Do stocks really rally after midterms? Testing 19 elections since 1950

So, do stocks go up after midterms? Historically, the S&P 500 has tended to rise in the 12 months after midterms. But the ride wasn’t always smooth. In 1986–87, the index fell 33% from peak to trough before finishing just 1.1% higher. With the next vote on Tuesday 3 November 2026, this article sets out the full record of the stock market after midterm elections – every year, the method, the misses and the drawdowns – so you can judge it for yourself.

For context, the average across every 12-month period starting between 1950 and 2018 was 8.9%, and 74% of those periods ended higher. So the year after a midterm has done better than a typical year. But 19 observations is a small sample, and other forces, such as interest rates, earnings and recessions – were at work in every one of them.

This table shows the S&P 500’s price return 3, 6 and 12 months after each midterm election day, the largest peak-to-trough fall within those 12 months, and who controlled Washington afterwards. The methodology box at the end explains every convention. You can also download the data as a CSV.

Table: S&P 500 after midterm elections, 1950–2022 (price return)

*The NYSE was closed on Election Day, so this is the previous trading day’s close. Returns are price returns and exclude dividends. Source: Pepperstone analysis of Yahoo Finance ^GSPC daily closes; full method below.

Past performance is not a reliable indicator of future results.

A few things stand out. Every 12-month return was positive, ranging from 1.1% (1986) to 33.2% (1954). Over three months, 17 of 19 were positive. The two exceptions were 2002 (–7.8%) and 2018 (–0.9%). And a positive end point often hid a rough path: in 11 of the 19 years, the index fell more than 10% from a peak at some point in the following 12 months.

Search for midterm stock returns and you’ll find several different headline figures. Most aren’t wrong. They answer slightly different questions:

• 95% positive since 1938: The Motley Fool, citing Fidelity research, reports that the S&P 500 was positive one year after the midterms 95% of the time since 1938.¹ On our data, the one miss is 1938 itself (–5.1%). Start in 1950 and there are none.
• 12.4% average since 1900: U.S. Bank Asset Management Group Research puts the average 12-month return at 12.4% across 31 midterms from 1900 to 2025, and 17.5% since 1980.² Its series uses the Dow Jones Industrial Average before 1928 and includes the 1930s, when the S&P 500 fell 36.8% in the year after the 1930 midterm on our data.
• 14.5% in year three since 1950: The Motley Fool also reports a 14.5% average annual return in year three of the presidential cycle.¹ That’s a calendar year, January to December after the vote, not 12 months from election day. Measured our way on calendar years 1951–2023, it’s 17.2%.
• 6.5% three months after, since 1970: Hartford Funds, using Morningstar data, reports an average 6.5% gain in the three months after election day.³ Our figure for the same period is 6.3%. Small gaps like this usually come from date conventions and data providers.

Three choices drive most of the differences: the start year, the window measured, and whether returns include dividends. We use price returns, which leave out dividends. Including them would add a few percentage points a year over this period.

No single cause is proven. These are the explanations analysts most often give, with what our data shows for each.

Uncertainty resolves

Before the vote, investors and traders don’t know who will control Congress or what that means for tax, spending and regulation. Once results are in, that uncertainty narrows, whichever way they go. The idea is that markets price risk, and less uncertainty can mean less risk to price. It’s plausible, but hard to separate from everything else happening at the same time.

Year three of the presidential cycle

The presidential election cycle theory holds that administrations tend to take unpopular steps early in a term and favour growth-friendly policy ahead of the next presidential election. The data fits the shape: across midterm calendar years from 1950 to 2022, the S&P 500 averaged 4.6%, against 9.1% for all calendar years from 1950 to 2019. Year three of the presidential cycle averaged 17.2%. A weak midterm year also means a lower starting point. In 1962, 1966, 1970, 1974, 2002 and 2022, the index had fallen sharply before election day.

Divided government and stocks

A common explanation is that split control of Congress and the White House makes sweeping policy change less likely. In our sample, 14 of the 19 midterms produced divided government. Divided government has coincided with an average 12-month return of 15.1%, against 15.0% for the five years that produced unified government. On these numbers, the outcome made little difference to the average.

Midterm year seasonality: why is September often weak in midterm years?

September has been the S&P 500’s weakest month on average, falling 0.4% across all years from 1950 to 2019. In midterm years it has been slightly weaker, averaging –0.8%, and it was negative in 10 of 19 midterm Septembers, including a 9.3% fall in 2022. Pre-election positioning is one explanation. October to December has been stronger: the S&P 500 rose in 16 of 19 midterm years over that period, by an average of 6.6%, against 3.9% for all years.

A positive 12-month return says nothing about what happened in between. These years show how much depends on the wider backdrop:

• 1986: The S&P 500 was 17.5% higher six months after the vote. Then came the October 1987 crash. From 25 August to 19 October 1987, it fell 33.2%, and the 12-month gain shrank to 1.1%.
• 2010: The index fell 19.4% between April and October 2011, through a US debt-ceiling standoff, a US credit-rating downgrade and the euro-area debt crisis. It finished the window 3.7% higher.
• 2018: It fell 16.4% between 7 November and 24 December 2018, as the Federal Reserve raised rates and US–China trade tensions escalated. It was still 0.9% lower after three months.
• 2002: The midterm came weeks after the low of the dot-com bear market. The index was 7.8% lower after three months and 12.5% below its election-day level by March 2003, ahead of the invasion of Iraq.
• 1974: In a recession, weeks after the low of the 1973–74 bear market, the index fell 13.4% below its election-day level by 6 December 1974 before recovering.

U.S. Bank’s research reached a similar conclusion on reliability: ‘Statistical testing did not find differences large or consistent enough to establish a dependable midterm-election effect.’²

The S&P 500 closed at 3,828.11 on 8 November 2022, then fell 2.1% the day after the vote. The following day it rose 5.5%, after US inflation data for October came in lower than expected – a reminder that economic data can matter more than the election itself. Republicans won the House, in a result not called until more than a week later, and Democrats kept the Senate. Twelve months on, the index was 14.5% higher, although it had fallen 10.3% between late July and late October 2023 as bond yields rose. For a closer look, read our 2022 midterms study.

On 3 November 2026, all 435 seats in the House of Representatives and 35 Senate seats are up for election, including two special elections. History can tell you what has happened after past midterms on average, and how wide the range of outcomes was. It can’t tell you what will happen this time.

Each of the 19 years in our table had its own backdrop of inflation, interest rates, earnings and valuations, and those forces often mattered more than the vote. The 2026 row will be added once the 12-month window is complete, and we’ll update the table as results come in.

EXPERT QUOTE – Chris Weston, Head of Research, Pepperstone

These are among the markets traders and analysts commonly watch around a scheduled market event like the midterms. They’re described here for context, not as recommendations.

• US500: tracks the S&P 500, the index used in this study, which covers 500 large US companies.
• US30: tracks the Dow Jones Industrial Average, a price-weighted index of 30 large US companies.
• NAS100: tracks the Nasdaq 100, which is weighted towards technology and growth companies.
• VIX: the Cboe Volatility Index, which reflects the volatility of the S&P 500 expected by options markets over the next 30 days.
• USD: the US dollar, which can respond to changing expectations for interest rates and government spending.

You can follow scheduled data releases on our economic calendar. And if you’re interested in how perpetual CFDs work around scheduled events, see our guide to perpetual CFDs and market events. CFDs are leveraged products and carry a high risk of losing money rapidly.

• Index: S&P 500 price index (Yahoo Finance ^GSPC), excluding dividends.
• Data source: daily closing prices from Yahoo Finance. 1950–2019 via a public archive of ^GSPC history; 2022–23 retrieved directly. Retrieved 23 October 2026.
• Start point: the S&P 500 close on election day. The NYSE was closed on Election Day from 1950 to 1966, so for those years we use the previous trading day’s close.
• Return windows: the close on the same calendar date 3, 6 and 12 months later, or the last close before it if markets were shut.
• Maximum drawdown: the largest peak-to-trough fall in daily closes between the start point and the 12-month date.
• Control: party control of the House, Senate and presidency after the election. Unified means one party holds all three.
• Comparisons: monthly and calendar-year figures use month-end and year-end closes. ‘All years’ baselines cover 1950–2019.