Recently, there has been a topic in overseas markets that everyone has been closely watching: with US Treasury yields reaching 5%, are tech stocks about to peak? This logic seems straightforward, but when you actually put recent market performance together, you realize things are not that simple.
First, let us clarify: the so-called "US Treasury new high" refers to rising yields (meaning Treasuries are being sold off and prices are falling), specifically re-entering multi-year highs. As of September 25, long-end US rates continued to hold at historic highs: the 10-year Treasury yield rose to about 5.21% (a new high since 2007), the 20-year at about 5.55%, the 30-year at about 5.5% (a new high since 2004), and the 10-year TIPS real yield approached 2.8%. In other words, the market is repricing long-term capital.
But what is more interesting is that weakening Treasuries did not automatically translate into weakening US tech stocks. Taking a rough look from the September 1 close to the latest trading days of September 24-25 across markets, the Nasdaq rose about 3.2%, the SSE STAR 50 fell about 1.6%, the Hang Seng Tech fell about 5.2%, the Taiwan Weighted Index rose about 2.3%, the Nikkei 225 rose about 0.4%, and the KOSPI rose about 3.6% through September 23. These markets differ in tech weighting, economic structure, and local liquidity, but at least one thing is clear: the same round of rising US Treasury yields can produce completely different answers across equity markets. In fact, a US Treasury bear market and a Nasdaq bull market have walked hand in hand for five years, which shows that Treasuries are not the single steering wheel for the Nasdaq.
First, we need to understand why US Treasuries are still rising. What truly deserves attention is that this round of long-end yield increases is no longer just about whether the Fed will hike rates. On September 16, the Federal Reserve raised the federal funds rate by 25 basis points to 3.75%-4.00%. But long-end government bonds are trading future years of inflation, economic growth, fiscal financing, term premium, and private capital demand. The US Treasury expects net marketable borrowing of $739 billion in the third quarter of 2026 and still $628 billion in the fourth quarter. J.P. Morgan's recent judgment is worth noting: the drivers of this round of US Treasury yield increases are shifting from "the Fed" to "fiscal deficits + hyperscaler financing + energy prices + the US economy itself." BlackRock's framework is similar: sovereign financing demand and private financing demand brought by AI are competing for capital at the same time. In other words, US tech stocks themselves have become a driver of rising US Treasury yields, and long bonds now look more like they are repricing the entire US economic system. The Fed determines the short-term price of money, but 10-year, 20-year, and 30-year Treasury yields increasingly resemble the sum of America's future financing costs.
This also explains why the yield curve has recently returned to a positive slope, but one cannot simply explain all market action with the phrase "bear steepening." On September 24, the 10-year was about 5.18%, the 2-year about 4.87%, and the 10Y-2Y spread about 31 basis points, indeed no longer inverted. But compared with the long-term pressure truly driven by fiscal supply, inflation, and term premium, the shape of the curve itself is not the most critical variable. What is truly critical is: why are long-end yields rising?
This is where Treasuries and US equities must be understood separately. Bonds first price the time value of money; stocks price future earnings, cash flows, risk premium, and valuation. The 10-year Treasury is an important denominator for equity valuation, but it is not the entirety of equity pricing. If a tech company will only realize profits ten years from now, and rates rise from 4% to 5%, its valuation will naturally come under pressure. But if a company's revenue growth, profit growth, and free cash flow are all being continuously revised upward, then earnings themselves are racing against the discount rate. This is why the combination of "Treasury bear, Nasdaq bull" has been entirely possible in recent years.
From the second half of 2022 to the first half of 2023, the market was similarly worried that high rates and financial liquidity would hit tech stocks, and the Silicon Valley Bank incident amplified that fear. What the market truly repriced later was not that "rates do not matter," but that the AI industry trend was too strong and earnings expectations began to be revised upward again. In historical tech bull markets, the places where they truly tend to end are often industry bottlenecks, not rates themselves. Rates amplify and compress valuations, while industry trends determine whether earnings can continue to grow.
But there is one obvious difference between today and 2023. AI has already moved from a software narrative to a capital expenditure cycle involving data centers, GPUs, switches, optical communications, storage, power, and liquid cooling. The capital structure of tech companies is also beginning to change. According to Reuters, over the past year, hyperscale tech companies such as Microsoft, Alphabet, Meta, Amazon, and Oracle issued about $220 billion in bonds; AI-related bond spreads have widened to about 115 basis points, versus about 78 basis points for the overall market, and the market expects hyperscaler bond issuance could reach $420 billion in 2027. What does this show? It is not that AI is immediately in trouble, but that AI's sensitivity to rates is shifting from the "valuation side" to the "financing side." Previously, tech stocks mainly feared a higher discount rate; now some AI infrastructure companies also face a practical question: can the future cash returns generated by massive capital expenditure outrun increasingly higher financing costs?
As a result, tech stocks must be layered, and investing is more about finding structural opportunities. Mature platform companies have strong cash flow and relatively limited debt pressure; rising rates first pressure valuation, while profitability can provide a buffer. AI infrastructure companies have large capital expenditure and high financing needs, and are affected by both valuation and financing costs. There is also a category of companies whose profits are mainly realized far in the future and whose valuations depend heavily on terminal market space; they are most sensitive to real rates and risk premium. Therefore, the statement that "tech stocks are not afraid of rates" is outdated. The reality is: high growth can cover high rates, but only if the growth is real and the growth rate is high enough.
This is also why 5% itself is not a number that requires mechanical liquidation. What truly deserves vigilance is: if US Treasury yields continue to rise, AI capital expenditure begins to slow, corporate earnings forecasts are revised down in tandem, and credit spreads begin to widen again. Then it is not simply a valuation adjustment, but pressure on both valuation and earnings at the same time.
It should also be noted that dividend stocks are not a natural safe harbor. This is an easily overlooked point in this round of asset allocation. When the 10-year US Treasury yield rises above 5%, it means risk-free assets themselves begin to offer higher coupons. Utilities, REITs, telecom, and other high-dividend, long-duration assets have their biggest advantage in stable cash flow and dividend yield; once the risk-free rate rises, the yield gap between them and Treasuries will naturally be compressed. So rising rates are pressure from "discounting future cash flows" for tech stocks, and more pressure from "dividend yield relative to risk-free yield" for dividend stocks.
But banks, insurers, and energy are different. Banks may benefit from a higher rate center, but if high rates ultimately translate into credit risk, the logic reverses; insurers may benefit from higher long-term investment yields, while also bearing bond price volatility; energy and metals trade more on inflation, supply and demand, and commodity cycles. This means: dividend is not an industry, but a cash flow attribute; tech is also not an industry, but a growth attribute. (This indirectly explains why the Magnificent Seven and momentum stocks have often been stronger than other sector themes recently.) What the market is really doing is repricing the duration, certainty, and growth rate of different cash flows.
Going forward, what is most worth watching is not the round number of 5% on US Treasuries, but whether several variables deteriorate at the same time. First, watch the 10-year nominal yield and the 10-year real rate. The nominal rate determines the overall price of capital in the market, while the real rate more directly affects growth stock valuations. Second, watch term premium and the 10Y-2Y curve. The curve has now returned to a positive slope, but more important than whether it is steep or flat is whether the rise in long-end yields comes from economic growth, inflation, or fiscal supply. Third, watch the MOVE index and credit spreads. Bond market volatility and investment-grade and high-yield credit spreads are key to judging whether the "rate problem" has evolved into "tightening financial conditions." Changes in AI bond spreads relative to the overall market are especially worth observing. Fourth, watch AI capital expenditure and earnings expectations. Whether capex at Microsoft, Google, Amazon, Meta, and Oracle continues to be revised upward, and whether orders for GPUs, storage, optical modules, optical communications, and power equipment continue to be fulfilled, is more valuable than staring at the Nasdaq's daily moves. BlackRock still believes that supply bottlenecks and earnings growth in AI infrastructure can support risk assets, but it also reminds us that financing structure will become increasingly important. Fifth, watch the relative strength of growth versus value, Nasdaq versus Dow, and semiconductors versus high-dividend sectors. This indicator can help determine whether the market is trading "rates" or "industry." If the future brings "Treasuries keep rising + real rates keep rising + AI earnings expectations unchanged," it is more of a valuation rebalancing; if it becomes "Treasuries keep rising + real rates rise + earnings revised down + credit spreads widen," the nature of the risk is completely different. These two types of market action may both look like rising US Treasury yields on the surface, but the asset allocation answer cannot be the same.
Asset allocation is essentially a balance of risk and return across time and space. Macro beta addresses when to take risk, industry beta addresses which trend risk to take, and individual stock alpha addresses who can truly retain profits within the same industry chain. Stock investing certainly cannot ignore macro beta, nor can it disregard rates, the dollar, oil prices, and liquidity; but if all attention is placed on these variables, it is easy to fall into a short-term loop of guessing the Fed and guessing Treasury levels every day. A truly cycle-crossing investment framework ultimately returns to three questions: what is the cost of capital, where is the industry cycle, and can corporate profits be realized? The biggest lesson from this round of rising US Treasury yields is not that "tech stocks are in danger," but that the valuation logic previously amplified by low rates is gradually giving way to a more realistic standard: can a company's earnings growth outrun its cost of capital? For growth stocks that can do this, rising rates are more of a valuation disturbance; for companies that cannot, rising rates will truly change their investment logic.
The market has never been a simple seesaw between bonds and stocks. Rates determine the water level of valuation, industry determines the space for earnings, and time will ultimately reveal the gap between the two, while the crux of America's "high growth, high inflation, high rates" ultimately rests on the return of the AI industry. If you are still confused about investment positioning after the double holiday; if you are puzzled about industry investing; if you are unsure how to find industry leaders and potential companies; if you feel lost about capital rotation and sector volatility; then please scan the QR code to join us, obtain more detailed industry chain analysis and investment strategies, and follow Gelonghui Research Institute to seize opportunities in the wave of investing and share in the dividends of the industrial revolution.
Risk warning: Investing involves risk, and decisions must be made cautiously. The above analysis is only a reference for industry research and market strategy, and does not constitute specific investment advice. To obtain the latest updates and operational strategies on targets in a timely manner, and to capture signals, identify risks, and position for opportunities as soon as possible, please scan the QR code below to access more timely and detailed professional investment strategies.
Disclaimer: The copyright of this article belongs to the research team of Beijing Gelonghui Investment Advisory Co., Ltd. (Duan Yuehan: A0160625050003). This report is produced based on the principles of independence, objectivity, fairness, and prudence. All information comes from public sources and has been collected, summarized, and edited lawfully, reasonably, and appropriately. The stock market carries risk, and caution is required when entering. None of the investment advice in this article should be used as the basis for your investment buying or selling decisions. You must make investment decisions independently and bear the risks yourself. Any unit or individual that publishes, copies, distributes, or posts or reprints this content online without the company's permission shall be deemed to have committed infringement, and the company will pursue legal liability in accordance with the law.
Comments