Macro Model · 2026
A Bayesian framework for estimating the probability of a U.S. recession in the next 12 months, combining four macroeconomic indicators weighted against historical expansion and recession regimes.
Download model XLSXFramework
Bayesian
Prior Probability
15%
Indicators
4
The model starts with a historical base rate prior of 15%, the unconditional probability of a U.S. recession in any given 12-month window. It then updates this estimate using the likelihood ratio of each indicator, comparing current values to historical averages observed during Expansion and Recession regimes. The four indicators are weighted equally at 25% each, and their combined signal produces a Weighted Signal Score that drives the final posterior probability.
Yield Curve (10Y–2Y Spread)
25%Measures the difference between long-term and short-term interest rates. A narrowing or inverted spread, where short-term rates exceed long-term, is a classic warning of an impending recession, signalling that markets expect future rate cuts in response to economic weakness.
ISM Manufacturing PMI
25%A sentiment index based on surveys of manufacturing managers. Readings below 50 signal a contraction in industrial activity, often acting as a leading indicator for the broader economy given manufacturing's sensitivity to credit conditions and global demand.
Sahm Rule Indicator
25%A real-time recession signal that triggers when the 3-month average unemployment rate rises by 0.5% or more above its 12-month low. Designed to capture the early stages of a labor market downturn before it becomes visible in lagging indicators.
BBB Credit Spread
25%The premium investors demand to hold moderate-risk corporate debt over safe Treasuries. Widening spreads indicate rising financial stress and tightening credit conditions, both headwinds to investment and growth.