Recession in 2027?

The core question is not whether risks exist, but whether those risks are large enough to translate into an actual recession during calendar 2027. On the facts provided, the answer is plausible but not the central forecast. Most baseline outlooks still point to positive growth in 2027, including the ECB, Oxford Economics, KPMG, and Japan-focused forecasts, which collectively argue that the economy can slow materially without crossing the recession threshold. That keeps the starting point below 50%, because a recession needs a concrete shock or policy error rather than just elevated fragility.

That said, the downside case is real and unusually concentrated. The sources repeatedly flag a narrow set of triggers that could propagate quickly through the economy: an unwind of AI-related valuations and capital spending, an energy or geopolitical shock, or tighter financial conditions if inflation stays sticky and central banks do not ease enough. Fitch’s stress scenario is especially relevant because it shows how a severe AI correction could push the U.S. into recession with materially negative quarterly growth in 2027, even if that is explicitly not the base case. In other words, the risk is less about a slow drift into recession and more about a sharp discontinuity triggered by crowded markets.

Relative to the market price of 22% yes, I think the market is somewhat too optimistic. A 22% implied probability feels low given the number of credible downside channels and the fact that 2027 sits far enough out for one or two macro shocks to matter. Still, the market is not wildly wrong: the baseline remains growth, and recession forecasts are not the consensus. My 34% view reflects a meaningful premium over the market for the combination of late-cycle vulnerability, possible AI-asset repricing, and the historical tendency for recessions to arrive after policymakers and forecasters have already become comfortable with a soft-landing narrative.