The Federal Reserve raising interest rates this year is still not the baseline scenario, but it is no longer just a tail risk. Disruptions in the Strait of Hormuz are pushing up commodity price pressures, while AI-related capital expenditures are straining parts of the global supply chain. Markets are beginning to reprice a "hawkish pivot": the probability of a rate hike by the December 2026 FOMC meeting now exceeds 60%, and a full 25-basis-point hike is fully priced in by March 2027.
According to analysis from Barclays FICC economic research by Jonathan Millar et al. in a May 18 Fed commentary, the core assessment is: "While our base case also does not anticipate a hike before the end of 2027, the upside risks to the policy rate have intensified." The three potential paths triggering a hike are clear: long-term inflation expectations becoming unanchored; core inflation remaining stubbornly high even after tariff impacts fade; and demand outpacing supply, particularly if the AI investment cycle and wealth effects materialize before productivity gains.
The baseline path remains relatively dovish: the Fed holds rates steady through 2026, with the next move being a 25-basis-point cut in March 2027. This outlook hinges on two premises: disturbances in the Strait of Hormuz end relatively quickly, with tariff pass-through and energy-related price pressures receding; simultaneously, consumer spending slows, leading to an overall cooling of aggregate demand.
However, the probability distribution has shifted significantly away from a dovish stance. In subjective probability scenarios, a 25-basis-point cut in 2027 has a 35% probability; maintaining rates unchanged through end-2027 has a 30% probability; a hiking scenario (approximately 50-100 basis points) has a 25% probability; and a recession triggering significant cuts has a 10% probability. In other words, the most likely deviation from the baseline is not an immediate hike, but a delay in rate cuts.
Markets No Longer Treat Hikes as a Tail Risk The shift in interest rate markets is direct. Before the Iran conflict, markets were pricing in more cuts; since then, pricing has rapidly shifted toward hike risks.
The underlying variable is not a single oil price shock. The ISM manufacturing and services prices paid indices, along with the New York Fed's Global Supply Chain Pressure Index, all indicate rising cost pressures. Concurrently, the US labor market has not deteriorated significantly, with the unemployment rate remaining low and the three-month average of non-farm payrolls still giving an impression of relative stability.
This presents a conundrum for the Fed. If it were merely a one-time shift in the price level, policy could "look through" it. But if the shock persists long enough, inflation expectations, wages, and corporate pricing behavior begin to adjust, the problem transforms from a supply shock into inflation persistence.
The Shortest Path to a Hike: Long-Term Inflation Expectations Becoming Unanchored The most direct trigger would be long-term inflation expectations starting to loosen.
The focus should not be on one or two months of CPI data, but on 5- to 10-year market-based inflation expectations, particularly the 5y5y inflation breakeven rate. If such indicators rise persistently and their movements become disorderly or decoupled from short-term inflation changes, the Fed would likely interpret this as a signal of damaged credibility for its 2% inflation target. A concurrent rise in long-term inflation expectations from surveys like those from the University of Michigan and the New York Fed would reinforce this assessment.
We are not at that point yet. Long-term inflation breakeven rates do not currently show the Fed's credibility being questioned, and survey-based long-term expectations are only somewhat elevated. The real risk is if disruptions in the Strait of Hormuz and commodities last too long or are too severe, leading markets to doubt the Fed's willingness to bear the growth and employment costs for the 2% target.
If such signs emerge, policy communication would likely turn hawkish first. The Fed would not wait for all data to confirm the shift before alerting markets.
The Real Trouble: If Core PCE Doesn't Fall After Tariffs Fade The second path is slower but more realistic: core inflation repeatedly exceeding expectations.
A key threshold in the framework is a core PCE monthly rate around 0.18%, roughly consistent with the 2% annual target. If core PCE persists above this level while tariff-related pressures are theoretically fading, the Fed's room for continued patience would shrink.
Several components are more critical: core goods not showing the expected disinflation; core services excluding housing (the "supercore") cooling only modestly; global supply chain pressures rising again; and trimmed-mean or median inflation measures strengthening, indicating pressures are no longer concentrated in a few items.
The keyword for this path is "persistence." A single month's data is insufficient, and a few months may not be enough. But if it persists for several quarters, it becomes difficult for the Fed to dismiss it as a temporary disturbance.
There is also a fork in this path: if high inflation coincides with strong demand, the policy bias would shift toward tightening. If demand has already weakened significantly while inflation remains high, the Fed's dual mandate would come into conflict. However, after the 2021-2022 inflation shock, the bar for not prioritizing price stability is very high.
AI Might Boost Demand Before Immediately Lowering Inflation The third path is less related to Middle East conflicts and stems more from domestic US demand.
The AI investment cycle is accelerating. Private domestic final purchases have not slowed significantly, AI-related investment has reaccelerated this year, and financial conditions, as measured by Fed models, remain supportive of growth. If the capital expenditures and stock market wealth effects from AI materialize before gains in productivity and cost efficiency, the result might not be lower inflation, but demand surging first.
This differs from the technology narrative of the Greenspan era. Back then, productivity gains were harder to identify in real-time, with supply improvements occurring first and demand responding later. This time, the potential productivity benefits from AI are already widely anticipated by markets and reflected in financial conditions and spending behavior.
The data is also two-sided. Nonfarm business sector productivity grew nearly 3% over the four quarters ending Q1 2026, about double the pre-pandemic pace. However, a San Francisco Fed measure adjusted for capacity utilization suggests this growth rate might be overestimated by about 1.5 percentage points. In other words, the apparent supply improvement may not be sufficient to support the demand that has already been pulled forward.
The Fed would look to several traditional signals to judge overheating: growth exceeding trend, the unemployment rate falling below the 4.0%-4.3% NAIRU range, wages reaccelerating and outpacing real productivity growth. Evidence of wage reacceleration is not strong yet, but this is a line to watch closely.
Baseline Remains a 2027 Cut, But the Premises Are Stringent The base case remains unchanged: an extended pause, followed by a 25-basis-point cut in March 2027. The logic is that by then, energy, tariff, and supply chain-related price pressures will have receded, and core PCE will have slowed noticeably, giving the Fed room to move the policy rate toward its long-run neutral level.
This path is highly sensitive to the duration of Strait of Hormuz disruptions. If disturbances are brief and tariff pass-through fades gradually, the window for cuts could still open. If disruptions are prolonged, core inflation and inflation expectations would likely close that window first.
Consumption is also crucial. Real household disposable income growth has slowed markedly over the past year, primarily due to slower employment growth. If labor supply growth continues to slow, consumer spending should also cool. If this assumption fails, and AI capital expenditures and wealth effects continue to support demand, the Fed will face a tougher judgment: whether current policy is sufficiently restrictive.
Therefore, the Fed risk in 2026 is not simply a shift from "rate cuts" to "rate hikes." A more accurate description is: the path to rate cuts is being squeezed simultaneously by supply shocks, sticky core inflation, and AI-driven demand spillovers. Rate hikes still require harder data to trigger, but they have returned to the policy discussion table.
Comments