Is This Time Different? LEI and Yield Curve Inversion as Recession Predictors
Coming out of the pandemic, economists and investors have been vexed by premature recession calls in the US. Indicators that have historically helped predict recessions, such as the Conference Board Leading Economic Index (LEI) and the yield curve, have been signaling that one is coming soon. And yet, the strength of the US economy has time and again surprised forecasters. The Atlanta Fed GDP Now is tracking an above-trend 3.1% real GDP growth for Q2, and consensus expectations for the rest of 2024 remain solid. Many economists who had previously predicted a recession have either removed their calls or pushed them further out. And perplexingly, a number of indicators continue to suggest a recession is around the corner, rather than on a distant horizon.
Have these indicators lost some of their luster in predicting recessions? Before drawing that conclusion, let’s dig into what the LEI and yield curve have been telling us:
Let’s start with the LEI, which aggregates 10 components* to provide early signs of turning points in the business cycle. Following declines over the past year, in April 2024 we saw the LEI falling to its lowest level since April 2020, continuing to suggest softening US economic conditions ahead. The drivers of this predicted economic weakness, spread across major components of GDP, include a deteriorating consumer outlook on business conditions, weaker new orders, a negative yield spread, and a drop in new building permits. The LEI’s six-month and annual growth rates still point to serious growth headwinds: elevated inflation, high interest rates, rising household debt, and depleted pandemic savings, which are all expected to continue weighing on the US economy in coming quarters.
Historically, the LEI has peaked ahead of recessions (on average, a little over 10 months before the start of a recession), and bottomed toward the end of recessions (see Figure 1, below). In the current cycle, the LEI peaked in 2021, 28 months ago. Putting this in a historical context, the longest lead time from peak to recession has been 20 months, as happened prior to the 2008-2009 downturn. While the average lead time is around 10 months, the historical dispersion from peak to recession is quite wide, ranging from one to 20 months. In that light, although it’s been 28 months since the last peak, which is higher than the next-longest lead time of 20 months, it could perhaps still be too soon to dismiss the LEI recession signal.
Figure 1: LEI as Predictor of Recessions
Let’s next turn to the yield curve. Yield curve inversions, whether looking at the 10y-2y spread or the 10y-3mo spread, have historically signaled looming recessions (see Figure 2, below). Currently, both spreads have been inverted for approximately 22 and 19 months, respectively. Historically, the time between inversion and recession is around 13 to 14 months for either curve. Similar to the LEI, the longest stretch from inversion to recession was about 22 months, ahead of the 2008-2009 downturn., We are now approaching that interval.
Figure 2: Yield Curve Inversions as Predictors of Recessions
Another useful yield curve indicator is the maximum inversion to recession (though we acknowledge that this may be difficult to identify in real time). As market participants reassess Fed activity the yield curve inversion weakens , and reprices bonds across the curve ahead of recessions. This lead time averaged around seven to eight months. Currently, the 10y-2y and 10y-3mo spreads are 10 and 12 months from peak inversion, respectively. Though these may be somewhat stretched versus the average, they are at levels similar to those seen before the 2008-2009 recession.
Have traditional indicators such as the LEI and yield curve lost their recession-signaling power? While we don’t have a definitive answer, we recognize that their lead-lag times are variable and we therefore remain open to a scenario of slower growth, if not outright recession, in the coming quarters as we continue to assess cyclical/structural changes in the economy that may impact the efficacy of these signals.
As we continue to watch classical indicators, our team has constructed alternative measures of recession sentiment which we use to inform our investment process about US business cycle dynamics. Details are in William Liang, Is No News Good News? News Sentiment and U.S. Recessions (pgimquantitativesolutions.com).
For compliance use only PGIM Quant- 20240611-2487
