Authored by Simon White, Bloomberg macro strategist,
The clamor for a US recession has grown louder in recent weeks, but forward-looking data shows the risk of one is easing not rising.
Despite the abrupt turn in the mood music in recent weeks it is looking increasingly likely the official arbiter of US recessions, the National Bureau of Economic Research, won’t call one, leaving stocks priced for too negative an outcome.
Warning sounds about a US recession have been growing louder. But that has been fueled by principally lagging economic data telling us where the economy’s been, not where it’s going next. Leading data is inflecting higher, suggesting the US may skirt a NBER-recession for most of next year.
Academic disputes are said to matter so much as they mean so little. The same can be said for recession prediction — after all, it’s what the market does that ultimately counts for investors. Nonetheless, the recent flurry of recession chatter deserves a reply, as a close look at the data shows that a slump in the medium term is no longer the base case.
I come at this as a soft-landing skeptic. Multiple reliable data points had been pointing toward a recession, but there is now enough contrary data to sow sufficient doubt.
Before going any further we have to set the terms of reference. It can be argued the US has experienced or is already in a recession, for instance by looking at GDI, manufacturing, goods GDP, or earnings. But with no hard-and-fast definition of an economic contraction (the technical definition of two-consecutive quarters of negative growth is too simplistic), having a referee in the NBER is the best option.
The downside is the NBER doesn’t announce recessions until after they have started, and thus is impractical for investors in real time. The utility instead comes from noting that the times the NBER deems to be recessions are when stocks have experienced their worst downturns — thus trying to figure out ahead of the NBER when there will be a contraction can save investors from considerable losses.
Since March this year, the S&P has been trading as if an NBER-recession will be avoided. Lately stocks have sold off, but they are still a long way above the median S&P in bear markets that had a recession (the blue line in the chart below). That means there is still sizable downside risk today if there is a downturn; but equivalently there is the risk from missed opportunity as stocks could eclipse their recent highs, or more, if a slump is avoided.
NBER recession dating is not an exact science. As the bureau itself puts it, “there are no fixed rules or thresholds that trigger a determination of decline.” But the research body requires a recession be durable, diffuse and deep. It also gives four of the key variables it uses as recession determinants:
Let’s start there. In the chart below we can see the four series over the last 50+ years. All of them contracted on an annual basis at some point during each NBER recession since 1970, apart from the one in 2001, where real PCE didn’t contract (recessions are vertical gray bars in the chart below).
Zooming in to the present in the next chart, we can see that none of the four indicators are currently contracting. Industrial production is flat-lining around zero, payrolls growth is turning lower from a high level, while real PCE and real personal income are positive and have been rising.
Not only is the NBER very unlikely to call a recession based on the current state of play, leading data shows that is unlikely to change in the next 6-9 months.
All four indicators are coincident-to-lagging, and most of them are released with a delay.
That’s why we must turn to leading data series to pre-empt the NBER. It is increasingly pointing in the direction that the current slowdown won’t last quite long enough (have the duration) or fall hard enough (have the depth) to trigger an NBER recession.
First, there is the recent upturn in the manufacturing ISM, which anticipates that industrial production should soon recover. October’s ISM disappointed to the downside, and if sustained this would cause some concern. But leading data for the ISM, such as the rising new orders-to-inventory ratio and the steepening in the global yield curve (shown below), intimate the ISM has bottomed and should continue its rise.
Second, there is the strong upturn in the US Leading Indicator (composed of leading data series such as building permits), which points to a continued rise in real retail-sales growth, and thus real PCE.
Third, real wage growth has been growing positively, keeping real income less transfer payments supported. However, of the four indicators, real income faces the most downside risk. Leading indicators for wages are rolling over, pointing to lower wage growth next year.
Finally, an inflection lower in the annual growth of unemployment claims points to eventual support for payrolls growth. WARN data (which leads claims) is also now flat again, indicating any claims growth should be contained. The jobs market should continue to slow, but leading data is showing the slowdown may not be of sufficient depth for the NBER to deem it recession worthy.
None of the indicators are screaming recession. The NBER has stated it gives more weight to real income growth and payrolls in its assessment, but even if these two end up contracting, the data is showing that the other series do not look like following them down soon. Remember, the bureau has never called a recession in the last 50 years when fewer than three of the four indicators have been contracting, and then only once.
Instead, the biggest risk to a soft landing come from two other factors: geopolitical, and credit.
I won’t dwell on the first – an unexpected escalation in global hostilities would likely hit sentiment enough to tip the US into recession.
Credit is the biggest endogenous risk facing the economy and the market. Bankruptcy filings and charge-off rates are rising, indicating underlying stress not reflected in credit spreads. Moreover, the opacity in rapidly growing private-credit markets is becoming a greater concern. Credit should be monitored closely as a fast unraveling would swiftly take the economy into recession territory.
Until then an NBER-recession looks less likely than so over the next 6-9 months. Which is not how the situation looked earlier in the summer. Abnormally high fiscal deficits; the asynchronous recovery after the pandemic disrupting the traditional interplay between the goods and services economy; and money illusion to a degree most haven’t experienced before are all factors why this time is not following the usual script.
When the leading data turns sufficiently down, the NBER will make a recession announcement in due course. But we are unlikely to hear from them for a while yet.