TL;DR

• Bank of America warns that there are too many risk signals in U.S. stocks, suggesting investors take profits and manage risks.
• AI leaders still have revenue and capital expenditure support, but the market has already bought into much of the future growth.
• Related assets: SPY, QQQ, NVDA, MSFT, GOOGL, AMZN, META, AVGO, AMD, SOXX.

U.S. stock investors are now facing issues that extend beyond simply being bullish or bearish.

On one side is the Bank of America U.S. equities and quantitative strategy team, led by Savita Subramanian, which published a client report titled "Too many red flags. Take profits." on June 5. According to Axios on June 9, the report argues that the risk signals in U.S. stocks have become excessive and provides a more direct position recommendation: take profits.

On the other side are the still strong fundamentals of AI. Microsoft, Google, Amazon, and Meta are continuing to increase their capital expenditures on AI and data centers, and Nvidia's data center demand remains the core anchor of the semiconductor cycle. Unlike the internet bubble of 2000, this round of market leaders has shifted to a group of giants with cash flow, profits, cloud revenue, and chip orders.

Therefore, the real question has shifted from "Is there a bubble in AI?" or "Is Bank of America calling a top?" to a more difficult question to answer: How should investors understand the current risks in U.S. stocks when historical top signals and real AI growth coexist?

The answer may be more uncomfortable than simply being bearish: the AI bull market may not have ended, but it has transitioned from the "buying growth" phase to the "testing the speed of growth realization" phase.

Bank of America warns about deteriorating odds

The value of Bank of America's report lies in placing the current market within a historical risk structure rather than providing a precise timing for a market top.

According to multiple financial media sources citing the Bank of America report, approximately 70% of the 10 bear market warning signs it tracks have been triggered. This ratio approaches the average level seen before several previous tops in the S&P 500 since 1990. The Bank of America framework also shows that 17 out of 20 valuation indicators for the S&P 500 indicate statistical overvaluation, with 8 of them above the peaks of the 2000 internet bubble. The CAPE (Cyclically Adjusted Price-to-Earnings ratio) or P/E10 is around 40, positioned in a historically high range.

These numbers can be argued against when viewed in isolation. High valuations do not mean a decline will occur tomorrow. Historical signals being valid does not mean they are accurate every time. AI companies’ stronger profits do make today different from 2000. However, when extreme readings of valuation, market breadth, style differentiation, and momentum appear simultaneously, the main point Bank of America wants to express is closer to this: the market can continue to hold, but the odds have worsened.

Market breadth is the key here. The index remains high, but the gains are increasingly reliant on a few AI and technology leaders. The current U.S. stock market exhibits narrow leadership characteristics similar to historical top phases: a small number of stocks contribute to most index gains, the proportion of S&P component stocks above key moving averages has retreated, and many stocks are not close to their own highs. The strength at the index level masks the decreasing internal participation.

Style differentiation reinforces the same signal. Bank of America mentions that the median return of the top quintile stocks in technology and the bottom quintile stocks is approximately 120 percentage points apart, the highest since February 2000, approaching the 130 percentage points seen before the top in March 2000. This resembles a concentrated bet on a few certain narratives, with no broad dispersion typical of a healthy bull market.

For investors holding SPY, QQQ, NVDA, or SOXX, the most dangerous aspect of this structure is the reduced margin for error. The index could certainly continue to rise, but as gains become increasingly dependent on a small number of stocks, any deviations in profits, guidance, capital expenditure returns, or valuation assumptions from any leading company can be magnified into a pullback for the entire portfolio.

AI cannot simply be equated to 2000

If one only looks at Bank of America's valuation and breadth signals, it might be easy to directly equate the current situation to that of 2000. However, this analogy is only partially correct.

A typical characteristic of the 2000 internet bubble was that many companies lacked mature business models, and investors were primarily trading on the imagination of "the internet changing the world." Today's AI leaders are different. Microsoft, Google, Amazon, and Meta's cloud and AI businesses are already reflected in real revenues, capital expenditure plans, and data center demand. Nvidia is not only a narrative center but also a highly concentrated provider of profits and cash flow.

Nvidia's latest financial report provided the strongest support for bulls. The company's Q1 FY2027 financial report, announced in May 2026, showed a quarterly revenue of $81.6 billion, with data center revenue of $75.2 billion, growing 92% year-over-year. In the face of these numbers, simply labeling the AI market as "a bubble with no fundamentals" is not very convincing.

AI optimists, including large tech company management and growth investors, are rebutting the bubble theory based on this point. They believe that this round of increases resembles an infrastructure cycle: training and inference demand are driving GPU, networking, storage, power, and data center construction, while cloud vendors are using higher capital expenditures to exchange for future AI service revenues, and enterprises further integrate AI into software, advertising, search, office, and development processes.

This framework has a factual basis. In the past few earnings seasons, major cloud vendors have continued to emphasize strong AI demand, and cloud business growth has been maintained. Nvidia's data center revenue has become an important pillar of the earnings growth narrative for U.S. stocks. Broadcom, AMD, data center, and power infrastructure companies are also included in the same investment chain. The market is willing to assign these companies higher valuations not only because their stories are appealing, but also because orders, revenues, and profits are indeed materializing.

This is also why Bank of America's signals cannot be crudely interpreted as "the AI bull market has ended." If the underlying fundamentals are still improving, a high valuation trend can persist longer than historical experiences. Especially in a market where passive funds, index weights, and institutional allocation collectively reinforce the leaders' positions, the strong getting stronger is part of the capital flow mechanism.

However, just because AI is real does not mean valuations are safe. Here a common misconception can arise: as long as the technological revolution is genuine, prices are not high. Historically, many bubbles have indeed been built on real technologies being overvalued too soon and too much. The internet truly changed the world, but investors who bought many internet stocks in 2000 still experienced a long period of valuation compression.

The core divergence in the current AI market is shifting from "Is AI useful?" to "How many years has the market already discounted?" The importance of Bank of America's historical signals lies in reminding investors that even when the fundamentals are real, if prices have already reflected too many positive future developments, risks will still rise.

Pressure shifts to income and cash flow

The AI bull market is entering its most challenging phase, not because demand has suddenly disappeared. The real change is that the market is beginning to demand more proof.

Over the past two years, investors were willing to pay high valuations for AI leaders because the growth path seemed clear: cloud vendors increased capital expenditures, chip companies sold more high-end GPUs, data center and networking equipment companies received orders, and future enterprise applications would release even greater revenues. Moving into 2026, the market needs to see not only continued investments but also whether these investments can translate into sufficiently high revenues, profit margins, and free cash flow.

Capital expenditure is the focal point of this issue. Microsoft, Google, Amazon, and Meta are continuing to increase their investments in AI and data centers, with the direction being essentially clear, but there is considerable variation in the estimates of the specific scale by different institutions and media. More importantly, investors have started to worry about the pressure of higher capital expenditures on free cash flow and return on investment. This cannot simply be written off as "AI investments cannot be recovered," but after the investment curve becomes steeper, the market's requirements for the return curve will also increase.

For Microsoft, Google, Amazon, and Meta, continuing to increase AI investment is strategically necessary. Those who stop might fall behind in cloud, search, advertising, office, models, and developer ecosystems. But from a shareholder's perspective, the higher the capital expenditures, the more future financial reports need to prove that these investments can bring incremental revenues, stable profit margins, and cash flow resilience.

For Nvidia, Broadcom, AMD, and the semiconductor chain represented by SOXX, the logic is slightly different. They are direct beneficiaries of the AI investment cycle, with orders and profits materializing sooner. However, because the market has regarded them as the core winners of the AI infrastructure cycle, if downstream cloud vendors slow capital expenditures, delay procurement, or begin to emphasize investment discipline, semiconductor valuations will react first.

This will create a more fragile feedback loop. Cloud giants increase capital expenditures, supporting chip company revenues. Chip companies' high growth supports index rises. Rising indexes and upward earnings revisions further strengthen the market's confidence in the long cycle of AI. If any link in this chain slows down, what the market faces may not be "the end of AI," but rather the need for valuations to realign with the speed of realization.

The second half's financial reports need to prove that risks can be covered by growth

Bank of America's 70% bear market signals will not automatically turn into a top, and the strong financial reports of AI leaders will not automatically eliminate valuation risks. What truly needs to be validated next is whether sustained growth can cover these valuation and market structural risk signals.

The most direct observation window will be the financial reports of the second half of 2026. Investors need to see that the AI revenues of large tech companies continue to grow, while profit margins are not being significantly eroded by capital expenditures and depreciation pressures. While cloud vendors continue spending, they also need to prove that customer demand is strong enough. Orders and guidance from semiconductor companies like Nvidia, Broadcom, and AMD will reflect whether downstream investment rhythms have slowed down.

Another variable is market breadth. If the S&P and Nasdaq continue to reach new highs but the number of stocks participating in the rise decreases, and high P/E stocks continue to systematically outperform low P/E stocks, the historical top structure mentioned by Bank of America will become harder to ignore. Conversely, if earnings spread to more industries, and the index no longer relies solely on a few AI leaders, there will be opportunities for the risk signals to be gradually absorbed by time and performance.

For ordinary investors, it is now more suitable to perform checks on position and concentration. Simply calling it "bullish on AI" or "bearish on U.S. stocks" does not solve the problem. AI may still be the most crucial investment theme for the coming years, but as pressures from valuation, breadth, and capital expenditures rise simultaneously, continuing to hold it has shifted from early trend discovery to a bet on the speed of realization.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。