A Recession Model That's Never Been Wrong Just Hit 49%. That Was Before the Iran War.

The first few months of 2026 have been difficult for investors. The S&P 500 (SNPINDEX: ^GSPC) is down 4.6% (it was down nearly 8% before a partial recovery this week), while the tech-heavy Nasdaq Composite has fallen further, down 7.1% (it was down more than 11% before a partial recovery).

There's a reason to believe the losses could get worse.

Moody's just released its latest recession odds: 49%. That's just one percentage point below the threshold rating that has correctly indicated a recession every time in 80 years of backtested data. Even worse, that rating came out before the start of the war in Iran.

Image source: Getty Images.

Economic warning signs were flashing before the Iran War

The report from Moody's is something new: The model was introduced in 2025 as an artificial intelligence system (AI) trained on 80 years of economic data. It has a perfect track record of determining prior recessions because it was designed to figure out what caused the past ones -- that's how training a model works. It was designed to identify the conditions that preceded past recessions, and continuously adjusted until it best fit real data perfectly.

There have been no instances yet where crossing the 50% line didn't result in recession, but just because it crosses that threshold, it's by no means a guarantee of a recession. But it should be concerning nonetheless.

What's driving Moody's recession odds

Mark Zandi, the Moody's economist behind the model, told Euronews that deteriorating labor market figures are the primary driver of the high reading. But he emphasized that virtually every major economic indicator has weakened since late last year.

The most recent numbers are certainly concerning: 92,000 jobs lost in February, unemployment ticking up, and gross domestic product (GDP) at a paltry 0.7%, all while inflation continues to stay above its 2% target.

$100-plus oil price is likely to push recession risk even higher

But Moody's 49% figure was calculated before the U.S.-Iran conflict shut down roughly a fifth of global crude production and sent oil rocketing above $100 a barrel. As Zandi admits, there is good reason to believe the odds will cross the 50% line.

Energy costs are a critical part of the model. Every recession since World War II was preceded by a spike in fuel prices, save for the brief recession in the wake of COVID-19.

Don't panic, but don't ignore this either

No model is perfect, and there is a reason 50% odds are not 100% even if backtesting shows a perfect record. But I am concerned. I think we're very likely headed for a recession and a serious market downturn.

Still, panic selling is almost certainly the wrong move. Timing the market is exceptionally difficult, and for long-term investors, time is your friend. The market has always recovered and set new highs following a bear market, no matter how deep.

Johnny Rice has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Moody's. The Motley Fool has a disclosure policy.