Is the AI frenzy facing a "financing kill line" test? A well-known Wall Street strategist warns: Unless the dollar and U.S. Treasury yields peak, risk appetite will struggle to recover.

Title context: Is the AI frenzy facing a "financing kill line" test? A well-known Wall Street strategist warns: Unless the dollar and U.S. Treasury yields peak, risk appetite will struggle to recover. Text: A team led by Michael Hartnett, a senior strategist at Bank of America who has been dubbed "Wall Street's most accurate strategist," published a research report saying that until the dollar's rally shows clear signs of peaking and the surge in long-dated U.S. Treasury yields of 10 years and above, driven by energy inflation, retreats from historically high levels, risk assets will still struggle to shake off deleveraging and selling pressure. It is reported that the Bloomberg Dollar Index has rebounded about 3% from its September low, reflecting investors reducing holdings of risk assets such as stocks and cryptocurrencies and rebuilding cash positions; at the same time, the yield on the "global asset pricing anchor," the U.S. 10-year Treasury note, briefly touched 5.34% on October 1, a 24-year high (the highest since 2002), before later easing. Therefore, BofA strategist Hartnett, while remaining cautious, advised investors to start buying bonds that have been shunned by the market, and proposed a policy-level support expectation: if yields continue to climb and threaten the AI investment boom before the November midterm elections, the U.S. government may increase Treasury buybacks. He is especially focused on whether the decline in bank stocks spreads broadly to small- and mid-cap stocks in the equity market, because that would mean economists' and investors' growth optimism about a "soft landing" for the U.S. economy has been seriously shaken, and enormous selling pressure could eventually be transmitted to tech stocks. In the view of the strategist team led by Hartnett, the investment prosperity of the AI computing power industry chain still needs to withstand a dual stress test from the dollar and the long-end yield curve, and whether the explosive expansion of strong demand for AI computing resources can asset price gains still depends on whether financing conditions and valuation discount pressure can ease simultaneously. The Middle East energy inflation situation that has recently caused a surge in long-dated U.S. Treasury yields of 10 years and above presents a pattern of "diplomatic and military pressure proceeding in parallel, oil prices pulling back but the geopolitical war premium still present." Qatar continues to mediate a "seven-day mutual trust plan" between the United States and Iran, with the two sides differing on the sequence of actions; the United States is deploying more forces to the Middle East, and Iran is also preparing to respond to a possible resumption of large-scale U.S. strikes. As of 16:40 Beijing time on October 2, the international crude oil pricing benchmarkBrent crude oil futureswas last reported at $99.48 per barrel, down 2.77% on the day, while WTI crude oil futures were reported at $89.52 per barrel, down 3.61%; based on the settlement prices of $72.48 and $67.02 on February 27, the last trading day before the war broke out on February 28, the two benchmarks are still up about 37.3% and 33.6%, respectively. This comparison uses the front-month futures price benchmark at each point in time, which is enough to show that after the short-term oil price pullback, energy prices are still significantly higher than before the war. BofA senior strategist Michael Hartnett: Risk aversion may persist until the dollar index peaks The strategy team led by Michael Hartnett, a senior strategist at Bank of America, said investors will continue to avoid riskier trades until the dollar's recent sharp rise shows signs of peaking. In addition to needing to wait for the key signal of a dollar peak, the strategist's team also said in a latest report that market unease and anxious selling sentiment may persist until rising bond yields retreat from their highest levels in more than 20 years. He recommended "buying the assets the market despisesthat is, assets that have recently suffered major selloffs," and has begun leaning toward adding some allocations to recently battered long-dated U.S. Treasury assets in portfolios. As financial market investors exit riskier asset positions and begin rebuilding cash buffers, the Bloomberg Dollar Index has risen 3% from its September low. At the same time, bond yields have also climbed, driven by inflationary pressure from the Iran war, expectations of further monetary tightening ahead, and strong corporate earnings growth. As shown in the chart above, the dollar index and U.S. Treasury yields have surged recently, and the stock market rally has tended to stall. Hartnett said that although recent price action shows the market is reducing leverage and cutting exposure to risk assets such as stocks and cryptocurrencies, larger U.S. government Treasury buybacks could provide downside support for the marketespecially when rising yields could threaten the AI investment boom before the November U.S. midterm elections. Hartnett said that if small- and mid-cap stocks also join bank stocks in sharp declines, downside risk would become even more concerning. He stressed that such a trend would clearly indicate that market optimism about strong economic growth has peaked and would ultimately drag down tech stocks. From energy transportation bottlenecks to the AI frenzy's "financing kill line" The series of cautious views recently put forward by Bank of America senior strategist Michael Hartnett has consistently revolved around funding, positioning, and the bond market. On September 11, his team pointed out that U.S. equity funds had seen net outflows of $14.2 billion over the previous three weeks, while average weekly inflows into global equity funds had fallen from $52 billion in July to $7 billion. A subsequent Bank of America September fund manager survey showed that although cash allocation rose to 3.9%, it was still at a low level that he identifies as triggering a contrarian sell signal for risk assets, while disorderly rises in bond yields became the tail risk respondents worried about most; on September 25, he also warned that the bond volatility gauge MOVE had risen 33% in two days. The thread running through these latest Hartnett views means that even if strong earnings and economic growth driven by the AI computing power theme remain resilient, thin cash buffers, rising bond volatility, repeated stage highs in U.S. Treasury yields, and a stronger dollar may also compress investors' ability to continue taking risk. Energy transportation is recovering, but there is still a long way to go before transport costs return to normal. The restart of Saudi Arabia's East-West pipeline and the resumption of loadings at Yanbu port have increased export channels bypassing the Strait of Hormuz; its designed transport capacity is 7 million barrels per day, while Reuters cited an actual throughput estimate of only about 2 million to 2.65 million barrels per day on September 29. Although statistics show that traffic through Hormuz has rebounded somewhat, with LNG outbound cargoes reaching 19-21 shipments in September, three oil tankers were still attacked by unidentified projectiles on September 29; the Bab el-Mandeb Strait also has escort needs, with the French military saying on October 1 that about 10 commercial vessels had been escorted through over the past week. The differences in energy routes among the major oil-producing countries of the Persian Gulf are especially critical: Yanbu can go north to Europe via the Suez Canal, while southbound routes to Asia usually still must pass through the Bab el-Mandeb Strait, and insurance, escort, and diversion costs continue to constrain transport efficiency. The volume of supply recovery and how cheaply energy can be delivered are two variables the market needs to price simultaneously. Against the backdrop of the AI frenzy's "financing kill line" amid energy inflation driving 10-year U.S. Treasury yields to a more than two-decade high, the surge in the dollar index has brought rising safe-haven demand and persistently weak risk appetite, and the AI investment frenzy's "financing kill line" seems to be getting closerthat is, as the "global asset pricing anchor," the 10-year U.S. Treasury yield, hits its highest since 2002, benchmark financing costs that remain near historic highs are beginning to constrain financing progress critical to AI capital expenditure and the returns on AI infrastructure projects, and the market is therefore increasingly questioning whether the expected returns of many large AI data center projects can continue to cover rising capital costs. When the expected cash return of a new computing power project after deducting expenses such as electricity, operations and maintenance, and equipment renewal cannot cover cost indicators including financing costs, continued expansion will find it difficult to create economic value, and a single large AI infrastructure project may begin to collapse from that point; it first constrains marginal projects with weak cash flow and financing that has not yet been locked in. The 10-year U.S. Treasury note is called the "global asset pricing anchor" because of its benchmark status in the dollar financing system and in the valuation of medium- and long-term cash flows. The U.S. Treasury market is massive and actively traded, and the dollar is widely used in international financing and reserves, so changes in its yield have cross-market effectsdollar corporate bonds usually reference Treasury yields of similar maturity plus a credit spread, housing mortgage rates are affected by the pricing of Treasuries and mortgage-backed securities, and stock and real estate valuations are highly sensitive to the discount rate applied to future cash flows. When this benchmark rises while earnings and rent expectations do not improve in tandem, asset prices face downward pressure. The impact is also transmitted overseas through dollar financing costs, currency hedging, and cross-border capital flows; different currencies, maturities, and credit risks determine the extent to which specific assets are hit. From the perspective of data center projects, GPU servers, power connections, and cooling facilities require upfront investment, while computing power service revenue is recovered period by period; rising long-term risk-free rates and credit spreads will simultaneously raise financing costs and lower the valuation of forward cash flows, while a stronger dollar will also increase the debt servicing and equipment procurement burden of some non-U.S. borrowers. Therefore, strong computing power demand and tighter project financing conditions can occur at the same time, and the first to be tested are usually expansion plans that rely on external financing and have more distant payback periods. The policy support Hartnett hopes for is precisely to ease this capital cost constraint: the Treasury has already expanded buybacks to support long-term Treasury liquidity, but its official goal is to improve market liquidity, and further stepping up protection for AI investment remains his policy judgment. For investors' overall strategy, the dollar's trajectory, U.S. Treasury yields of 10 years and above, and the market performance of banks and small- and mid-cap stocks are becoming important signals for testing whether the AI computing power super bull market can continue to spread to broader equity market sectors.