For most of 2026, a recession was said to be on the way. It kept not arriving. This is the story of a number that chased the news all year, and of what it cost anyone who obeyed it.
Four times, The Wall Street Journal asked dozens of economists the same question: what are the odds of a recession within 12 months? Here are all four answers, with the papers they ran in.
| Survey | Recession odds | Where it ran |
|---|---|---|
| January 2026 | 27% | cited in the Mar. 20 and Apr. 13 papers |
| March 2026 | 32% | Mar. 20 paper, 50 economists |
| April 2026 (conducted Apr. 3–9) | 33% | Apr. 13 paper |
| July 2026 | 25% | Jul. 13 paper — lowest since the start of 2025 |
What the number actually said
Start with an honest observation. The economists never predicted a recession. Even at the April peak of 33%, no recession was still the two-to-one favorite.
The real problem is which way the number moved, and when. It rose through the spring — 27, then 32, then 33 — precisely while the danger was fading. The March 20 paper’s own headline said economists doubted the war would spark a slump, even as their number climbed.
In the April 16 paper they allowed that persistently high oil could push the odds above 50%. Oil never obliged.
Then the number fell to 25% in July, after the all-clear. The July 13 paper summed the year up: the war with Iran did not hurt the U.S. nearly as much as economists feared. The forecast trailed the headlines like a caboose. It told you where the economy had been, never where it was going.
The economy, meanwhile, did this. The 2026 growth forecast rose to 2.1% in the July survey. Unemployment fell to 4.1% by the August 8 paper, from 4.3% in the February 12 paper. David Berson of Cumberland Advisors put it plainly in the July 13 paper: “Who would have thought this, with a trade war and an oil-price shock?”
Credit where due. Brian Mulberry of Zacks said in the April 13 paper — the gloomiest reading of the year — that there was still a lot more good than bad. He was right, on the day it was hardest to say.
Also worth knowing: this was a rerun. The January 24, 2025 paper recalled that economists once put recession odds above 60%, two years earlier, and that downturn never materialized either.
What obeying the forecast cost
Suppose you sold stocks whenever the number went up. The arithmetic below uses the S&P 500 levels printed beside each survey and the 7,730.99 close in the August 28 paper.
| If you sold on | S&P 500 then | Gain missed by Aug. 27 |
|---|---|---|
| The Mar. 20 paper’s 32% reading | 6,606.49 | +17.0% |
| The Apr. 13 paper’s 33% reading | 6,816.89 | +13.4% |
| The Jan. 27 record high | 6,978.60 | +10.8% |
The round trip punished anyone who blinked. Robert Pozen noted in the May 22 paper that the S&P 500 was down 7.33% for the year as of March 30, then back to plus 5.62% by May 1. The bottom lasted roughly a month.
Miss that month and you missed the year. The index clinched its 26th record close of 2026 on August 7, per the August 8 paper, and set an all-time high of 7,798.99 on August 13.
The cash math nobody enjoys
Spencer Jakab’s May 29 column supplied the long-run version. About $5.6 trillion, roughly 10% of Americans’ liquid wealth, sits in low-yielding bank deposits.
Fidelity once priced the cost of that comfort. Invest $5,000 a year in U.S. stocks from 1980 through 2023 with perfect timing every year, and you end with nearly $5.6 million. With the worst possible timing every year: $4.3 million. Leave it all in cash instead: $350,000.
Read that middle number again. The unluckiest stock buyer in the experiment beat the cash holder by roughly $4 million. Jakab added a sharper cut: holding 25% in cash while timing the rest perfectly still leaves you behind the all-in buyer with the worst timing.
Two habits explain the damage. Money-market funds take in their biggest flows right after selloffs, which is historically when stocks are set up to run. And Morningstar’s Mind the Gap study found fund investors earned 1.2 percentage points less than their own funds over a decade, mostly by moving at bad moments.
Jakab’s fair caveat belongs here too. Cash has a real job: emergencies, and money you will spend soon. As a reaction to a 33% survey reading, it is merely expensive. The best five-year CDs in the August 26 paper paid 4.35% to 4.50%. The S&P 500, in the August 28 paper, was up 12.9% on the year.