# Midterm election year stock market performance: dated observations and limits

Sources checked 2026-09-29.

## What the annual claim measures

The claim that the S&P 500 was higher a year after every midterm since 1950 describes an election-to-one-year-later window, not the next session. LPL Research reported that pattern in its 2022 midyear discussion, with an average gain of 14.5%. At that publication date the completed 1950 through 2018 sample contained 18 midterms (n = 18, counted every four years). It did not yet include the year following the November 2022 election. A positive frequency in this selected sample is not a probability for 2026.

[LPL Research, 2022 midyear discussion](https://www.lpl.com/join-lpl/why-choose-lpl/news-and-insights/lpl-myo-2022-navigating-turbulence-offers-financial-forecasts.html)

## Keep the baseline and sample attached

U.S. Bank's October 3, 2022 note reports a 16.3% average over the 12 months following 15 midterms since 1962 (1962 through 2018, n = 15). It compares a -1.1% average in the preceding 12 months (n = 16) with an 8.0% average for all 12-month periods starting October 31 since 1961. That is a different sample from LPL's 1950 start. Neither annual window measures the first session after voting. These are source-reported historical returns, not SquawkFlow price calculations.

[U.S. Bank, October 3, 2022](https://www.usbank.com/dam/documents/pdf/wealth-management/midterm-election-update-october-10-3-2022.pdf)

## Two more windows, not two more confirmations

Carson's 2022 bear-market discussion reports a 14.1% average for the year after midterms since World War II; its text does not state the observation count, so no count is inferred here. Separately, AAII's May 10, 2018 report attributes a 1.1% average decline in May through October of midterm years to CFRA's Sam Stovall, using data starting in 1946. The same report cites LPL's 1.5% average gain for all May-through-October periods from 1950 through 2017. Those start dates and pre-election windows differ from the post-election claim. They must not be pooled into one average or treated as independent replications.

[Carson, 2022 bear-market questions](https://www.carsongroup.com/insights/blog/10-answers-to-questions-about-the-bear-market/)

[CFRA findings reported by AAII, May 10, 2018](https://www.aaii.com/investor-update/article/Midterm-Elections-and-the-Worst-Six-Months)

## 2010: policy on the first post-election day

The midterm was November 2, 2010. On November 3 the Federal Reserve announced the additional Treasury-purchase program commonly called QE2. The first post-election session therefore also contained a monetary-policy announcement. A close-to-close change on that date cannot establish an election-only effect.

[Federal Reserve Bank of New York, QE2 program record](https://www.newyorkfed.org/markets/fast_facts_lsap.html)

## 2018: the FOMC met in the same week

The November 6, 2018 midterm was followed by the November 7 to 8 FOMC meeting. The first post-election session and the policy-decision session are separate dates within that week. A weekly return includes both windows and other market information.

[Federal Reserve, November 7 to 8, 2018 minutes](https://www.federalreserve.gov/monetarypolicy/fomcminutes20181108.htm)

## 2022: CPI and FTX are separate dated events

The November 8, 2022 midterm was followed by the November 9 session. The October CPI report was released November 10, and FTX's bankruptcy petition date was November 11. The FTX crisis and the inflation release belong in the surrounding-week context; neither justifies naming the whole interval an election effect. A next-session return and a return ending after CPI answer different questions.

[BLS, October CPI released November 10, 2022](https://www.bls.gov/news.release/archives/cpi_11102022.pdf)

[FTX Trading Ltd., Kroll Chapter 11 case record (22-11068)](https://restructuring.ra.kroll.com/FTX/)

## What this cannot tell you

A historical pattern is not a forecast. Small, selected samples, changes in market structure, different return definitions and overlapping economic events limit comparisons. None of these observations identifies what caused a return or says what any market will do after November 3, 2026. This page reproduces no price history, chart or Yahoo-derived table. The Election Center carries the separate dated options measurement and its own limits.

[Election Center](https://squawkflow.com/elections/2026)