In the first full trading week of August 2026, the US stock market presented a fascinating picture. On one side, cloud giants surged collectively—Alphabet, Google's parent company, surpassed Apple in market capitalization to become the second globally; Amazon's market cap reached $3 trillion for the first time. On the other side, the semiconductor sector, which had been booming over the past two years, faced overall pressure, with leading memory chip stocks experiencing successive sell-offs. This bifurcated trend, characterized by "ice and fire," is being interpreted by more and more institutions as a profound "gear shift" in the AI investment logic: the market is no longer satisfied with the "shovel seller" computing power narrative but is beginning to scrutinize who can truly turn AI into profit.
Cloud Giants Reach the Summit, Semiconductors Face Pressure: Why Are They Diverging in the Same Sector?
On August 3rd, the Van Eck US Technology Seven Giants Index rose by 3.03% in a single day. Among them, Meta rose by over 6%, Microsoft by nearly 5%, Amazon and Google rose by 4.58% and 4.44% respectively, Tesla by 3.49%, and Nvidia by nearly 3%, while Apple declined by 1.78% against the trend. With this strong rally, Google's total market capitalization suddenly surpassed Apple's to become second globally, and Amazon's market cap broke the $3 trillion mark for the first time, quietly rewriting the global tech landscape.
In sharp contrast to the cloud giants, the semiconductor sector has been under overall pressure recently. From the sharp fluctuations in the Philadelphia Semiconductor Index to the continuous pullbacks in memory chip stocks, the market's divergence in valuing the hardware segment has significantly increased. Some analyses point out that the semiconductor sector has undergone a significant deleveraging process since July: hedge funds have drastically reduced their positions in core semiconductor stocks, and momentum traders like CTAs have also exited in large numbers, rapidly dismantling the previously crowded trading structure.
Shen Meng, Executive Director at Songsun Capital, commented that the US stock market is currently in a sector rotation phase of the AI industry chain—the semiconductor sector rose significantly earlier, while the cloud sector is in a valuation trough, creating a natural momentum for "high cuts and low buys." This rotation is not a simple style switch but a revaluation of the business models across various segments of the AI industry chain.
Narrative Shift: From "Speculating on Shovels" to "Validating Gold Mines"
Why are stock prices diverging in the same technology sector? Tian Lihui, a finance professor at Nankai University, offers a very insightful interpretation: the market is shifting from "speculating on shovels" to "validating gold mines." Over the past two years, funds frantically bought stocks of "shovel sellers" like Nvidia and Micron, betting on the explosive growth of AI computing power demand; but now, funds are voting with their feet, flocking to the "gold diggers" who are actually making money from AI. This indicates that the market is no longer satisfied with the "having computing power" story but wants to see "who can monetize that computing power."
Earnings data confirms the rationality of this trend. Alphabet, Google's parent company, saw a 24% year-on-year revenue increase and a near threefold increase in net profit in the second quarter; Microsoft's latest quarterly revenue grew by 18%, and net profit increased by 31%; and Amazon's second-quarter revenue grew by 20%, with net profit surging by 245% year-on-year. The performance reports from the three major cloud providers have convinced the market that the cloud business is on an upward trajectory—they control data entry points and paying users, possess pricing power, and the monetization path for AI applications is far clearer than simply selling hardware. In contrast, the logic in the hardware sector is relatively fragile. Once demand growth slows, pure hardware manufacturing is prone to falling into a price war; if hardware expansion outpaces the digestion rate of application demand, temporary overcapacity may emerge. This is the underlying reason why capital is shifting from semiconductors to cloud computing—the narrative of computing power is grand, but the realization of capital expenditure takes time, and the market, after two years of frenzy, has lost patience.
Performance is the dividing line
Market participants generally believe this is not a short-term fluctuation of a few days, but a structural turning point. AI is gradually moving from the infrastructure investment phase to the application implementation phase, and capital will accelerate from stocks lacking performance support to leading companies with AI computing power implementation capabilities and high certainty growth. Investors are no longer using a general upward trend approach; instead, they are conducting fine-grained evaluations of corporate fundamentals, and the market's tolerance for high-risk, concept stocks with no clear monetization logic is rapidly decreasing. In the future, capital will avoid hardware hoarders who "only burn cash without producing," opting instead for software and service giants who can turn AI into profit—a shift from "abstract to tangible," which is the most distinct feature of current US stock capital flows.
Valuation "Draining Water" Enters the Second Half: Premium Returns to Six-Year Low
It is worth noting that this round of adjustments is also reshaping the valuation coordinates of US stocks. Data shows that the forward P/E ratio of the S&P 500 has fallen from 22.2 times at the end of last year to 20.4 times, and the P/E ratio of the technology sector has retreated from 26.5 times to 22.1 times—the rapid growth in earnings is effectively digesting the previous high valuation. The JPMorgan strategy team points out that the de-leveraging speed of the technology sector, including memory chips, far exceeds previous expectations; based on position distribution, there is relatively limited room for further selling.
What is more significant is the change in the valuation premium of US stocks relative to global markets. Currently, the valuation premium of US stocks over global stocks has narrowed to around 22%, which is far below the 31% average over the past 10 years and is a six-year low. Goldman Sachs strategists also found that the valuation premium of the top five weighted stocks in the US is "almost gone" compared to the other components, with their forward P/E only slightly higher than the other 495 components of the S&P 500. This return to valuation advantage creates space for long-term capital to re-enter.
However, "reasonable" valuation does not mean "cheap," nor does it mean an immediate rebound. RBC Capital Markets strategist Lori Calvasina warns that the period of the Federal Reserve's policy shift is often accompanied by severe market volatility. Based on historical experience, the performance of the S&P 500 during policy-sensitive periods often exhibits non-linear characteristics, and the path to recovery will not be a smooth ride.
Capital Side: De-leveraging Nears End, But August Lacks "Fuel" for Upward Movement
From a capital perspective, US stocks in August are more likely to continue a pattern of "volatile digestion" rather than "unilateral attack." A recent Goldman Sachs "Fund Flow" report indicates that US stocks have experienced noticeable momentum pullback and position unwinding in recent weeks, and while the overall institutional positions have been cleaned up, the substantive risk reduction is not complete. Constrained by seasonal fund outflows and insufficient institutional offensive intent, US stocks lack "fuel" for upward movement in August, and the market may continue to maintain a range-bound pattern.
Historical data also suggests seasonal headwinds: in election years, the average return of the S&P 500 in August and September is often negative. Coupled with the Federal Reserve "doing nothing" and the continuation of a high-interest-rate environment, the market lacks the strong impetus of new money, and the index is more likely to digest positions with wide fluctuations at the current level, trading time for space.
Persistent Risks: Interest Rates, Inflation, and Policy Variables
Risk factors cannot be overlooked. Driven by continuously rising oil prices, concerns about "secondary inflation" are heating up, with US 10-year Treasury yields steadily climbing toward 4.7%. In an environment where real yields are at a cycle high and US Treasury volatility is increasing, stock market volatility is unlikely to remain low. Goldman Sachs derivatives experts warn that as the post-earnings season buzz fades, market focus will shift back to the interest rate and inflation trajectory. The strategy team at FaXing Bank holds a more aggressive view: considering tariff policy risks, AI-driven capital expenditure cycles, and persistent fiscal deficits, the future inflation backdrop will be more complex than what the market is currently pricing in.
Meanwhile, hawkish voices within the Federal Reserve are also noteworthy. Although the recent FOMC meeting maintained interest rates, three regional Fed presidents voted in favor of a rate hike, indicating the decision-making body's vigilance regarding inflationary stickiness. With the prospect of high interest rates remaining for longer, the valuation ceiling for growth stocks still exists, and any inflation data exceeding expectations could trigger sharp market fluctuations.
Conclusion: Structural Opportunities Emerge, Stock Picking is More Important Than Betting
For investors, this round of internal divergence in technology stocks offers several insights.
- First, AI investment is moving from the "storytelling" phase to the "performance validation" phase. Free cash flow and the proportion of AI revenue will become the core metrics for screening targets; mere capital expenditure narratives are no longer convincing.
- Second, intra-sector divergence signals the end of the era of broad-based gains. Fine-grained stock selection is more important than betting on indices, and the low correlation among US stock sectors in 2026 further highlights the value of diversified allocation.
- Third, as valuation premiums revert to the mean and leveraged positions are significantly deleveraged, structural opportunities are emerging in US stocks, but volatility risks from fluctuating interest rates and inflation must be guarded against.
In terms of allocation strategy, some institutions suggest focusing on weighted indices like the S&P 500 at this stage—as it more genuinely reflects the overall picture of the real economy, rather than being dominated by the weight of a few giants. At the industry level, cloud service and software leaders capable of realizing AI performance, as well as high-quality semiconductor stocks benefiting from AI capital expenditure expansion, may become the next focus areas for capital. Institutional strategies are shifting from "full-market bullishness" to "structural allocation," highlighting the current market divergence.
Ultimately, the market always rewards investors who remain sober amid the noise. As the halo of the "shovel seller" gradually fades, the quality of the "gold digger" will be the key variable determining the height of the next cycle in US stocks. At the crossroads of the AI narrative shift, whoever first presents a profit report that convinces the market will win the next round of capital votes.
