Nominal GDP vs S&P 500

The Bureau of Economic Analysis publishes nominal GDP quarterly under the National Income and Product Accounts framework. The advance estimate releases roughly 30 days after quarter-end, with second and third estimates following at one-month intervals. The series available on FRED as GDP is the seasonally adjusted annualized rate in current dollars; for the December 31, 2025 vintage, that figure was approximately $31.3 trillion. SPDR S&P 500 ETF (SPY) holds the 500 large-cap U.S. equities in the index, with $560 billion plus in assets under management and an expense ratio of 0.0945%.

The Wilshire 5000 Total Market Index, used as the numerator in the Buffett Indicator, captures all publicly traded U.S. equity. For the December 31, 2025 print, the Wilshire 5000 was $72.14 trillion. The ratio (Wilshire 5000 / GDP) was 230% per Current Market Valuation. The textbook reference is Warren Buffett's December 2001 Forbes interview where he called the ratio probably the best single measure of where valuations stand at any given moment, with 75 to 90% being reasonable and over 120% suggesting overvaluation.

The Buffett Indicator hit 230% on December 31, 2025, then 232.6% briefly in early 2026, and 226.9% on April 29, 2026, the highest sustained level in the post-1971 series. Prior peaks: the dot-com bubble peaked at approximately 145% in March 2000; the 2007 cycle peaked at approximately 105%; the post-2020 reflation reached 200% by November 2021 before retracing to 152% in October 2022 during the bear market; the 2024 to 2026 cycle has now extended above 220% for sustained windows. The 2025 peak is roughly 60% above the dot-com peak and roughly 120% above the 2007 peak.

The long-run trend regression (since 1971) sits around 95 to 100%. The 230% reading is approximately 2.4 standard deviations above trend. Historically, readings more than 1.5 standard deviations above trend have produced negative 10-year forward S&P 500 real returns in 7 of 10 instances since 1900 per multiple long-horizon studies (Hussman, Shiller CAPE comparisons). The current 2.4 standard deviation reading has no historical precedent, so forward-return base rates are extrapolated rather than empirical.

Three structural drivers explain extended SPY-GDP decoupling. First, corporate profit share of GDP has risen from 6 to 8% in the 1980s to 12 to 14% in 2024 to 2026 per BEA NIPA Table 1.10. Higher profit share means each dollar of GDP supports more equity-market value than in prior eras. Second, S&P 500 firms derive a growing share of revenue from outside the U.S.: roughly 40% of S&P 500 revenue is international per FactSet, so global GDP rather than U.S. GDP is the relevant denominator for that revenue stream. Third, the post-2008 era has compressed real interest rates structurally, with DFII10 averaging roughly 50bp in 2010 to 2021 versus 200 to 300bp in 1980 to 2007, supporting higher equity multiples through the discount-rate channel.

The 2000 to 2002 episode is the cleanest test of when decoupling reverses. The Buffett Indicator peaked at 145% in March 2000 with corporate profit share elevated and trade-weighted dollar strong. SPY fell -49% peak-to-trough through October 2002 while nominal GDP grew at 3.5% annualized over the same window. The reconciliation came entirely through the equity leg, not through GDP. The 2008 to 2009 episode is similar: Buffett Indicator peaked at 105% in October 2007, SPY fell 57%, nominal GDP grew at 1.7% annualized over October 2007 through March 2009 (briefly contracting in the recession quarters). Reconciliation again came through equity decline, not through accelerated GDP.

Convex Net Liquidity Impulse (CNLI), constructed from Federal Reserve H.4.1 weekly data (reserves plus reverse repo plus Treasury General Account), is the most relevant Convex composite for the GDP-SPY pair because liquidity provision is what supports above-trend Buffett readings. The 2009 to 2014 first QE cycle, the 2020 to 2021 emergency QE, and the 2023 BTFP plus reverse-repo unwind all produced positive CNLI impulses that lifted SPY relative to GDP. The 2022 QT plus rising-rates episode produced the most negative CNLI on record and pulled the Buffett Indicator from 200% in November 2021 to 152% in October 2022.

Convex Risk Appetite Index (CRAI) provides the faster signal. Buffett Indicator extremes (above 200%, below 80%) are consistently accompanied by CRAI prints in the corresponding quartile. April 2026 CRAI sits in the upper half but not the extreme top quartile, which suggests the 230% Buffett reading reflects a combination of structural factors (profit share, foreign revenue, low real rates) and risk-on positioning rather than pure speculative euphoria. That is a less brittle configuration than 1999 to 2000, when CRAI was in its top decile alongside Buffett at 145%.

The current configuration has two competing historical analogs. Bear analog: the 1999 to 2000 dot-com peak. Buffett Indicator 145% in March 2000, SPY fell 49% over the next 32 months, nominal GDP grew 3.5% annualized. Reconciliation came through the equity leg. Bull analog: the 1995 to 1996 mid-cycle. Buffett Indicator rose from 75% to 105% as the post-1994 Greenspan tightening ended and tech multiples re-rated upward. SPY went on to triple over the following five years before the 2000 peak. Reconciliation came through GDP catching up partially and equity continuing to outperform.

The difference between the two analogs is whether the current readings reflect a sustainable shift in the equity-claim share of economic value (bull case, supported by higher profit share, foreign revenue mix, and low real rates) or a positioning extreme that will revert (bear case, supported by 2.4 standard deviation distance from trend and no historical precedent for sustained 220%-plus). The 2025 to 2026 configuration is closer to the bull analog on CRAI positioning (mid-range rather than top decile) but closer to the bear analog on absolute valuation extremity (2.4 standard deviations above trend). The 12-month forward read is therefore high tail-risk but not high modal-drawdown.

Long-horizon allocators (pension funds, endowments) at current Buffett readings have historically reduced strategic equity exposure by 5 to 15 percentage points relative to the policy benchmark, sized against the historical drawdown distribution at similar valuations. The Norwegian Government Pension Fund Global, the Yale endowment under David Swensen's framework, and CalPERS have all publicly reduced U.S. equity weights in 2024 to 2026 relative to their 2010s benchmarks, citing valuation extremity.

The trade-structure for tactical allocators is to maintain core SPY exposure while sizing tail hedges proportional to the historical drawdown at similar valuations. The 1999 to 2000 analog produced a 49% drawdown over 32 months; the 2007 analog produced 57% over 17 months. Both fall within the range that systematic tail-hedge programs (e.g., 5% annual budget on rolling 90-day SPY puts at 90% of spot) target. CRAI and CNLI together provide the timing overlay: when both deteriorate concurrently, the tail hedges are most likely to monetize. The current April 2026 configuration shows neither composite in deterioration, but the absolute Buffett reading at 226.9% means the structural setup remains the most extended on record.