Recession Probability Sits at 40% to 50%—Here Is How to Size That Risk

Recessions are famously hard to predict. Economists have called nine of the last four, as the joke goes, and the persistent failure of recession forecasts over the past several years — particularly in 2022 and 2023, when virtually every major forecaster saw contraction as near-certain — has made many investors appropriately skeptical of recession calls. But there's a meaningful difference between predicting a recession and sizing the risk of one in a portfolio. The former requires a forecast. The latter requires a probability and a framework for acting on it without requiring certainty.

---

Right now, the major Wall Street banks have recession probability estimates clustered in the 40% to 50% range, up sharply from below 20% at the start of the year. The Atlanta Fed's GDPNow model entered negative territory in late Q1 2026 for the first time since the pandemic. That doesn't make a recession inevitable — or even the base case. But it does mean the tail risk has moved significantly closer to the center of the distribution, and portfolios built for a soft landing deserve to be pressure-tested against a harder outcome.

What the Leading Indicators Are Actually Saying

Reading recession risk correctly requires watching a dashboard of indicators rather than any single metric, because no single indicator is reliable enough to carry the weight. Here is what the current picture looks like across the instruments that have historically had the most predictive value.

The yield curve, specifically the spread between 2-year and 10-year Treasury yields, has spent much of the past two years inverted or flat — a configuration that has preceded every US recession of the past 50 years, with variable lead times typically ranging from 12 to 24 months. The inversion has been resolving, which some analysts interpret optimistically as a return to normalcy. But historically, the recession signal tends to arrive not during the inversion but after it steepens, as short rates fall in anticipation of Fed cuts. That resolution is occurring now.

Credit spreads have widened from historically compressed levels. The spread between high-yield corporate debt and Treasuries moved from roughly 2.65% to over 3% in Q1 2026. That's not yet at recessionary levels — spreads typically move above 6% or 7% in genuine recessions — but the directional move from a tight starting point carries signal.

The ISM Manufacturing PMI has been in contraction territory for extended periods, and while services have been more resilient, the gap has been narrowing. Consumer sentiment surveys have deteriorated sharply, particularly among lower-income households absorbing higher prices on imported goods. And the labor market, which has been the primary shock absorber keeping the expansion intact, has begun showing signs of softening — not collapse, but the kind of gradual deceleration that historically precedes broader weakness.

The Rolling Recession Versus Broad Recession Distinction

One reason recession calls have repeatedly failed in recent years is that the US economy has not been experiencing broad cyclical downturns — it has been experiencing a series of rolling sector recessions. Housing entered a severe contraction in 2022 and 2023 while consumer services boomed. Manufacturing contracted while services remained resilient. This rolling pattern can generate GDP figures that look marginally positive in aggregate while sectors of the economy experience genuine recessionary conditions.

For investors, this distinction matters because a rolling recession has very different portfolio implications than a broad one. In a rolling recession, some sectors are under severe pressure while others thrive, and the cross-sectional dispersion of returns is high. Concentrated positions in the contracting sectors generate real losses even as the index holds up. In a broad recession — one that hits multiple sectors simultaneously — correlations rise, most assets fall together, and the primary protection mechanisms are cash, short duration fixed income, and assets that historically serve as stores of value.

The current macro setup has characteristics of both. Tariff disruption is creating genuine sectoral contractions in trade-exposed industries, while domestic services and energy remain relatively robust. Whether this remains a rolling recession or becomes a broad one depends primarily on whether the labor market continues to hold and whether the Fed has room to respond with aggressive easing if conditions deteriorate.

How Asset Classes Have Historically Behaved in Different Landing Scenarios

Not all recessions are created equal, and the asset class behavior in a mild recession is meaningfully different from that in a severe one. In the mild recessions of 1990 to 1991 and 2001, equities declined 20% to 30% peak-to-trough and recovered within 12 to 18 months. Investment-grade bonds provided genuine ballast, as falling rates supported bond prices. In the severe recessions of 2008 to 2009 and 2020, equity drawdowns reached 50% and 34% respectively, and even high-quality assets came under pressure during the acute phase before monetary policy intervention arrived.

The current probability-weighted scenario is somewhere in between. A 40% to 50% recession probability means the base case remains expansion — barely — but a significant probability-weighted expected drawdown exists. The more actionable frame is to ask: if a mild recession occurs, how much would a given portfolio drawdown, and can the investor stay the course through that drawdown? If the answer to the second question is no, the portfolio is too risk-concentrated regardless of one's point view on recession probability.

Portfolio Adjustments That Don't Require a Recession Call

The useful insight here is that many prudent risk management steps make sense regardless of whether a recession actually materializes. Reviewing concentration in highly cyclical, highly tariff-exposed, or heavily leveraged sectors is reasonable in any environment where recession probability has risen substantially. Ensuring adequate short-duration liquidity — cash and short-term Treasuries that can be deployed opportunistically if equities fall — is a perennial discipline that becomes more valuable when drawdown risk is elevated.

Duration management in fixed income deserves attention. In a genuine recession, long duration Treasuries tend to perform well as rates fall and investors seek safety. In a stagflationary scenario — where growth slows but inflation remains sticky — the bond ballast function is less reliable, as 2022 demonstrated. Understanding which recession scenario is more likely in the current tariff-driven context informs how much duration is appropriate in the fixed income allocation.

Recessions are not predictable with precision. But they are priceable, and at 40% to 50% probability, the risk deserves portfolio-level attention. The investors best positioned for whatever happens next are those who have sized the risk clearly rather than either ignoring it or reacting to it with conviction they don't have.

By Jason Kirsch, Contributor