Corporate profits as a share of the economy — where the share stands against its own trend, and how much the current print deserves to be trusted.
Corporate profits are 1.35σ above their own trend, the second-highest of five levels — they have stood here 17 times before, since 1963.
How far profits are from trend has never said anything about what comes next: across sixty years, the level a period started at made no difference to the year that followed.
A period is the consecutive quarters the share spends at one level, a single quarter included.
Log scale. The trend and envelope are expanding fits — each point uses only history to that quarter (today's revised values; the point-in-time honesty check lives in the methodology). Left of the divider the underlying data is revised-basis; true vintages begin 1991-12. Historical bands change only when BEA's annual update revises the past, each September; otherwise only the right edge is live. The chart opens on the last 40 years; the full record from 1947 is in Expand.
Across the 55 completed periods since 1963, a level change was followed by a losing year in the S&P 11% of the time — against 22% for any quarter in the same years. High profits have not been a warning. Six periods did end in a losing year. One condition did stand out: when the housing cycle had already turned past its expansion, 4 of 19 periods ended in a losing year, against 1 of 23 otherwise — across three separate episodes (1987, 2007–08 and 2021). It was one of eighteen conditions tested, so read it as a lead worth watching, not a rule. The profit share's own level still says nothing about what follows.
Measured October 2026 over 1963–2025. The full record →
Across four elevated episodes since 1991, margins never closed half the gap back to their historical median in the modern era — zero reversions across all three post-2000 episodes at any horizon. Not a usable timing tool since 2000; the register accrues as new episodes complete.
4 episodes — the rule →
All three qualifying recoveries (2001, 2009, 2020) were first reported weaker than they proved: episode revisions of +1.09, +1.57, +1.27 points of GDP share against an all-quarter median of +0.85. Three episodes is a small base; the claim is stated only that strongly.
feeds the trust dial · the trough quarter itself revised down in 2001
Where this sits in the whole record: higher than 100.0% of every quarter since 1947. It has been in the top fifth of its own history for 88 of the last 91 quarters — it describes a regime, not a signal.
how much first prints later moved (first release → today, all 140 vintage-true quarters)
Elevated periods: 1997–2000 · 2003–2009 · 2009–2019 (42 quarters) · 2020–present (open).
S&P figures are total return · Shiller monthly convention — dividends reinvested monthly (see "The total-return basis" on the Confluence methodology). How it's measured →