The year’s strongest months are beginning. Will the rally reach your ETF too?
November and December have historically been among the best months for stocks. But October split the market in two, and this time it won’t be the calendar or the economy that decides where the rally goes, but the budgets of a handful of companies for data centers.
Key points
The Nasdaq has just had its worst month since March 2025, while the Dow Jones ended October in positive territory.
Three of the four biggest AI investors reported negative free cash flow for the last quarter.
November in an election year is historically the Nasdaq’s strongest month, with an average gain of 3,5 %.
Earnings at 493 smaller S&P 500 companies are growing faster than those of the Magnificent 7, for the first time since the end of 2022.
The Fed’s December meeting will, for the first time, directly affect the costs of the companies keeping the entire index afloat.
October: one half of the market fell, the other rose
The Nasdaq Composite closed October down 5,2 %, its worst month since March 2025. On Friday, October 30 alone, it shed 2,9 %, or more than 700 points, and it is down 7,8 % from its October record. The S&P 500 lost just 1,2 % for the month, while the Dow Jones actually gained 1,5 %.
Those figures side by side say more than any commentary could. This wasn’t a market sell-off, but a sell-off in one group of companies. Alphabet $GOOG fell 12 % on Thursday after its results, wiping out roughly half a trillion dollars in market value; October ended as the stock’s worst month in history. Intel $INTC plunged 28,6 % on Friday, its biggest one-day drop since 1974. The semiconductor index lost 7,2 %, and Nvidia $NVDA 4,2 %. The Dow, where tech giants carry less weight, meanwhile ended the month in positive territory.
So the market isn’t abandoning stocks as a whole. It’s abandoning one specific trade: the bet that the hundreds of billions of dollars poured into AI infrastructure will pay off for investors. And this is exactly when the period that statistics call the best part of the year begins.
The calendar promises the best months of the year
History is on the bulls’ side. According to the Stock Trader’s Almanac, November in congressional election years is the best month for the Nasdaq, with an average gain of 3,5 %; for the S&P 500 and the Dow, it’s the second-best month, with average gains of 2,7 % and 2,6 %. Since 1934, the Dow has gained an average of 2,8 % during the eight trading days around elections, which this year fall on Tuesday, November 3. The Nasdaq rose in 9 of 13 November election-year periods.
This is the beginning of the so-called “sweet spot” in the four-year cycle: from the fourth quarter of an election year to the second quarter of the following year, the S&P 500 has gained an average of 20,2 % and the Nasdaq 29,4 %. The reasons are prosaic. Political uncertainty fades after the election, funds add to their holdings before year-end, tax-loss selling ends, and the holiday season brings strong retail figures.
But an average is no guarantee. The sample includes thirteen election cycles, and a single exceptional year can rewrite the story. 2018 was also a congressional election year; November played out by the book, and then came December, when the S&P 500 lost more than 9 % because the Fed kept raising rates and the market feared what that would do to the economy. Seasonality works like a tidal wave: it lifts boats, but it won’t fix a hole in the hull. And this year, the biggest boat in the harbor has several holes.
Why five budgets, not the calendar, will decide things this time
The ten biggest companies make up roughly 40 % of the S&P 500’s weight, and an even larger share of the Nasdaq 100. Imagine a house whose roof rests on five beams. All five owners have just announced they’ll take out another mortgage to build an extension. Their data center investment plans look like this:
Alphabet: 195 to 205 billion dollars this year, including 51,4 billion in the third quarter alone. Free cash flow for the quarter was -9,1 billion dollars; over the past twelve months, it was 19,7 billion versus 73,6 billion a year ago. For 2027, the company promises a “significant increase.”
Amazon $AMZN : around 220 billion dollars this year, including about 76 billion in the quarter. Free cash flow over the past twelve months stands at -42,4 billion dollars, the second negative quarter in a row.
Meta $META : 135 to 145 billion dollars this year, including 34,9 billion in the third quarter. Free cash flow was -2,1 billion dollars, compared with +10,6 billion a year ago.
Microsoft $MSFT : 45,2 billion in the quarter, more than 50 billion next quarter, and around 170 billion dollars for the full year. Free cash flow was 18,6 billion, down 27 % year over year. It’s the only one of the four still in positive territory.
Three of the four biggest AI investors have just reported negative free cash flow. Together, these four companies will spend more than 700 billion dollars this year, and all are signaling that next year will be more expensive. When the index depends on them, calendar statistics become a secondary factor.
The way we think about it is that this is a once-in-a-lifetime opportunity. I think what we’re seeing with AI is the biggest transformation of our lives.
Andy Jassy, Amazon CEO, conference call on third-quarter 2026 results
The Fed is back in play and will hit the companies it used to miss
On October 28, the U.S. central bank kept rates in a range of 3,75 to 4 %, but two committee members voted for an immediate increase, and its statement noted that inflation remains elevated. Based on interest-rate futures prices, the market puts the probability of a rate hike at the December 9 meeting at roughly 73 to 85 %.
In the past, this would have left Big Tech unfazed. Alphabet, Meta, and Microsoft sat on tens of billions in cash and had virtually no net debt. Higher rates were an abstract discount factor in analysts’ models, not a budget line item. That changed in a single year.
At the end of September, Amazon had long-term debt of 173,1 billion dollars, up from 65,6 billion at the end of last year, and had also secured a supplier credit line of up to 80 billion. In June, Alphabet announced an 80 billion dollar stock issuance, and its total debt exceeded 100 billion. Meta announced plans to sell 97,5 billion dollars’ worth of shares, with the first 37,5 billion immediately and the rest through a new class of non-voting shares starting next year; shortly afterward, S&P lowered its rating outlook to negative.
It’s the difference between a tenant paying out of savings and a buyer with a mortgage. The first doesn’t care about interest rates; for the second, every quarter-point move changes the monthly payment. So a December rate hike will, for the first time, hit exactly the companies keeping the index afloat—not just their valuations, but their costs directly.
Over the course of this investment cycle, we expect attractive returns on invested capital that remain significantly above our cost of capital, and we anticipate strong operating cash flow.
Anat Ashkenazi, Alphabet CFO, conference call on third-quarter 2026 results
The rest of the market has its best numbers in years. You can’t see it in the index
Now for the strange part. S&P 500 companies’ third-quarter earnings are growing 30,2 % year over year, and 82 % of companies beat estimates. But the Magnificent 7’s earnings are growing 22,6 %, while the remaining 493 companies’ are growing 32,6 %. The seven largest companies missed analysts’ estimates by a combined 14,7 %. For the first time since the end of 2022, the rest of the market is driving index earnings, and estimates for the fourth quarter (22,6 % versus 30,8 %) suggest that this will continue. Meanwhile, the index as a whole is trading at 18,7 times expected earnings, below its ten-year average.
For holders of an index ETF, this makes for uncomfortable math. A typical S&P 500 fund is market-cap weighted: Nvidia has a weight of around seven percent in the index, while the smallest company has a hundredth of a percent. When 493 companies rise and seven fall, the index can decline even while most of the portfolio is making money. Imagine an election in which seven people out of a hundred hold half the votes. An equal-weight version of the S&P 500 gives each company 0,2 %, and that’s where the “rest of the market” is showing up this year.
So the rally suggested by seasonal statistics will most likely arrive. But if it reaches the 493 companies rather than the seven, your index fund will largely miss it.
While we remain constructive on the structural growth of the AI theme, high expectations and potentially wide divergences in individual companies’ results underscore the risks of concentration in individual stocks.
Ulrike Hoffmann-Burchardi, CIO Americas and global head of equities, UBS Chief Investment Office
Three scenarios for year-end—and the calendar that will tell you which one is unfolding
The next ten weeks are unusually packed, and each item belongs to one of the scenarios:
November 3: congressional elections, which could bring data center regulation.
This week: Apple $AAPL results, the only company in the Magnificent Seven without massive capex.
November 17: Nvidia results. The company’s own guidance calls for revenue of around 108 billion dollars for the quarter. This is the one place to measure whether the hundreds of billions spent by other companies have actually shown up in someone’s revenue.
December 9: the Fed, with new rate projections for 2027.
End of January: Alphabet and Meta will publish specific data center budgets for 2027.
Scenario A: a classic rally. Nvidia confirms demand and raises its guidance; the Fed raises rates but signals that it will be a one-and-done hike. The market accepts the argument that capex is investment, not spending, and the seasonal rally reaches everywhere. The S&P 500 and Nasdaq 100 gain, with tech ETFs rising faster. You’ll recognize it if Alphabet and Amazon stop falling before Nvidia reports.
Scenario B: rotation. The most likely version. The rest of the market’s earnings draw money into industrials, financials, health care, and smaller companies. The Dow and equal-weight indexes rise, the S&P 500 treads water because the seven companies’ weight holds it back, and the Nasdaq 100 lags. The rally happens, just not where most investors are invested. Signal: the equal-weight S&P 500 beats the traditional index by several percentage points in a month.
Scenario C: capex becomes a burden. Nvidia disappoints or merely “meets” expectations, the Fed takes a tougher tone, and borrowing costs for hyperscalers rise. Data center budgets then become a liability for the market, not a promise. Concentration drags the index down regardless of the calendar, with Nasdaq 100 ETFs hit hardest, the S&P 500 less so, and the Dow least. Signal: how stocks react to Big Tech’s share and bond offerings—whether the market welcomes or punishes further issuance.
One thing holds true in all three scenarios. Your ETF’s return through year-end won’t be decided by the economy or a historical table, but by five budgets voted on by five boards. Seasonality can tell you how strong the tide will be. Whether you feel it depends on which boat you’re in.
