Surging Wall Street profit forecasts fuel fears of ‘earnings bubble’
Wall Street's 12-month S&P 500 earnings consensus has been revised upward by nearly 20% in six months. That pace of upgrades has no precedent outside a post-crisis recovery.
Russell Cobb·updated July 06, 2026
The Consensus Is Running Ahead of the Cash Flow
Who's Driving the Upgrades — and Who's Skeptical
The upward drift is concentrated in two pockets: semiconductor names and hyperscale cloud operators benefiting from AI infrastructure spend. That narrowness matters. When consensus revisions are this aggressive but this concentrated, you're looking at a single thesis propping up the aggregate multiple — not broad-based earnings momentum.
Ben Inker, Co-Head of Asset Allocation at GMO, flagged the structural risk directly: forecasts for the next two years are climbing at a rate "unprecedented outside of crisis recovery periods." His point is arithmetic, not directional. When the sell-side upgrades earnings this far in advance, the burden of proof shifts to actual reported results — and the runway for disappointment widens.
Michel Lerner, head of UBS's HOLT analytics platform, went further, labelling the setup an "earnings bubble." His argument: AI supply-chain equities have already fully priced in the assumption that abnormally high profitability persists. The probability of sustaining that level, in his view, is low.
The Low P/E Illusion
The S&P 500 forward P/E sits near 20x — well below the peaks of the 2020 rebound or the 2000 dot-com era. On the surface, that looks benign. Sarah Ketterer, CEO of Causeway Capital Management, offers the forensic counter-read: a low forward P/E does not mean stocks are cheap. It may simply mean the "E" denominator is near a cyclical peak. If you're reverse-engineering intrinsic value, the question is not whether 20x is historically expensive. It's whether the "E" in that ratio is sustainable two or three quarters out.
Capital Economics flagged that AI-related markets may be approaching a tipping point. If corporate capital expenditure assumptions prove overstated or monetisation timelines slip, the gap between forecasted and actual cash conversion will show up fast.
What the Math Demands You Check
Three things deserve a line-by-line review before the Q2 print cycle begins. First, the gap between GAAP and adjusted earnings across the AI capex beneficiaries — watch for aggressive capitalisation of costs that inflate the reported number. Second, the cash conversion cycle: are these upgrades translating into actual free cash flow, or are they sitting in accruals? Third, the concentration risk embedded in the index itself — if the top 10 contributors to S&P 500 earnings growth miss by even a modest margin, the aggregate consensus reprices faster than most risk models anticipate.
The revision speed is the signal. Not the direction.