We are just three months away from the U.S. midterm elections. The Federal Reserve is debating whether its next move may be an interest-rate hike, conflict continues in the Middle East, and investors are questioning whether artificial intelligence companies can deliver on elevated expectations. And the S&P 500 has reached another all-time high.
From Wall Street’s perspective, conditions appear remarkably strong. The blended earnings growth rate for the S&P 500 is tracking at 47.4% in the second quarter! If that pace holds, it will mark the strongest year-over-year earnings growth since the second quarter of 2021, with Alphabet and Amazon leading the gains.
Although the headline number is flattered by unusually large gains at those two companies, even when excluded, S&P 500 earnings would still be growing by 28.8%. The second consecutive quarter of earnings growth above 20%.
Figure 1. S&P 500 Earnings-Per-Share Growth (Year-on-Year Percent Change)
Sources: Factset and Potomac. Data as of July 31, 2026. All forecasts are expressions of opinion and are subject to change without notice and are not intended to be a guarantee.
On the surface, corporate America is thriving, and the stock market appears to be in very good shape. But elections are not decided by earnings estimates or stock-market records. They are decided by voters.
So, as the midterms approach, what should investors be watching to judge the state of the union? The answer may ultimately come down to just four Senate seats.
Start With the Maps
Although we would all like to believe that our vote is the only thing that matters, the electoral landscape matters as well.
Thirty-five Senate seats are being contested in 2026, including 22 currently held by Republicans and 13 held by Democrats. Republicans currently control the Senate with 53 seats, while Democrats hold 47 seats when the two independents who caucus with them are included. Because the vice president would break a 50-to-50 tie in favor of Republicans, Democrats need a net gain of four seats to take control.
At first glance, having 22 Republican-held seats on the ballot compared with only 13 Democratic-held seats would appear to favor Democrats. However, most of those races are not competitive and the Senate has also historically been less susceptible to large midterm swings than the House.
Since 1950, the president’s party has lost an average of just three Senate seats during midterm elections, compared with an average loss of roughly 24 seats in the House.
Figure 2. Average Midterm Seat Losses by the President’s Party Since 1950
Sources: Brookings Institution and Potomac. Data as of 2022.
To win control of the Senate, Democrats will likely need to hold Michigan and Georgia and then flip four Republican seats. The key races to watch are in North Carolina, Maine, Ohio, and Alaska.
Republicans, meanwhile, can probably maintain control by winning just one of those four races, assuming the less competitive races go as expected.
This leaves Democrats with a narrow path. Betting markets currently put the odds of Republicans maintaining control at about 54%.
Interestingly, those odds have moved closely with oil and retail gasoline prices. That relationship may help explain the administration’s growing urgency to reopen the Strait of Hormuz with Treasury Secretary Bessent moving financial markets this week by suggesting that a deal restoring freedom of movement through the Strait may be near.
Figure 3. Republicans Senate Control Odds vs. Oil Prices
Sources: Polymarket, Bloomberg L.P., and and Potomac. Data as of August 4, 2026. Note: Oil price is the spot price of West Texas Intermediate crude oil All forecasts are expressions of opinion and are subject to change without notice and are not intended to be a guarantee.
Looking at the House, Republicans currently control 218 seats, while Democrats hold 212. An independent holds one seat and generally caucuses with Republicans, and four seats are vacant. All 435 seats are up for reelection because House members serve two-year terms.
Republicans need to win at least 218 seats to maintain control. Democrats need a net gain of just three seats from their 2024 total to reach the 218 seats required for control.
Of the four vacant seats, two of the races are solely between Democrats, in California and Georgia. The race in Florida is seen as favoring the Democrat, while the race in Texas is seen as favoring Republicans.
As Figure 2 shows, outside of 1998 and 2002, the president’s party has always lost more than three House seats in a midterm election since 1950. In fact, the president’s party has gained House seats only twice since World War II.
That history helps explain why betting markets heavily favor Democrats to retake control of the House.
As it stands now, betting markets place the odds of Democrats controlling at least one chamber at about 87%, with the odds of a full Democratic sweep at approximately 46%.
Figure 4. Democrat House of Representatives Control Odds (%)
Sources: Polymarket, Bloomberg L.P., and and Potomac. Data as of August 4, 2026. All forecasts are expressions of opinion and are subject to change without notice and are not intended to be a guarantee.
There are still three months to go, which is forever in politics. But if those odds prove accurate, the final two years of Trump’s presidency could look very different.
A Democratic House would largely end Trump’s legislative agenda and subject the administration to additional oversight and investigations. House Democrats could also impeach the president, although impeachment alone would not remove him from office.
If Democrats also win the Senate, they would gain the ability to delay or block judicial, cabinet and other executive-branch nominees. They could also pass legislation seeking to restrict presidential tariff authority, military operations and other executive powers.
However, their authority would still have limits as removing President Trump would require a two-thirds vote in the Senate and legislation would be subject to Trump’s veto.
In other words, Democratic control of Congress would not give Democrats full control of the government. It would, however, replace unified Republican control with increased oversight and legislative gridlock.
Figure 5. What Democratic Control of Congress Could Mean
Sources: U.S. Constitution, Congressional Research Service, U.S. Senate, and Potomac. Data as of August 4, 2026.
The election may therefore come down to two different numbers: three net House seats could end unified Republican control of Washington and four Senate seats could determine whether Democrats sweep Congress entirely.
Three months. Four seats. And potentially a very different Washington.
Market Implications
Historically, midterm election years have produced somewhat below-average returns ahead of Election Day, followed by a boost once the uncertainty clears. Presidential election years often show a similar pattern.
This time, however, we suspect the artificial intelligence investment cycle will remain in the driver’s seat.
Political headlines are likely to get louder as the election approaches. At the same time, corporate earnings remain robust and continue to rise. That should remain supportive of risk assets, although increased volatility would not be surprising given the ongoing push and pull between the Federal Reserve, the conflict with Iran, the election, and the AI investment narrative.
Investing often comes down to climbing the wall of worry and focusing on data over feelings. In that vein, it may help to remember that the average one-year return for the S&P 500 following midterm elections since 1950 has been an impressive 18.8%.
There has also never been a negative return one year after a midterm election over that period. The weakest return occurred during Ronald Reagan’s second term, a period that included the 22.6% one-day decline on Black Monday.
Figure 6. S&P 500 Return One-Year After Mid-Term Elections Since 1950
Sources: Standard and Poor’s and Potomac. Data as of November 2023. Past performance is no guarantee of future results. It is not possible to invest directly in an index.
Politics will dominate the headlines over the next three months. Four Senate seats may determine whether Democrats can complete a sweep, while just three House seats could end unified Republican control.
That would matter for policy, oversight, regulation, and near-term volatility. But unless the election changes the earnings path or the AI investment cycle, it is unlikely to replace them as the market’s primary drivers.
The headlines will get louder, but the data still matter more.
Remember: Data over feelings.
