Editor's note: After the Federal Reserve re-entered a rate-hiking cycle, the focus of market discussion is shifting from "why raise rates in September" to "how far will this round of rate hikes actually go." Oil prices remain high, inflation has become sticky again, and AI infrastructure investment and defense spending remain strong—all of which appear to support higher interest rates. But at the same time, housing, autos, small businesses, and low- and middle-income consumption are already clearly being suppressed by high financing costs. As "inflation is still above target" gradually becomes consensus, a more fundamental question begins to emerge: is today's U.S. economy still suitable for understanding through the traditional framework of "economic overheating—rate hikes—cooling demand"?
In the latest episode of Goldman Sachs Exchanges, Goldman Sachs Vice Chairman and former Dallas Fed President Robert Kaplan discussed the policy path after the Fed's first rate hike in three years. He endorsed the necessity of a September rate hike and also considered another hike within the year reasonable, but compared with the market's more aggressive pricing, Kaplan leans toward the view that the Fed will then need to pause and reassess the economy, rather than mechanically entering a continuous rate-hiking cycle.
In this conversation, Kaplan actually breaks down "how many more times will the Fed hike" into a set of more fundamental structural questions: Does inflation come from demand or supply? Which sectors can policy rates actually suppress? And how much of long-end rates has already detached from the Fed itself and is instead determined by fiscal deficits, energy prices, and bond supply?
First, the nature of inflation is changing, and what the Fed faces is not traditional demand overheating. In the past, when employment was strong, wages were rising, and consumption was expanding, rate hikes could curb inflation by suppressing demand. But current pressures come simultaneously from supply factors such as oil prices, tariffs, and labor supply constraints. The Fed cannot increase oil supply by raising the federal funds rate, nor can it eliminate tariffs, but if a supply shock lasts long enough, it may gradually spread to transportation, goods, and services prices. Therefore, the purpose of raising rates now is not entirely to directly eliminate the initial price shock, but to prevent it from evolving into broader second-round inflation. This means that the key to future policy judgments is not just whether oil prices are high, but whether oil prices are changing the entire price system.
Second, the sectors of the economy that monetary policy can affect are misaligned with the current strongest sources of growth. In the past, rate hikes targeted a relatively synchronized economic cycle; now, however, clear divergence is appearing within the U.S. economy. Housing, autos, small businesses, and low-income consumption are highly sensitive to short-term rates and have already been significantly squeezed; but AI infrastructure and defense investment remain strong. More importantly, large AI projects mainly rely on long-term bonds, credit markets, and equity financing, rather than directly relying on the federal funds rate. This means that if the Fed continues to raise rates, the first sectors hit may not be the hottest parts of the economy, but those that have already cooled. The further policy tools are pushed, the more asymmetric their marginal effects may become.
Third, the market is cramming two different problems into the single price of "rising interest rates." Short-term rates mainly reflect the Fed's future policy path, but 10-year and longer-term U.S. Treasuries are increasingly affected by fiscal deficits, Treasury supply, energy prices, and term premiums. Kaplan therefore emphasizes that higher long-end yields cannot simply be understood as "the market expects the Fed to hike more times." If fiscal deficits still do not improve significantly even with relatively strong nominal growth, then long-term bond investors demanding higher returns is itself a pricing logic independent of the Fed. In other words, the U.S. Treasury yield curve is shifting from a single monetary policy trade into an outcome shaped jointly by monetary, fiscal, and supply risks.
Fourth, the market's current rate-hike pricing itself also includes an uncertainty premium regarding the new policy framework. After Kevin Warsh took office, the market has not yet fully understood his reaction function—that is, under what inflation, employment, and financial conditions the Fed will act with what degree of force. When the policy reaction function is not yet stable, markets often proactively leave a buffer for the "unknown." Therefore, the degree of tightening priced into the current rate curve does not necessarily equal investors' confidence that these hikes will ultimately materialize; it may simply be a risk premium paid for uncertainty about oil prices, war, and the new Fed decision-making framework.
If this conversation is compressed into one judgment, it is this: the real difficulty of this policy cycle is not whether the Fed is willing to continue raising rates, but that traditional interest rate tools are facing an increasingly unconventional economic structure.
In this sense, the subject discussed in this article is no longer just whether the next FOMC meeting will raise rates, but how much monetary policy can still rely on a single policy rate to manage a highly divergent economic cycle when AI CapEx, fiscal expansion, and multiple supply shocks exist simultaneously.
The following is the original content (edited for easier reading and comprehension):
Key Points
Goldman Sachs Vice Chairman and former Dallas Fed President Robert Kaplan said that because clear divergence is emerging within the U.S. economy, the market's pricing of the Fed's subsequent tightening may already exceed what is actually needed.
On the Goldman Sachs Exchanges podcast, Kaplan pointed out that on one hand, artificial intelligence (AI) infrastructure and defense spending continue to boom; on the other hand, rate-sensitive sectors such as housing and autos are already under clear pressure in a high-rate environment. The intertwining of multiple forces may make further policy tightening by the Fed less necessary than current market pricing reflects.
· The Fed may still hike, but only modestly: Kaplan expects the Fed may raise rates one more time, pushing the federal funds rate to about 4%–4.25%, and then pause and reassess economic conditions. But the market currently prices in more hikes. Kaplan believes this includes a certain risk premium, possibly reflecting investors' uncertainty about oil price trends and how Fed Chair Warsh will adjust policy in response to economic data.
· The neutral rate remains important: Kaplan said that although the neutral rate is not the "only criterion" for determining monetary policy, it still has reference value. In his view, one more hike may push the Fed's policy rate above the neutral level, thereby creating a "mildly restrictive" effect on the economy.
· A shock with no ready policy playbook: Kaplan described the current economy as a "low fire, low hire" labor market: labor supply constraints from tariffs and tighter immigration, along with oil price shocks, are appearing alongside a historic capital expenditure boom. This is a supply-side combination with no textbook precedent. Kaplan believes that in this environment, the Fed's focus is no longer on eliminating the initial supply shock, but on limiting as much as possible the "spillover" of inflation into more goods and services prices.
Main Text Points
On September 16, the Fed raised the target range for the federal funds rate by 25 basis points to 3.75%–4.00%. In its statement, the Fed said U.S. economic activity was still expanding at a "solid pace," but inflation remained elevated, so further support was needed for inflation to return to the 2% target.
For Robert Kaplan, this rate hike itself was not surprising. The real question is: is this a preventive adjustment, or the beginning of a new round of continuous rate hikes?
Kaplan's answer is closer to the former.
One more hike may be reasonable, but Kaplan does not think many consecutive hikes are needed
Kaplan believes the Fed's choice to raise rates in September was reasonable.
In the fall of 2025, the Fed cut rates three times in a row. Entering 2026, the economic environment changed: fiscal incentives were still at work, AI infrastructure investment was expanding rapidly, and geopolitical conflicts were pushing oil prices higher. The Fed initially chose to wait, hoping energy prices would fall on their own, but this process did not happen as quickly as expected.
What really worried Kaplan was not just year-over-year inflation, but that recent monthly inflation remained strong.
Data from the U.S. Bureau of Labor Statistics showed that in August, CPI rose 0.4% month over month and 3.4% year over year; core CPI rose 0.3% month over month. This means that although inflation is far below its previous peak, at least in recent data, it has not yet steadily returned to a pace consistent with the 2% target.
But this does not mean Kaplan supports a round of substantial tightening.
The latest FOMC dot plot itself is relatively restrained. September projections showed the median federal funds rate at 4.1% by the end of 2026 and still 4.1% in 2027. From the distribution of dots, among 18 participants, 12 expected the median year-end rate to be 4.125%, 4 expected 4.375%, and only two expected it to remain near the current level.
Kaplan himself leans toward a "one step at a time, then watch" path: one hike has already happened in September; if there is no new inflation shock in October, it can be skipped for now; then in December, another 25 basis point hike can be considered.
In that case, the rate range would rise to about 4.00%–4.25%.
Kaplan estimates that the U.S. nominal neutral rate is roughly around this level as well. The so-called neutral rate refers to the interest rate level that in theory neither clearly stimulates nor clearly suppresses the economy; it is not directly observable data, but a policy estimate.
Therefore, in Kaplan's view, if there is another hike in December, the Fed may already be near neutral. Whether it then needs to move further into restrictive territory requires reassessment, rather than assuming in advance that more consecutive hikes are needed.
Why does the market still bet on more rate hikes? The key is that the "policy reaction function" is not yet clear enough
The real disagreement between Kaplan and the market is not about "whether there will be one more hike," but about how many more there will be after that.
He believes that the degree of tightening currently priced into the interest rate market may be higher than the amount of rate hikes that will actually ultimately be needed.
One important reason is that the market is still paying a risk premium for uncertainty.
First, the policy reaction function of new Federal Reserve Chair Kevin Warsh is not yet fully understood by the market.
The so-called policy reaction function, simply put, is what the market tries to judge: when variables such as inflation, employment, and financial conditions change, with how much force and at what speed the Federal Reserve will adjust interest rates.
When the market is familiar with a central bank governor, investors can often roughly estimate the policy path based on past speeches and decision-making patterns. But if this set of patterns has not yet formed, the bond market usually needs to add an extra "buffer"—that is, a risk premium.
Kaplan therefore believes that part of the rate hike expectations currently in the curve does not necessarily mean investors are convinced that the Fed will definitely hike that much, but may instead be pricing in "we do not yet fully know how Warsh will react."
He also believes that the market is paying a premium for another risk: oil prices remaining high for a long time.
If the energy shock ends quickly, the Federal Reserve can usually choose to "look through" short-term price increases. But if a supply shock lasts six months, eight months, or even longer, it may gradually spread to more price items such as transportation, food, manufacturing, and services.
This is also Kaplan's revision of the view that "supply shocks should not lead to rate hikes."
Raising the federal funds rate certainly cannot increase oil supply, nor can it directly bring oil prices down. But it can suppress demand in other areas, thereby slowing the spread of energy price increases into broader inflation.
Therefore, for the Fed, what really matters is not how much energy prices themselves have risen, but whether an exogenous shock will evolve into broader second-round inflation.
The biggest challenge: AI is booming, but interest-rate-sensitive sectors are no longer hot
This is also why Kaplan does not favor mechanically raising rates continuously.
Today's U.S. economy is not overheating across the board in the traditional sense, but has shown very obvious divergence. AI infrastructure investment is still growing rapidly, and defense spending remains strong; but housing, automobiles, and businesses serving low- and middle-income consumers have already weakened noticeably. This creates a monetary policy dilemma: the sectors most easily hit by the federal funds rate are precisely no longer the hottest sectors of the economy.
Real estate is the most direct example.
High mortgage rates reduce households' ability to buy homes, while at the same time real estate developers themselves often need to rely on short-term financing to support inventory and project turnover. Therefore, when the Fed raises short-term rates, these businesses may face both weakening demand and rising financing costs.
Small businesses and individuals relying on floating-rate loans also face similar pressure. But the financing structure of AI infrastructure investment is not exactly the same.
Large technology companies and infrastructure projects rely more on corporate bonds, the long-term bond market, and equity financing. Therefore, compared with the federal funds rate, they may care more about the entire U.S. Treasury yield curve and credit spreads.
Kaplan's judgment is that merely raising the short-term policy rate is not enough to significantly prevent the expansion of AI infrastructure.
This also explains why the U.S. economy can simultaneously show two seemingly contradictory phenomena: on one hand, a capital expenditure boom and resilient corporate profits; on the other hand, housing and some consumer sectors already feeling obvious pressure.
He summarizes the current labor market as "low fire, low hire"—low layoffs, low hiring.
This is different from the typical economic overheating of the past. Employment has not deteriorated on a large scale, but companies' willingness to hire is also not strong, so the labor market is temporarily not the clearest monetary policy signal.
Kaplan believes that what the Federal Reserve is actually facing now is a very rare combination: a historic capital expenditure cycle, together with multiple supply shocks such as energy, tariffs, and labor supply constraints occurring at the same time.
The traditional model of "economic overheating—rate hikes—demand cooling" has therefore become less useful.
10-year U.S. Treasury yield breaks above 5%, and the problem is no longer just the Fed
If short-term rates mainly answer "how much more will the Federal Reserve hike," then long-term rates are answering another question: at what price are investors willing to hold U.S. government debt for the long term.
This is also the point Kaplan believes is most easily overlooked in the current market.
U.S. Treasury data show that the 10-year U.S. Treasury yield was 4.96% on September 22 and rose to 5.11% on September 23; the 20-year and 30-year yields rose to 5.45% and 5.40%, respectively, over the same period.
Long-term yields are already significantly higher than the current policy rate.
Kaplan believes that this part of the move cannot simply be explained as "the market expects the Fed to continue raising rates." U.S. Treasuries with maturities of 10 years or even longer will increasingly be affected by fiscal deficits, bond supply, and long-term inflation risks.
CBO data for September show that in the first 11 months of fiscal year 2026, the U.S. federal budget deficit was about $2 trillion, superficially about $6 billion less than the same period last year; but after excluding payment date misalignments, the deficit for the same period this year was actually about $82 billion higher than last year.
This means that although nominal economic growth remains strong, the U.S. fiscal gap has not improved noticeably.
In Kaplan's view, this will force long-term bond investors to demand higher yield compensation. U.S. Treasury bond buybacks can improve market liquidity and debt management to some extent, but Kaplan emphasizes that this is not the same thing as truly reducing the fiscal deficit. Therefore, he believes that the core issue for long-term rates has increasingly shifted toward fiscal policy, rather than monetary policy itself.
This is also why, after the Fed's September rate hike, the U.S. Treasury curve did not simply reprice according to "another 25 basis points"—a large amount of tightening expectations had already been reflected in the curve beforehand, while the long end also has its own fiscal logic.
What to watch next? Not "whether the Fed is hawkish," but whether inflation continues to spread
Kaplan's baseline judgment for the next meeting still leans toward "pausing."
The Federal Reserve's next FOMC meeting will be held on October 27–28. Before that, the market will also receive two sets of key inflation data: the BEA will release August personal income and outlays data on September 30, which includes the PCE price index; the BLS will release September CPI on October 14.
If monthly inflation begins to cool, oil prices do not further spread to other goods and services prices, and housing, automobiles, and consumption continue to bear the pressure of high interest rates, then the path Kaplan describes of "one more hike in December, followed by a pause to observe" will be easier to establish.
Conversely, if PCE and CPI accelerate noticeably again, or business surveys show that energy and other supply shocks are transmitting to more and more price items, the Federal Reserve may not have room to wait until December.
Therefore, what truly needs to be observed now is no longer a simple question of "hawkish or dovish."
The more critical variable in this policy cycle is: whether the supply shock has merely pushed up a few prices, or has begun to change the entire inflation process.
If it is only the former, the market may indeed have bet on too many rate hikes; if it is the latter, then today's seemingly high interest rate pricing may gradually be validated by fundamentals.
