Originally published 7/16/2019, updated 8/15/2019
The economy is widely expected to continue to slow, and yield curve inversion has intensified concerns we’re heading into a recession. Yet following the U.S. economy involves checking in on hundreds of indicators – some of which can give conflicting signals. While a recession is not our base case, risks around the trajectory of the economic slowdown are heightened. Over the years I’ve found that two indicators are particularly good at signaling a meaningful downshift in growth: initial jobless claims and consumer confidence.
Reading the tea leaves of economic data
It may seem surprising that we need timely indicators of a recession. After all, in the rearview mirror, recessions are obvious. However, in the moment it can be hard to distinguish a moderate economic slowdown from an actual recession, which is defined as a significant decline in economic activity. Remember, economic data is released with a lag and can be conflicting.
The National Bureau of Economic Research (NBER) is the organization that decides the official start and end dates of recessions, but this typically happens with quite a lag. Since the 1970s, NBER has announced the start of a recession 6–21 months after it began, in some cases after the recession is already over!
Yet keeping tabs on the economy is perhaps more pressing now than at any time in the past 10 years, given the length of the current business cycle. Most forecasts expect the economy to slow throughout 2020, but a “soft landing,” where growth gradually decelerates to match the pace of underlying potential growth, can be difficult to navigate. At present, the yield curve is inverted according to the often-watched 3M-10Y spread, which is considered a warning sign that a recession is likely within the next 4–6 quarters. Concerns about trade tensions and slowing growth in other developed countries have ratcheted up market jitters that a recession is looming.
#1: Initial jobless claims
One of my favorite leading indicators of the economy is initial jobless claims. The data is very timely and released weekly. While this data series can be volatile in reaction to one-off events, like a hurricane or a government shutdown, it’s a powerful leading indicator of recession with few incidents of giving a false-positive signal.
Initial claims for unemployment insurance, when a worker applies for state-sponsored unemployment benefits, typically bottoms 9–24 months before a recession begins. Often, the path from weaker business sentiment to the broader economy comes through a slowdown in hiring. A 20% increase in initial claims over an 8-week period would be a clear warning sign that broad-based uncertainty is causing companies to increase layoffs.
At present, initial claims rest near multidecade lows. In 2019 to date, claims have averaged 218,000, a level that reflects a tight labor market.¹ Recently, the manufacturing ISM dropped to a 33-month low, causing concern to spike about the health of the economy. But initial claims have remained at exceptionally low levels. Should claims rise to 275,000 over several months, I would become more pessimistic about our growth outlook.
#2: Consumer confidence, present situation vs. expectations
The consumer is the single most important sector for the economy, making up 69% of GDP. I look at a simple spread of the two main components of the Conference Board’s consumer confidence numbers: the difference between the Present Situation Index and the Expectations Index. This has a slightly better track record of timing the start of a recession than the headline index.
At present, this spread rests near cycle highs. This is partly due to the tight labor market, which has propelled confidence and the present situation measure close to a 20-year high. Should the economy meaningfully falter, the difference between low growth and a contraction in growth is the consumer, and the deterioration in household sentiment is what causes spending to dry up quickly.
Although this spread has retreated modestly from its cycle high in January, it remains elevated. But should business confidence waver and hiring start to slow, I will be watching this measure to see whether households are entering bunker mode and radically cutting back spending, which could push our entire economy into a recession.
Where’s the yield curve in all this?
There is no shortage of economic data. Economists spend their days absorbing, analyzing and reporting on hundreds of data series. But for those who don’t have the luxury – or the desire – to wade into this deep ocean of data, initial jobless claims and consumer confidence are two data series well worth keeping on the radar.
Some may wonder why I haven’t put yield curve inversion on my watchlist. I’ve written about yield curve inversion extensively, and it’s critical to be mindful of the signals the bond market is sending about the economy. However, yield curve inversion covers a wide variety of yield pairs, which right now show a varied picture. While the 3M-10Y is inverted, the 2Y-10Y is flat. The Fed funds futures curve, also an important yield curve to consider, is deeply inverted, but it’s harder for the layperson to keep track of (I check it on my Bloomberg).
In addition, like any product in finance, the fixed income market is about both economics and financial market conditions. Yield curve inversion may be amplified by low international rates, for example. My quest is to find indicators of Main Street, not Wall Street.
Consumer confidence reflects the outlook of Main Street, arguably a more important guide to the underlying economy. Initial jobless claims likewise is a timely indicator showing businesses’ willingness to hire and retain workers, a critical forward signal of how business sentiment will feed into household spending habits.
History can serve as a useful guide for how these indicators act around turning points, but it is no guarantee. Yet while every expansion is unique, we haven’t had a contraction in growth without a contraction in household spending, making the consumer a critical sector to track. Likewise, with a decline in business sentiment already apparent, it will be important to see whether this translates into more conservative hiring practices. I don’t expect a recession in the near term, but any slowdown generates uncertainty and heightened volatility. As the anatomy of the slowdown further materializes in the second half of the year, I’ll keep you updated on my two favorite recession indicators.