Where the profits come from

The Kalecki–Levy identity, computed on today's data: an accounting decomposition that must balance — where each dollar of corporate profit arithmetically came from. This page is not a timing tool — the timing lenses live on the main page.

AS OF 2026Q2 · estimate of Sep 30, 2026 · revision stage 3 of 3 · all sources through 2026Q2 — 2026Q3 data has begun arriving (4 of 8 sources); profits and the full mix follow with the ~late-Nov release — the display shows 2026Q2, the latest complete quarter.

The read

CHARACTERIZATION, NEVER A SIGNAL — PERSISTENCE DESCRIBES; IT DOES NOT PREDICT

Today's mix is mixed — no single family of sources carries it; the cards below give the full accounting.

  • This quarter, the state & local balance crossed above its band.
  • Closest to change: money flowing abroad (0.58 points of GDP from its band edge) and the federal deficit (0.82), both narrowing toward their bands.

These columns sum to reported corporate profits exactly — every quarter since 1947; the statistical discrepancy is a real, small leg, not a plug.

The live era (2020–) is federal-deficit-dominant, and the main page reads the margin at +1.35σ versus trend — that period rides on a mixed base.

Today's mix

Today's level by source — each source's share of GDP against its own middle-50% band, with its move vs four quarters ago; this quarter's change by source is the main page's "why" line.

−9% of GDP0+9%
one shared axis, all rows
ADDS — SORTED BY SIZE
Government (all in)
deficits + government investment
+7.04% of GDP
sum of its members — bands live on the member rows
Dividends recycled
+6.75% of GDP
+2,198.2 $B
p93 of own history+6.75 now vs +7.04 four quarters ago63q above band
Business investment (net)
+4.15% of GDP
+1,352.9 $B
p13 of own history+4.15 now vs +4.29 four quarters ago5q below band
The statistical discrepancy
+0.58% of GDP
−189.5 $B
p90 of own history+0.58 now vs +0.04 four quarters ago3q above band
DRAINS — SORTED BY SIZE
Money flowing abroad
−3.37% of GDP
−1,096.0 $B
p16 of own history−3.37 now vs −3.76 four quarters ago24q below band
Household saving
−3.24% of GDP
+1,056.4 $B
p92 of own history−3.24 now vs −4.35 four quarters ago7q above band

Bars: today's contribution (dark marker) against the source's own middle-50% historical band (grey underlay), % of GDP, zero aligned on the one shared axis. Direction reads are vs four quarters ago — stated on every chip. Latest-basis data, labeled; September's annual update revises history and the methodology says so.

The periods

periodavg margincharacter (from the computed dominants)
2020–LIVE10.67%federal-deficit-dominant + the capex boom
THE FULL MIX (%GDP) net investment +4.31 · govt investment +0.78 · dividends +7.34 · household saving −6.10 · govt federal +7.87 · govt state local +0.16 · current account −3.53 · stat discrepancy −0.15 n=26q
2009–20199.88%dividends-led (the deficit narrative for this period fails its own arithmetic — 5.73 vs 5.33)
2001–20087.83%investment-led
1980–20005.91%investment-led
1947–19796.26%investment-led

Era boundaries frozen at the charter; a new era row only by recorded amendment. Only the top row is live.

The capex drill

1%2%3%4%5%6%19501960197019801990200020102020
  • info-processing equipment
  • software
  • structures
  • boom windows
  • component peaks

"Never-seen levels" is true of software only (2.52% of GDP vs the 2001Q1 peak of 1.55%, still rising); equipment (2.83%) sits below its dot-com peak (2.91%, 2000Q4); structures (2.93%) below its own 2023Q4 peak (3.39%). Latest-basis data, labeled. The acceleration rule tested here was a documented pass — the methodology carries the full record. The 2000–02 unwind is shown as history, not forecast: one corporation's capex is another's revenue, so the unwind pulled the investment source down. Profits did not follow: the federal budget's swing from surplus to deficit more than made up for it, and the profit share rose from late 2000 to the end of 2002.

How it's measured → · ← back to Corporate Profit Share