Based on CME data, the market currently assigns a near 90% probability to a rate hike at the September meeting. The U.S. Treasury market is also pricing in such a move. The firm argues that a rate hike is not necessarily a negative event, unless it becomes a series of consecutive hikes; conversely, holding rates steady does not automatically guarantee positive outcomes. In terms of key assets, aside from a relatively optimistic view on U.S. equities, particularly technology stocks, most other asset classes have already fully priced in the September hike expectations.

For U.S. Treasuries, short-end yields are likely to rise, while the term premium on long-end bonds may initially decline, with the possibility that long-end yields could eventually peak and drift lower. The outlook for U.S. equities is not pessimistic, as short-term disruptions could even present better entry points. If the Fed hikes, it could restore market trust and provide support for the U.S. dollar, and vice versa. For gold, the potential upside from a hold is greater than from a hike; currently, it resembles a call option with uncertain upside potential.

Where the Debate Stands Now

The question of whether the Fed should raise rates at all has been the most divisive, volatile, and perplexing issue over the past month. Warsh's ambiguous stance on inflation targets and interest rate reflexivity during the July FOMC meeting served as the catalyst for the Treasury market turmoil, even prompting the Treasury Department to intervene directly in buybacks without effect. The core issue is not merely an oversupply of bonds but rather the market's concern that the Fed cannot offer the reassurance needed to hold long-term Treasuries with confidence. From that point onward, the decision to hike transcended inflation and data alone. This is why Warsh felt compelled at the August Jackson Hole meeting to reiterate the Fed's commitment to the 2% inflation target and its responsibility to fight inflation with hawkish rhetoric, which did help stabilize expectations, evidenced by the term premium retreating by 20 basis points from its peak.

From the perspective of maintaining Fed credibility, hiking rates appears to be the best course of action unless forthcoming data consistently surprises to the downside. However, the data have not cooperated; subsequent releases, including non-farm payrolls, PPI, and CPI, have consistently exceeded expectations. The Fed has found itself backed into a corner, forced to deliver on its hawkish commitments. According to CME data, market expectations now imply a near 90% chance of a September hike. The Treasury market is also pricing this in, with the 2-year yield climbing 32 basis points from 4.34% to 4.66% since late August. The current 4.96% yield on the 10-year Treasury, after stripping out the 76 basis point term premium, implies a rate expectation of 4.18%, which is 60 basis points above the benchmark rate of 3.5%, pricing in more than two rate hikes.

The issue has now shifted from 'whether to hike' to 'what happens after the hike.' In the firm's view, the market often avoids the topic due to fear of the hike's impact and fails to prepare, yet when it becomes apparent that a hike is unavoidable, there tends to be excessive worry about the aftermath. The firm believes such concerns are largely unnecessary; the key is not the hike itself but the underlying reason for it.

Analyzing the September Decision

From a fundamental perspective, the data presents a close call, but from the angle of preserving Fed credibility, a hike is the preferable outcome. If the same rhetoric delivered at Jackson Hole had come from Powell, a September hike would almost certainly be on the table. Warsh reiterated the 2% inflation target and emphasized the Fed's duty to focus on price stability with a notably stringent tone. However, Warsh is not Powell, and his ambiguous style contrasts sharply with Powell's. He believes excessive market reliance on Fed guidance creates a 'hall of mirrors' effect and does not advocate for extensive communication with the market, and his coordination with the market remains inadequate. Consequently, market expectations have remained volatile; a dovish speech from Waller in early September even dampened rate hike expectations, as he pointed to signs of cooling inflation and suggested giving disinflation a chance.

From a fundamental standpoint, the evidence is mixed, and even with several data points exceeding expectations, one can find less robust signals. August's inflation reading was largely driven by oil prices, airfares, and hotel rates, factors unlikely to persist. The firm's calculations suggest that unless oil prices remain above $95 per barrel, year-end CPI is likely to fall from its current 3.4% level. Growth in the U.S. is also K-shaped, though with less extreme divergence; at such high interest rates, many traditional demand sectors will quickly feel the pressure of elevated costs, particularly real estate. As such, fundamentals may not be the deciding factor here. From the perspective of maintaining Fed credibility, a hike is the best course, and this is the crux of the decision. Warsh's deliberate ambiguity at the July FOMC threw market expectations into disarray, pushing the term premium from 0.65% to 0.9%. If the data were to weaken, the Fed might have more room to maneuver, but with non-farm payrolls and inflation coming in above forecasts, the Fed has been maneuvered into a corner. Although these indicators have limitations, they are all that is available before the FOMC meeting. If the Fed were to refrain from hiking despite its hawkish signals, the market's perception of its credibility would suffer more severely, potentially leading to a more serious trust crisis and loss of control over the Treasury market.

Post-Hike Market Dynamics

A short-lived or 'preventive' rate hike often triggers a counterintuitive reaction, as the news of the hike itself can mark a clearing of the air. The market tends to avoid preparing for a hike before it happens, only to overestimate the shock once it is imminent. When analyzing the impact, the reason for the hike may be more critical than the hike itself. A sustained and aggressive hiking cycle would inevitably raise financing costs and tighten financial liquidity, leading to prolonged and substantial impacts on growth and markets, as seen in 2022 and 2023 when the Fed hiked by 525 basis points over 16 months. Conversely, a small or preventive hike causes limited disruption, and often, when the expectation is fully priced in, the actual hike can be seen as an 'exhaustion of bad news.' Historically, a modest or preventive hike tends to be a cue for markets to rebound, as the event is already priced in. A case in point is the 1997 rate hike, where the environment was similar: inflation was rising but not out of control, financial conditions were manageable, and growth was robust. As the market had gradually priced in the hike before March 1997, Treasury yields peaked and declined shortly after the hike, and U.S. equities resumed their rally within about a week.

Conversely, the 'preventive rate cuts' of 2019, 2024, and 2025 followed a similar script. In these cycles, the economy faced pressure but not a stall, necessitating only modest policy adjustments. As easing expectations were usually pre-priced, the post-cut trading theme quickly shifted from 'easing expectations' to 'growth improvement.' Treasury yields and the dollar index rebounded shortly after the cuts, and gold performed better before the cut than after, unless a major narrative, as in 2025, sharply elevated prices. U.S. equities gained somewhat after the cuts as the easing transmitted to corporate earnings. Hong Kong stocks benefited to some extent from improved denominator effects, but domestic fundamentals were the decisive factor, as seen in September 2025 when the post-cut rally peaked.

If multiple hikes become necessary, the market adjustment is more significant and prolonged, as exemplified by the 2022-2023 tightening cycle. In early 2022, U.S. CPI was running at 7-8% year-over-year, surpassing 9% by June, amid pandemic supply chain disruptions and the Russia-Ukraine conflict. Monetary policy was clearly behind the curve, compelling the Fed to hike rapidly from March 2022 through July 2023, totaling 525 basis points. This tightening exerted a more sustained drag on asset prices. The 10-year Treasury yield only peaked in October 2023, up 278 basis points from the start of the hiking cycle, while the rate expectation peaked in March 2023, up 171 basis points. The S&P 500 and gold bottomed out in October-November 2022, corresponding to the confirmed inflection point for U.S. inflation, with declines of 18% and 16%, respectively. Excluding the support to risk assets from the AI trend since 2023, and focusing on the most intense period of tightening shock from March to December 2022, all major assets, except for the dollar index and short-end Treasuries, declined.

Possibility of a Sustained Hiking Cycle

At present, the foundation for a series of large hikes does not exist, unless oil prices spiral out of control and stay above $95 per barrel. Conversely, with the K-shaped U.S. economy and high rates suppressing traditional demand, the fundamentals may not withstand a sustained tightening, which could paradoxically constrain further hikes. Traditional demand is already constrained by high rates, with the August ISM manufacturing PMI falling and new orders slipping from 56.7 to 53.7. Rate-sensitive real estate data has weakened again, with existing home sales declining. Additionally, ongoing layoffs in IT and financial services could disrupt consumption. AI, a major growth engine, is also becoming more sensitive to financing conditions. With free cash flow at cloud vendors turning negative, their reliance on external financing is growing. Although the return on invested capital at 20% for cloud vendors clearly exceeds the weighted average cost of capital at 9%, a significant rise in financing costs could dampen their capex appetite.

What could lead to a shift toward multiple consecutive hikes? First, if oil prices run amok and stay at elevated levels; in the base case, oil prices are expected to average $80-90 per barrel in H2, and the firm calculates CPI could fall to around 3.0% by year-end. In an extreme scenario where oil prices stay at $100 per barrel or more, year-end CPI could exceed 3.6%, putting significant pressure on the Fed. Second, if AI capital expenditure once again exceeds expectations, it could push up hardware prices and core inflation, as seen in the August PPI readings, where electronic components were a notable contributor. Moreover, since AI already contributes 40% to the U.S. GDP growth rate, a rapid acceleration in capex would make it difficult for the Fed to justify rate cuts based on weak fundamentals.

Asset Allocation Implications

The firm differs from market consensus in that it does not view rate hikes as inherently detrimental, unless they become consecutive. Conversely, not hiking might not be beneficial. A one-off hike, often pre-priced, tends to result in a counterintuitive market reaction, as seen in 2024 and 2025 with rate cuts and in 1997 with a hike. Therefore, any disruption from a hike could present better buying opportunities. In contrast, a forced pause might provide temporary relief but could sow more significant problems down the line. Examining the pricing of future hikes across various assets, rate futures imply 3.7 hikes, more than Treasuries and copper at 1.7 hikes, gold at 1.2 hikes, the Fed's dot plot at 0.7 hikes, the Dow at 0.6 hikes, and the S&P 500 at 0.2 hikes, with the Nasdaq at -0.4 hikes. Compared to the July FOMC conclusion, the 10-year yield expectation has risen by 15 basis points, and the 2-year yield by 30 basis points. This suggests that, apart from U.S. equities, especially tech stocks, most assets have already fully priced in the September hike, akin to the situation before the March 1997 hike. Specifically:

For U.S. Treasuries, short-end yields are set to rise, while the long-end term premium may initially decline, and overall long-end yields could gradually peak and fall. A single hike would correspond to a midpoint of 4.5-4.7%, and current long-end yields already reflect about 1.7 hikes, fully capturing 'preventive hike' expectations. A hike that restores confidence could also narrow the term premium, making the risk-reward profile for Treasuries quite compelling. For U.S. equities, the outlook is not pessimistic, and short-term disruptions may provide better entry points. The firm raised its S&P 500 target to 7800-8000 in early June. After reaching that level, the market has been in a consolidating mode due to the need for new catalysts for earnings driven by AI and macro disruptions. With a rate hike out of the way, if the AI industry sees new progress, U.S. equities could have room to rise, and any significant pullback would offer a better buying opportunity. For the U.S. dollar, a rate hike would restore market trust and lend support, while the opposite would apply if the Fed holds. The firm's dollar model projects the dollar index to trade in a 96-98 range in H2. For gold, the upside from a hold is greater than from a hike, and it currently resembles a call option with uncertain upside. Based on static calculations from Treasury yields and the dollar, gold's support level is around 4400-4600. Unless there are consecutive hikes, downside pressure on gold is relatively contained, but its upside requires grander narratives. If the Fed does not hike, it could create a narrative of trust erosion and de-dollarization, whereas a hike would undermine that narrative, so the upside may be less than would be the case without a hike, awaiting more catalyst.