Since 1950, the US has held 19 midterm elections, and the S&P 500 was higher twelve months later after every single one. That's a clean record, but the path to those gains was rarely clean. With the 2026 midterms coming on November 3, I'm more interested in what happened during the first three months, and how the result changed the balance of power in Congress.
What the average hides
The 19-for-19 record includes 1974, with Watergate and a brutal bear market in the background, and 2022, when inflation and rate hikes had investors bracing for worse. You'll see this statistic a lot over the next few weeks. It's true, but it doesn't promise a smooth ride after November.
Measured from the close of election week, the S&P 500 gained about 14.5% on average over the following 52 weeks. Start from the election-day close instead and the figure moves closer to 15%. Remove the best and worst cases, and the average still sits around 14%. The result isn't being carried by one or two exceptional years.
The timing tells a different story. From the Friday before the election to the Friday after, the index gained about 1.8% on average. The following week was slightly negative, and the first month averaged just +0.5%. By month three, the average reached +6.3%, and by month six it was +12.6%.
Election week itself was positive in 17 of the 19 cases, but that didn't eliminate the risk of a serious decline later. In 11 of the 19 periods, the index experienced a drawdown of at least 10%, measured from the end of October. Yet all 19 periods finished positive at both six and twelve months.
The uncomfortable part is how much damage could happen inside a winning year. The worst example came after the 1986 midterm. That window included Black Monday, when the S&P 500 lost about 20% in a single day, yet the full period still finished higher.
That's why I find the historical record useful over a year, but much less useful as a guide to the first few weeks. A positive destination tells you very little about the journey.
Congress changes the path
To understand the early differences, I think it's more useful to look at what happened to Congress than which party occupied the White House. The 19 elections fall into three groups based on how much the balance of power changed.
In five elections, the same party kept both chambers and the president's party lost fewer than 20 House seats. In 1998, it actually gained five. Another five saw the president's party lose 20 or more House seats without losing control of either chamber. In the remaining nine, the House or Senate changed hands.
The last group doesn't necessarily mean the president's party lost ground. In 2002, Republicans gained House seats and took back the Senate, which still counts as a flip. Every midterm from 2002 through 2022 changed control of at least one chamber. The last quiet outcome, with the same control and only small changes, was 1998.
The first month is where these groups separate most clearly. Quiet outcomes averaged +4.6% initially and +10.8% after three months. Large House losses without a flip produced the weakest first month at -3.5%. When a chamber flipped, the index was nearly flat at +0.4%, and the three-month average was still only +4.0%.
That difference makes some sense. When the balance of power barely changes, investors have fewer political decisions to reprice. A shift in Congress raises questions about committee leadership, legislation, spending and taxes. Those questions don't necessarily determine the market's direction, but they can complicate the first months.
Consider 2018. Democrats took the House, and over the following seven weeks the S&P 500 slid into Christmas Eve, ending almost 20% below its September peak. The Fed and the trade war mattered more than the election itself, but the vote clearly didn't calm the market. Twelve months later, the index was trading at record highs.
Why the gap closes
Give these outcomes more time and the differences narrow. The group that started with a negative first month was up about 15% after six months. By month twelve, all three groups had average gains between roughly 13% and 18%.
Congress can enable or block tax and spending decisions, but over a full year the market tends to return to its usual drivers: earnings growth, bond yields and liquidity. Political uncertainty still matters. It just competes with much larger forces as the months pass.
The election result may help explain the first few months of market behavior. It does much less to explain the entire year.
There's also a sample problem. Eight of the nine congressional flips happened from 1986 onward, while the other two groups are mostly older elections. Some of the difference could simply reflect a different market era. Five or nine observations per group aren't enough to treat these averages as a reliable trading model.
Where 2026 fits
Republicans currently control the White House, the House and the Senate, but their House majority is only a few seats wide. That makes the historical middle group effectively unavailable this time. Republicans can't lose 20 House seats and still hold the chamber when even a small net loss could flip it.
So I see two realistic categories for November 3. Republicans either retain both chambers and produce a relatively quiet outcome, or at least one chamber changes hands. The more useful question is how much of that second outcome is already reflected in market expectations.
Prediction markets were heavily leaning toward a change in control. As of October 12, the average of Kalshi and Polymarket priced Democrats at about 90% to win the House and 62% to take the Senate. A Republican hold on both chambers was trading around 10%.
At those prices, the most heavily favored result belongs to the historical group with a nearly flat first month and the weakest three- and six-month averages. Even that group finished with a twelve-month average near 13%. It doesn't tell me to be bearish for a year. It tells me not to assume the first quarter will be easy.
And the broader setup isn't typical. Since 1950, midterm years have been the weakest part of the four-year presidential cycle, averaging a gain of about 4.6%. This year, the S&P 500 is already up around 14%, trading near a record a few weeks before the vote, with the 10-year yield above 5%.
The calendar adds another complication. The Fed meets on October 27 and 28, while the Treasury's next quarterly refunding announcement falls on November 4, the day after the election. Markets will have monetary policy, fiscal financing and congressional control to process within a short window.
What this means for crypto
For crypto, I separate the implications into two levels. The first is macro. Bitcoin trades much more like a macro asset than it did a few cycles ago, with a significant share of demand coming through ETFs and corporate treasuries. Whatever happens to yields, liquidity and risk appetite around the election can reach Bitcoin through those channels.
The second is regulation, and that could matter for much longer. The CLARITY Act, the market structure bill the industry has been waiting on, failed its Senate cloture vote 49 to 50 on September 15. It was eleven votes short of the 60 required.
Anything that doesn't pass before the new Congress is sworn in on January 3 has to start over. Senator Lummis has already warned that if this Congress misses the opportunity, the next realistic window might not come until 2030.
A Democratic House would also control the House committees responsible for writing these bills. That makes congressional control directly relevant to the regulatory process, not just indirectly relevant through broader market sentiment.
For stocks, the most visible historical effect looks concentrated in the first few months. For crypto, the market structure question could stretch across the next two years. Those are different timelines, and I wouldn't treat them as the same trade.
How I'm reading it
The 19-for-19 statistic makes the post-midterm year sound much cleaner than it actually was. What interests me is the first three months after November 3, because that's where congressional outcomes produced the clearest differences. After that, earnings, yields and liquidity become harder to separate from whatever happened at the ballot box.
I'm not treating November 3 as a standalone trade. I'll watch which chambers change hands, how the market responds through the first month, and whether the Fed and Treasury announcements shift expectations for yields and liquidity. With the 10-year above 5%, those reactions could matter as much as the election result itself.
The historical record gives me a framework, not an entry signal. If the early market reaction follows the congressional pattern, that's worth watching. If macro conditions overwhelm it, I would give those conditions more weight than a sample of 19 elections.