What Recession Probability Models Actually Measure

Recession probability models are statistical tools that estimate the likelihood of a recession occurring within a specified time horizon. The most widely cited model is the Federal Reserve Bank of New York's recession probability index, which uses the ten-year to two-year Treasury yield spread as its primary input. Other models incorporate unemployment claims, industrial production, consumer sentiment, and credit conditions.

The yield curve model works because an inverted yield curve reflects the market's expectation that future short-term rates will be lower than current rates. This expectation typically arises when the market believes the Federal Reserve will need to cut rates in response to economic weakness. The historical accuracy of yield curve inversions as recession predictors is approximately seventy to eighty percent, with a lead time of six to eighteen months.

The limitation of these models is that they measure the probability of recession, not the timing or severity. A model might indicate a forty percent probability of recession within twelve months, which means that in sixty percent of historical scenarios with similar conditions, no recession occurred. The model is not saying a recession is certain or even likely. It is saying the conditions that have historically preceded recessions are partially present.

For investment research, the practical application is to treat recession probability as one input among many, not as a standalone signal. When probability models are elevated but the labor market is still strong, corporate earnings are growing, and credit conditions are normal, the signal is less actionable than when multiple indicators are deteriorating simultaneously. The models work best when they confirm trends already visible in real economic data.

The research methodology for using recession models effectively is to track the trend rather than the absolute level. A model rising from ten percent to thirty percent over six months is signaling increasing risk even if thirty percent is not alarming in isolation. Conversely, a model declining from forty percent to twenty percent is signaling improving conditions even if twenty percent is not negligible. The direction of the trend is typically more informative than the point estimate.

This research note is not financial advice. It is meant to help readers build a watchlist, compare market conditions, and think through risk before making independent decisions.