S&P 500 Average Monthly Return: What the 1928-2023 Data Shows

What Is the S&P 500 Average Monthly Return, Historically?

The S&P 500's average monthly return from 1928 through 2023 is approximately +0.74% on a price-return basis, or roughly +0.90% including dividends, according to Federal Reserve FRED database historical data. The standard deviation of those monthly returns sits near 5.5%, which means a single month swinging ±11% is statistically unremarkable. That context matters more than the average itself.

For a $5M portfolio, a one-standard-deviation down month is a $275,000 paper loss. Knowing that is normal, not alarming, is the difference between behavioral discipline and panic selling at the worst possible time.

The long-run annualized total return averages approximately 10% per year, as documented by Morningstar's 2024 U.S. Market Outlook, which translates to that ~0.8% mean monthly figure before inflation adjustment. See inflation-adjusted performance for what that number looks like after the CPI haircut.

S&P 500 Average Monthly Return by Month: The Actual Numbers

The table below reflects historical average price returns by calendar month, drawn from Federal Reserve FRED data covering 1928 through 2023. These are arithmetic means across all available years.

A few things stand out. April and November share the top spot at +1.5% average returns. September is the clear outlier at -1.1%, with a positive-month rate of only 44%. December has the highest win rate at 73%, though its average is pulled down by occasional sharp year-end drawdowns.

The critical caveat: month-to-month variance is so high that calendar month alone explains less than 2% of actual return variation in any given year. These averages describe historical tendencies, not reliable forecasts.

Which Month Has the Highest Average S&P 500 Return?

April and November are statistically tied at the top. Both average +1.5% in price return across the full historical dataset.

April's strength has several proposed explanations: tax refund reinvestment, strong Q1 earnings season, and institutional rebalancing after Q1 closes. None of these explanations has been definitively proven causal. The pattern exists; the mechanism is debated.

November's strength is better documented. Bouman and Jacobsen's peer-reviewed study in the Journal of Finance found statistically significant evidence of a November-through-April seasonal premium across 37 countries, lending academic credibility to what traders call the "Best Six Months" period. The November-April window has historically outperformed the May-October window by a meaningful margin over multi-decade periods.

December's 73% positive-month win rate is the highest of any month, even though its average return is slightly below April and November. That consistency reflects year-end institutional buying, window dressing, and the well-documented "Santa Claus rally" in the final trading days of the year.

For context on how these monthly patterns aggregate into annual return benchmarks, the month-level data is useful but the annual picture is what actually compounds in your account.

Is the September Effect in the S&P 500 Statistically Significant?

Yes, with important qualifications.

September's -1.1% average is the only calendar month with a negative long-run mean return. Its 44% positive-month rate is the lowest of any month. By those measures, the September effect is real and persistent across the full 1928-2023 dataset.

Whether it is exploitable is a different question entirely.

French's foundational research in the Journal of Financial Economics established that calendar-based return anomalies are real but frequently arbitraged away once widely known. The September effect has been documented for decades. If it were cleanly exploitable, institutional capital would have eliminated it. The fact that it persists suggests either that the arbitrage is too costly (transaction costs, tax drag) or that the underlying cause is structural rather than behavioral.

The proposed structural explanations include mutual fund fiscal year-end selling (many funds end their fiscal year in October, triggering September liquidations), post-summer portfolio repositioning by institutional managers, and historically elevated geopolitical risk events clustering in September. None of these fully explains the persistence.

For practical purposes: September's negative average is real, but its standard deviation is wide enough that any given September can easily be positive. The historical drawdown patterns show that September's worst months have been severe, but its median month is only modestly negative.

The actionable implication for high-net-worth investors is not to exit equities in August. It is to use September's historically elevated volatility as a systematic trigger for tax-loss harvesting reviews, which is covered in the next section.

How FATFIRE Investors Should Actually Use S&P 500 Seasonal Patterns for Tax-Loss Harvesting

This is where monthly return data becomes genuinely useful for a $5M+ taxable portfolio, and it has nothing to do with market timing.

The standard retail interpretation of seasonal patterns is: buy in November, sell in May. For investors in the 37% marginal income tax bracket, that strategy converts long-term capital gains into short-term gains or triggers unnecessary taxable events. The pre-tax return advantage of seasonal timing would need to be substantial to survive the after-tax math. Schwab's research found that even perfect market timing added only marginally more value than consistent dollar-cost averaging over a 20-year period. Imperfect timing, which is the only kind available in practice, frequently underperformed a fully invested strategy.

The better use of seasonal data is systematic tax-loss harvesting in historically weak months.

Consider the mechanics for a $5M taxable portfolio:

The 23.8% rate applies to long-term capital gains above $250,000 AGI, combining the 20% LTCG rate with the 3.8% Net Investment Income Tax under IRC Section 1411. IRS Publication 550 governs the wash-sale rules that apply when you harvest losses and reinstate similar positions, requiring a 30-day window before repurchasing substantially identical securities.

The practical execution: maintain a list of correlated-but-not-identical positions (e.g., swap a total market fund for a large-cap value fund) so you can harvest the loss, maintain market exposure, and avoid triggering the wash-sale rule. September's historically elevated volatility simply creates more harvesting opportunities than calmer months.

This reframes the entire premise. Monthly return patterns are not trading signals. They are a scheduling framework for tax management, which is the primary lever high-net-worth investors can actually control.

What Is the Best Month to Rebalance a High-Net-Worth Investment Portfolio?

There is no universally optimal rebalancing month, but the question has a more useful answer than "it depends."

Threshold-based rebalancing (rebalancing when allocations drift beyond a set band, say 5 percentage points) consistently outperforms calendar-based rebalancing in academic literature. For a $5M+ portfolio, a 5% drift in equities represents a $250,000 allocation shift. Waiting for a calendar month to rebalance when you are already 7 points overweight equities is a behavioral convenience, not a financial optimization.

That said, if you are going to rebalance on a calendar schedule, the evidence slightly favors combining rebalancing with tax-loss harvesting opportunities. That points toward late September or early October, after the historically weak September period has created potential losses to harvest and before the historically strong November-December window begins.

The seasonal market patterns data supports a rebalancing review in late Q3, not because September is reliably bad but because the volatility creates both harvesting opportunities and natural drift in equity allocations.

For taxable accounts specifically, the sequencing matters:

1. Review allocations in late September after any drawdown
2. Harvest losses in positions with embedded losses using correlated substitutes
3. Rebalance back to target allocation using new contributions or proceeds from harvested positions
4. Avoid triggering short-term gains on positions held less than 12 months

Vanguard's research consistently demonstrates that strategic asset allocation, not calendar-based timing, drives long-term portfolio returns. Rebalancing serves the allocation target, not the calendar.

Does Dollar-Cost Averaging Outperform Seasonal Market Timing for Portfolios Above $5 Million?

The evidence is clear, and the answer is mostly yes, with nuance.

Schwab's "Does Market Timing Work?" analysis found that even a hypothetical investor with perfect foresight (buying at the annual low every year for 20 years) outperformed a consistent dollar-cost averaging investor by a relatively modest margin. The investor who waited for the "right time" and stayed in cash significantly underperformed. The cost of being wrong about timing is asymmetric: missing the market's best days destroys compounded returns in ways that avoiding the worst days cannot fully offset.

Dimensional Fund Advisors' research reinforces this. Investors who attempt to time the market based on seasonal patterns frequently miss the market's best-performing days, which tend to cluster around periods of maximum pessimism, often in the same months (October, November) that follow historically weak September periods.

For portfolios above $5M, there is an additional complication. Large taxable positions often cannot be tactically repositioned without triggering substantial capital gains. A $3M embedded gain in an S&P 500 position does not get sold in August because September is historically weak. The tax cost alone makes seasonal timing strategies largely academic for investors with concentrated, appreciated positions.

The rolling return analysis shows that over any 10-year rolling period since 1928, the S&P 500 has been positive roughly 94% of the time. The compounding argument for staying invested overwhelms the marginal benefit of seasonal positioning.

Dollar-cost averaging wins not because it is optimal in theory but because it is executable in practice without triggering tax events, behavioral errors, or the execution risk of being wrong about timing.

The January Effect: What It Means for Large-Cap Portfolios Over $5 Million

The January Effect is largely a small-cap phenomenon, and it has weakened substantially since the 1980s.

NBER research indicates that the January Effect, historically characterized by outsized small-cap returns in January, has substantially diminished as institutional awareness and tax-loss harvesting activity increased. Once the pattern became widely known, institutional capital began positioning in December to front-run January buying, effectively pulling the return forward and compressing the January premium.

For large-cap S&P 500 exposure, the January effect is statistically weaker. The index's +1.1% January average is solid but not exceptional relative to April (+1.5%) or November (+1.5%). The positive-month rate of 58% in January is roughly in line with the index's overall historical average.

The mechanism behind the January effect in large caps is primarily tax-loss harvesting reversal: investors sell losing positions in December for tax purposes, then repurchase in January, creating temporary selling pressure in December and buying pressure in January. For investors who are systematic about tax-loss harvesting throughout the year (particularly in September and October), this December-January dynamic is already partially priced into their strategy.

The practical implication for a $5M+ portfolio: do not build a January allocation strategy around the January Effect. The academic evidence for large-cap January outperformance is weak, the pattern has diminished over time, and any tactical repositioning to capture it would likely cost more in taxes and transaction friction than the expected return premium.

Sell in May and Go Away: The After-Tax Math for High-Net-Worth Investors

The "Sell in May" strategy has genuine academic support. Bouman and Jacobsen's study found statistically significant evidence of a November-through-April seasonal premium across 37 countries. The pattern is real.

The after-tax math for investors in the top bracket makes it largely unworkable.

Consider the execution for a $5M S&P 500 position with a $2M embedded long-term gain:

• Selling in May triggers $2M in long-term capital gains
• At 23.8% (20% LTCG + 3.8% NIIT), the tax bill is $476,000
• That $476,000 must be recovered by the May-October underperformance before the strategy breaks even
• Historically, the May-October period has averaged roughly +2% to +4% in total return, meaning the strategy would need the market to lose more than 9.5% in that period just to justify the tax cost

The Sell in May strategy would have underperformed a buy-and-hold approach in 7 of the 10 years from 2013 through 2022. For investors in the 37% marginal income tax bracket, any strategy that converts long-term gains into short-term gains or triggers taxable events requires a substantially higher pre-tax return to justify the after-tax cost.

The 10-year performance trends make the compounding cost of being out of the market even more apparent. Missing the best-performing months in any given year has historically been more damaging than participating in the worst-performing months.

This is not an argument against all tax-motivated selling. It is an argument for precision: harvest losses systematically, not gains opportunistically.

Seasonal Patterns and Portfolio Volatility: What the Distribution Actually Looks Like

Averages obscure the distribution, and the distribution is what matters for behavioral discipline.

The standard deviation of monthly returns near 5.5% means that a "bad" September is not a signal of anything structural. It is a data point within a wide distribution. The 10th percentile for September is -9.4%, which on a $5M portfolio is a $470,000 paper loss. That is within normal statistical range, not a crisis.

December's notably lower standard deviation (3.9%) reflects the month's historically consistent, if modest, positive bias. It is the most predictable month in the dataset, which is why it has the highest win rate at 73%.

For investors managing large taxable portfolios, understanding the distribution rather than just the average is what separates reactive decision-making from systematic strategy. A September drawdown of 5% is not a reason to sell. It is a reason to open your tax-loss harvesting spreadsheet.

The 20-year rolling returns data contextualizes these monthly fluctuations within the longer compounding arc, which is ultimately what determines outcomes for investors with multi-decade time horizons.

Long-Term Compounding vs. Monthly Noise: Keeping the Right Frame

The S&P 500's monthly return data is genuinely useful. It is also genuinely easy to misuse.

The index's ~10% annualized total return, documented across multiple market cycles by Morningstar, does not arrive smoothly. It arrives through a sequence of months where roughly 60% are positive and 40% are negative, with a standard deviation that makes any individual month's result nearly uninformative about the next month's result.

Calendar anomalies like the September effect and the November-April seasonal premium are statistically real. They are documented in peer-reviewed academic literature. They are also too small relative to transaction costs, tax drag, and execution risk to generate net alpha for taxable high-net-worth investors who attempt to trade around them.

The frame that actually serves a $5M+ investor:

• Monthly patterns inform tax-loss harvesting timing, not trading decisions
• Rebalancing should be threshold-driven, not calendar-driven
• Dollar-cost averaging with systematic tax management outperforms tactical seasonal positioning in after-tax terms
• The valuation metrics over time and dividend yield history matter more to long-term return expectations than which month you happen to be in

The election year market behavior adds another layer of seasonal complexity in relevant years, though the evidence for election-cycle effects is similarly mixed when examined rigorously.

Modern portfolio theory, as extended by Fama-French factor research, holds that persistent, exploitable calendar anomalies are inconsistent with semi-strong market efficiency. The anomalies that do persist are generally too small relative to friction costs to generate net alpha for taxable investors. That is the intellectually honest resolution to the tension between "these patterns exist" and "you should not trade around them."

They exist. They are not reliably actionable as trading strategies. They are occasionally useful as a scheduling framework for tax management. That is the complete picture.

References

• Federal Reserve Bank of St. Louis (FRED) -- "S&P 500 Historical Data (SP500)" (2024)
• Morningstar -- "2024 Morningstar U.S. Market Outlook" (2024)
• Vanguard -- "Vanguard's Principles for Investing Success" (2023)
• Journal of Finance (Bouman & Jacobsen) -- "The Halloween Indicator, 'Sell in May and Go Away': Everywhere and All the Time" (2002)
• Journal of Financial Economics (French) -- "Stock Returns and the Weekend Effect" (1980)
• NBER (National Bureau of Economic Research) -- "The January Effect Revisited" (2003)
• Dimensional Fund Advisors -- "Pursuing a Better Investment Experience" (2023)
• IRS -- "Publication 550: Investment Income and Expenses" (2023)
• Schwab Center for Financial Research -- "Does Market Timing Work?" (2021)