Stats › Bear markets and drawdowns › Unemployment
Unemployment and stock returns
Unemployment goes up, stocks go down. That's the intuition almost everyone starts with, and it's wrong about half the time. The best known peer reviewed work on this says the sign flips depending on where you are in the business cycle: rising unemployment is usually good news for stocks during expansions and bad news during contractions. Same news, opposite reaction, and which one you get depends on something you can't observe directly. The reason is that an unemployment number carries two messages at once. It hints at lower interest rates ahead, which helps stocks, and lower company earnings ahead, which hurts them. In expansions the rates message wins. In contractions the earnings message wins.
This page has where unemployment stands now, the post war extremes, and what the research actually supports. It also does something a bit unusual: it spends time on a claim that doesn't exist. The line about buying at peak unemployment producing a big forward return circulates constantly and I couldn't trace it to any peer reviewed source. That absence is a finding, and I'd rather report it than quietly fill the gap with the best available blog post.
Four sources, including two academic papers and the official BLS series, are listed at the end. Where a figure's attribution is uncertain, or where a result sits behind a paywall so I could only report the direction and not the size, I've said so in the notes instead of pretending otherwise.
US unemployment was 4.2% in June 2026, inside a narrow 4.0% to 4.5% band that has held since early 2025. The post war record high is 14.8% in April 2020 and the low 2.5% in May 1953. The rigorous finding on unemployment and stocks is that the sign depends on the cycle: Boyd, Hu and Jagannathan found rising unemployment is good news for stocks in expansions and bad news in contractions. The popular claim that buying at peak unemployment produces a large forward return has no traceable peer reviewed source, and the reason is structural: there have been about a dozen unemployment peaks since 1948 and each is only identifiable in hindsight.
Members only
The rest of this page is for members
Below this point there are 8 sections, 1 chart and 4 named sources, roughly 1650 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.