Midterm Election Year 2026 — the cycle's weakest phase, right before the turn?
The midterm election year stock market carries a split reputation for 2026. Historically the midterm year is the weakest and most volatile phase of the U.S. presidential cycle — and it is followed by one of the strongest recovery windows the four-year rhythm has to offer.
We didn't just retell this; we ran the numbers: 56 years of the S&P 500 (^GSPC, since 1970), 15 midterm years. The result lines up with the classic cycle literature — and it pins down exactly where 2026 stands.
What is a midterm election year?
Every four years the U.S. elects a president. Exactly in the middle of the term, in the second year, the midterm elections for Congress take place. 2026 is one of those years: the president elected in November 2024 faces his first big report card at the ballot box in November 2026.
The cycle position is easy to compute: year mod 4. For 2026 that leaves remainder 2 — a midterm year. Yale Hirsch first documented this pattern in the 1960s in the Stock Trader's Almanac; his son Jeff Hirsch continues the work today.
The four positions in the cycle:
| `year mod 4` | Position | Examples |
|---|---|---|
| 1 | Post-Election Year | 2025, 2021 |
| 2 | Midterm Year | 2026, 2022, 2018 |
| 3 | Pre-Election Year | 2027, 2023 |
| 0 | Election Year | 2024, 2020 |
The data: midterm years are the weakest — and the riskiest
For every calendar year since 1970 we computed the largest decline from the yearly high to the yearly low (peak-to-trough drawdown) and the annual return, each on a normalized basis (year start = 100). Sorted by cycle position, the picture is clear:
| Cycle year | Avg annual return | Avg max drawdown | Win rate (positive) |
|---|---|---|---|
| Post-Election | +11.3% | –13.2% | 71% |
| Midterm | +0.9% | –18.0% | 53% |
| Pre-Election | +16.5% | –12.3% | 86% |
| Election | +8.7% | –14.3% | 86% |
The midterm year finishes last on both counts: the lowest average annual return (+0.9%) and the deepest average drawdown (–18.0%, median –16.6%). Barely more than one in two midterm years even closed positive. That matches the reference figures from the cycle literature (Stock Trader's Almanac / Jeff Hirsch), which put the average midterm drawdown near –17%.
The seasonal path makes the pattern visible — it sets the average midterm-year path against the overall average and the current 2026 track:
The chart shows the normalized average path over the past 20 years (each year starts at 100), including the ±1σ band. The familiar pattern is visible: a summer-to-autumn dip with the seasonal low in late summer/early autumn — followed by the year-end rally. In midterm years this autumn dip runs deeper than average.
When does the low come — and how strong the recovery?
The most interesting part for 2026 isn't the drawdown itself but what followed. For each of the 14 completed midterm years we located the exact low and measured how the S&P 500 performed over the 12 months from that low:
| Midterm year | Date of low | Return 12 months later |
|---|---|---|
| 1974 | Oct | +38.0% |
| 1982 | Aug | +58.3% |
| 1990 | Oct | +29.1% |
| 1998 | Aug | +37.9% |
| 2002 | Oct | +33.7% |
| 2010 | Jul | +31.0% |
| 2018 | Dec | +37.1% |
| 2022 | Oct | +21.6% |
Across all 14 cases: an average of +30.7% in the 12 months after the midterm low, median +32.4%. And the striking part: all 14 cases were positive — a 100% hit rate. That aligns with the roughly +31% recovery figure cited by market observers such as Stan Wong (BNN Bloomberg, Aug 27, 2026).
Two drivers sit behind this. First, political uncertainty resolves with the November vote — the market knows where it stands and prices in the relief. Second, the recovery bleeds into the transition to the pre-election year (2027), which at +16.5% average annual return is the strongest cycle phase of all.
Where exactly does the low sit historically?
The lows of midterm years cluster clearly: five of the 15 yearly lows landed in October, with more in August and September. The classic "autumn low" character of the midterm year is data-backed — the turning window typically sits in Q3/Q4.
September: the weakest month, worse in an election year
The calendar month we're in right now is statistically the weak spot of the trading year:
The chart shows the average S&P 500 return by calendar month over 30 years, with the current month, September, highlighted. Across all years September averages –0.69% — the only month with a clearly negative expected value.
In midterm years it gets worse: September averaged –1.95% (14 observations since 1970). It fits the overall picture — the weakest monthly seasonality meets the weakest cycle position. This very window (September/October) is historically where the midterm low forms before the post-midterm rally kicks in.
Where exactly does 2026 stand?
Precision matters here, because this year departs from the textbook timing. The lowest point of 2026 so far didn't come in autumn but back in late March — the S&P 500 was down roughly –9.1% from its yearly high by then. Since then it has recovered and sits around +11% above the year's start as of early September.
In other words: the early spring pullback is worked off, and 2026 is running milder than the average midterm year (–18% drawdown). That's not unusual — 2006 (–7.7%) and 2014 (–7.4%) were mild midterm years too.
Important: the early bottom does not rule out a renewed autumn dip. Right now (early September) the historically weakest calendar window begins, and the November vote still lies ahead. The statistical expectation: elevated volatility into October, then the typical post-midterm relief. What actually happens depends on the economy, the rate path and the election outcome — not the calendar alone.
Limits: a pattern, not a schedule
The cycle effect is a statistical average across decades, not a signal for any single year:
- Large outliers shape the average. 1974 (–37.6%) and 2002 (–33.8%) were driven by external crises — the oil shock and Watergate, and the dot-com aftermath. The midterm year causes no crashes; it statistically coincides with elevated nervousness.
- Small sample. 15 midterm years is statistically modest. The 100% hit rate on the recovery is impressive, but 14 cases are not a law.
- No trading signal. These numbers calibrate an expectation; they don't replace analysis of valuation, macro and politics. Use the cycle as a frame, not a trigger.
Putting it to work on SeasonAlpha
- Open seasonalpha.ai → Yearly Cycle and pick ^GSPC.
- In the sidebar enable the Cycle filter → Midterm Years only. The seasonal path recomputes — you see the deeper autumn dip of midterm years directly.
- Scroll to Drawdown & Risk to overlay the current 2026 path on the historical midterm average.
- The trading-day header in the top right shows the current cycle position ("MidTerm" for 2026).
Conclusion
The midterm election year stock market in 2026 follows one of the most robust patterns in the presidential cycle: the weakest return (+0.9% avg) and the deepest drawdown (–18% avg) — followed by a recovery that was positive in 14 of 14 cases in our data (+31% avg from the low). 2026 already has its spring low behind it and is running milder than average; the seasonally weakest window of September/October still lies ahead. Track it live with the Cycle filter on seasonalpha.ai.
Frequently asked questions
What is a midterm election year for the stock market?
A midterm year is the second year of a U.S. presidency, when the congressional midterm elections take place. For the market it is historically the weakest and most volatile phase of the four-year presidential cycle — with the deepest average drawdown.
How strong was the recovery after the midterm low historically?
In our data the S&P 500 rose an average of +31% (median +32%) in the 12 months after the yearly low of a midterm year — and it did so in all 14 completed cases since 1970. One reason is the resolution of political uncertainty after the November vote, plus the transition into the strong pre-election year.
Is September especially weak in midterm years?
Yes. September is already the only month with a clearly negative expected value in the S&P 500 (avg –0.69% since 1970). In midterm years it deepens to –1.95% on average. This is precisely the window (September/October) where the midterm low tends to form.
Does the pattern mean 2026 automatically rises in Q4?
No. The cycle effect is a statistical average across 15 years, not a signal for any single year. 2026 already saw its lowest level in March and is running milder than the historical average. The expectation of elevated autumn volatility followed by recovery is a frame — not a schedule, and not investment advice.