CICC: What will happen if the Federal Reserve raises interest rates?
The Zhitong Finance App learned that CICC released a research report saying that according to CME data, the current market expects the probability of interest rate hikes in September to be close to 90%. The US bond market is also raising prices and interest rates. The bank believes that raising interest rates is not necessarily a bad thing, unless it is a continuous rate hike; conversely, not raising interest rates is not necessarily a good thing. In terms of key assets, with the exception of US stocks (especially technology stocks), which are relatively optimistic, the rest of the assets have been fully included in the September interest rate hike expectations. US debt: Interest rates on short-term bonds are higher, long-term debt maturity premiums fall first, and even long-term bonds as a whole may gradually peak and fall; US stocks: not pessimistic, short-term disturbances or even provide better buying points; US dollars: if the Fed raises interest rates, it will support the US dollar, and vice versa; gold: there is more room for no interest rate hike than interest rate hikes. Currently, there is more room for upward uncertainty.
CICC's main views are as follows:
Whether the Federal Reserve actually wants to raise interest rates is the most fragmented, fluctuating, and tangled question in the market over the past month. Walsh's “vague” statement about inflation targets and market interest rates at the FOMC meeting at the end of July became the starting point of a “storm” of US debt, forcing the US Treasury to go directly to the market to buy back treasury bonds. The core is not because of the sudden increase in bond supply that cannot be digested, but because he is worried that the Federal Reserve will not be able to give everyone peace of mind to hold long-term US bonds.
Chart 1: The US debt maturity premium once climbed from 0.65% at the end of July to 0.90% in mid-August
From that moment on, not raising interest rates was not just a matter of inflation and the data itself. Therefore, at the Jackson Hole conference at the end of August, Walsh had to re-emphasize the 2% inflation target and the hawkish statement that anti-inflation is the Federal Reserve's unshirkable duty (rather than self-fulfilling by raising market interest rates) to stabilize market expectations. This did have some effect, measuring that investors' willing term premiums fell 20 bps from a high point. From the perspective of maintaining the Federal Reserve's credibility, raising interest rates is the best option, unless subsequent data “cooperates” to fall short of expectations.
However, the biased data did not match. Since then, several data, such as non-agricultural, PPI, and CPI, have successively exceeded expectations, and the Federal Reserve has indeed been “unfavorable” and has been forced step by step into a “corner”, forcing it to fulfill its hawkish promises.
According to CME data, the current market expects the probability of interest rate hikes in September to be close to 90%. The US bond market is also raising interest rates: since the end of August, interest rates on 2-year US bonds have risen from 4.34% to 4.66%; currently 4.96% of 10-year US bonds have deducted 76 bp of maturity premium, and the remaining 4.18% interest rate, which reflects interest rate hikes, is 60 bps higher than the 3.5% benchmark interest rate, implying that interest rate hikes will exceed expectations of two rate hikes.
Chart 2: The current market expects the probability of interest rate hikes in September to be close to 90%
Will interest rates be raised in September? Fundamentals are in between, but from the perspective of maintaining the Federal Reserve's credibility, it is best to add
As a frame of reference, if the same statement on JacksonHole comes from Powell's words, then it's almost certain that interest rates will rise in September. At the Jackson Hole meeting, Walsh reiterated the 2% inflation target, stressing that the Federal Reserve must perform its duties and focus on prices. The tone cannot be described as unharsh. However, after all, Walsh is not Powell. His vague style is the exact opposite of Powell's. He believes that excessive reliance on the Federal Reserve's guidelines will bring about a “mirror office effect” and does not advocate full communication with the market; furthermore, there is insufficient run-in with the market. As a result, market expectations are still repeated, and Waller's dovish statement in early September suppressed expectations of interest rate hikes for a while. He pointed out that recent inflation has shown signs of falling, energy prices have not spread to other price segments, and “give inflation a chance.”
From a fundamental point of view, it is “between the two”. Even if several figures, including inflation, exceed expectations, it is still possible to find “not that strong” evidence. 1) Inflation: The increase in August data was mainly driven by factors such as gas prices, air tickets, and hotels, which seem unlikely to continue. The bank estimates that unless the oil price center is maintained and exceeds 95 US dollars, the CPI is likely to fall from the current 3.4% year-on-year level at the end of the year. 2) Growth: The growth of the United States is also K-type. Although the differentiation is not that extreme, under such high interest rates, many traditional demands will also quickly bear the pressure of high costs, especially real estate, so fundamental factors are not a decisive factor.
Chart 3: US growth is also K-type, although it is K-type, which is less extremely differentiated 
From the perspective of maintaining the Federal Reserve's credibility, it is best to add. This is the key to deciding whether or not to raise interest rates. The FOMC Wash's “deliberate vague” in July disrupted market expectations. The US bond maturity premium rose from 0.65% at the end of July to 0.9% in mid-August, forcing Walsh to make hawkish promises at the Jackson Hole meeting. If the data weakens, the Federal Reserve is estimated to have room for maneuver, but as non-agricultural agriculture and inflation continue to exceed expectations, the Fed will gradually be forced into a “wall corner.” Although these short-term indicators all have limitations, they could only rely on these data before the FOMC. Just imagine, if this meeting were still “not forced”, how would the market view previous hawkish statements? I'm afraid it will only cause a more serious crisis of trust and US debt getting out of control.
What will happen after the rate hike? Short-term “preventative interest rate hikes” are often “doing the opposite” and are a negative cashout
The problem with the market is that in the past, they avoided interest rate hikes because they were afraid of the impact of interest rate hikes and were unprepared, but when they saw that the risk of interest rate hikes was inevitable, they were overly worried about the impact after the rate hike.
When analyzing the impact of interest rate hikes, why are interest rate hikes more important than interest rate hikes themselves: 1) Long-term and drastic interest rate hikes will inevitably have a longer-term and major impact on actual growth and financial markets by raising financing costs and tightening financial liquidity. The most typical example is that after the outbreak of the Russian-Ukrainian situation in 2022, the Federal Reserve started 16 months to deal with high supply inflation caused by the mismatch between supply and demand due to the pandemic and high oil prices. 2) In contrast, small or even precautionary interest rate hikes, because of the small magnitude and limited disturbance, and even in many cases expectations are taken into account early, the implementation of interest rate hikes often means that the profit is exhausted. There are many “classic cases” in history:
If only a small interest rate hike is even a “preventative rate hike”: since expectations have been taken into account in advance, the implementation of interest rate hikes may mean running out of profit. This is not necessarily bad for the market; they are all “doing the opposite.” A typical example is the interest rate hike in 1997: the US environment at that time was similar to today. Inflation was heating up but not out of control, financial conditions were limited, and economic growth remained strong, supported by technological trends. However, since the market had gradually absorbed expectations of interest rate hikes before the FOMC in March 1997, US bond yields peaked and fell quickly after the policy was implemented, and US stocks resumed their rise in about a week.
Chart 4: US bond interest rates hit a phased high of 6.9% shortly after the implementation of the 1997 rate hike
Chart 5: US stocks began to rebound one week after the 1997 rate hike was implemented
In turn, “preventative interest rate cuts” in 2019, 2024, and 2025 were also staged in a similar “script”. In these three rounds of interest rate cuts, although the economy faced some downward pressure, there was no risk of stalling, so the policy only needed minor adjustments to provide support. Since expectations of interest rate cuts are often taken into account by the market in advance, the main trading line will quickly switch from “loose expectations” to “improved growth” after actual implementation. The US debt and the US dollar index bottomed out soon after the interest rate cut was implemented, and the response was most sensitive. Gold also performed better before the interest rate cut than after the rate cut, unless there was a grand narrative like 2025 that drastically boosted the price of gold; US stocks waited for interest rate cuts to be transmitted to the molecular side, so there was still a certain increase after the interest rate cut; Hong Kong stocks benefited from improvements on the denominator side to a certain extent, but domestic fundamentals were the decisive factors, such as September 2025 Interest rates rose for a while after cutting, but it also became a high point in this round.
Chart 6: In the “preventive interest rate cuts” in 2019, 2024, and 2025, after the interest rate cuts are implemented, the main trading line will quickly switch from “loose expectations” to “improved growth”
If interest rates need to be raised more than once: The market adjustment is larger and takes longer, as is typical of the 2023-2023 austerity cycle. The US CPI reached 7-8% year-on-year in early 2022, and more than 9% in June due to the double impact of the pandemic supply chain disruptions and the Russian-Ukrainian conflict. At this point, monetary policy clearly lags behind the inflation situation (fall behind the curve), and the Federal Reserve had to raise interest rates quickly from March 2022 until July 2023, with a cumulative increase of 525 bps.
This round of austerity has been suppressing asset prices for significantly longer. Interest rates on 10-year US bonds only peaked in October 2023, rising 278 bps from when interest rate hikes began, and interest rates are expected to peak in March 2023, rising 171 bps; the S&P 500 and gold bottomed out in October-November 2022, corresponding to when the inflection point in US inflation was confirmed, with declines of 18% and 16%, respectively. Further excluding the support of risk assets from AI industry trends since 2023, looking only at March to December 2022, when the impact of austerity is most concentrated, it can be found that with the exception of the US dollar index (4.5%) and short-term US bonds (1.4%), which represent cash, all major assets such as stocks, long-term bonds, and gold have declined.
Chart 7: Interest rates on 10-year US bonds will only peak in October 2023, and interest rates are expected to peak in March 2023
Chart 8: S&P 500 bottomed out in October 2022, down 18%
Chart 9: Gold bottomed out in November 2022, falling 16%
Chart 10: From March to December 2022, with the exception of the US dollar index and short-term US bonds, which represent cash, all other major assets fell
Can interest rates be raised continuously and drastically? There is currently no such basis unless oil prices get out of control
At present, there is no basis for continuous and drastic interest rate hikes, unless the oil price center remains above 95 US dollars or higher, making it difficult for CPI to fall or rise year on year. Conversely, the current K-type division of the US economy, and high interest rates are in turn suppressing traditional demand, and US fundamentals are not necessarily able to “bear” the continuation of interest rate hikes. This will also become a “backlash” to curb repeated interest rate hikes.
Traditional demand is already constrained by high interest rates. The ISM manufacturing PMI declined in August, when high interest rates were suppressed, and the forward-looking new order segment slipped from 56.7 to 53.7. Real estate data that is sensitive to interest rates has also been affected, and sales of existing homes have declined again. Furthermore, continued layoffs in the IT and financial industry in recent years may also disrupt consumption.
Chart 11: The recent rise in interest rates is driving the previously repaired real estate data back down again
As an important growth engine, AI is also becoming more sensitive to financing conditions. As cloud vendors' free cash flow began to turn negative, the US AI investment's dependence on external financing gradually increased. Although the ROIC (20%) of cloud vendors still clearly exceeds WACC (9%), in the context of AI demand facing phased bottlenecks, if financing costs rise sharply, it may weaken cloud companies' willingness to spend capital.
Chart 12: Free cash flow from cloud vendors other than Microsoft and Meta has turned negative in the second quarter
Figure 13: Currently 19.9% of cloud vendors have a weighted ROIC higher than WACC's 9%
Under what circumstances will interest rates be raised multiple times and continuously? 1) Oil prices are out of control and have been high for a long time: Under the benchmark situation, oil prices were at the center of $80-90 in the third and fourth quarters, and the bank estimates that the CPI will fall back to around 3.0% year over year. However, in extreme cases, assuming that the oil price center remains at $100 (average since September) or even higher, then the CPI may exceed 3.6% year over year, which will put more pressure on the Federal Reserve at that time. 2) When AI capital expenditure once again exceeds expectations: On the one hand, it will cause technology and hardware prices to rise, driving up core inflation. For example, in the August PPI month-on-month ratio, AI already contributed 40% of the US GDP month-on-month. If capital expenditure accelerates drastically and even spreads to other fields, then it will be difficult for the Federal Reserve to use “poor fundamentals” as a reason to cut interest rates.
Figure 14: CPI YoY Path Estimation
Chart 15: AI contributed 40% of US growth in the first quarter
Implications for assets: Raising interest rates is not necessarily a bad thing; after venting, they provide a better point to buy; not raising interest rates is not necessarily a good thing
Based on the analysis of the reasons for interest rate hikes and the impact of different interest rate hikes in the above, the difference between the bank and the market's idea is that they don't think that interest rate hikes are a monster; raising interest rates is not necessarily a bad thing, unless it is a continuous rate hike; conversely, not raising interest rates is necessarily a good thing. That's not to say that interest rate hikes won't cause disturbances, but short interest rate hikes are more likely to be recorded in advance, often in the opposite way. This is the case with the 2024 and 2025 rate cuts, as was the case with the 1997 rate hike, so if they cause disturbances, they may provide a better buying point. Conversely, if forced interest rate hikes provide a short-term boost, it could cause more trouble in the future.
Looking at interest rate hikes for the next year, which includes various types of assets, interest rate futures (3.7 times) > US bonds and copper (1.7 times) > gold (1.2 times) > Federal Reserve bitmap (0.7 times) > Dow Jones (0.6 times) > S&P 500 (0.2 times) and NASDAQ (-0.4 times). Compared with when the FOMC ended in July, interest rates on 10-year US bonds are expected to rise by 15 bps, and 2-year US bonds are expected to rise by 30 bps. This means that with the exception of US stocks (especially technology stocks), which are relatively optimistic, the rest of the assets have fully taken into account the September interest rate hike expectations, similar to the situation before the interest rate hike in March 1997. More specifically:
Chart 16: Judging from the interest rate hike expectations for the next year, including various types of assets, interest rate futures (3.7 times) > US bonds and copper (1.7 times) > gold (1.2 times) > Federal Reserve bitmap (0.7 times) > Dow Jones (0.6 times) > S&P 500 (0.2 times) and NASDAQ (-0.4 times)
1) US debt: Interest rates on short-term bonds rise, and the maturity premium on long-term bonds falls first. Even long-term bonds as a whole may gradually peak and fall. On the one hand, one rate hike corresponds to 4.5 to 4.7% of the central bank. Currently, interest rates on long-term bonds have been included in 1.7 interest rate hikes during the year, which fully reflects the expectation of a “preventive interest rate hike.” On the other hand, after interest rate hikes restore confidence, the term premium that pushes up interest rates on long-term bonds will also narrow. As a result, US bonds have shown high “odds” characteristics.
Chart 17: If the Federal Reserve raises interest rates once, corresponding to 4.5 to 4.7% of the US debt center, there is a trading opportunity
2) US stocks: Not pessimistic; short-term disturbances may even provide better buying points. At the beginning of June, the bank further raised the S&P 500 point level to 7800-8000. After the market reached this point, it also began to struggle and fluctuate. First, the profit space driven by AI will have to wait to be catalyzed in the future, and the second is disturbances due to macroeconomic factors such as interest rate hikes. With the implementation of interest rate hikes, if there is a new catalyst for the development of the AI industry, there is still room for growth in US stocks, and it can also provide better buying points if there is significant disturbance in the short term.
Chart 18: Under the benchmark situation, the bank raised the S&P point level in 2026 to 7800-8000 in mid-year
Chart 19: The bank predicts that the US dollar index may fluctuate in the 96-98 range in the second half of 2026, and will not weaken significantly
Chart 20: The bank's static support level is around 4400-4600 based on US bond interest rates and the US dollar
