According to Woofun AI, the divergence between the U.S. stock and bond markets is approaching a critical threshold. Goldman Sachs (GS.US) strategist Louis Miller has issued a stern warning: credit spreads for AI mega-cloud providers have widened to their highest levels this year, yet corresponding stock prices have not fully reflected this pressure, leaving the market with a binary choice—either the bond market is overly pessimistic, or the stock market still has approximately 15% of downside left to be absorbed.
Alert signals in the credit market are rapidly spreading from the core to the periphery. The Goldman Sachs AI Hyperscale Cloud Provider Credit Spread Index (GSUCHS30) has reached its widest level this year; historically, when this spread was last close to its current level, related stock prices were approximately 15% lower than they are now.
Meanwhile, Goldman Sachs’ broader basket of AI beneficiary stocks has declined approximately 25% from its June high relative to the S&P 500 ex-AI components. The pressure is not limited to large cloud providers, whose credit default swaps (CDS) have just reached record highs. At lower levels of the capital structure, risks are even more severe: of 20 junk-rated AI data center bonds, seven rely on a single tenant, CoreWeave (CRWV.US), whose own bonds have already fallen into distressed territory.
According to data compiled by Woofun AI, the AI sector suffered widespread declines this week, with NeoClouds falling 24.8% over five days, optical networks dropping 21.4%, and storage chips declining 16.6%. The only theme to record a notable gain was "AI-at-Risk" (traditional industries impacted by AI), which rose 3.5%.
The core source of current pressure lies in the turning point of capital expenditure growth. Miller warned that the widening of spreads across the sector forces investors to confront a critical question: What happens when AI spending growth slows? AI capital expenditures have likely peaked, with financing sustainability and capital returns now replacing near-term earnings forecast revisions as the central focus of market debate. The year-over-year growth rate of capital expenditures by hyperscale cloud providers peaked at 116% in the third quarter of 2026, and the market broadly expects a sequential decline thereafter, falling to approximately 20% by the end of 2027.
The impact of this trend on S&P 500 earnings has been clearly quantified: Goldman Sachs estimates that the net capital expenditures of hyperscale cloud providers (after depreciation) contributed approximately 11 percentage points to S&P 500 EPS growth in 2026, declined to 7 percentage points in 2027, and turned negative at -1 percentage point by 2028 due to the onset of significant data center depreciation and amortization. In other words, capital expenditures that once boosted earnings will become a drag once their growth slows.
Internal market indicators are simultaneously signaling warnings, indicating that pressure on the AI sector is not an isolated phenomenon. On the sentiment front, put option volume on Friday reached the second-highest level of the year, surpassed only by June. On the risk appetite front, Goldman Sachs’ Beta Factor Hedge Portfolio (GSP1BETA) has declined 15% from its recent high, with high-beta cyclical stocks now trading below their uptrend channel. On the growth expectations front, the implied economic growth forecast from Goldman Sachs’ Cyclical-to-Defensive Stock Ratio (GSP1CYDE, excluding commodities) has fallen to approximately 2%, suggesting market pessimism on growth exceeds what fundamentals support. On leverage and positioning, hedge funds’ beta exposure exceeds levels implied by their leverage—beta-adjusted net exposure sits at the 67th percentile over the past five years, while net leverage remains at only the 17th percentile. Miller also noted that financial conditions have “shifted from a positive driver of growth to a negative drag,” as higher interest rates and wider spreads are directly pressuring debt-dependent stocks, with Goldman Sachs’ Debt-Sensitive Stock Basket (GSX1DEBT) continuing to underperform the broader market.
Despite the aforementioned warnings, Miller did not predict a full market collapse; he believes the market is still absorbing pressure through rotation rather than broad index declines, and he proposes five trading strategies. First, go long on a "de-AI" broad market: the S&P 500 ex-AI index (SPXXAI) has outperformed the S&P 500 by 586 basis points since October 6, and the correlation between the AI sector and the rest of the market has turned negative. Second, go long on the quality factor but be mindful of technical risks: the Goldman Sachs Quality Factor Hedge Portfolio (GSP1QUAL) rose 20% over three months, with its RSI exceeding 80, indicating a risk of sharp reversal; the European Quality Factor (GSPEQUAL) has gained 7.5% over the past month.