Recession indicators that also warn of a stock market crash

What can recession data tell you about stock market crash risk?

Four families of data carry most of the information: the yield curve, the unemployment rate, broad activity data such as income and production, and statistical models that blend them. Each one measures the economy from a different side. A recession lowers company profits, and profits support share prices, so each indicator doubles as an indirect crash signal.

The link has limits. A recession is a decline in economic activity, and a crash is a sharp fall in prices. They can overlap, as in 2020, or follow separate schedules. The useful question is the speed of each indicator, because speed decides whether it prepares you, alerts you, or only confirms what has already happened.

Which recession indicators lead, coincide, or confirm a stock market crash?

The inverted yield curve leads. Since 1970 it has preceded every US recession, with leads of 6 to 17 months. The Sahm rule coincides: it typically triggers about 3 months into a recession, before GDP data makes the downturn clear. The official dating by the NBER confirms, because it arrives a number of months after the turning point.

This ordering gives a simple use. Leading indicators justify a review of plans, coincident indicators justify action on those plans, and confirming indicators close the file. A reader who waits for the confirming indicator acts after prices have moved, while a reader who acts on a leading indicator alone risks early alarms.

How does the Sahm rule work as a recession indicator?

Economist Claudia Sahm built the rule in a 2019 Brookings Institution chapter on fiscal policy, as a trigger for early stimulus payments. It signals a recession when the 3-month moving average of the national unemployment rate rises 0.50 percentage point or more above its lowest level of the previous 12 months. FRED added the series in October 2019.

Since 1959 it has produced only two false positives, in 1959 and 1969, both a little early. An August 2024 trigger with a value of 0.57 had no recession attached when Wikipedia recorded it, and Sahm herself and other economists have questioned its reliability in the current economy. The rule is fast and simple, and it needs a second opinion.

How does the NBER date a recession, and why does it matter for crash risk?

The NBER committee defines a recession as a significant decline in economic activity spread across the economy that lasts more than a few months. It looks at employment, real personal income less transfers, real consumption, manufacturing and trade sales, and industrial production, and it gives the most weight to income and payroll employment.

Announcements are retrospective, because the committee waits for enough data to avoid major revisions. This matters for crash risk: the official label arrives after prices have reacted. It still provides the reference calendar. The committee dates the 2001 recession from March to November, the 2007-2009 recession from December 2007 to June 2009, and the 2020 recession from February to April.

Which statistical recession indicators add to a stock market crash dashboard?

FRED publishes smoothed US recession probabilities from a dynamic-factor Markov-switching model that uses industrial production, payroll employment, real personal income, and trade sales. The model turns four monthly series into one probability, which is easy to chart and to compare from month to month.

Other readings add detail. The JOLTS report on job openings and labor turnover shows the labour market from the hiring side. William Dudley suggested in 2000 that a rise in unemployment of over one-third of a percentage point signals a recession. Each reading offers a different threshold, and the agreement between several is the stronger signal.

How do recessions and stock market crashes overlap in practice?

The 3 recent recessions show three overlaps. The 2001 recession lasted 8 months, while the Nasdaq Composite fell 78% between March 2000 and October 2002, so the fall began before the recession and outlasted it. The 2007-2009 recession lasted 18 months, with GDP down 4.3% and unemployment rising from under 5% to 10%, and the Dow lost 54%.

The 2020 recession lasted 2 months, and the S&P 500 fell 34% in about a month. In the first case, the market fell before and beyond the recession. In the second, the two ran together. In the third, the crash was almost the whole event. No fixed rule links them, which is why recession data and crash signals serve different decisions.

How do you use recession indicators to prepare for a stock market crash?

Build a three-line routine. Check the yield curve each month for the lead, the unemployment rate for the Sahm reading, and the NBER calendar for confirmation. When two lines turn red in the same quarter, review cash needs, debt maturities, and position sizes. This page is general information, not investment advice.

A 30-minute pre-diagnostic with EQOS Asia adds the structural view that recession data cannot give, and the EQOS Asia pricing page lists the formats. Teams in Singapore, ASEAN, and Hong Kong pair the economic cycle with the φ(S) score to see how much shock their exposure can absorb.

Frequently asked questions

About the author

Stéphane Jambu

President, EQOS Forecast Asia

Stéphane Jambu leads EQOS Forecast Asia and works as a consultant in e-commerce, search visibility, and AI. LinkedIn profile