Silicon Valley's Wang Chuan: When the neighbor Lao Wang invests in storage stocks and makes 30 times profit, how can he not feel anxious?

null

Author: Silicon Valley Wang Chuan, investguru

1/ People’s anxiety often stems from the stress response of the brain’s amygdala. To alleviate anxiety, simple admonitions like “Don’t worry, worrying is useless,” or comforting words won’t work. These cannot change the underlying mechanism that triggers the stress response. The truly effective method is for anxious individuals to slowly write down the things and reasons causing their anxiety. The process of writing is about transferring control from the amygdala, which only knows how to be tense and alert, to the rational thinking prefrontal cortex. The more carefully and deeply you analyze while writing, the more thoroughly you transfer control, and naturally, anxiety decreases. Even if you can’t finish in one go, just starting to write and having a rough idea of how long it will take to organize your logic can help reduce anxiety. Don’t force yourself to write slowly; just drinking mental chicken soup won’t achieve this control transfer, and you won’t fundamentally change your anxiety.

2/ One thing that made people anxious in May 2026 was the overall surge in semiconductor stocks. The famous Philadelphia Semiconductor Index increased by 150% compared to a year earlier. Various semiconductor memory company stocks also rose across the board, with SNDK closing at $1,562 on May 8th, which is 38 times higher than a year ago. I don’t have investments in semiconductor companies, but I attempt to analyze the underlying logic behind these stock increases in this article and share my thoughts with readers.

3/ This story begins with the rapid rise in valuation of various AI startup companies. Many speculators only care about stock price fluctuations and lack some basic knowledge:

First, most AI startups are severely unprofitable, but many people completely confuse the concepts of revenue and profit (some even do so intentionally), and this misunderstanding persists and spreads.

Second, when a company says its ARR (Annual Recurring Revenue) reaches one billion dollars, it doesn’t mean its total revenue for the past year was one billion dollars; at most, it indicates that its most recent monthly revenue is about $833,000,000. If you misunderstand this as a total annual figure, it’s your own problem.

Third, a company that doesn’t say much but reveals more information than it states is telling you something. If it reports ARR but doesn’t mention profitability, it means it’s seriously losing money. If it announced six months ago that its ARR was $1 billion but hasn’t spoken out since, it’s likely that its revenue has shrunk and losses have worsened.

4/ Take the currently popular large model company Anthropic (hereafter referred to as Anth) as an example. Some media reports at the end of April disclosed that the company’s ARR had approached $40 billion. In other words, the company’s monthly revenue in April was close to $3.3 billion ($40 billion divided by 12). In March, the company’s ARR was still below $30 billion, meaning its March revenue was under $2.5 billion. Anth’s CFO disclosed in a public document in March that from its founding in 2021 until March 2026, its cumulative revenue exceeded $5 billion. In other words, the total revenue (again, not profit) accumulated over the past five years since Anth’s founding is about $10.8 billion.

5/ There is also a definition issue regarding revenue here. A large proportion of Anthropic’s income comes from Google Cloud and AWS. Cloud providers take at least 20% commission from what they charge their clients, and the remaining goes to Anthropic. Therefore, the ARR that Anthropic publicly promotes needs to be discounted by 20%. But we won’t dwell on this, because Anthropic still loses a lot of money every month.

6/ How much loss? I asked several different AI companies, and based on limited public information, assuming their monthly revenue reaches $3.3 billion, estimates of monthly losses range from $1.1 billion to $1.7 billion. This is also the fundamental reason why Anthropic continues to seek external financing. By the end of April, Anthropic had raised a total of $72.3 billion over five years, but its accumulated revenue was only $10.8 billion, and it still needs to raise more funds. People who only read headlines mistakenly believe that Anthropic has already made a fortune, but they don’t understand the enormous pressure on management to constantly seek new external funding to sustain operations.

7/ Anthropic is not without competition. Its CEO recently publicly admitted that U.S. AI models are only behind Anthropic by one to three months, while Chinese AI models lag behind by six to twelve months. The products of various companies wax and wane, and changes happen very quickly, making it difficult to determine the ultimate winner based on one or two years of performance. For example, Cursor was the most popular programming assistant among AI developers from 2023 to early 2025, and at one point, it was extremely hot. But in the first half of 2025, Anthropic launched Claude Code, which can automatically generate large amounts of code, allowing many non-programmers to participate in coding. This became the core driver of Anthropic’s rapid growth in revenue and valuation over the past year. Recently, some programmers have complained that Claude Code’s experience and cost-effectiveness are inferior to OpenAI’s Codex, so they switched to the latter. For many users, switching costs are not high. Besides OpenAI, SpaceX’s X.ai also started working closely with Cursor last month, developing a competitor to Claude Code.

8/ Investors optimistic about Anthropic often cite scenarios where Claude Code’s inference profit margin reaches 70%. However, such claims ignore two key points: first, the huge costs of continuously training new models; second, the assumption that competitors do not exist and there is no pressure to lower prices. Clearly, this is unrealistic. When Anthropic raised funds in February, its valuation was $380 billion. Just three months later, it sought new financing with a valuation of over $900 billion. Recently, in a secondary private equity market, there were reports of a valuation reaching $1.2 trillion. (Some podcasts even mentioned a long-term valuation of $5 trillion.) This is a company with only five years of history, $10.5 billion in cumulative revenue (not profit), yet still losing over $1 billion each month. Despite only a year of rapid growth, some believe it can last forever. With $720 billion raised and five years of losses, generating only $10.5 billion in revenue, it can be valued at $1 trillion in the capital market. Such optimism, with so many entrepreneurs longing for it, is understandable!😊

9/ Berkshire Hathaway (hereafter BRK), under Warren Buffett, had $370 billion in revenue in 2025, operating cash flow of $45.9 billion, after-tax profit of $66.9 billion, nearly $400 billion in cash on hand, but a market value of only about $1 trillion. From another perspective, ten days of BRK’s revenue equals Anthropic’s total revenue since inception, yet some investors still believe that Anthropic, which still heavily relies on external investment to survive, should have a higher valuation than BRK. The enormous courage needed to hold such a belief is astonishing.

10/ The valuation growth of Anthropic and OpenAI has been crucial for the entire AI industry’s capital expenditure (capex) surge and the subsequent boom in the storage industry. Looking back over the past few years, this logical chain is very clear. To see what happens next, stay tuned for the next installment.