Walsh is actively setting the long-term outlook for AI, and the Federal Reserve's interest rate path is still blank in October! Is the cooling of US bond yields opening a window for technological counterattack?
The Zhitong Finance App learned that during the noon session of Beijing time on October 16, Federal Reserve Chairman Kevin Walsh held talks with Georgieva during the annual meetings of the IMF and World Bank in Bangkok, reaffirming that the Federal Reserve will stabilize prices, and described the September rate hike as removing “some easing.” Regarding the Federal Reserve FOMC monetary policy meeting to be held from October 27 to 28, he made it clear that he would not predict the Fed's monetary asset path, nor would he announce the actions that may be taken during the remaining meetings of the year in advance. He emphasized that every decision made by the Federal Reserve will form policy judgments based on output, employment, and inflation, rather than a predetermined policy path, and reiterated that he personally favors the Fed not providing any forward-looking guidance.
It is worth noting that Walsh's series of speeches at the IMF strengthened the judgment that cutting-edge AI technology will change the economy in the long term, but he did not promise to suspend interest rate hikes in October. As for global semiconductor sector stocks, which have recently experienced sharp sell-off and have fallen into a pullback trajectory, the positive point is that he has not further strengthened expectations for recent interest rate hikes; however, whether US bond yields can continue to cool down, AI computing power-themed stocks that have been hit hard recently, and whether the broader semiconductor and technology-related sectors can launch a counterattack still depends on whether inflation, corporate financing conditions, and AI profit can be realized at the same time.
More noteworthy than “not predicting monetary policy,” he said that since the interest rate meeting in September, economic evolution has generally been in line with expectations, and there is currently structural growth and a surge in investment. This means that he has yet to equate the recent adjustment of semiconductor-led global technology stocks with a breakdown in the economic growth logic.
At the same time, he emphasized that most regions of the world are simultaneously tightening policies, and all countries need to maintain price stability; US policies are transmitted overseas, and overseas policies will also affect the US in reverse. Its policy framework remains to maintain inflation and institutional credibility while retaining room for judgment. Prior to the speech, MUFG recently revealed that the probability of an interest rate futures market price increase in October was about 27%, but it was still fully included in the 25 basis points of interest rate hike before the end of the year — so before and after Walsh's speech, the interest rate futures market was always trading that the interest rate hike time might have moved backwards, rather than the end of the austerity cycle.
Wall Street giants' recent predictions about the Federal Reserve are being further focused on “continuing to raise interest rates and extend the austerity cycle.” The latest clear adjustments come from Citibank and UBS. The Federal Reserve raised interest rates by 25 basis points in September. This round of institutional revisions focuses on how much more will be added later. Citi withdrew its forecast of “no interest rate adjustments during the year” on October 14, and instead expected to raise interest rates by another 25 basis points in December, then suspended until June 2027. From October 12 to 15, UBS Global Wealth Management extended the original “December Plus One More Time” approach to raising interest rates by 25 basis points each in December of this year and March of next year; along with the actions already implemented in September, interest rates were raised by a total of 75 basis points throughout the entire round, and is expected to maintain interest rates until near the end of 2027 thereafter.
The core driving the revaluation of Wall Street's rate hike predictions is that the energy shock has begun to spread to a wider range of prices, and the economy has not weakened enough to force the Federal Reserve to prioritize growth. The core CPI rose 0.3% month-on-month in September, exceeding market expectations of 0.2% for two consecutive months. The year-on-year growth rate rose from 2.4% to 2.6%; the combined price increase of housing, air tickets, automobiles, clothing, and technology-related products made it more difficult for policymakers to view the rebound in inflation as an isolated oil price disturbance. The PPI for September, which was later announced, rose 1.3% month-on-month and 6.3% year-on-year, further strengthening concerns about upstream costs being transmitted to terminals.
AI boosts demand and then releases productivity: Why does a boom in computing power coexist with high interest rates
In his speech, Walsh viewed AI as the most profound economic transformation of his life, and the key to understanding this judgment is that he also emphasized that AI's impact on the demand side will come before supply-side improvements. From the perspective of data center engineering and AI computing power industry chain investment, enterprises need to first purchase GPUs, TPUs, ASIC, HBM, and network equipment to invest in electricity, cooling, plant and financing; while the cost reduction and efficiency improvement formed by AI being widely embedded in R&D, production, and enterprise processes require deployment, organizational adjustments, and commercial verification. This time difference enables AI to boost semiconductor orders while increasing demand for capital and electricity, driving up some prices and financing costs in the short term.
Therefore, Walsh's AI optimism does not naturally correspond to falling interest rates. Productivity improvements can help expand supply and curb unit costs, but higher return on investment expectations may also attract more companies to compete for capital and push up actual interest rates. As far as the 10-year and 30-year US Treasury yield trajectory is concerned, the market must also take into account future policy interest rates, fiscal financing, and maturity risk compensation at the same time. Strong demand for computing power and high long-term capital costs are probably due to the same round of AI investment expansion; semiconductor profit growth and pressure on valuations can also occur at the same time.
His latest statement also revealed how policy research will follow up on this change: an AI live hearing is scheduled to be held in the San Francisco Bay Area on November 16, and the Federal Reserve Task Force will form new research understanding to assist the FOMC in making decisions; AI is also expected to improve the central bank's ability to understand data, but the short-term impact on employment is still uncertain. When it comes to private technology investments, the Federal Reserve will not judge who is right and who is wrong, but will focus on financial tail risks that can be addressed through policies. These signs all mean that central banks around the world, such as the Federal Reserve, will pay more and more attention to AI's contribution to growth, are also reviewing financing structures and risk transmission, and are not promising to protect AI asset prices.
US debt has cooled down, and the semiconductor counterattack still needs to cool down credit financing costs and resonate with profits
Long term yields have indeed begun to fall back. According to the unified caliber of the US Treasury, the 10-year and 30-year US bond yields on October 15 were 5.10% and 5.43%, respectively, down 9 basis points on the same day; as of the 15th of this week, they were down 14 and 17 basis points respectively, down 19 and 21 basis points from the end of September. However, the two are still 35 or 18 basis points higher than at the end of August, indicating that the recent cooling has only partially reversed the previous impact on financing costs.
It is worth noting that an important driving force behind this round of decline was safe-haven debt purchases caused by AI stock and credit market shocks. TD Securities believes that investors are lowering their growth expectations, and US bonds are once again playing the role of stock hedging tools. As a result, a seemingly contradictory market mix was formed. The pressure on computing power stocks is driving down US bond yields, and the decline in yield is not enough to dispel concerns about computing power financing for the time being. The cost of corporate debt is composed of a combination of the benchmark interest rate and the credit spread; if the former falls and the latter expands, and data center projects that rely on external financing, actual borrowing costs may remain high.
Global stock markets were also in a recovery window after a sharp decline at the time of Walsh's speech. As of October 15, the Philadelphia Semiconductor Index fell 0.88%, with a cumulative decline of about 12.4% for seven consecutive trading days, down about 7.9% this week, and a cumulative decline of about 8.3% in October. Although it still rose by about 63.5% during the year, short-term momentum has been clearly damaged. On October 16, during the Asian midday trading period, the Korean stock market benchmark, the South Korean KOSPI Index, which has the “AI memory chip” trend, rebounded 3.17% in a single day. Samsung and SK Hynix rebounded simultaneously. This rebound also predated Walsh's speech, but the upward trend continued.
To a certain extent, Walsh's statement provides support for the recovery of sentiment in the global semiconductor sector, which has been hit hard, but the combination that can actually drive a continuous counterattack on semiconductors is “long-term yield decline, stable credit spreads, and continued fulfillment of computing power orders and cash flow.” If financing pressure eases and AI demand remains strong, the market will be in a position to repay higher valuations for growth; among them, chip and device leaders with abundant cash flow, stable customer funding sources, and strong order conversion capabilities are more likely to benefit first. The long-term transformation of AI emphasized by Walsh provides a macro explanation for this main line of growth, and whether the short-term rebound can escalate into a trend still depends on whether the decline in capital costs can actually be transmitted to industrial investment.
Wall Street financial giant Goldman Sachs recently described the “three major reinforcements” (the three forces of repurchase, retail investors, and foreign capital), which are expected to improve the financial environment for AI assets to be re-priced. The Goldman Sachs capital flow team pointed out in the latest report “Witching Hour” that according to “Witching Hour” estimates, about 48% of the S&P 500 constituent stocks will re-enter the open repurchase window at the end of October; historically, November contributed about 10.1% of the annual repurchase execution volume, combined with the subsidence of mutual fund tax sales pressure at the end of the fiscal year, retail capital inflows at the end of the year, and overseas allocation requirements after the midterm elections, and the relationship between supply and demand for stocks may gradually improve. Among them, overseas purchases of about 190 billion US dollars are a potential scale extrapolated by historical laws. For semiconductors that have experienced concentrated position reductions, buying recovery means that the company may trigger position recovery even if it only proves that orders and profits have not deteriorated significantly. However, the technology sector's buyback execution volume fell 40% year over year, indicating that capital expenditure is competing with shareholders to return capital; if AI stocks are to continue to attract incremental allocations, it is still necessary to prove that capital investment related to investing in key AI infrastructure resources such as GPUs, network infrastructure, and data center optical interconnects and CPUs can form higher future cash flows.