The Stock Market's "Best Six Months" Start November 1. Here's What 76 Years of Data Actually Show

Every November 1, the same statistic circulates: since 1950 the Dow Jones Industrial Average has averaged a 7.3% gain from November through April against just 0.8% from May through October. The claim is easy to repeat and hard to check, because the two most-cited sources publish different numbers, neither publishes the hit rates that get quoted alongside them, and the arithmetic example that usually follows does not reconcile. So this version is built out of comparison tables: one table per claim, with a column for what we could verify ourselves.

The New York Stock Exchange on Wall Street. Photo: Arnoldius, CC BY-SA 3.0, via Wikimedia Commons. File photo, not tied to any specific trading session discussed in this article.

Key takeaways

Fidelity, citing FactSet data since 1945, puts the S&P 500 at roughly 7% for November-April against about 2% for May-October. Those are price returns, dividends excluded.
Our own count on 81 November-April windows since 1945 (Shiller monthly closing-price series) gives +5.07% against +1.65% for May-October. Same direction, smaller gap - and the measurement method explains most of the difference.
The winter half did not win more often. It was positive in 52 of 81 windows (64.2%); the summer half was positive in 55 of 82 (67.1%). The widely quoted 75%-versus-66% hit rate does not appear on the page it is usually attributed to.
Over the last 20 cycles the gap nearly closed: November-April averaged +3.97% (positive 14 of 20), May-October +3.03% (positive 16 of 20).
$10,000 run through the actual windows from November 2006 to October 2026 ends at about $33,187 if left invested all year and about $20,556 if held only through the winter halves with cash earning nothing.

Figures checked: November 3, 2026.

Table 1: the same six months, measured four ways

The pattern has a name and a long history, so it has also been measured by several people who did not use the same index, the same start year, or the same price convention. Putting the measurements in one table is the fastest way to see how much of the "Best Six Months" is a robust finding and how much is an artifact of who did the counting.

Who measured	Index and start year	Nov-Apr average	May-Oct average	What the source actually publishes
Fidelity / FactSet	S&P 500, since 1945	roughly 7%	about 2%	Exact wording: an "average price return of roughly 7% from November to April, compared to about 2% from May to October." Chart note credits FactSet as of October 22, 2026.
Fidelity / FactSet	Russell 2000, start year not stated	about 9%	not reported	The page says the Russell 2000 "has shown even greater seasonal strength, averaging 9% returns during the best 6-month window." It gives no May-October counterpart and no start year of its own, so the 1945 start cannot be confirmed for this line.
Stock Trader's Almanac	DJIA, since 1950	+7.3%	+0.8%	The figure we could only reach through a third-party summary. The Almanac's own site is a subscription product and its article text sits behind a login, so we could not read the original table.
This article's own count	S&P 500, 1945-2026, monthly closing-price series	+5.07% (median +5.06%)	+1.65% (median +2.62%)	81 winter windows and 82 summer windows, computed for this piece from the Shiller monthly series. Method and its limits are in the note below.

Why our numbers come in lower than Fidelity's: the series we can read ourselves is monthly, and each monthly value is an average of that month's closing prices rather than the last close of the month. A November-average starting point and an April-average ending point cut roughly half a month off each end of the window, which trims whatever trend was running. That compresses both halves, which is why the winter figure falls from about 7% to 5.07% and the summer figure from about 2% to 1.65%. The direction of the gap survives the change of method; its size does not. Anyone quoting a one-decimal seasonal average should be asked which price convention produced it.

Table 2: how often, which is the number nobody publishes

An average can be carried by a handful of enormous winters. The more useful question for someone deciding whether to sit out a summer is how often each half was positive at all. This is where the commonly repeated version of the story breaks down: the hit rates that circulate with the seasonality charts are not on the Fidelity page they are credited to, and we found no primary publisher for them. Counted directly, they also point the other way.

Window	Periods counted	Positive	Hit rate	Average
Nov-Apr, 1945-2026	81	52	64.2%	+5.07%
May-Oct, 1945-2026	82	55	67.1%	+1.65%
Nov-Apr, 1990-2026	36	25	69.4%	+4.59%
May-Oct, 1990-2026	37	26	70.3%	+1.64%
Nov-Apr, last 20 cycles	20	14	70.0%	+3.97%
May-Oct, last 20 cycles	20	16	80.0%	+3.03%

Read the table rather than the headline. The winter half earns more per window; the summer half shows up positive at least as often, and in the recent samples more often. Those two facts are not in conflict - they describe a winter distribution with fatter tails in both directions. An earlier version of this article reported a 75% winter hit rate against 66% for summer, and a May-October average of about -2% since 1990 with losses in 56% of periods, both credited to Fidelity. Neither figure is on that page, and our own count contradicts the second one outright: May-October since 1990 averaged +1.64% and was positive in 26 of 37 periods. Those sentences have been replaced with the counts above.

Table 3: $10,000 in the tidy model against $10,000 in the actual record

The standard illustration multiplies the two averages together and compounds. It is worth doing, because it shows that even taking the seasonal averages at face value, sitting out the summer loses money - and then it is worth redoing with real windows, because the real answer is considerably worse than the model suggests.

Version	Invested all year	Winter halves only	Gap
Averages model, 20 years (1.073 x 1.008 compounded against 1.073 compounded)	about $48,000	about $40,926	about $7,100
Actual windows, Nov 2006 - Oct 2026, cash earning 0%	$33,187	$20,556	$12,631
Same 20 years, cash earning 1% per summer half	$33,187	$25,083	$8,104
Same 20 years, cash earning 2% per summer half	$33,187	$30,546	$2,641
Actual windows, Nov 1945 - Oct 2026 (81 cycles), cash at 0%	$1,001,725	$423,576	$578,149

Three things fall out of that table. First, the arithmetic in the tidy model: (1.073 x 1.008) compounded 20 times turns $10,000 into about $48,000, not the $48,180 or $48,200 an earlier version of this article gave, and the winter-only leg lands at $40,926 rather than $40,900 or $40,940. Second, the real 20-year record is far less kind to the seasonal switch than the model, because the summer halves of the last two decades were not the small positive drag the long-run average implies - they averaged +3.03%. Third, the cash rate does most of the work in the comparison. At 2% per half-year the strategy nearly keeps up, which means the seasonal trade is in large part a bet on short-term interest rates rather than on seasonality. All of these figures are price returns; dividends, which you forfeit for half of every year under the switching rule, are not counted on either side, and neither are commissions or the capital-gains tax an April sale can trigger in a taxable account.

Table 4: the explanations, and which ones a table can test

No one has established why the calendar split exists, and Fidelity says so plainly - it files the pattern alongside the September and January effects as a calendar anomaly. The candidate explanations are worth laying out next to what would have to be true for each one, because several of them are not testable from price data at all.

Explanation offered	Who offers it	What would have to show up
Year-end portfolio rebalancing moving cash into stocks	Named on the Fidelity page	A concentration of the winter excess in November-December rather than spread across six months
Optimism tied to higher holiday consumer spending	Named on the Fidelity page	Consumer-linked sectors leading the winter half, not the index as a whole
Election cycles, with November votes shifting sentiment and policy expectations	Named on the Fidelity page	A visible split between election and non-election years, which a six-month average hides
Sector composition: cyclicals lead winter, defensives lead summer	Attributed by Fidelity to CFRA	Sector-level data, which price-index seasonality cannot confirm or refute
The effect has faded as technology and globalization smoothed trading patterns	Raised by Fidelity as a caution	A shrinking gap in recent samples - which is what our last-20-cycles row shows

Only the last row is settled by the data in this article, and it cuts against the pattern rather than for it. The 1945-2026 gap between the halves is 3.42 percentage points; since 1990 it is 2.95 points; over the last 20 cycles it is 0.94 points. That is a pattern on its way to indistinguishable, measured with the same series throughout.

Where the current market sits against the 81-year record

Seasonality is the weakest of the inputs available right now, and it is pointing into a market that several other measures already describe as stretched: valuation ratios near the top of their recorded range, index concentration in the largest companies at record levels, and light household cash positioning. A winter-half average of 5.07% is a statement about 81 overlapping windows, not a forecast for one; roughly one winter in three since 1945 finished flat or lower, and the individual windows run from double-digit gains to double-digit losses. Nothing in the record tells you which kind this one will be, and the one thing the record does say clearly - that staying invested through both halves beat switching, in every sample we ran - is an argument against acting on the calendar at all.

FAQ

What exactly are the "Best Six Months" in the stock market?

The November-through-April half of the calendar year, which historical data shows has carried most of the long-run gain in U.S. stock indexes. The split was popularized by the Stock Trader's Almanac. It is a statistical regularity, not a rule, and it is measured differently by different publishers.

Is "sell in May and go away" the same thing?

It is the trading version of the same calendar split - sell at the start of May, return at the start of November. Fidelity notes the precise seasonal trading rule is no longer in vogue, and our 20-year run of the actual windows shows why: holding only the winter halves turned $10,000 into $20,556 against $33,187 for staying invested, with cash earning nothing.

Does this pattern mean stocks will go up this November?

No. The winter half was positive in 52 of 81 windows since 1945, which is 64.2% - a majority, and nothing more. Roughly one in three finished flat or negative.

Has the pattern gotten weaker in recent years?

By our count, yes, and sharply. The Nov-Apr minus May-Oct gap is 3.42 percentage points across 1945-2026, 2.95 points since 1990, and 0.94 points over the last 20 cycles, when the summer half was positive more often than the winter half.

Would shifting money seasonally have beaten just staying invested?

Not in any sample we ran. Over the last 20 cycles staying invested led by $12,631 per $10,000 with cash at 0%, and by $2,641 even with cash earning 2% each summer half. Over 81 cycles the gap is $578,149 per $10,000. None of these figures include dividends, commissions, or taxes.

How were this article's own figures calculated?

From the monthly S&P 500 closing-price series that multpl.com publishes from Robert Shiller's dataset. Each November-April window is the April value divided by the November value, each May-October window the October value divided by the May value. The series states that "other than the current price, all prices are monthly average closing prices," so every window is measured average-to-average rather than close-to-close. Price returns only; no dividends.

Sources

This article is for general investment education only and is not personalized investment, tax, or legal advice. Past performance, including the seasonal averages discussed above, does not guarantee future results. Consult a qualified financial professional before making investment decisions.

Search This Blog

USStockInsights