Reality Gap
Home Macro RG Correlation Analysis

Macro RG × Buffett Indicator — Correlation

How closely do the Shiller CAPE-based Macro RG and the US Buffett Indicator (market cap / GDP) move together — and where do they diverge?

r = 0.916 over 50 years Annual series · 1975–2024 CAPE: Yale / Shiller · Buffett: World Bank

1. Core Finding

Pearson r (1975–2024)
0.916
Very strong positive correlation
Dec CAPE/10 vs. annual Buffett, 50 paired observations
M(S&P 500)RG10 — now
4.13×
CAPE/10 · 2026-06
Overlap-period avg: 2.24×
US Buffett Indicator — 2024
216.3%
Market cap / GDP · World Bank
Overlap-period avg: 104.4%

A Pearson correlation of 0.916 means the two metrics move in the same direction roughly 84% of the time (R²). They share the same macro drivers — interest rate cycles, economic expansions, and risk appetite — but measure different dimensions of the same phenomenon.

What each captures: Macro RG (CAPE/10) reflects how expensive corporate earnings are relative to a smoothed 10-year baseline. The Buffett Indicator reflects how large the equity market has grown relative to the total economy. High correlation means both signal excess simultaneously — but the magnitude and timing of their movements differ, and that divergence carries information.

2. Historical Comparison (1975–2024)

Annual series. Left axis: M(S&P 500)RG10 = CAPE ÷ 10. Right axis: US Buffett Indicator (%). Dashed lines = overlap-period averages.

M(S&P 500)RG10 (left) US Buffett Indicator % (right) Period averages

CAPE source: Robert Shiller / Yale (December monthly reading used for annual alignment). Buffett Indicator source: World Bank — CM.MKT.LCAP.GD.ZS.

3. Why They Correlate — and What Each Adds

Shared drivers (why r = 0.92)
  • Interest rates. Lower rates raise equity valuations (CAPE rises) and expand market cap relative to GDP (Buffett rises). Simultaneously.
  • Economic expansions. GDP growth and corporate earnings growth tend to coincide, pulling both metrics in the same direction.
  • Risk appetite. Bull markets compress risk premia, pushing up both the earnings multiple (CAPE) and the overall market-to-economy ratio.
What each captures uniquely
  • Macro RG / CAPE. Earnings-cycle valuation. Smooths over a full business cycle, detecting whether prices are expensive relative to long-run earnings capacity.
  • Buffett Indicator. Structural concentration. How large is the listed equity sector relative to the whole economy? Captures financialization — the growing share of GDP represented by listed corporations.
  • The gap between them signals whether earnings are expensive (CAPE problem) or whether markets have simply grown larger than the economy can justify (structural problem).

4. Key Historical Episodes

Periods where the two metrics converged, diverged, or gave conflicting signals.

Period Episode RG10 Buffett Key observation
1975–1994 Structurally cheap era 0.9–2.0 37–71% Both metrics far below today. Buffett Indicator below 100% every year; CAPE at or below long-run average.
1995–2000 Tech-boom surge 2.5→4.4 91→153% Synchronized explosion upward. CAPE peaked at 44.2 (Dec 1999). Buffett hit 153% — first time above 100%.
2001–2002 Dot-com unwind 3.1→2.3 132→101% Both deflated together but remained elevated. CAPE slow to fall because earnings had already been inflated.
2008 Financial crisis 1.5 79% Buffett crashed 43% in one year (equity value wiped out). CAPE fell less — earnings lag effect.
2009 Post-crisis recovery 2.0 104% Buffett recovered +32% as markets rebounded before earnings. Divergence: Buffett surged, CAPE lagged.
2020–2021 COVID-era expansion 3.4→3.8 197→208% Both at historic highs. Buffett disproportionate: fiscal stimulus expanded GDP slowly, equities exploded.
2024 Current: new all-time high 4.13 216.3% Buffett at all-time high (216.3%). CAPE/10 at 4.13× — elevated but below 2000 peak (4.42×). Structural divergence: equities now 216.3% of GDP, never seen before.

5. Current Reading (2026-06 / 2024)

M(S&P 500)RG10
4.13×
84% above overlap-period average (2.24×). Below the 2000 dot-com peak (4.42×). Historically in the 85th–90th percentile.
US Buffett Indicator
216.3%
107% above overlap-period average (104.4%). All-time high — no prior year in the dataset exceeded this level.

The divergence in extremity matters. CAPE is elevated but not at 2000 bubble levels. The Buffett Indicator is at a genuine all-time high with no historical precedent in 50 years of data. This asymmetry suggests the excess is partly structural — equities have grown as a share of the economy beyond what earnings multiples alone would imply.

One explanation: the post-2010 era combined zero-rate policy (which expands CAPE) with tech-sector concentration and share buybacks (which expand market cap without proportional GDP growth). The result is a Buffett Indicator that has drifted structurally higher even as CAPE periodically corrected.

6. Limitations

Correlation ≠ causation. Both metrics rise in bull markets because they share macro drivers, not because one causes the other.
Structural shift in Buffett baseline. The US Buffett Indicator has risen secularly since 1990 as the corporate sector financialized. The 1975–2024 average of 104.4% includes a long low-value period (1975–1995) that may not be representative of the current economy.
CAPE timing lag. CAPE uses 10-year trailing earnings. In sharp recessions (2008–09), earnings collapse fast while the 10-year average moves slowly — making CAPE appear lower than actual conditions warrant. Buffett reacts more immediately to market cap changes.
Annual vs. monthly granularity. The Buffett series is annual (World Bank); CAPE is monthly. This analysis uses December CAPE to align, losing intra-year detail.
Not a predictive model. High correlation between two valuation metrics does not predict near-term returns. Both can remain elevated for years.

CAPE source: Robert Shiller, Yale University — ie_data.xls, monthly series 1881–present. Buffett Indicator source: World Bank Open Data — CM.MKT.LCAP.GD.ZS, annual 1975–2024. Correlation computed on 50 aligned annual observations (December CAPE/10 vs. annual Buffett). For research purposes only. Not investment advice.