (arithmetic mean) and medians for September versus other months; and
• Frequency or probability that September posts a positive versus negative return.

Related terminology:

• Seasonality / calendar anomalies: patterns in returns that appear tied to calendar dates (months, days of week, or trading days). The September Effect is one example, alongside "Sell in May" and the January Effect.
• Price return vs total return: price return ignores dividends; total return includes reinvested dividends. Many studies use total return series when assessing long-term averages.
• Index scope: much of the literature focuses on U.S. broad-market indices (S&P 500, Dow Jones Industrial Average). International evidence is mixed.

When evaluating "is september historically a bad month for stocks," readers should note which series (price or total return), date range, and index are being analyzed—these choices materially affect conclusions.

Historical evidence

Researchers and financial media commonly cite long-run datasets (for example, the S&P 500 series dating back to the 1920s) when assessing whether "is september historically a bad month for stocks." Broadly, the headline findings in many public analyses are:

• Long-run arithmetic average for September returns on the S&P 500 is often slightly negative (commonly reported near −0.7% to −0.85%).
• The frequency of positive Septembers typically falls below 50% in many multi-decade samples, but estimates vary by sample period and index (commonly reported win rates in the ~44%–52% range depending on period and metric).
• Median September returns are often closer to zero, indicating that a small number of large negative Septembers can pull down the mean.

Below we summarize the evidence in more detail and by granularity.

Long‑term averages and medians

Analyses that use long samples (S&P 500 data from the 1920s onward) typically report a negative arithmetic mean for September, often cited roughly between −0.7% and −0.85% per month. However, medians for September are frequently nearer to zero, which shows the mean is influenced by large negative outliers. In practical terms, when you ask "is september historically a bad month for stocks," part of the answer is that the average result looks mildly negative, but the typical (median) September is far less extreme.

Important methodological notes that affect the numbers:

• Total-return series (including dividends) reduce the magnitude of negative monthly averages compared with price-only series.
• The start and end dates of the sample matter: including periods with large crises (Great Depression, 2008) increases negative bias.
• Arithmetic versus geometric averaging changes the reported long-run result; most public articles use simple arithmetic monthly averages for month-to-month comparison.

Frequency of positive/negative Septembers

When asking "is september historically a bad month for stocks," many readers expect a simple frequency answer: how often is September down? Common findings:

• In long samples for the S&P 500, September has been positive in roughly 44% to 52% of calendar-year Septembers, with precise estimates varying by period and data source.
• Compared with other months, September often ranks among the weaker months by frequency of positive returns, but it is not uniformly the worst across every subperiod.

These frequencies reflect the same point as the averages: September is more often negative than some months, but the difference from other months is modest and time-period dependent.

Weekly and daily patterns

Some research drills down to weekly and daily seasonality. Findings referenced in industry summaries include:

• Week-level patterns: analyses (including wealth-management blogs) have noted that the calendar week numbered around week 38 (which falls in September) has historically been weak in U.S. equities.
• Day-level patterns: certain calendar dates within September have shown outsized negative returns in some samples; for example, September 26 is sometimes flagged among weaker days, while October 19 historically ranks as one of the single worst trading days (a reminder that large negative days can cluster around crisis months, including September and October in some years).

These intra-month details are statistical curiosities that may or may not persist going forward.

Notable Septembers and outliers

Major historical negative Septembers often coincide with systemic crises rather than calendar-driven behavior. Examples commonly cited when addressing "is september historically a bad month for stocks" include:

• September 1931: deep bear-market dynamics during the Great Depression drove extreme monthly losses.
• September 2008: the global financial crisis produced very large negative returns in September 2008 as the banking system and credit markets seized up.
• September 2022: equity drawdowns occurred amid inflation, rate-hike cycles, and geopolitical friction.

There have also been strong Septembers (e.g., rebound months after sharp sell-offs), which illustrates that the month contains both severe negative and positive outcomes depending on macro context.

Theories and possible explanations

Multiple hypotheses attempt to explain why September sometimes underperforms. These remain hypotheses—none is a universally accepted causal mechanism that fully answers "is september historically a bad month for stocks." Below are prominent theories discussed in academic and industry writing.

Institutional portfolio rebalancing and fiscal‑year effects

Some institutional investors and funds have reporting or fiscal-year conventions that create trading pressure around certain calendar windows. Rebalancing at quarter- or fiscal-year turns (and the timing of mutual-fund flows and reporting) can increase selling pressure in late summer/early autumn for some institutions, which could contribute to September weakness in aggregate.

Tax‑loss harvesting and year‑end positioning

While tax-loss harvesting is most active near year-end, preparatory portfolio moves and strategy positioning in autumn (especially as managers reassess tax and performance positions) have been suggested as partial drivers of seasonal patterns. This explanation helps for some years but does not fully explain persistent September underperformance in multi-decade averages.

Post‑summer reappraisal and liquidity effects

Lower trading volumes during summer months can conceal problems or inflate valuations. When market participants return in September, reassessments of risk and refreshed trading activity can lead to outsized moves and higher volatility. In that sense, a "reality check" after summer may be one reason some negative outcomes congregate in September.

Behavioral and psychological factors

Market participants aware of the "September Effect" may act preemptively—selling or hedging because they expect weakness. This self‑fulfilling dynamic and confirmation bias can amplify price moves that follow the calendar.

Macro and event clustering

Finally, it is important to note that major macro shocks, geopolitical events, or credit shocks sometimes happen to occur in September (or close to it). Those events, rather than the month itself, can generate the large negative returns that make September look weak in long‑run averages.

Empirical studies and notable analyses

A variety of industry and research pieces have examined the September Effect. As of January 15, 2026, several widely referenced pieces and their general takeaways include the following:

Industry articles and data analyses

• Wealthfront: long-run analysis using S&P series often cited in articles shows a negative average September and identifies weak weeks/days such as week 38 and some late‑September dates. As of Jan 15, 2026, Wealthfront and other advisory blogs continue to show a modestly negative average September when analyzing data since the 1920s.

• Investopedia: provides a definitional overview and cautions that results depend on period choice and data type. Media entries on Investopedia stress that historical correlation does not imply a causal relationship for "is september historically a bad month for stocks."

• MarketWatch & Money.com: journalistic coverage summarizes the statistical patterns and interviews market commentators who emphasize variability and context. Reports commonly note that headline September weakness is often concentrated in crisis years.

• Fisher Investments: commentary from asset managers often frames the effect as a statistical curiosity and stresses that the signal is noisy—advisors caution against mechanical timing based on month of the year.

• Morningstar / SentimenTrader: some short-term seasonal strategies have been observed in their data (for example, patterns involving the first three trading days of September versus the rest of the month). These short-term rules are sometimes proposed as tactical trades but are sensitive to data-snooping and trading frictions.

Robustness and sample‑selection issues

Key empirical caveats highlighted across studies:

• Index selection changes results: small‑cap indices, international indices, and S&P 500 may show different seasonal profiles.
• Total return vs price return alters magnitudes: including dividends typically makes long-run monthly averages less negative.
• Start and end dates matter: including large crisis months in the sample (Great Depression, 2008) biases the arithmetic mean downward; excluding them changes the headline.
• Survivorship and data-snooping biases can overstate the reliability of reoccurring patterns.

Taken together, empirical studies show a consistent but modest negative bias for September in many U.S. samples, while emphasizing that the signal is neither stable nor obviously exploitable without risk.

Criticisms, limitations, and statistical caveats

When exploring whether "is september historically a bad month for stocks," several methodological and economic criticisms are relevant.

Correlation vs. causation

Calendar correlations—even persistent ones—do not establish a causal mechanism tied to the month. Many critics point out that absent a clear economic channel, a calendar correlation is unreliable as a trading signal.

Outliers and skewness

A handful of extreme negative Septembers (e.g., 1931, 2008) exert outsized influence on arithmetic averages. The median is often closer to zero, which suggests skewness matters when interpreting the data.

Time‑period dependence and potential disappearance

If market participants attempt to trade on a calendar anomaly, arbitrage and changing market structure can erode predictable patterns. Historical patterns can therefore diminish over time.

Transaction costs, taxes, and implementation friction

Even if a backtest shows outperformance from avoiding September, real-world costs—transaction fees, market impact, taxes, and the opportunity cost of sitting out positive months—can nullify theoretical gains. This practical limitation is frequently emphasized by professional investors answering "is september historically a bad month for stocks."

Practical implications for investors and traders

How should individuals respond to the evidence on September seasonality? The pragmatic guidance below reflects mainstream advisory practice and the literature's caveats.

Long‑term investors

For most long-term buy-and-hold investors, responding to the calendar alone (for example, selling at the end of August to avoid September) is typically discouraged. The main reasons:

• The historical September weakness is modest relative to long-term expected returns.
• Missing a small number of positive months (including positive Septembers) can materially reduce long-run portfolio performance.
• Rebalancing to strategic allocations and maintaining diversification are higher‑value actions than calendar timing.

If you manage crypto or digital-asset exposure alongside equities, consider that liquidity and volatility characteristics differ—use risk management primitives (size limits, stop-loss frameworks, and strategic rebalancing) rather than calendar trades.

Short‑term traders and seasonal strategies

Short-term traders sometimes test tactical rules tied to September (for example, the idea of exiting after the first three trading days if they are negative). These approaches face material challenges:

• Backtest overfit and data-snooping biases.
• Execution slippage and fees can erase small edge estimates.
• Tax consequences from frequent trading.

Short-term seasonal tactics may work episodically but require rigorous testing, robust risk management, and an acceptance that patterns can disappear.

Risk management and portfolio review

A practical, non‑speculative use of the September discussion is to treat the period as a prompt for portfolio hygiene:

• Reassess risk exposures and whether your asset mix still aligns with goals.
• Rebalance to target allocations if drift has occurred.
• Review liquidity needs and ensure margin, leverage, or concentrated positions are appropriate for your time horizon.

These actions improve preparedness without relying on the calendar as a predictive tool.

Bitget note: for investors who trade digital assets or manage hybrid crypto-equity portfolios, consider using reliable platforms and wallets. Bitget exchange offers spot and derivatives functionality and Bitget Wallet provides a way to hold on-chain assets securely while you implement risk controls and portfolio reviews.

International evidence and cross‑market comparisons

When broadening the question "is september historically a bad month for stocks" beyond U.S. markets, the evidence is mixed:

• Some international indices show similar modest September weakness in some samples, but not uniformly.
• Emerging markets and smaller-cap indices can have different seasonal profiles due to idiosyncratic flows and local institutional behaviors.
• Other asset classes (bonds, commodities) do not consistently mirror the September pattern seen in U.S. equities.

Therefore, while the September question is most commonly associated with U.S. equities, cross-market comparisons show heterogeneity and weaken the case for a universal calendar effect.

Related calendar effects

The September Effect is one among several commonly-discussed calendar anomalies:

• Sell in May ("Sell in May and go away"): a seasonal strategy that posits summer months underperform; mixed empirical support and strong sensitivity to sample and region.
• January Effect: historically, small-cap stocks have sometimes outperformed in January, linked to tax and window-dressing effects; the effect has attenuated over time.
• October Effect: October is sometimes highlighted due to several historical crashes occurring in October; however, October also contains large positive rebounds in some years.

Comparing these anomalies shows that calendar effects can appear and then weaken over time as market structure and participant behavior evolve.

Summary and consensus view

To answer the question directly: "is september historically a bad month for stocks?" — the historical record for broad U.S. indices often shows a modestly negative average for September and a slightly below‑average frequency of positive months. However, the signal is:

• Sensitive to dataset choices (index, total vs price return) and sample period;
• Driven in part by a few extreme negative months associated with crises;
• Not a reliably causal or predictive factor for most investors; and
• Poorly suited to mechanical, transaction-cost-sensitive timing strategies.

Most financial advisors therefore recommend treating September-seasonality observations as an informational curiosity rather than a trading rule. Use the month as a reminder to review risk and rebalance rather than as a calendar trigger to time markets.

See also

• Calendar anomalies
• Market seasonality
• Sell in May
• Behavioral finance
• Tax-loss harvesting

References

As of January 15, 2026 these public sources and commonly used datasets are frequently cited in discussions of the September Effect and were used as background for this article:

• Wealthfront — analysis of monthly and weekly seasonality (coverage of long-run S&P series; industry blog analysis).
• Investopedia — "September Effect" overview and methodological cautions.
• MarketWatch — journalistic summaries and market commentary.
• Money.com — consumer-facing explanation and context.
• Fisher Investments — commentary critiquing overreliance on calendar signals.
• Morningstar / SentimenTrader — short-term pattern observations (first trading days vs the remainder of September).
• Long-run market datasets commonly referenced: S&P 500 historical series, Kenneth R. French data library, Dow Jones historical data.

Notes on data and methodology: where numeric claims are cited above (for example, average September returns around −0.7% to −0.85% or positive‑month frequency ranges of ~44%–52%), these reflect commonly reported figures from public analyses of the S&P 500 series across samples dating back to the 1920s. Exact numbers vary with return definition (price vs total return), sample start/end, and whether outlier months are included.