The AI frenzy withstands the "5% US Treasury yield"! Nasdaq hits new highs, strengthening the "80/20" pattern as Wall Street reassesses stock-bond allocation

According to Zhitong Finance APP, the profit potential brought by the expansion of AI agents and the valuation pressure caused by the rising yields of long-term US Treasuries are becoming the two most prominent opposing forces on Wall Street. However, the answer given by market funds in the US stock market, at least so far, is that the unprecedented expansion in AI computing power and the frenzy of AI applications has outweighed a series of major negative factors, including the persistent rise in yields of US Treasury bonds with maturities of ten years and above.

This is why some top Wall Street hedge funds have started to abandon the traditional "60/40" investment portfolio and turn to a more aggressive 80/20 variant—where equity ETF assets account for about 80% of the entire ETF asset portfolio, fixed income ETFs focusing on treasury assets account for about 16%, and the remaining small portion may focus on cash management assets such as treasury reverse repos, making this stock structure closer to "80/20".

The 10-year US Treasury is known as the "anchor of global asset pricing", due to its benchmark position in the US dollar financing system and mid-to-long-term cash flow valuations. The US Treasury market is huge and highly liquid, and the US dollar is widely used in international financing and reserves, so its yield changes have cross-market effects—corporate dollar bonds typically reference US Treasury yields of similar maturity with added credit spreads, mortgage rates are influenced by US Treasuries and mortgage-backed securities pricing, and the valuations of stocks and real estate are highly sensitive to the discount rate of future cash flows.

Theoretically, the 10-year US Treasury yield essentially serves as the risk-free rate (r) in the denominator of the important DCF valuation model in the stock market. If other metrics (especially the expected cash flows in the numerator) have not changed significantly—for example, during earnings seasons when there is a lack of positive catalysts, the numerator is in a vacuum period—if the denominator is higher or stays above 5%, operating at historically extreme high levels, the valuations of those tech stocks closely associated with AI, high-yield corporate bonds, and cryptocurrencies, all risk assets at historically high valuations, are at risk of collapsing.

However, as AI skeptics once again surrender, traders bullish on AI computing power themes are actively exploring the future trend of tech stocks—mainly because risk assets are once again withstanding yield pressures. This also reflects investors’ continued belief that the commercialization of AI applications and profit growth can partially offset the impact of rising financing and discount costs. The Nasdaq 100’s journey to its first record high since June has been full of twists and turns. Now, equity market bulls driving this rally are seeking bullish evidence that the upward momentum from the unprecedented AI infrastructure boom will not fade.

Recently, Wall Street giant Jefferies stated that, powered by the dual engines of the AI investment boom and AI-related companies' earnings significantly beating expectations, the S&P 500 index is expected to soar to 8,000 by the end of 2026, and further reach 9,000 by 2027. Jefferies’ core logic is clear and powerful: in a cycle where AI-driven profit growth exceeds the historical average by more than double, fighting the earnings trend is dangerous. Their baseline forecast for the S&P 500 index at 8,000 in 2026 is based on EPS reaching $373 (35% year-on-year growth, much higher than the market consensus of 29%) and a price-earnings ratio of 21.5.

AI Profit Expectations Face Off Against High-Yield Curve: Why US Stocks Can Withstand Long-Term Treasury Yield Pressure

For the week ending September 25, the Nasdaq 100 rose 3.3%, the largest weekly gain since early August, and it even hit an all-time high by Tuesday’s close. Boosted by prospects for Muse, Meta climbed nearly 13% for the week. Meanwhile, the 30-year US Treasury yield once reached about 5.53%, the highest since 2004, and Brent crude oil still hovered above $100 per barrel. Risk assets withstood the pressure once again, reflecting investors’ enduring faith that the commercialization of AI applications and corporate profit growth can, to some extent, offset the shock of rising financing costs and discount rates. However, Friday’s oil price decline relied on hopes for progress in US-Iran talks, and a true normalization of energy supply has not yet occurred.

Muse and OpenAI’s GPT-6 Astra provide more concrete application pathways to these profit expectations. Muse can execute cross-application tasks via dedicated cloud VMs and browsers, and continue to operate even after users close the application; Astra enhances capabilities in computer operations, software engineering, and multi-step professional tasks. When AI can handle research, coding, shopping, and office workflows, its monetizable value expands from generating an answer to delivering a work result. From a business growth perspective, as task success rates rise and completion costs fall, enterprises and consumers are likely to expand usage, supporting expectations for growth in subscriptions, metered billing, and related infrastructure revenues.

This application expansion’s impact on the computing side involves a complete computing system: GPUs and other AI accelerators handle model computation, CPUs handle browsers, code execution, VM operations, tool invocation, and task orchestration; increases in model weights, context, and concurrent sessions drive high-speed memory demand, and larger task files, long-term memory, and tiered caches push up DRAM and enterprise-class SSD needs. Nvidia’s engineering analysis also notes that CPU tool execution speed affects agent workflow throughput, while KV cache can be tiered between GPU memory, CPU memory, and storage based on access requirements.

This is why Wall Street giants like Goldman Sachs and Jefferies are bullish on the hardcore logic behind the accelerated penetration of AI applications and the AI computing power boom, supporting continued profit expansion in the US stock market. The penetration rate of AI agents is set to soar, simultaneously expanding demand in core AI infrastructure fields such as computing, memory, storage, and optical interconnection networks.

Stock Rally Injects New Momentum into Wall Street’s Shift to “80/20” Offense-Defense Configuration

Wall Street has gone through another week like this: bond yields at multi-decade highs, oil staying above $100 per barrel, and stronger justification for the Federal Reserve to further tighten policy. Yet risk assets have once again withstood the pressure.

Despite the global bond selloff intensifying for much of the week and inflation worries returning to the forefront, stocks closed not far from record levels, and corporate bond credit spreads remained manageable.

However, confidence in AI driving profit growth and economic resilience, at least temporarily, has constrained these concerns, and holding equities continues to generate returns. News of a possible breakthrough between the US and Iran, which could open the Strait of Hormuz, led to an oil price retreat on Friday, helping the Nasdaq 100 post a 3.3% weekly gain, its biggest rise since early August. Meta Platforms stood out, its Muse AI tool launch driving nearly a 13% rise in its stock price.

Nonetheless, behind it all lurks the shadow of weakening consumer confidence and persistent inflation risk. The latter dealt a heavy blow to US Treasuries this week, which are normally seen as a safe haven from stock market risk.

Some funds have flowed into new defensive tools. Since September, about $2.7 billion has gone into the Schwab US Dividend Equity ETF, putting it on track for a tenth consecutive month of net inflows; JPMorgan Chase’s two large option income ETFs attracted a combined $800 million. The iShares Global Infrastructure ETF drew an additional $150 million. Meanwhile, some fixed income fund managers are keeping allocations at the shorter end of the yield curve.

These instruments are not traditional bond substitutes. But as yields on shorter-maturity bonds approach 5%, they offer some of the features investors now seek: sustainable income, without taking on long duration risk.

Adrian Helfert, Chief Investment Officer of Westwood, which manages about $18 billion, said, “Investors are actively reviewing that 40% fixed income portion of the traditional 60/40 portfolio. Through dividends or option income, I can obtain significant, sustainable income without taking on duration risk. As such, we’ve reduced our portfolio’s long-end risk.”

As shown above, the yield on US 30-year Treasuries has risen to the highest level in over twenty years.

At least in the ETF space, the textbook 60/40 mix now looks more like 80/20. According to Bloomberg Intelligence, equity ETFs account for about 80% of all ETF assets, one of the highest proportions on record, while fixed income ETFs account for about 16%. This tilt largely stems from divergent performances between asset classes and a shift in allocations by large institutional investors, rather than the active, uniform choices of retail investors. Equity returns have outperformed bonds, but some new inflows still follow a configuration closer to traditional proportions.

No matter the intention, ETF investors’ assets remain heavily weighted toward equities, even as some speculate that with bond yields at levels not seen since the global financial crisis, the bond market might siphon funds from stocks.

Sima Shah, Chief Global Strategist at Principal Asset Management, said, “If yields are rising because the US economy is growing stronger, the bond market’s traditional diversification advantage may be less reliable—especially during periods when growth and inflation expectations both rise. That’s one reason investors are considering broader equity opportunities.”

So-called “wealth destroyers”—as shown above, long-duration bond ETFs (10 years and longer) have become investment pits for severe capital losses. Note: The long-duration bond ETF basket 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 senior analyst Helfert as an example; he believes that at current yield levels, bonds offer highly attractive returns. But he is less willing to take on significant duration risk for that income. He said, “US Treasury yields are appealing—10-year yields are now markedly above 5%. But the speed at which yields are changing is alarming.”

This is because rapidly rising yields have quickly depressed long-term bond prices. The long-duration bond ETF basket tracked by Baird Strategas has seen about $224 billion in cumulative inflows since inception but now holds only about $182 billion in assets, reflecting the impact of broad market selloffs. Among the four categories tracked for this calculation by the firm’s Todd Sohn, the long-duration category has done investors the most harm.

This week, the 30-year US Treasury yield reached 5.53%, its highest since 2004; the 5-year yield surpassed 5% for the first time since 2007. Brent crude spiked above $106 per barrel, and bond market volatility surged. Meanwhile, the S&P 500 had its first gainful week since early this month.

Garrett Aild, Vice President of Investment Management Research at Northwestern Mutual Wealth Management—managing $461 billion—said his team has adjusted fixed income allocations in several managed 60/40 portfolios. They have been increasing allocations to real estate investment trusts, as well as so-called liquid alternative investments, including managed futures, merger arbitrage, and long-short strategies.

In an interview, he said, “Our entire portfolio construction approach is to offer greater diversification in the fixed income allocation. The key is to find strategies that provide uncorrelated sources of returns—that is, strategies that aren’t tied to equities or fixed income. That’s crucial.”

The Logic Behind “80/20” Fund Choices: Retain Stock Growth, Rethink Long Bond Defensive Logic

Notably, equities outperforming bonds naturally lifts the proportion of equity assets, meaning the Wall Street shift toward the 80/20 allocation reflects both passive weighting changes driven by rising stock prices at asset management institutions and hedge funds, and some active institutional choices to reduce long-duration Treasury exposure.

But what is clear is that capital is seeking tools that can provide yield without taking on a lot of long-term rate risk. Since September, the Schwab US Dividend Equity ETF has drawn about $2.7 billion, JPMorgan’s two large option income ETFs have attracted about $800 million combined, and the iShares Global Infrastructure ETF has brought in about $150 million. Their shared appeal lies in dividends, operating cash flow or options premiums, but these income streams can still be accompanied by falling share prices; selling covered calls for income usually means giving up some upside as well.

The problem facing long-duration Treasury allocations of 10 years or more is that duration risk remains high and continues to rise. Rapidly rising yields cause more pronounced price loss in bond allocations, and coupon income may not be able to offset this shock in the short term.

From this, it appears that Wall Street asset management giants and hedge funds are attempting to re-decompose the three core functions of the traditional 60/40 equity-bond allocation: provide sustainable growth returns, reduce volatility, and hedge equity shocks. Some investment institutions are now using shorter-duration bonds to manage interest rate sensitivity, using high-momentum, AI-focused strategies paired with dividend and option strategies for additional income, and seeking different return sources through managed futures and other leveraged strategies. Whether these tools will achieve true diversification still depends on the specific holdings and market environment.