Why November 3rd matters

The Liquidity Map · Stock market report

Why November 3rd matters

Only a few trading days are left until the US midterm elections on November 3rd.

Since 1990, the drops in this stretch have stayed under 4% on closing prices.

And small caps have already given a good early sign.

October 29th, 2026 · 4 min read

Key takeaways

Since 1990, between October 23rd and election day, the S&P?'s biggest drops on closing prices were 3.4% in 1990 and 3.6% in 2018.
On October 23rd, the S&P, the Nasdaq 100? and the Magnificent 7? made new lows, but the Russell 2000? held above its October 16th low.
Marathon and CleanSpark have lost around 60% since the summer and hit new lows on October 26th.
Anthropic has filed to list on the Nasdaq, and the US press says it could be the biggest listing yet.

Last week I showed what the S&P did after past midterm elections: from the vote to June 30th, up in 21 years out of 21.

Today I'm looking at the days before November 3rd. They're the last ones where I can still buy at a discount before the window that runs into 2027.

On Black Friday, we line up for jeans at half price. In the stock market, when stocks are at half price, we keep our distance.

Before the vote

How deep the pre-vote discounts went

I measured the S&P from the October 23rd close to election day, in the nine midterm years since 1990. On closing prices, the biggest drops were 3.4% in 1990 and 3.6% in 2018.

Counting intraday lows, the deepest point came on October 29th, 2018, 5% below the October 23rd close. That same day, the index closed at -3.6%.

In 8 of 9 years, election day found the S&P higher than on October 23rd.

The ninth year, 1990, closed practically flat, at -0.2%. On average, the index gained 1.4% into the vote, and the best year, 1998, brought 3.8%.

Yes, I know, 3.6% doesn't sound like the apocalypse. That's exactly why it never makes the headlines.

-3.6%The record on closing prices before a vote, in 2018.

The S&P 500 from October 23rd to election day, in every midterm year since 1990.

This year started the same way.

This year

The low came right on October 23rd

The S&P hit $7,845 on October 6th, then fell 3%, down to $7,612 on October 23rd. On Wednesday it closed at $7,744.

If the 2018 drop were to repeat, the S&P would close around $7,347, 3.6% below the October 23rd close of $7,624. That's the worst case in recent history, not a forecast.

Between October 23rd and the vote, recent history left room for small discounts, not for crashes.

When the big drops came, they came after the vote. And 2018 had a specific reason: analysts were cutting next year's earnings.

At the end of September 2018, they expected 10.3% earnings per share growth for 2019. By December 21st, the estimate was down to 7.9%.

Now it's the opposite. For 2027, the estimate rose from 15.4% on September 30th to 16% on October 23rd.

The market pays today for tomorrow's earnings. As long as estimates keep rising, a 2018 repeat has less fuel. It's one more reason backing the window after the vote, into year-end and through June 30th.

-3%How far the S&P 500 fell from its October 6th high to its October 23rd low.

The S&P 500 from September to today, with the levels a 2018-style drop would take it to.

The S&P 500 in fall 2018 and now, versus the fall high. The big 2018 drop came after the vote.

Estimated earnings per share growth for the following year: falling in fall 2018, rising now.

What's more interesting is who didn't go lower.

Small caps

The giants made new lows. Small caps didn't

On October 23rd, the S&P, the Nasdaq 100 and the Magnificent 7 all fell below their October 16th lows. Small caps stopped 1.3% higher. The Russell 2000 now sits 8% below its high, versus 9.7% last week.

Small caps, usually the first to get scared, didn't show up to the panic this time.

One higher low? doesn't make a trend. But it's the kind of sign I want to see before a good window for individual stocks.

+1.3%How far above its October 16th low the Russell 2000 stopped on October 23rd.

The October 23rd low versus the October 16th low, for the big indices and for small caps.

At the other extreme are the miners.

The miners

The miners, 60% cheaper than in the summer

Marathon fell from $16.43 in June to $6.62 on October 26th. CleanSpark, from $19.17 to $7.36. From its 2026 high, the S&P gave up 3% at most. The miners gave up 20 times as much.

If you're wondering who buys a stock that has lost 60% since the summer: hi, that's me.

These companies mine Bitcoin, so they move harder than it does, in both directions. For me they're the high-risk part of the portfolio, not its core.

-60%How much Marathon and CleanSpark lost from their summer highs to their October 26th lows.

How much each one gave up from its 2026 high.

Meanwhile, one company is getting ready to ask the market for more money than anyone before it.

Anthropic

Anthropic is coming to the Nasdaq

The stock market story of the week is Anthropic, the company behind the Claude assistant. On October 26th, it publicly filed to list on the Nasdaq, under the ticker ANT.

The US press says it could raise more than $100 billion. That would top SpaceX's June listing, the biggest so far.

A listing this size needs a lot of buyers, which means a lot of liquidity.

If it comes after the elections, it lands right in the window that has historically been good for stocks. The share price hasn't been set yet, so no, I don't know what it'll cost either.

$16.3BAnthropic's revenue in the first half of 2026, 14 times what it was a year earlier.

Which brings me to positioning.

How we're positioned

What's new this week: until November 3rd, I'm buying more, with the cash I set aside. Last week I was only buying gradually.

I've added Marathon and CleanSpark, alongside Nike and the other individual stocks I'm accumulating?. My trading money stays separate and doesn't go into this.

Conclusions

Before the vote, the discounts were small. After the vote, the window to June 30th ended higher every time. Small caps have given the first sign that sellers are getting tired. I'm using these days to buy, not to guess the bottom.

Why November 3rd matters

Keep reading

JPMorgan moves onto Ethereum while the funds sell

21 out of 21. What comes after the US midterms

Ichimoku flagged the altcoin reversal. 2027 is next