The conventional 2026 setup — fiscal lift, AI capex, easing policy, a 7,400 print on the S&P — describes a Goldilocks year that depends on one thing being true: that the current profit cycle is structural rather than borrowed. The earnings data say it is. The valuation data say to be careful which earnings you are paying for. The interesting question is no longer whether equities can rally through geopolitical noise. They already have. It is whether the profit engine can absorb a wage cycle that has not yet fully decelerated and an oil tape that has not yet fully calmed.
Start with the numbers that actually matter. Aggregate S&P 500 earnings rose roughly 19 percent year-on-year in the first quarter, and the consensus is for forward twelve-month profits to come in 24 percent above last year. That is not a late-cycle print. FactSet analysts note that bullishness of this order is normally reserved for recoveries out of deep recessions — and the United States has not had one. The implication is that something other than the business cycle is driving margins: a combination of post-pandemic pricing power, industry consolidation, and an AI capex wave whose beneficiaries are concentrated but whose spillover into broader productivity is now visible in the tape.
The concentration problem, restated
The bullish case is heavily intermediated by a small number of names. Nvidia is forecast to grow earnings by nearly 80 percent in 2026, with Broadcom close behind; Micron and Sandisk are penciled in for seven-fold and sixteen-fold earnings increases respectively. The Magnificent Seven will, on James Surowiecki's reckoning, clear more than half a trillion dollars in profit this year. That is the engine. It is also the single point of failure. Maverick Equity Research's observation that the S&P trades at 22x forward earnings — outside the interquartile and 10th-to-90th percentile ranges — and that the index ex-Big Tech still trades at 19x, tells you the market has already paid for the spillover. The capex must convert.
The counter-trade is more interesting than the headline trade. If you believe AI productivity is real and broadening, the right exposure is not another marginal dollar into the names already discounting 80 percent earnings growth. It is the rest of the index — the 19x cohort — where margin expansion from the same productivity wave is not yet priced. Compounding Quality's portfolio math, which combines roughly 12 percent EPS growth with a 5.8 percent earnings yield to imply a 17.8 percent expected return, is one version of that trade. The Cassandra Unchained observation — that recent rallies have come from revenue growth rather than multiple expansion — is another way of saying the same thing: the durable part of this market is earnings, not re-rating.
The market has already paid for the spillover. The capex must now convert.
The risk that breaks this is not a recession. It is an energy shock that compresses real consumption while wages are still decelerating, leaving the consumer-facing half of the S&P caught between input costs and pricing fatigue. Surowiecki flags the scenario directly: sustained high energy prices into summer feeding through to weaker profits and softer spending. It is the cleanest tail to short against a 7,400 base case, and the one most under-hedged at current implieds. A Goldilocks call on 2026 is defensible. A Goldilocks call that does not budget for the oil tape is not.
Briefings are synthesised by the Ledger Desk from multiple sources cited in the sidebar. They are distinct from Articles, which are written by named contributors and carry a tracked Calibration Index. The Desk does not currently carry a Brier score; this is a deliberate choice for the v0.1 editorial layer and will be revisited.
