Goldman Sachs warns: Whether the Fed raises rates in December or not, the market may not be prepared.
US Treasury yields have rapidly retreated from their highs, and the two forces that once jointly pushed interest rates higherthe AI investment boom and the energy price shockare now diverging.
Title context: Goldman Sachs warns: Whether the Fed raises rates in December or not, the market may not be prepared.
Text:
US Treasury yields have rapidly retreated from their highs, and the two forces that once jointly pushed interest rates higherthe AI investment boom and the energy price shockare now diverging. Goldman Sachs warns that the market faces greater uncertainty in judging the Fed's policy path, and whether or not there is a rate hike in December, it could trigger a new repricing of risk.
On Thursday, the 10-year US Treasury yield fell 17 basis points in a single day, the largest one-day drop since March 2023, and fell back below 5%. Cosimo Codacci-Pisanelli, a strategist in Goldman Sachs' Fixed Income, Currency and Commodities division, noted in a latest report that what drove yields lower was precisely the AI trade that had previously helped push rates higher.
As AI stocks came under selling pressure, capital flowed back into long-duration US Treasuries, the stock-bond correlation reversed, and bonds once again became a tool for equity investors to hedge risk. But this does not mean the upside risk to rates has been lifted: the US Treasury supply pressure brought by AI capital expenditure and debt financing still exists, and energy prices and inflation trends also make the Fed's next move harder to predict.
Goldman Sachs still regards a December rate hike as its base case, but believes the market will find it difficult to easily reprice a full 25-basis-point hike. At the same time, the interaction among AI growth prospects, the energy shock and bond supply and demand is changing, and the market must both guard against a deterioration in growth expectations and not ignore the tightening risk brought by persistently high inflation.
AI trade reverses, US Treasuries regain safe-haven bids
Codacci-Pisanelli attributes the recent decline in US Treasury yields to three factors: high rates are beginning to weigh on AI-related assets; the tail risk of a sharp slowdown in AI-driven growth has risen somewhat; and the stock-bond correlation has changed, with long-duration US Treasuries once again playing a role in hedging equity risk.
This change contrasts with the situation earlier this month when US inflation data came in hot and stocks and bonds fell in tandem. Nomura strategist Charlie McElligott also previously noted that as the market reassesses AI-driven growth prospects, bonds have once again become a portfolio hedging tool.
Meanwhile, AI-related credit spreads have widened to their widest level of the year, drawing more attention to financing risk. Goldman Sachs had previously warned that the credit market could remain calm until the wave of debt financing driven by AI capital expenditure restarts. Now, the pullback in AI stocks and the widening of credit spreads are testing this growth model that relies on continuous financing and capital expenditure expansion.
However, Goldman Sachs cautions that short-term shifts in capital flows are not enough to reverse the structural supply-demand pressure in the US Treasury market. AI capital expenditure is expected to reach about $1.6 trillion by 2027, and the share of debt financing is also rising. As long as the AI investment outlook does not substantially collapse, related financing demand may still bond supply pressure. Outflows from AI stocks can temporarily support US Treasuries, but sustained buying still requires a broader stock market decline or further deterioration in growth expectations.
December rate hike remains Goldman Sachs' base case, but market pricing is not easy to keep up
Before the Fed's October meeting, despite hot inflation data, market expectations for a rate hike this month had cooled significantly. Goldman Sachs strategist Codacci-Pisanelli noted that the inflation performance provided a basis for a rate hike, but Fed Chair Warsh had previously refused to give clear guidance on subsequent rate decisions, and the market gradually leaned toward the view that the Fed would hold off on a rate hike in October and leave the policy adjustment to December.
Goldman Sachs still regards a December rate hike as its base case, but believes it will be difficult for the market to fully reprice a 25-basis-point hike. As stock market volatility intensifies, investors need to weigh the further impact of a rate hike on risk assets and the risk that economic and financial conditions may deteriorate accordingly.
As of October 23, Goldman Sachs data showed that the market priced about a 90% chance of a December rate hike and about 2.7 cumulative hikes through September 2027. Although a rate hike has been reflected in prices to a considerable extent, there is still significant uncertainty about the future policy path. Goldman Sachs' judgment is that it remains extremely difficult to predict where the Fed will go next.
This also means that the decline in US Treasury yields does not equal a peak in interest rates. If a slowdown in AI investment drags down growth expectations, US Treasuries may continue to benefit from safe-haven demand; but if energy prices remain high and inflation does not fall for a long time, policy rates and bond supply could still push yields higher.
This article is reprinted from "Wallstreetcn", author: Li Jia; GMTEight editor: Chen Siyu.
Related Articles

Nomura: There is a 75% probability that the AI arms race will continue, and the 2027 memory chip shortage is basically locked in.

After record hedge fund selling, Goldman Sachs recommends buying the dip: allocate to AI "selectively," but with limited confidence.

Paul Chan: Corporate confidence in Hong Kong is gradually being reflected in the real economy
Nomura: There is a 75% probability that the AI arms race will continue, and the 2027 memory chip shortage is basically locked in.

After record hedge fund selling, Goldman Sachs recommends buying the dip: allocate to AI "selectively," but with limited confidence.

Paul Chan: Corporate confidence in Hong Kong is gradually being reflected in the real economy






