Sunday 1 November 2026International EditionRSS
The financial daily of the startup economy
Investing

October splits Wall Street in two as the calendar's favourite six months begin

The Dow gained 1.7% in October while the Nasdaq lost 5.3%. Now comes the November to April stretch that history favours, in a midterm year, which raises the stakes on a famous statistic.

· · 6 min read
Technical drawing for: October splits Wall Street in two as the calendar's favourite six months begin
Illustration: Invest Times

October has a reputation it rarely deserves, and this year it delivered a strange version of it. When trading closed on Friday, October 30th, the Dow Jones Industrial Average stood at 51,770.31, about 1.7% above its September close of 50,906.05. The Nasdaq Composite finished at 25,437.00, down 5.3% from 26,861.06, its worst month since March 2025. The S&P 500 sat between the two, at 7,580.07, a loss of just under 1% for the month and about 3% below the record of 7,818.93 it had set on October 6th.

A gap of seven percentage points between two indices that are supposed to describe the same market is unusual enough. It arrives at a moment when the calendar becomes a talking point in its own right. November opens what seasonal investors call the "best six months", and in 2026 it opens the day before the midterm elections on November 3rd. The statistics attached to both are striking. They also deserve to be read carefully.

A month that told two stories

The split in October was not subtle. The Nasdaq 100, the narrower index of the largest non-financial companies on that exchange, fell 6.65% in the month, while the Dow rose 1.70%, a spread of 8.35 percentage points. The small companies of the Russell 2000 ended slightly higher, up about 0.5%. The S&P 500's total return, with dividends included, was minus 0.91%, which ranks 24th of the 33 Octobers since 1994 in one tracker's tables.

Inside the S&P 500 the picture was even more lopsided. Nine of its 11 sectors rose in October. Energy gained 13% as West Texas Intermediate crude climbed more than 15% in the month to $104.12 a barrel. Health care, utilities and consumer staples were each on course for gains above 10%. Against that, the technology sector fund XLK lost 7.5%, communications services lost more than 9% and the PHLX Semiconductor index fell about 15.3%.

The losses were concentrated in the companies that had carried the market. Alphabet fell 17.22% in October, Meta Platforms 15.30% and Nvidia 11.36%. A long list of names linked to artificial intelligence infrastructure did worse: Intel lost 45.20% in the month, CoreWeave 38.05% and Nebius 33.65%. At the other end of the table sat Lockheed Martin, up 26.00%, Eli Lilly, up 23.28%, and Walmart, up 18.46%. Apple and Microsoft, two of the largest companies in the index, both rose.

That is the arithmetic of a capitalisation-weighted index. Most sectors can rise and the headline number can still fall, because the few largest companies carry a weight that the many smaller ones cannot offset. For an investor who owns a broad index fund, October was a mild disappointment. For one who owned a portfolio of chip and AI infrastructure stocks, it was among the worst months in years. For one heavy in health care or energy, it was excellent.

The bond market had a better October

The other notable number of the month came from Treasuries. The 10-year yield ended October around 4.91%, down 36.4 basis points in the month, its largest monthly fall since December 2023. The two-year yield fell 30.6 basis points to 4.579% and has declined for five straight weeks. Gold, by contrast, ended the week at $3,898.90 an ounce for December delivery, a 52-week low and roughly 30% below its January record.

What the calendar claims

The "best six months" idea was identified in 1986 by Yale Hirsch, whose son Jeffrey now edits the Stock Trader's Almanac. Its claim is simple: stocks have earned most of their returns between the end of October and the end of April. According to Almanac figures cited by Yahoo Finance on October 28th, the Dow has averaged gains of about 7% from November to April since 1950, against less than 1% from May to October. On those numbers, $10,000 invested only in the winter halves grew to about $1.3m, while the same sum invested only in the summer halves shrank to about $4,300.

Academics have taken the pattern seriously. Sven Bouman and Ben Jacobsen gave it a formal name in a 2002 paper in the American Economic Review, "The Halloween Indicator, 'Sell in May and Go Away': Another Puzzle". The word puzzle was well chosen: a pattern this simple should, in an efficient market, be traded away.

The midterm version is more dramatic still. Yahoo Finance's own calculations show that the November to April stretch following a midterm election has risen 19 consecutive times since 1950, with an average gain of about 15%. If the S&P 500 ends April 30th above its October 30th close, the record becomes 20 out of 20. A loss would be the first in a post-midterm winter since 1946. Almanac data also show that October and November have been the S&P 500's two best months in midterm years since 1950. This October has already broken that pattern.

Why the statistics deserve scepticism

The more recent evidence is less flattering to the calendar. On Yahoo Finance's figures, the S&P 500 averaged about 7.3% from November to April and 0.9% from May to October between 1950 and 1987. From 1988 to 2025 the gap narrowed to 6.6% against 3.3%. Over the past 10 years the two halves were almost identical, at about 6.8% and 6.9%, and the summer half won in five of those 10 years, including a gain of about 23% in 2025. The S&P 500 is up about 6% since the end of April this year, so the "bad" half of 2026 was not bad at all.

There are plausible reasons for the fading. Regular contributions from index funds and workplace retirement plans now flow into equities every month regardless of season. A pattern that has been printed in an almanac every year since the 1980s is also, by now, very widely known.

The midterm record carries a different problem: it rests on 19 observations. A sequence that short can be produced by a handful of strong economic cycles that happened to line up with the electoral calendar. The streak tells investors something about the past behaviour of markets after political uncertainty is resolved. It is not a mechanism, and it says nothing about valuations, interest rates or the earnings of the companies that dominate today's indices. It also measures a single index from a single closing level, which this year happens to follow a sharp Friday sell-off.

Finally, average returns hide the distribution. The same Yahoo analysis notes that the summer half produced losses of about 15% in 2001, about 18% in 2002 and about 30% in 2008. October itself illustrates the point: since 1994 the S&P 500 has risen in 63% of Octobers with a median return of 2.3%, yet the worst October in that span, 2008, cost 16.5%.

Our view

The most useful fact from October is not seasonal at all. It is the gap between a rising Dow and a falling Nasdaq, and the reminder that a single index number can conceal very different experiences. Investors who believed they were diversified because they owned many funds may find, on inspection, that they owned the same handful of AI-linked companies several times over. October showed what that concentration costs when sentiment turns.

The calendar statistics are worth knowing and worth discounting. The "best six months" effect has weakened markedly over the past four decades, and the midterm streak is a small sample attached to a large claim. What to watch over the coming months is whether October's rotation into energy, health care and defensive sectors persists, whether the fall in Treasury yields continues, and whether the 7,580.07 close of October 30th proves to be a low or a staging point. The almanac will record the answer at the end of April; it cannot supply it in advance.

This article is analysis, not investment advice.

Sources

Filed under Investing. Written by Invest Times Editorial under our editorial standards. Commentary and analysis for information only, not investment advice.

More in Investing

All Investing analysis