Midterm Elections and the Stock Market: How Markets Have Historically Reacted

How do markets historically react to midterm elections? It is one of the most common questions investors ask in a year like this one, and the historical record gives a reasonably clear answer. Midterm elections and the stock market have a recognizable relationship: returns tend to be below average and volatility above average in the run-up to the vote, particularly in September and October, and then markets have historically recovered strongly in the twelve months that follow.

That pattern is real. It is also weaker and less reliable than the headlines suggest, and the cost of acting on it has historically exceeded the volatility anyone was trying to avoid. Both things are true at once, and holding both is the difference between using history and being used by it.

With election day set for November 3, 2026, what follows is what the data actually shows, what it does not show, and the handful of things that are genuinely worth doing between now and then.

How do markets historically react to midterm elections?

Midterm years have been the weakest leg of the four-year presidential cycle. The S&P 500 has averaged roughly 7.5% in midterm election years against about 12.4% across all years in BlackRock’s data, with presidential election years near 12.1% and the odd-numbered years between them closer to 14.9%. [4] Fidelity’s series puts the midterm figure lower still, around 5% for the second year of a presidential term, the weakest average of any year in the cycle since 1950. [2]

The two most recent midterms were uncomfortable in exactly that way. The S&P 500 fell 18.1% in 2022 and 4.4% in 2018. [4] And since 1961, the average peak-to-trough decline inside a midterm year has run near 19%. [2]

What defines a midterm year is the number of sharp days, not the ending number. Policy uncertainty rises, positioning turns defensive, and the market moves more on less news. That is an unpleasant experience in a way no annual return figure captures.

Worth holding onto, though: uncomfortable years are not unusual years. Since 1980, the S&P 500 has fallen an average of roughly 14% at some point inside every calendar year and still finished positive in most of them. [5] Since 1942, a decline of 5% or more has arrived about three times a year, and a drop of 10% or more about every 16 months. [5] The pullback that gets blamed on the election in October is, statistically speaking, an ordinary Tuesday.

Why September and October are the historically rough stretch

If there is one genuinely timely fact in this article, it is this one. Since 1930, the S&P 500 has slid 5% or more during the September to October stretch in 63% of the past 24 midterm election years. [1] That is not a coin flip. It is the clearest seasonal tendency in the whole midterm dataset, and we are standing in the middle of it right now.

September carries its own baggage independent of politics. Between 1928 and 2025 it was the weakest month of the year for the S&P 500, averaging a decline of about 1.1%, finishing lower 55% of the time, and averaging a 4.7% loss in the years it fell. [1] Stack a midterm on top of that and you have the market’s least hospitable six weeks.

Why this happens is not mysterious. Markets dislike unresolved questions more than they dislike bad answers. In the run-up to a close election the range of plausible policy outcomes is at its widest, so capital that could wait, waits. That shows up as thinner participation and larger swings on ordinary news.

The practical implication is a matter of expectation-setting rather than action. If the market drops 6% in the third week of October, that is not a signal about 2027. It is the single most historically normal thing that could happen in that particular week of a midterm year. Knowing that in advance is most of what keeps people from selling into it.

The post-election pattern, and the honest caveat

The clearer signal is what comes after. Looking back to 1900, the S&P 500 has averaged roughly 2.9% in the twelve months leading into a midterm, well below the 8.9% average for all years in that study, then roughly 12.4% in the twelve months after. [3] Fidelity’s data shows post-midterm twelve-month gains in the large majority of cycles with an average near 14%. [2] And on the most striking version of the statistic, the index has risen in the twelve months following a midterm 95% of the time since 1938 and 100% of the time since 1950, averaging about 14.5% since 1950, though that particular figure circulates widely without a named research house behind it. [1] Other research finds the rally beginning roughly a month before election day, once the range of outcomes narrows, with an average gain near 14.1% in the six months after. [4]

The mechanism is the mirror image of the one above. An election is a scheduled reduction in uncertainty. Whatever the result, on November 4 investors know something they did not know on November 2, and money that was waiting has one less reason to wait.

Now the caveat, and it matters more than the statistics. U.S. Bank’s researchers ran formal significance tests on this data and concluded that the differences between midterm years and other years were not large or consistent enough to establish a reliable election effect. [3] Of the 31 elections in their study, 11 coincided with inflation shocks, rising interest rates, wars, or financial crises. [3] Those were the forces moving markets. The election happened to be standing nearby.

A “100% since 1950” record sounds like a law of nature. It is 19 observations. That is a small sample describing a period that also contained the most powerful sustained bull market in human history, and it would take only one exception to reduce it to a good but ordinary batting average. I would not build a plan on it. I would use it, as I do here, to argue against panic.

One more note on the numbers in this article. The baselines come from different studies with different start dates and are not interchangeable. BlackRock’s 12.4% all-year average is measured from 1970 forward; U.S. Bank’s 8.9% all-year average runs from 1900. Each is accurate within its own dataset. Comparing them to each other is not meaningful, and anyone who does it fluently is probably selling something.

Midterm elections and the stock market: where 2026 actually stands

This midterm year has been calmer than average, not stormier. Through September 24, the S&P 500 was up roughly 13.5% on a total return basis for the year. [6] As of Fidelity’s mid-August review, the deepest pullback of 2026 was about 9%, back in March, well short of the 19% average midterm drawdown since 1961. [2] For a year that was supposed to be the difficult part of the cycle, that is a notably benign result.

The political arithmetic is genuinely close, which is worth naming plainly. All 435 House seats and 33 Senate seats, plus two special elections for the seats vacated by Marco Rubio and J.D. Vance, are on the ballot November 3. Republicans currently hold the House 219 to 213 with three vacancies, and hold 53 Senate seats. Democrats would need a net gain of three House seats and four Senate seats to take control of each chamber. [7] Narrow margins mean a wider range of plausible outcomes, and markets generally price a wider range as more volatility rather than less.

The wider backdrop is doing more work than the election. Consensus expectations call for strong corporate earnings growth in 2026, core inflation has continued to moderate, and Fed pricing has turned more hawkish even as inflation and labor data have softened, which introduces its own meeting-to-meeting rate volatility. [8] Those are the variables that will still matter in March 2027. The election will not be.

Two things can be true at once. The setup argues for expecting more noise between now and November. Nothing in the setup argues for changing a long-term allocation because of it.

Figures compiled from U.S. Bank (1900 to 2025), BlackRock (Morningstar data from 1970), and Cantor Fitzgerald data reported by The Motley Fool. [1][3][4] The studies use different start dates and samples, so the rows are not strictly comparable to one another. Averages are historical and are not predictions. Past performance is no guarantee of future results. An index cannot be purchased directly by investors.

Divided government and what markets really price

Markets care less about who wins than about what can pass. A common belief is that one party’s control is inherently better for stocks. The historical record does not support it. What the record suggests is that markets respond to the range of policy outcomes narrowing, not to any particular party doing the narrowing.

One useful data point: when a party lost the political trifecta (the presidency plus both chambers) in a midterm, six-month post-election returns averaged about 10.4%, compared with roughly 16.1% when control was gained or Congress remained divided. [4] That looks like the market repricing legislative capacity rather than rewarding an ideology.

It is also worth remembering that the current Senate majority sits below the 60 votes generally needed to end debate on most legislation. [3] Sweeping change is harder to legislate than to campaign on, in either direction. Portfolios built on the assumption that a campaign promise becomes law tend to age badly.

Sector effects are noisier than they look. In midterm years, health care has averaged about 10.7% and energy about 8.9%, while industrials averaged 0.6% and financials about −0.6%. [4] Before acting on that, note that technology has ranked among the top two sectors across the last four presidencies, spanning both parties. [4] Sector leadership has tracked earnings and innovation far more closely than it has tracked the party in power.

What sitting out an election has cost

Going to cash requires two correct decisions, not one. You have to be right about when to leave and right about when to return. First Trust puts it plainly in its client research: stepping aside does not merely risk missing a rebound, it risks missing all the growth on that money going forward. [5]

The arithmetic is unforgiving. A $10,000 investment in the S&P 500 from the end of 1979 through June 30, 2026 grew to roughly $2,218,960. Miss only the five best days and it falls to about $1,370,542. Miss the ten best days and it falls to about $962,494, less than half. Miss the fifty best days and roughly $148,043 remains. [5] The best days cluster near the worst days, which is precisely when a nervous investor is most likely to be out of the market.

The election-specific version is equally sobering. BlackRock found that $100,000 invested in the S&P 500 and left alone from 2013 to the start of 2026 grew to about $398,000. Investors who moved to cash whenever their preferred party was out of power ended with roughly $214,000 or $186,000, depending on which party they favored. [4] Both sides underperformed by roughly half. That is a thirteen-year window rather than a century, and it is not a political finding. It is a behavioral one.

What the long record shows. Since 1937, the S&P 500 has produced a positive return 53.4% of the time over one day, 77.7% over one year, 93.2% over five years, and 97.5% over ten years. [5] The average bull market since 1942 lasted 4.4 years and returned 152.8%; the average bear lasted 11.1 months and lost 31.7%. [5] Even after the fifteen worst single days since 1960, which averaged a one-day loss of 8.85%, the index averaged a 30.29% gain over the following year. [5]

None of that guarantees anything about November. It describes an asset class that has repeatedly rewarded patience and repeatedly punished timing, across every political configuration this country has produced.

How we approach an election year at Dominion

We start with your plan, not the news cycle. The Dominion Wealth Architecture™ framework runs on a simple sequence: Decide, Design, Build, Steward. An election is a stewardship question, not a redesign question. If your allocation was appropriate in July, an election result rarely makes it inappropriate in November.

What we look at in a year like this is specific and mostly unglamorous:

• Cash runway. Are the next two to three years of planned withdrawals held in something that does not care about volatility, so an October drawdown never forces a sale at a bad price?
• Drift. After a year up roughly 13.5%, has your equity exposure climbed above where you intended? If so, rebalancing is warranted on its own merits, not on a forecast.
• Tax opportunities. Volatility creates tax-loss harvesting and Roth conversion windows that a calm, straight-up year never offers. This is the constructive way to respond to a decline.
• Concentration. Several strong years in a row have a way of quietly turning a diversified portfolio into a bet on a handful of names.

Those are responses to your circumstances. They are not bets on an election, and that distinction is essentially the whole of it.

Because Dominion Growth Advisors sits alongside the wealth management practice, we can look at the tax side of a volatile year in the same conversation rather than three months later. For business owners in particular, an election year raises real questions about timing a sale, a Roth conversion, or a charitable gift. Those deserve a planning answer, and they are usually driven far more by your own timeline than by the calendar in Washington.

The one thing I would ask you to avoid. Do not let a strongly held political view become an investment thesis. It is the single most expensive mistake I watch intelligent people make, and it is expensive in both directions.

Frequently asked questions

Sources

Every figure and date cited above comes from the source listed below. Links verified current as of September 25, 2026.

1. The Motley Fool: Prediction: September Will Be Volatile, but Stocks Will Rally After November Midterms (September 17, 2026).
   https://www.fool.com/investing/2026/09/17/prediction-september-will-be-volatile-but-stocks/ Supports: the 5%+ September to October decline in 63% of the past 24 midterm years since 1930, which that article credits to Cantor Fitzgerald; September as the weakest month since 1928 (−1.1% average, lower 55% of the time, 4.7% average loss in down years), credited to Macrobond; and gains in the twelve months after a midterm 95% of the time since 1938 and 100% since 1950, averaging about 14.5% since 1950. Note that the source article does not name a research firm behind that last figure, so it is reported here as secondary rather than primary data.
2. Fidelity: How Might Midterm Elections Impact the Stock Market?
   https://www.fidelity.com/learning-center/trading-investing/the-surprising-truth-about-midterms-and-stocks Supports: the roughly 5% average return in the second year of a presidential term; the approximately 19% average midterm-year drawdown since 1961; the roughly 9% peak-to-trough decline in March 2026; the roughly 14% average post-midterm twelve-month return.
3. U.S. Bank: How Midterm Elections Affect the Stock Market.
   https://www.usbank.com/investing/financial-perspectives/market-news/stock-market-performance-after-midterm-elections.html Supports: the 2.9% average in the twelve months before a midterm versus 8.9% for all years in the study (1900 to 2025); the 12.4% average in the twelve months after; the statistical significance testing; the finding that 11 of 31 elections coincided with inflation shocks, rate increases, wars, or crises; the Senate cloture threshold point.
4. BlackRock: 2026 Midterm Elections and Market Performance.
   https://www.blackrock.com/us/financial-professionals/insights/2026-midterm-elections-and-market-performance Supports: the midterm-year average of 7.5% versus 12.4% for all years (Morningstar data from 1970); presidential-year and odd-year averages; 2022 and 2018 calendar returns; the 14.1% six-month post-election average; the trifecta comparison of 10.4% versus 16.1%; midterm-year sector averages; the 2013 to 2026 cost of moving to cash by party.
5. First Trust Portfolios L.P.: Client Resource Kit: Markets in Perspective (data as of June 30, 2026; source First Trust, Bloomberg). https://www.ftportfolios.com Supports: the approximately −14% average intra-year decline since 1980; frequency of 5% and 10% declines since 1942; growth of $10,000 from 12/31/1979 to 6/30/2026 and the missing-best-days figures; probability of positive returns by holding period since 1937; average bull and bear market duration and return since 1942; performance after the fifteen worst single days since 1960.
6. Slickcharts: S&P 500 YTD Return (as of market close September 24, 2026).
   https://www.slickcharts.com/sp500/returns/ytd Supports: the S&P 500 total return of approximately 13.5% year to date through September 24, 2026.
7. Ballotpedia: United States Congress Elections, 2026.
   https://ballotpedia.org/United_States_Congress_elections,_2026 Supports: the November 3, 2026 election date; 435 House seats and 33 Senate seats plus two special elections for the Rubio and Vance seats; the 219 to 213 House margin with three vacancies and the 53-seat Senate majority; the net gains required to flip each chamber.
8. Goldman Sachs Asset Management: US Market Pulse, September 2026 (published September 8, 2026; data as of August 31, 2026). https://am.gs.com/en-us/advisors/insights/article/market-pulse Supports: 2026 consensus earnings growth expectations; moderating core inflation; the more hawkish Fed pricing against softening inflation and labor data, and the resulting rate volatility.

Justin Kauffman, CFP®, CEPA®

Founder & Financial Advisor, Dominion Private Wealth

Justin is a Certified Financial Planner™ and Certified Exit Planning Advisor® who helps families and business owners in Goodyear and across the West Valley coordinate investment, tax, estate, and legacy strategy into a single plan.

Published September 25, 2026 · Updated September 25, 2026 · Reviewed by Dominion Private Wealth