AI frenzy defies the "5% Treasury yield"! Nasdaq hits a new high, reinforcing the "80/20" allocation, as Wall Street reassesses the stock-bond ratio.

Understanding that the profit imagination brought by the expansion of AI agents, and the valuation pressure caused by rising long-term U.S. Treasury yields, are becoming the two most striking opposing forces on Wall Street. Yet at least so far, the answer given by U.S. equity market capital is that the unprecedented AI computing power expansion and AI application frenzy brought by AI agents have outweighed a series of important bearish factors, including the rise in yields on long-term U.S. Treasuries with maturities of 10 years and above. This is also why some top Wall Street hedge funds have begun abandoning the traditional "60/40" portfolio and shifting toward a more aggressive 80/20 variant allocationthat is, stock ETF assets account for about 80% of the total ETF portfolio, fixed-income ETFs focused on Treasury assets account for about 16%, and the remaining very small share may focus on cash management assets similar to reverse Treasury repurchases, making this stock structure closer to "80/20." The 10-year U.S. Treasury is called the "anchor of global asset pricing" 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 influencedollar 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. From a theoretical perspective, the 10-year U.S. Treasury yield is equivalent to the risk-free rate indicator r on the denominator side of the DCF valuation model, an important valuation model in the stock market. If other indicators (especially cash flow expectations on the numerator side) do not change significantlyfor example, during earnings season, when the numerator side is in a vacuum period due to a lack of positive catalyststhen if the denominator level is higher or continues to operate in the historically extreme high range above 5%, valuations of risk assets closely tied to AI, such as technology stocks, high-yield corporate bonds, and cryptocurrencies, which are already at historically high levels, face the threat of collapse. However, as AI skeptics once again declare surrender, traders bullish on the AI computing power theme are actively exploring the future trajectory of technology stocksmainly because risk assets have once again withstood yield pressure, while also reflecting that investors still believe AI application commercialization and corporate earnings growth can to some extent offset the impact of rising financing costs and discount rates. The process by which the Nasdaq 100 index set its first record high since June was full of twists and turns. Now, the stock market bulls driving this rally are looking for bullish evidence that the upward momentum based on the unprecedented AI infrastructure boom will not fade. Wall Street financial giant Jefferies recently said that, driven by the dual engines of the AI investment frenzy and AI-related corporate earnings rising above expectations, the S&P 500 index is expected to soar to 8,000 by the end of 2026 and further reach 9,000 in 2027. Jefferies' core logic is clear and powerful: in a cycle where AI-driven earnings growth exceeds the historical average by more than twice, it is dangerous to fight the earnings trend. Jefferies' baseline forecast of 8,000 for the S&P 500 in 2026 is based on earnings per share (EPS) reaching $373 (up 35% year over year, far above the market consensus of 29%) and a price-to-earnings ratio of 21.5 times. AI earnings expectations versus a high yield curve: why can U.S. stocks still withstand pressure from long-term U.S. Treasury yields? In the week ended September 25, the Nasdaq 100 index rose 3.3%, the largest weekly gain since early August, after closing at a record high as of Tuesday. Meta rose nearly 13% for the week amid optimism over Muse. At the same time, the 30-year U.S. Treasury yield briefly touched about 5.53%, the highest level since 2004, while Brent crude oil remained above $100 per barrel. Risk assets once again withstood pressure, reflecting that investors still believe AI application commercialization and corporate earnings growth can to some extent offset the impact of rising financing costs and discount rates. However, Friday's oil price pullback relied on hopes for progress in U.S.-Iran negotiations, and a return to normal energy supply has not yet become a certainty. Muse and OpenAI's GPT-6 Astra provide a more concrete application path for this earnings expectation. Muse executes cross-application tasks through a dedicated cloud virtual machine and browser, and can continue working after the user closes the application; Astra strengthens computer operation, software engineering, and multi-step professional work capabilities. When AI can complete research, programming, shopping, and office workflows, the chargeable value expands from generating an answer to delivering a work result. From the perspective of business growth logic, higher task success rates and lower completion costs are expected to drive enterprises and consumers to expand usage, thereby supporting growth expectations for subscriptions, usage-based payments, and related infrastructure revenue. The transmission of this application expansion to the computing power side involves a complete computing system: GPUs and other AI accelerators handle model computation, while CPUs are responsible for browsers, code execution, virtual machines, tool invocation, and task orchestration; model weights, context, and concurrent sessions increase demand for high-speed memory, while task files, long-term memory, and caches suitable for tiered storage increase demand for DRAM and enterprise-grade SSDs. Nvidia's engineering analysis also points out that CPU tool execution speed affects agent workflow throughput, and KV cache can be managed in tiers among GPU memory, CPU memory, and storage according to access requirements. This is also why Wall Street giants such as Goldman Sachs and Jefferies are bullish on the hard-core logic supporting continued U.S. equity earnings expansion under accelerated AI application penetration and the AI computing power frenzy. A major increase in AI agent penetration is expected to simultaneously expand demand in core AI infrastructure areas such as computing, memory, storage, and optical interconnect networks. The stock market rebound injects new momentum into Wall Street's "80/20" offensive-defensive allocation shift Wall Street has just experienced another week like this: bond yields hit multi-decade highs, oil prices remained above $100 per barrel, and the case for further Federal Reserve tightening became stronger. Yet risk assets once again withstood the pressure. Although the global bond selloff intensified for most of the week and inflation concerns once again became the market's focus, stock market closing levels remained not far from record highs, and corporate bond credit spreads remained within controllable ranges. Still, investors' confidence in AI-driven corporate earnings growth and economic resilience has at least temporarily suppressed these concerns, and holding stocks has therefore continued to be rewarded. News that the United States and Iran might make a breakthrough and promote the opening of the Strait of Hormuz caused oil prices to fall on Friday, helping the Nasdaq 100 index rise 3.3% for the week, its largest weekly gain since early August. Meta Platforms was particularly prominent, with its stock rising nearly 13% after the release of its Muse AI tool. However, behind all this still lurks the shadow of weakening consumer confidence and lingering inflation risks. The latter dealt a heavy blow this week to U.S. Treasuries, which are usually one of the safe havens for avoiding stock market risk. Some capital flowed into new defensive tools. So far in September, about $2.7 billion flowed into the Schwab U.S. Dividend Equity ETF, putting it on track for a 10th consecutive month of net inflows; two large JPMorgan option income ETFs together attracted about $800 million. The iShares Global Infrastructure ETF attracted another $150 million. At the same time, some fixed-income fund managers are keeping allocations at the shorter end of the yield curve. None of these tools are bond substitutes in the traditional sense. But with shorter-term bond yields near 5%, they offer some of the characteristics investors now seek: sustainable income without taking on long-duration risk. Adrian Helfert, chief investment officer at Westwood, which manages about $18 billion in assets, said: "Investors are actively examining that 40% in the traditional 60/40 portfolio. Through dividends or option income, I can obtain considerable sustainable income without taking on that kind of duration risk. Therefore, we have reduced the long-end risk of the portfolio." As shown in the chart above, the U.S. 30-year Treasury yield rose to its highest level in more than two decades. At least in the ETF space, the textbook 60/40 allocation already looks more like 80/20. According to statistics from Bloomberg Intelligence, stock ETFs account for about 80% of all ETF assets, among the highest proportions on record, while fixed-income ETFs account for about 16%. This tilt stems largely from divergent asset performance and allocation shifts by large institutional investors, rather than a unanimous active choice by ordinary investors. Stocks have clearly outperformed bonds, while some new capital is still flowing in at ratios closer to the traditional allocation. Regardless of intent, ETF investors' assets remain clearly skewed toward stocks, despite speculation that the bond market may siphon funds away from the stock market as bond yields remain at levels unseen since the global financial crisis. Sima Shah, chief global strategist at Wall Street's Principal Asset Management, said: "If yields are rising because U.S. economic growth is stronger, then bonds' traditional diversification advantage may be less reliable, especially during periods when inflation expectations are also rising. This is one reason investors are considering broader equity investment opportunities." A veritable "wealth destroyer"as shown in the chart above, long-duration bond ETFs of 10 years and above have become an investment destination that has dealt heavy losses to capital. Note: The long-duration bond ETF group includes TLT, IEF, SPTL, TLH, VGLT, EDV, ZROZ, GOVI, XTEN, UTEN, SCHQ, XSVN, GOVZ, XTWY, UTWY, UTHY, TYA, BBLB, USIN, LLDR. Source: Baird Strategas. Take Westwood's veteran analyst Helfert as an example. He believes that at current yield levels, bonds offer quite attractive income. But he is reluctant to take on a large amount of duration pricing risk just to obtain that income. He said: "U.S. Treasury yields are attractivethe 10-year Treasury yield is now clearly above 5%. But the speed of yield changes is frightening." This is because the rapid rise in yields has quickly pushed down long-term bond prices. A basket of long-duration bond ETFs tracked by Baird Strategas has attracted about $224 billion in cumulative inflows since inception, but now holds only about $182 billion in assets, a side reflection of the impact of widespread market selling. Among the four categories tracked by Todd Sohn at the firm for this calculation, the long-duration category has caused the greatest damage to investor capital. This week, the 30-year U.S. Treasury yield touched 5.53%, the highest since 2004; the 5-year Treasury yield broke above 5% for the first time since 2007. Brent crude rose above $106 per barrel, and bond market volatility surged. At the same time, the S&P 500 posted its first weekly gain since the beginning of the month. Garrett Aird, vice president of investment management research at Northwestern Mutual Wealth Management, which manages $461 billion in assets, said his team has adjusted the fixed-income allocation in several 60/40 portfolios it manages. They have been increasing allocations to real estate investment trusts and so-called liquid alternatives, which include managed futures, merger arbitrage, and long-short strategies. He said in an interview: "Our entire approach to building portfolios is to provide a bit more diversification in the fixed-income portion of the allocation. The most critical thing is to find strategies that can provide uncorrelated sources of returnthat is, to find strategies that are neither correlated with equities nor with fixed-income assets. This is crucial." The capital choice behind "80/20": retain equity growth, reassess the long-bond defensive logic It is worth noting that stocks outperforming bonds itself pushes up the equity asset share. The so-called shift in Wall Street asset allocation ratios toward an 80/20 portfolio includes both passive weight changes for asset managers and hedge funds caused by stock price gains, and active choices by some institutions to reduce exposure to long-duration Treasuries. But there is no doubt that capital is looking for tools that can provide income without taking on a large amount of long-bond interest rate risk. So far in September, the Schwab U.S. Dividend Equity ETF has attracted about $2.7 billion, two large JPMorgan option income ETFs have together attracted about $800 million, and the iShares Global Infrastructure ETF has attracted about $150 million. Their common appeal lies in dividends, operating cash flow, or option premiums, but these income sources may still be accompanied by stock price declines; earning income by selling call options also usually means giving up some upside. The problem facing the allocation logic for long-duration bonds of 10 years and above is that duration remains high and continues to rise. A rapid increase in yields causes more obvious price losses in bond asset allocations, and coupon-level income may not be able to cover this shock in the short term. From this reasoning, Wall Street asset management giants and hedge funds seem to be re-splitting the three core functions traditionally undertaken by the 60/40 stock-bond allocation: providing sustainable growth returns, reducing volatility, and hedging stock market shocks. Some investment institutions have begun using shorter-term bonds to control interest rate sensitivity, combining high-momentum AI-focused strategies with dividend and option strategies to supplement income, and then seeking different sources of return through leveraged strategies such as managed futures; whether these tools can achieve diversification still depends on specific holdings and the market environment.