The market is at record highs. Since the start of the year we have already seen 26 new all-time highs for the S&P 500, and the index crossed 7,700 points for the first time in history.

In such euphoria, however, it pays to look at the other side – what the market might face heading into the November midterm elections.

What do the numbers say?

History is not entirely on our side. The S&P 500 has existed since 1957, and since then 17 midterm elections have taken place in the US.

In 12 of the 17 midterm years, the index entered a correction, that is, at least 10% below its high. That is roughly 70% of cases.

In 6 years the decline was even deeper – 20% or more, meaning roughly a 35% probability of a bear market.

The average peak-to-trough decline in midterm years was about 18%, significantly larger than in other years of the presidential cycle.

And why such declines? The market simply prices in uncertainty around future fiscal and regulatory policy in advance.

But it’s not just about the elections.

According to Bank of America data going back to 1928, August through October is historically the weakest three-month period of the year. When this period ended in the red, the average correction was about 7%.

In midterm years, seasonality is even weaker. According to TIAA, from 1960, the S&P 500 has averaged a drop of 2.6% from May to October in midterm years, while in other years over the same period it averaged a gain of 3.1%.

And then there is September – historically the weakest month of the year. So purely based on seasonality, the market is holding some of the worst cards for the coming months out of the whole year.

But this year could be different.

The biggest argument against this scenario are the companies themselves.

Earnings of companies in the S&P 500 grew about 29% year-on-year in the first quarter, and roughly 24% growth is expected in the second quarter.

And above all, AI infrastructure continues to see massive investments. TSMC grew revenue by 36% and profit by 77% in the second quarter. It also raised its capex outlook for 2026 from $52–56bn to $60–64bn. ASML beat estimates and also raised its outlook.

So results are already visible today among chipmakers, equipment manufacturers, memory makers, or data infrastructure. Moreover, the market has already shown it can bounce back sharply. The Nasdaq fell nearly 10% from its high in late July, but erased most of the loss by early August.

But what may be even more important is what comes after the elections.

Historically, the post-midterm period is one of the strongest parts of the presidential cycle. Since 1948, the S&P 500 has risen by an average of 13.9% over the 12 months following midterm elections, versus 5.8% in comparable periods of other years.

And looking at the bounce from a correction trough, since 1962 the index has rebounded by an average of 31% over the following 12 months – and that happened in all 16 cycles observed.

In other words: the pre-midterm period is historically unpleasant, but the subsequent recovery ranks among the strongest parts of the cycle.

My view

These numbers are not a signal for me to sell and wait on the sidelines, but rather a reason to be prepared.

Keep some capital on the side for potential dips, don’t add unnecessary leveraged risk now, and focus on quality companies I want to own even in the face of a 20% drop.

How are you preparing for autumn – will you buy the dips, or do you prefer to wait on the sidelines? 👇