Is there a correlation between the US voting / election cycle and the stock market? Yes — but it is a weak, contested pattern, not a tradeable edge. Here is what the research actually shows, in plain terms and pictures.
One-line answer: a documented 4-year rhythm exists in the averages, but with only ~24 elections of modern data it is best treated as a seasonal bias / regime tag layered on the market — never a standalone buy/sell signal.
1 · The Presidential Election Cycle Theory
The main framework, from Yale Hirsch’s Stock Trader’s Almanac, claims a four-year rhythm tied to the presidential term:
YEAR 1
Post-election
WEAK ▾
Often soft — the president tackles the tougher campaign promises early.
YEAR 2
Midterm
WEAKEST ▾
Historically the weakest and most volatile year of the four.
YEAR 3
Pre-election
STRONGEST ▴
Historically the strongest year of the entire cycle.
YEAR 4
Election
POSITIVE ▴
Generally up as the economy is pushed ahead of the vote.
weak → weakest → strongest → positive↻ repeats every 4 years
Average S&P 500 return by year of presidential term
Illustrative of the classic cycle shape (Stock Trader’s Almanac framing). Bars are typical historical averages, not a forecast.
Weaker half of cycleStronger half of cycle
2 · Election years skew positive
Looking only at the election year itself, the long-run record leans green:
11%
Avg election-year return (S&P 500, 1928–2016)
83%
Election years that were positive
83%
Final cycle quarter positive
+2.5%
Final quarter avg return
Source (I). Morgan Stanley study of S&P 500 returns, 1928–2016. The final quarter of the cycle has the highest hit-rate of positive returns even though the average size is middle-of-the-road.
3 · The stronger link runs the other way
The more reliable relationship is the market predicting the election — not the election moving the market.
First, the key word: the incumbent is the party (or person) currently in power — the one defending the office. The challenger is whoever is trying to take it. The pattern below is about whether the incumbent keeps power or loses it.
The 3-month rule: since 1928, the direction of the S&P 500 in the three months before the vote has matched the incumbent party’s win/loss about 80% of the time. Stocks up → incumbent tends to win; stocks down → incumbent tends to lose.
How the signal works
Stock market mood in the 3 months before election day, and what it has tended to mean.
Held true ~80% of elections since 1928. Voters reward a rising market and punish a falling one.
Which signal is stronger?
Rough historical hit-rates. The market-as-predictor effect is more robust than the cycle-as-signal.
Source (II). LPL Research / Bloomberg and the Stock Trader’s Almanac (Yale Hirsch), measuring S&P 500 direction in the three months before the vote vs. the incumbent party outcome since 1928.
4 · Party control barely matters
Despite the headlines, which party holds power has had little bearing on returns.
In the 45 years the same party controlled Congress and the presidency, the S&P 500 averaged 7.45% (WSJ) — close to its long-run average. Congressional control was generally not a useful factor for projecting equity performance.
Source (III). The Wall Street Journal, average S&P 500 return across years of unified party control of Congress and the presidency.
5 · The big caveat — mostly coincidence
The consensus among researchers is that the cycle is largely statistical coincidence, not a true signal. The real drivers are economic:
Macro backdrop
Typical outcome
Rising growth + falling inflation
Above-average returns
Falling growth + rising inflation
Below-average returns
Why “weak Year 1” may be an illusion:54% of the 12-month periods following the 24 modern presidential elections overlapped an official US recession. Part of the early-term weakness is just bad timing with the business cycle, not the election itself.
Source (IV). T. Rowe Price and U.S. Bank analyses of post-election 12-month periods and their overlap with official US recessions; growth/inflation regime as the dominant return driver.
6 · What it means for a backtester
With only ~24 elections / ~6 full cycles in modern data, the sample is far too small to trust as a standalone signal. Practical takeaways:
Treat it as a regime tag (cycle-year 1 / 2 / 3 / 4) layered on SPY — not a standalone entry.
It is a known seasonal bias, not a backtestable system with enough trades to be statistically trustworthy.
Let economic regime (growth + inflation) carry the weight; use the cycle only as context.
Do not trade the election cycle on its own. Small sample, contested significance, and confounded by recessions. Do use it as a descriptive overlay alongside stronger, higher-frequency signals.
Sources are cited inline by section, marked Source (I)–(IV). Figures are historical averages and vary by source and date range.
Article rev 2 · last revised 24 Jun 2026knowbase-election-cycle
