Every four years the calendar hands investors the same quarter: the one where the country is about to tell Washington what it can't do. Spencer Jakab's Heard on the Street column Friday runs the numbers — since 1950, the fourth quarter of a midterm year has returned 6.6% on average for the S&P 500, a 29% annualized clip and the tallest bar in the paper's chart — and then does the useful thing, which is to ask whether the folk explanation is even right.
The folk explanation is gridlock. A sitting president's allies have lost seats in nearly every midterm of the past century, markets like a hobbled Washington, so stocks rise. Jakab's reading is that certainty beats paralysis. He points to a study published five years ago in the Journal of Financial Economics covering 145 years of data: the U.S. equity premium — the return over risk-free assets — averaged an annualized 15.4% in the five months after a midterm, against 3% in all the other months. The gains mostly come after the vote, when the outcome is known, whatever the outcome is. Treasurys, he notes, have done worse than usual after midterms, possibly because the flight to safety ahead of the vote unwinds once it's over.
His counter-example is 2002, the last time a president's party gained control of Congress in a midterm: no gridlock arrived, and the S&P 500 still rose 7.9% that quarter, climbing out of a bear market. The effect can also bleed forward into the weeks before an election when the result looks settled, and Jakab says this year looks like one of those — citing Polymarket's implied odds on Wednesday of 93% for a change of majority in the House and 65% in the Senate. Those are the column's numbers, and they are crowd prices: what bettors will pay, not what will happen. We take no view on either party and none on the odds. His closing line is the one worth keeping: what matters isn't so much who holds power as “knowing who's in power.”
An average is not a promise
What a 6.6% average doesn't tell you is the spread around it. Jakab's own 2002 example was a rebound from a sharp bear market; this year's fourth quarter has already produced a record close — 7,818.93 on Tuesday, Oct. 6 — and Thursday's 7,765.36 came with 268 NYSE stocks at new 52-week lows against 37 new highs. A seasonal that averages well across decades can land anywhere in a single year. And if this is one of the years when the effect leaks forward, as Jakab suggests, some of it may already be in the price.
Our read
Investments/Risk (IN04), and the behavioral trap first. The pattern in this column is exactly the kind of thing that tempts a reader to front-run it — the fourth quarter's good, so buy now. The study Jakab cites points the other way: the measurable edge arrives after Election Day, and the pre-election stretch is where the safety trade crowds in before it unwinds. His own caveat cuts the same direction — in a year when the result looks settled, part of the gain leaks into the weeks before the vote, so some of it may already be paid for. The desk's standing plan is a purchase the week of the Nov. 3 vote, a date written down in September, and it stays there; the overlap with the study is coincidence, not foresight. We said in print this week that our fall caution was wrong on timing when the record arrived late, and a seasonal that happens to agree with our calendar doesn't make the calendar right. The September letter's conditions for new money still fail two of three. Nothing is bought on a pattern.
The planning point is dates versus hunches. If you've been waiting for after the election to add, write the date down now, with the amount, and let the calendar do the deciding — a Polymarket figure or a chart with four bars is not a buy signal, it's weather. Treasurys' post-midterm softness is one more reason long duration stays avoided here, though we would hold that view without the study. And if you have no written plan with dates on it, that's the review to book — before the vote, not the morning after.