BIG PICTURE中文

Midterms and stocks: what 80 years of history say, and what markets are pricing now

By · AI-assisted, human-reviewed and edited · · Updated · 14 sources · Fact-checked · 13 min read

How do midterm elections affect the stock market?

Historically, midterm elections have been followed by gains: in our calculation, the S&P 500 was higher 12 months after election day in all 20 midterms since 1946, by a median of about 14%, whichever party won. In our 1950–2025 sample, midterm years had the lowest median calendar-year price return of the four-year presidential cycle. This time, prediction markets price Democrats as likely winners of the House and favorites for the Senate, while analysts cited by CNBC say divided government could limit major policy changes. History is a small sample, not a forecast.

Up 12 months later
20 of 20 midterms
Median 12-month gain
+13.7%
Median midterm-year return
+1.1% (n=19)
Democratic House odds
~94% (Nov 2)
Democratic Senate odds
~73% (Nov 2)

Americans vote on Tuesday, Nov 3, to choose all 435 House members and about a third of the Senate. Here is what eight decades of S&P index data (including its 90-stock predecessor before the S&P 500 launched in 1957) say about markets around midterms, and what investors and analysts say they are positioned for this time. Nothing here predicts who wins or where stocks go next.

What history says: gains after the vote, whoever won

The pattern is consistent: in our calculation, stocks were higher a year after every midterm since 1946, whichever party won. We measured price changes in the S&P 500 index, excluding dividends, from the election-day close (or the previous close when markets were shut for the vote) to the last close on or before the three-, six- and 12-calendar-month anniversaries.

  • The index was higher 12 months later in all 20 midterms, by a median of 13.7% (average 14.4%).
  • Six months later it was higher in 19 of 20 (median +15.5%); three months later in 18 of 20 (median +7.2%).
  • Two of the 12-month gains were close to zero: +0.1% after 1946 and +1.1% after 1986. So the often-quoted line that stocks rose in the year after every midterm since World War II holds in our data, but only just in 1946.

The S&P 500 around midterm elections, 1946–2022

The median indexed level was 96.5 thirty trading days before the vote, 100 on election day and 114.5 after 240 trading days (roughly a year). The chart shows selected points, not every move in between.

Median path, election day = 10090100110120−60: 98.2−30: 96.50: 100+30: 101.1+60: 107+90: 109.8+120: 113.4+150: 114.8+180: 116.9+210: 116.7+240: 114.5Share of midterms above the election-day level (%)050100−60: 35%−30: 15%0: 0%+30: 75%+60: 90%+90: 90%+120: 95%+150: 95%+180: 100%+210: 90%+240: 100%−60−300+30+60+90+120+150+180+210+240Median path, election day = 10090100110120−60: 98.2−30: 96.50: 100+30: 101.1+60: 107+90: 109.8+120: 113.4+150: 114.8+180: 116.9+210: 116.7+240: 114.5Share of midterms above the election-day level (%)050100−60: 35%−30: 15%0: 0%+30: 75%+60: 90%+90: 90%+120: 95%+150: 95%+180: 100%+210: 90%+240: 100%−600+60+120+180+240

20 midterms (1946–2022), every 30 trading days; 240 trading days is just under a year. The table also shows the average and the full range. Summary statistics use unrounded returns.

Our calculation from S&P 500 daily closes (Yahoo Finance), price index without dividends.
Show the data table
Trading days from election day Median (election day = 100) Average Lowest Highest Share of midterms above election-day level (%)
−60 98.2 99.2 75.0 117.7 35
−30 96.5 96.9 89.5 109.3 15
0 100.0 100.0 100.0 100.0 0
+30 101.1 101.3 89.5 108.7 75
+60 107.0 106.9 94.0 114.6 90
+90 109.8 109.3 94.7 119.6 90
+120 113.4 112.4 95.1 124.2 95
+150 114.8 113.7 96.7 125.7 95
+180 116.9 116.3 103.4 132.8 100
+210 116.7 115.8 95.5 135.8 90
+240 114.5 115.0 101.1 130.2 100

Every midterm since 1946

Every year was positive 12 months out, but the size of the gain varied widely, from almost nothing to a third.

S&P 500 change in the 12 months after each midterm

Positive after all 20 midterms since 1946, from +0.1% (1946) to +33.2% (1954).

12 months after election day (%)020401946: 0.1%1950: 16.2%1954: 33.2%1958: 11.1%1962: 24.8%1966: 12.9%1970: 12.7%1974: 18.7%1978: 6.4%1982: 19.9%1986: 1.1%1990: 25.1%1994: 27.1%1998: 22%2002: 14.9%2006: 6.7%2010: 3.7%2014: 4.5%2018: 11.7%2022: 14.5%1946195419621970197819861994200220102018202212 months after election day (%)020401946: 0.1%1950: 16.2%1954: 33.2%1958: 11.1%1962: 24.8%1966: 12.9%1970: 12.7%1974: 18.7%1978: 6.4%1982: 19.9%1986: 1.1%1990: 25.1%1994: 27.1%1998: 22%2002: 14.9%2006: 6.7%2010: 3.7%2014: 4.5%2018: 11.7%2022: 14.5%1946195819701982199420062018
Our calculation from S&P 500 daily closes (Yahoo Finance). Party control: House and Senate historical party-division tables.
Show the data table
Year President Lost House Lost Senate Divided after 3 months 6 months 12 months 12 months from the year’s low
1946 Truman (D) Yes Yes Yes +3.8% −3.1% +0.1% +8.1%
1950 Truman (D) No No No +13.6% +16.9% +16.2% +24.2%
1954 Eisenhower (R) Yes Yes Yes +15.2% +19.7% +33.2% +43.9%
1958 Eisenhower (R) No No Yes +6.8% +11.8% +11.1% +37.5%
1962 Kennedy (D) No No No +13.8% +19.2% +24.8% +32.7%
1966 Johnson (D) No No No +8.7% +17.2% +12.9% +32.9%
1970 Nixon (R) No No Yes +14.7% +22.6% +12.7% +43.7%
1974 Ford (R) No No Yes +5.1% +19.9% +18.7% +38.0%
1978 Carter (D) No No No +3.5% +5.5% +6.4% +12.6%
1982 Reagan (R) No No Yes +4.2% +17.9% +19.9% +58.3%
1986 Reagan (R) No Yes Yes +13.6% +17.5% +1.1% +34.6%
1990 G.H.W. Bush (R) No No Yes +14.9% +22.0% +25.1% +29.1%
1994 Clinton (D) Yes Yes Yes +3.3% +12.5% +27.1% +15.1%
1998 Clinton (D) No No Yes +14.5% +21.9% +22.0% +37.4%
2002 G.W. Bush (R) No No No −7.8% +1.2% +14.9% +33.7%
2006 G.W. Bush (R) Yes Yes Yes +4.9% +9.2% +6.7% +23.9%
2010 Obama (D) Yes No Yes +9.3% +14.0% +3.7% +31.0%
2014 Obama (D) No Yes Yes +1.5% +5.1% +4.5% +17.7%
2018 Trump (R) Yes No Yes −0.9% +6.4% +11.7% +37.1%
2022 Biden (D) Yes No Yes +7.6% +8.1% +14.5% +21.6%

Measured from the lowest close of each midterm year, the gains were bigger: a median of about 33% over the next 12 months, positive every time. That figure is often quoted, but it uses hindsight: nobody knows in real time that the low is in.

Midterm years have been the weak spot of the cycle

The year of the midterm itself has been the weakest of the four years in a president’s term, measured by the median. From 1950 to 2025, the median calendar-year price return was 1.1% in midterm years, against 18.9% in the year after (the pre-election year). Midterm years rose in 11 of 19 and fell in 8, including −29.7% in 1974 and −23.4% in 2002.

Median, not every year: S&P 500 calendar-year return by year of the presidential term, 1950–2025

Midterm years had the lowest median return; the following year had the highest. These are medians: down years occurred in every group.

Post-election: median +9.1%, up 12/19Post-election: median +9.1%, up 12/19: 9.1% (Other)9.1%Midterm: median +1.1%, up 11/19Midterm: median +1.1%, up 11/19: 1.1% (Other)1.1%Pre-election: median +18.9%, up 17/19Pre-election: median +18.9%, up 17/19: 18.9% (Other)18.9%Election year: median +11.8%, up 16/19Election year: median +11.8%, up 16/19: 11.8% (Other)11.8%Post-election: median +9.1%, up 12/19Post-election: median +9.1%, up 12/19: 9.1% (Other)9.1%Midterm: median +1.1%, up 11/19Midterm: median +1.1%, up 11/19: 1.1% (Other)1.1%Pre-election: median +18.9%, up 17/19Pre-election: median +18.9%, up 17/19: 18.9% (Other)18.9%Election year: median +11.8%, up 16/19Election year: median +11.8%, up 16/19: 11.8% (Other)11.8%

Individual years varied widely: the S&P 500 fell in 20 of the 76 years from 1950 to 2025 (2011 by less than 0.01%), for example −38.5% in 2008 and −19.4% in 2022. Midterm years fell in 8 of 19.

Our calculation from S&P 500 daily closes (Yahoo Finance), price index without dividends. 19 years in each group.
Show the data table
Year of the presidential term Years (n) Median calendar-year return (%) Average Up years
Post-election: median +9.1%, up 12/19 19 9.1 +8.4% 12 of 19
Midterm: median +1.1%, up 11/19 19 1.1 +4.6% 11 of 19
Pre-election: median +18.9%, up 17/19 19 18.9 +17.2% 17 of 19
Election year: median +11.8%, up 16/19 19 11.8 +8.1% 16 of 19

2026 hasn’t followed that script so far: the S&P 500 was up 10.7% for the year at the Oct 30 close, after a low on Mar 30.

Gains followed every kind of result

Gains followed every kind of outcome. When the president’s party lost at least one chamber of Congress, the median 12-month gain was 6.7% (9 midterms); when it lost none, 16.2% (11). After the vote produced a divided government, the median was 12.7% (15 midterms); a unified one, 14.9% (5).

These groups are small, and a handful of years drive each one, so they can’t show how much the result itself mattered. In this 20-election sample, every 12-month return was positive under every outcome; that doesn’t mean stocks avoided losses along the way, or that future returns will be positive.

Median S&P 500 return in the 12 months after a midterm, by outcome

Stocks rose in the following year under every outcome; the samples are small (n in each label).

All midterms, 1946–2022 (n=20)All midterms, 1946–2022 (n=20): 13.7% (Other)13.7%President’s party lost the House (n=7)President’s party lost the House (n=7): 11.7% (Other)11.7%President’s party lost the Senate (n=6)President’s party lost the Senate (n=6): 5.6% (Other)5.6%Lost at least one chamber (n=9)Lost at least one chamber (n=9): 6.7% (Other)6.7%Lost no chamber (n=11)Lost no chamber (n=11): 16.2% (Other)16.2%Divided government after the vote (n=15)Divided government after the vote (n=15): 12.7% (Other)12.7%Unified government after the vote (n=5)Unified government after the vote (n=5): 14.9% (Other)14.9%All midterms, 1946–2022 (n=20)All midterms, 1946–2022 (n=20): 13.7% (Other)13.7%President’s party lost the House (n=7)President’s party lost the House (n=7): 11.7% (Other)11.7%President’s party lost the Senate (n=6)President’s party lost the Senate (n=6): 5.6% (Other)5.6%Lost at least one chamber (n=9)Lost at least one chamber (n=9): 6.7% (Other)6.7%Lost no chamber (n=11)Lost no chamber (n=11): 16.2% (Other)16.2%Divided government after the vote (n=15)Divided government after the vote (n=15): 12.7% (Other)12.7%Unified government after the vote (n=5)Unified government after the vote (n=5): 14.9% (Other)14.9%
Our calculation from S&P 500 daily closes (Yahoo Finance); outcomes from the House and Senate party-division tables. A midterm can appear in several groups.
Show the data table
Outcome Midterms (n) Median 12-month return (%) Average Rose in 12 months
All midterms, 1946–2022 (n=20) 20 13.7 +14.4% 20 of 20
President’s party lost the House (n=7) 7 11.7 +13.8% 7 of 7
President’s party lost the Senate (n=6) 6 5.6 +12.1% 6 of 6
Lost at least one chamber (n=9) 9 6.7 +11.4% 9 of 9
Lost no chamber (n=11) 11 16.2 +16.8% 11 of 11
Divided government after the vote (n=15) 15 12.7 +14.1% 15 of 15
Unified government after the vote (n=5) 5 14.9 +15.0% 5 of 5

Analysts often add a related statistic about gridlock, when political divisions make laws harder to pass: JPMorgan’s Dubravko Lakos-Bujas wrote that congressional gridlock “has been associated with positive S&P 500 returns since 1950,” with the market up 21% over the two-year term after gridlock-producing elections versus 18% after elections that left one party controlling both chambers of Congress, CNBC reported on Oct 5. Morgan Stanley found that under Republican presidents, the S&P 500 averaged a 23% gain in the year after midterms that split Congress, versus 16% when Democrats won both chambers; its strategist Ariana Salvatore said, “We don’t think that blueprint holds up this time around” (CNBC, Nov 2).

What markets are pricing for Nov 3

Traders and analysts expect Democratic gains; the open question is the Senate.

  • Prediction markets favor Democrats. At 7:30 a.m. ET on Nov 2, Kalshi priced a Democratic House at about 94% and a Democratic Senate at about 73%. On Polymarket, a Democratic sweep of both chambers was at about 72%, a Democratic House with a Republican Senate at about 23%, and a Republican sweep at about 6%. These are traders’ prices, which change quickly, not polls.
  • Most respondents in Wolfe Research’s survey expected a Democratic House. In the survey of 138 investors published Oct 27, Wolfe Research found 96% expected a Democratic House and 58% a Democratic Senate; Wolfe’s own estimate for a Democratic Senate was 49%, CNBC reported. TD Securities said Democratic control of both chambers had become its base case in an Oct 20 report, CNBC reported.
  • The Senate is closer. Fox News’s Oct 29 power rankings left six Senate races as toss-ups that decide the majority, after moving South Carolina into that group.
  • Options imply only a modest reaction. S&P 500 options priced the day after the vote at about 1.2 times a typical day’s swing, roughly in line with past midterms and about half the move usually priced around a presidential election, Man Group calculated (CNBC, Oct 29).

What the recent moves suggest

Man Group said a Democratic sweep would likely help health-care and retail stocks and weigh on investment banks and brokers, and to a lesser extent defense (CNBC, Oct 29). Since the Sep 30 close, exchange-traded funds (ETFs) tracking those groups moved as follows, in price terms to the Oct 30 close: health care (XLV) +9.8% and retail (XRT) +8.0%, while capital markets (KCE) rose 0.1% and aerospace and defense (ITA) 1.2%; the S&P 500 fell 0.9% and technology (XLK) 8.3%. These moves don’t establish how much election expectations affected prices. AP linked the Oct 30 technology sell-off to concerns about AI valuations and spending.

At the same time, many analysts say the market’s preferred outcome is a split government: with President Trump in office for two more years, a Democratic Congress could block his agenda and his veto could block theirs, making big policy swings less likely (CNBC, Oct 30).

What each outcome could mean for policy

Analysts frame the stakes around a few areas markets care about. These are their views, not ours.

Spending and deficits

Piper Sandler estimated, using Congressional Budget Office data, that Democratic proposals “would add $5 trillion to the deficit over the coming decade,” assuming repeal of the tariffs and reversal of cuts to Medicaid, nutrition programs and Affordable Care Act subsidies (CNBC, Oct 30). President Trump could veto such bills. Senate rules generally require 60 votes to overcome a filibuster, which can delay or block a bill; eligible tax and spending bills can instead use budget reconciliation, a process that limits debate and allows passage by a simple majority.

Health care

JPMorgan sees hospitals as the “clearest beneficiaries” if Medicaid cuts are delayed or reversed under a Democratic sweep, and says gridlock also lowers the odds of Medicaid cuts (CNBC, Oct 5).

Tariffs and trade

Tariff repeal is one of the assumptions in Piper Sandler’s deficit estimate. Under a Republican president, a bill changing tariffs could face a veto.

The debt ceiling

Absent congressional action, the Treasury could reach the debt limit between late winter and midsummer 2027, the Bipartisan Policy Center estimated on Jun 4. BPC expected cash and special accounting measures to provide roughly another six to nine months before the Treasury could no longer pay every bill in full and on time. TD Securities says a Democratic Congress would make a debt-ceiling bill slightly more likely to pass than a split one. Raymond James notes Republicans may raise the limit in the lame-duck session, the weeks between the election and the new Congress, if they lose, which would reduce Democrats’ leverage later. Republicans are weighing a year-end increase large enough to cover the rest of Trump’s term, possibly through a budget reconciliation bill, law firm WilmerHale wrote on Oct 14; a separate deadline to fund the government falls on Dec 11.

The Fed and regulation

TD Securities says Democratic control of the Senate, which confirms Federal Reserve governors, would ease concerns about the central bank’s independence. Democratic control would also give the party committee leadership and more scope to compel documents or testimony in investigations, subject to committee rules (CNBC, Oct 30). JPMorgan sees gridlock lowering the odds of AI regulation, and Republican control as the most supportive backdrop for the AI infrastructure build-out.

A slow count

Man Group warns that markets assume the result is known on Nov 4. Close races have taken longer: in 2022, Democratic Senate control was clear on Nov 12, but Georgia’s December runoff set the final 51–49 balance, and in 2020 Senate control waited for Georgia’s January runoffs.

The bottom line

In our sample, the index was higher 12 months after every midterm, whichever party won, but 20 elections is a small sample, and each came with its own economy, interest rates and valuations. This year, traders price a Democratic House as very likely and a Democratic Senate as likely, while many analysts say a divided government would suit markets best. Neither history nor the betting odds tells us what happens next.

What to watch next

Nov 3, evening
Polls close and results come in through the night. Man Group notes that close races have taken weeks to settle before: in 2022, Democratic Senate control was clear on Nov 12, but Georgia's December runoff set the final 51–49 balance.
Nov–Dec (lame-duck session)
Whether the outgoing Congress raises the debt ceiling in the weeks between the election and the new Congress, an option Raymond James says Republicans may take if they lose.
Dec 11
The current stopgap government funding runs out; Congress must replace it. This is separate from the debt limit (WilmerHale).
January 2027
The new Congress convenes; the Constitution sets Jan 3 as the default date unless a law specifies another.
Late winter–midsummer 2027
When the Treasury could reach the debt limit, per the Bipartisan Policy Center. It expects special accounting measures to delay the 'X-date', when the Treasury could no longer pay every bill on time, by roughly six to nine months after that.

Quick questions

Do stocks always go up after midterm elections?

In our calculation the S&P 500 was higher 12 months after all 20 midterms since 1946, but two of those gains were tiny: +0.1% after 1946 and +1.1% after 1986. Twenty elections is a small sample, and the past doesn't guarantee the future.

Does it matter which party wins?

In our 20-election sample, stocks were higher 12 months later whether the president's party lost a chamber or not, and whether government ended up divided or unified. The groups are small (5 to 15 elections), so they can't show how much the outcome itself affected returns.

Why do analysts say Wall Street likes gridlock?

Gridlock means political divisions make laws harder to pass. With the White House and Congress split, big policy changes are less likely, which analysts say reduces uncertainty. JPMorgan noted that gridlock has been associated with positive S&P 500 returns since 1950 (CNBC).

What do prediction markets say about Nov 3?

At 7:30 a.m. ET on Nov 2, Kalshi priced Democrats at about 94% to win the House and about 73% to win the Senate; Polymarket was at about 95% and 73%. Those are traders' prices, which change quickly, not polls or certainties.

Quoting this? You may quote short excerpts under our reuse policy. Please credit WhyMyStockMoved.com and link to this page.