Goldman Sachs: The Markets
2026-10-26 · Hosted by — · Goldman Sachs
Executive Summary
Goldman Sachs' Rich Privorotsky, head of European equities trading, explained why the "long AI, short everything else" trade that dominated summer positioning is now unwinding painfully heading into Q3 earnings season. He identified three converging concerns: rising cost of capital across triple-C credit spreads, European sovereign spreads, and especially the levered neo-cloud segment of the AI trade; questions about whether companies can actually energize enough gigawatts of power to use the chips they've already ordered; and a broader de-risking mentality that has pulled growth expectations down even as bonds rally.
Key Stories & Changes
1. The "Long AI, Short Everything Else" Trade Is Unwinding
Summer 2026's dominant "pain trade" (long AI/semis/hyperscalers, short the rest of the market) has become "particularly painful" in reverse over recent weeks
Three converging issues: (1) rising cost of capital spilling into triple-C spreads, European sovereign spreads, and especially levered neo-clouds; (2) doubts about the ability to energize gigawatts of power to use already-ordered chips — power buildout estimated at just 20-40-50 gigawatts for next year against demand; (3) a broad de-risking mentality pulling down growth expectations
Bonds have rallied as growth expectations fall — a dynamic Privorotsky says "was on no one's scorecard"
2. Hyperscaler Earnings Must Prove Monetization, Not Just Spend
Market consensus sits around $1.1 trillion in hyperscaler capex for next year
Privorotsky: it's "not enough to just say you're spending a lot of money" — backlog must convert into revenue, and the speed of the bridge to free cash flow breakeven will be closely watched by increasingly sensitive credit markets
He noted compute spot prices are "literally making new highs every day," supporting a "glass half full" read on underlying demand
3. From "As Goes Oil" to "As Goes AI, So Goes Everything Else"
Privorotsky's prior 2026 mantra was "as goes oil, so goes everything else" — he now says, "at the risk of being overly reductionist," it's "as goes AI, so goes everything else"
Energy price pressure has driven a significant tightening in financial conditions (a measure of how loose or tight money is across rates, credit, and equities), pushing the Fed to deliver an entire hiking cycle it hadn't planned
Despite the S&P index itself holding up, the median stock is down "well in excess of 10%"
4. Credit Cycle Risk Is Rising From a Low Base
Privorotsky describes the backdrop shifting from a decade-long "savings glut" to a "spending glut," with over $1 trillion in AI capex and wide fiscal deficits globally pushing out the weakest borrowers
Triple-C spreads have moved materially (though he calls this "a bit idiosyncratic"); French sovereign spreads have also widened on fiscal deficit concerns
Three months ago he saw credit cycle risk as low; now he says it is "certainly a lot higher"
5. Trade Recommendations: Electrification, Aerospace, Transports
Favors electrification names that have "de-rated a ton" with visibility into 2027 providing a margin of safety
Likes civil aerospace assets that have also become cheap
Sees convexity (asymmetric upside potential) in transports if there's any energy resolution, given how depressed positioning ("nets") has become after what he calls the biggest tech-selling in Goldman's prime book on record last week
Trends Identified
1. The AI Trade Has Become the Market's Central Organizing Variable
Privorotsky's shift from "as goes oil" to "as goes AI, so goes everything else" crystallizes how thoroughly AI capex expectations now drive broader market direction — commodities, nominal assets, and credit spreads are all being repriced in lockstep with AI sentiment rather than their own fundamentals.
2. Power Availability, Not Chip Supply, Is the New Bottleneck Debate
The market's split between fearing under-delivery and over-delivery of compute reflects genuine uncertainty about whether power infrastructure can keep pace with chip orders — a nuance distinct from the simpler "AI bubble" narrative and a factor likely to recur in coming earnings commentary.
3. Credit Markets Are the Transmission Mechanism for AI Trade Stress
Widening triple-C and sovereign spreads show how equity market AI anxiety is spilling into credit, with the "spending glut" framing suggesting this could persist as a structural rather than transient feature of markets. ---
Sentiment Analysis
Overall Market Sentiment: Confused and De-Risking
Privorotsky's own words — investors fearing mutually exclusive outcomes (compute shortage and compute oversupply simultaneously) — best capture a market genuinely uncertain about direction rather than confidently bearish or bullish.
Risk Factors Highlighted
Neo-cloud financing stress: The most levered part of the AI trade faces the sharpest rise in cost of capital.
Power/gigawatt delivery bottleneck: Uncertainty over whether enough power can be energized to use already-ordered compute.
Credit cycle acceleration: A "spending glut" environment is pushing out weaker borrowers faster than expected three months ago.
Free cash flow breakeven timing: Credit markets are becoming "acutely sensitive" to how quickly hyperscalers reach FCF breakeven on AI spend.
European sovereign spread widening: French fiscal deficit concerns are compounding broader credit nervousness.
Market directional confusion: Simultaneous fears of compute under- and over-delivery suggest genuine uncertainty rather than consensus risk.
Record prime-book tech selling: Signals potential for both further downside and a sharp reversal, raising volatility risk either way.
This episode was covered in today's [The Market Signal — 2026-10-26](https://marketsignal.beehiiv.com/p/the-market-signal-2026-10-26), a cross-source synthesis of multiple podcast reports.