The last weekend of August, the Little River in Townsend is wall to wall inner tubes. You can’t find a place to park at the Wye. The rental line runs out into the gravel, and 321 backs up past the fire hall.
Ten days later it’s over. The tubes are stacked and chained. The water is the same temperature it was on Labor Day. The rocks haven’t moved. Nothing about the river changed.
You can just hear it again.
That’s my favorite week of the year in Blount County. And about six hundred miles northeast, the opposite thing happens on exactly the same calendar.
Wall Street spends August somewhere else – the Hamptons, the Cape, the Jersey shore. Desks run thin. Volume dries up. The market drifts along on light participation and whoever drew the short straw and had to stay in the office. Then Labor Day passes, the senior people come back to Manhattan, and the volume comes back with them.
So does the volatility.
Since 1950, September is the worst month of the year for all three major indexes: the Dow averages a 0.8% loss, the S&P 500 and NASDAQ about half a percent each. Over the last ten years the S&P’s September has run worse than that, averaging roughly a 2% decline. And Bank of America’s work adds a wrinkle worth knowing: the second half of September is consistently weaker than the first. The trouble doesn’t show up the Tuesday after Labor Day. It shows up about the time everyone has finished catching up on email.
One more stat fits this year uncomfortably well. When the S&P 500 sets a record high in August, September finishes lower about 61% of the time. We just did that. The index closed Friday at 7,674 after a losing week driven by bond market declines, with the long end snapping right back up after Treasury’s expanded buyback program failed to hold yields down. Secretary Bessent said he wanted to signal that yields don’t reflect the underlying fundamentals. The bond market listened politely and did what it wanted.
Now layer the election calendar on top of it.
2026 is a midterm year, and midterm years are the ugly stepchild of the four-year presidential cycle. September in a midterm year runs worse than a normal September. The Dow averages a 1% loss, the Russell 1000 the same. More to the point, the average intra-year peak-to-trough drawdown in a midterm year has historically run near 19%. In 11 of the last 16 midterm cycles, the S&P 500 gave up 15% or more at some point during the year.
That’s the bad news but there is always the other side of the trade to consider.
Almost 90% of midterm years have still finished the full year positive. And Jeff Hirsch, whose Stock Trader’s Almanac work is the best in the business on this subject, identified what he calls the sweet spot of the whole four-year cycle: the fourth quarter of the midterm year through the first quarter of the following year. That stretch has averaged 16.0% for the S&P 500, 15.3% for the Dow, and 23.3% for NASDAQ.
What are the lessons?
First, seasonality is real but it’s still not a trading signal. Hirsch’s Best Six Months research has November through April averaging a 7.3% Dow gain since 1950 against 0.8% for May through October . The S&P 500 finished September higher in eight of the last ten years anyway. A pattern that strong on average and that unreliable in any given year describes weather, not a forecast. I’ve watched people sell in August and then spend the whole fall arguing with themselves about when to get back in. That isn’t risk management. That’s tuition.
Second, most of what’s coming is a bond story rather than a stock story. When yields move the way they did last week, someone holding a bond fund watches the share price and feels every bit of it. Someone holding individual bonds with real maturity dates gets paid on schedule regardless of what the screen says in the meantime. That’s why our large fixed income portfolios are in individual investment-grade bonds instead of inside a fund.
The river in Townsend in October is the same river it was in July. Same water, same rocks, same current. All that changed is the crowd standing in it. September works about the same way. It doesn’t make companies worth less. It just fills the room back up with people who have opinions, screens, and a Q4 number to hit.
I’ll take the empty stretch of river. And I’ll take a portfolio that doesn’t much care who’s back at their desk.
At LeConte, Purpose-Built Planning means the hard decisions get made on a calm Tuesday in August, in writing, tied to what your money is actually for. When the volume comes back in the fall, there’s nothing left to decide. If the next few months are going to be loud, that’s a conversation worth having before they are.
