Over 400% of GDP: top analyst finds corporate equity double the level from the dotcom bubble — and triple 1987’s Black Monday

JPMorgan’s David Kelly warns U.S. corporate equity is over 400% of GDP (vs. 244% pre-pandemic and 74% pre–1987 Black Monday), implying valuations are stretched even as the S&P 500 is up 13% YTD. He attributes part of the rally to AI exuberance but flags that Q2 earnings included $150B of unrealized capital gains booked by two large tech firms, boosting pro forma EPS by 50%—while “stripped” growth is closer to 20%. Kelly expects the Fed to hold rates steady and inflation to drift toward 2%, yet urges investors to diversify away from a concentrated AI-stock bet as real-economy indicators (job and wage growth) remain uneven and “paper wealth” still outpaces Main Street.

Analysis

This is less a “crash” call than a regime warning: when equity value outruns the real economy by this much, marginal returns to passive index exposure usually shrink first in the most crowded growth cohorts. The immediate vulnerability is not the average stock but the handful of mega-cap AI names whose earnings power is being capitalized at a far more aggressive rate than the rest of the market; if even one reports softer AI monetization, index-level de-risking can hit faster than fundamentals would suggest.

The consumer read-through is more nuanced. A broadening of spending into lower-income households is supportive for battered domestic retail names and payment/credit exposures, but the aggregate economy still depends on the top cohort’s balance-sheet strength. That means any drawdown in tech wealth, bonus pools, or home-equity optimism can show up with a 1-2 quarter lag in discretionary demand, credit quality, and store traffic before it appears in macro data.

Banks are a subtle loser here: with growth slowing toward a low-trend pace and rates not falling sharply, the setup is mediocre for loan growth and NII expansion while still leaving them exposed if the wealth effect reverses. The contrarian point is that valuation extremes are poor timing tools when earnings are being flattered by non-operating gains and buybacks; the sharper trigger is whether payrolls reaccelerate or AI capex converts into durable free cash flow. Absent that, the likely adjustment is multiple compression and narrower market leadership, not an outright bear market.

Key Decisions for Investors

• Hedge cap-weighted tech exposure with a 3-month QQQ or XLK put spread on strength; target protection against a 5-10% multiple reset if AI leadership stumbles. Falsify if breadth improves materially and megacap earnings revisions keep accelerating.
• Pair trade: long TGT or GAP vs a broad consumer-discretionary proxy over the next 1-3 months. The bet is that lower-income spending convergence helps traffic before the affluent cohort weakens; cut if wage growth re-accelerates or monthly spending data rolls over.
• Reduce JPM/BAC beta or buy downside hedges on XLF into the next payroll/CPI window. The risk is not immediate credit stress but slower loan growth and flatter NII if the economy remains low-trend; thesis breaks if Fed easing starts sooner than expected.
• Treat ORCL as a watch item, not a buy yet: only get constructive if AI backlog converts into margin-accretive revenue. Otherwise it remains exposed to the same duration/multiple compression as the broader AI trade.