This spot-on recession indicator is flashing red
'Schannep Recession Indicator' typically triggers within six weeks of the start of a recession
A "recession is underway," and we're just waiting for corroboration by the National Bureau of Economic Research, the semi-official arbiter of when recessions begin.
That's the confident prediction I received a few days ago from Jack Schannep, former editor of the TheDowTheory.com investment newsletter. He bases his conviction on an indicator he created in 2000 called the "Schannep Recession Indicator" (SRI), which corresponds to trends in the U.S. unemployment rate.
According to Manuel Blay, Schannep's successor, in the U.S. since 1946 there have been 12 recessions; "all of them were spotted by the SRI." Blay adds that the SRI issued only one false signal - in October 1959, six months prior to the recession that commenced in April 1960. (The SRI's track record since the late 1940s is summarized here.)
The SRI is based on the three-month moving average of the unemployment rate. A recession signal is triggered when it rises by at least 0.4 percentage points from that moving average's prior cyclical low. The SRI's latest signal was triggered on June 7 with the report of May's unemployment rate of 4.0%, which brought the three-month moving average to 3.9%. That's 0.4 percentage point higher than where the moving average stood in April of last year.
Schannep says he developed his indicator after investigating news media reports that a recession occurs whenever the U.S. unemployment rate rises by more than three-tenths of a percentage point. Schannep found this to be close, with the provisions that the threshold is 0.4 of a percentage point; the focus is on the unemployment rate's three-month moving average, and the comparison is to that moving average's prior cyclical low.
There is a similarity between the SRI and the so-called Sahm Rule, introduced in 2019 by Claudia Sahm, a former Federal Reserve economist. There are two major differences, however: first, the Sahm Rule doesn't trigger until the three-month moving average rises by at least 0.5 of a percentage point, in contrast to 0.4 for the SRI. Second, the Sahm Rule compares the latest three-month average to its trailing 12-month low, while the SRI compares it to the most recent cyclical low - which could be more than 12 months in the past. The Sahm Rule has not indicated a recession is imminent or underway. Still, if a U.S. recession is not already here, it's near.
Recessions are notoriously hard to predict
You might wonder why we even need an indicator to predict when the National Bureau of Economic Research will declare that a recession has begun. Why not just rely on the NBER to tell us? The answer is that the NBER often takes many months, sometimes more than a year, to determine when a recession has begun. Sometimes the recession has been over by the time the NBER verifies that it began - the recessions of 1990-1991 and 2001, for example.
An indicator that accurately signals a recession closer to real time clearly would be valuable. Blay reports that the "average lag from the recession's onset and the [Schannep Recession Indicator] signal's date is 1.58 months."
A good illustration is the recession that accompanied the Global Financial Crisis, which the NBER subsequently declared began in December 2007. The SRI was triggered in early February 2008 upon the release of the January 2008 unemployment rate - two months into the recession and nearly a year prior to the December 2008 NBER announcement.
Humility remains a virtue in this business, however, since the graveyards of Wall Street are filled with those whose confident predictions of an imminent recession turned out to be failures. Take the inverted yield curve, which used to be considered a nearly foolproof indicator of an imminent recession. But not this time: the yield curve became inverted in July 2022. Or take a Bloomberg Economics recession model, which in October 2022 declared that the probability of a recession beginning within one year was 100%.
As Sahm conceded in a blog post last December, even if her model were to indicate that a recession were either already underway or imminent, "a recession [would still not be]... a done deal... This time could easily be different."
Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. TheDowTheory.com, the subject of this report, is not one of the newsletters Hulbert Ratings tracks. Hulbert can be reached at
mark@hulbertratings.com
More: The bond market is sounding a recession alarm for the longest period on record. What that may mean for the stock market.
Plus: 'U.S. growth is exceptional': World Bank sees America fueling the global economy in 2024
