How do you view the collective collapse of AI giants?

这篇文章从美股中半导体板块的大跌入手，谈到了一系列巨头公司的股价下跌情况，包括台积电、英伟达、VRT、ARM、超微电脑、博通、AMD、美光科技、迈威尔科技等。作者指出这一波大跌对市场情绪造成了严重影响，拖累了其他行业。文章还探讨了导致这一局面的原因，包括超微电脑不按惯例公布业绩预告，以及其发行新股所带来的影响。

文章中提到的一些原因可能会导致投资者对半导体板块产生担忧，特别是超微电脑发行新股的举动可能被解读为对未来业绩的不确定性。这种不确定性可能导致投资者的恐慌情绪，从而引发股价的下跌。

至于未来走势和投资建议，需要根据更多的市场信息和个人投资偏好做出决策。

However, one month later today, the institution did not exercise this option. This has led to investors being even more convinced that the institution does not recognize the current valuation of AMD.

The convergence of these two negative events has dealt a severe blow to AMD, causing it to collapse completely.

In fact, there are more negative factors fermenting in the market at the same time.

Last Tuesday, ASML released its financial report, with a dramatic 61% decrease in the total value of new orders in the first quarter, far below expectations. The main reason for the decline in new orders is the significant decrease in demand for the most advanced EUV lithography machines. The stock price plummeted by over 7% the following day.

Then, a day later, TSMC released a mixed financial report. Although its net profit in the first quarter increased for the first time in a year, it lowered its global wafer foundry industry growth expectations.

Lithography machines and chip manufacturing are the most important upstream sectors in the corresponding industry chain. Now, with both sectors experiencing a downturn, it has significantly intensified market concerns about the entire industry chain.

Another significant signal comes from ARM, a chip design company that doubled in value in just a few days in February. ARM, which represents the most upstream company in the entire chip industry, also experienced a significant drop in stock price after ASML's performance disclosure, plummeting by 12% in a single day, causing the candlestick chart to completely break, bringing extremely negative stimuli to the market.

02 How do you see it?

From a macro perspective, although the overall performance of the current U.S. economy is strong, the global economic and political environment is not friendly.

In particular, the repeated delay in interest rate cuts by the Federal Reserve and the ongoing escalation of tensions in the Middle East have significantly undermined confidence in global economic development, causing capital markets to refrain from risky investments.

Stock markets in the U.S. and Japan have been continuously surging for a long time, especially with the valuation bubble of technology companies significantly inflated. At the same time, funds have been crowded into these areas for a long time.

So, some people are planning to exit first.

Data shows that in the past two weeks, U.S. stock investors have redeemed $21.1 billion from stock funds, marking the largest two-week record since 2022.

These funds have been redirected to buy gold, energy, and bulk commodities, leading to these assets continuously reaching historical or interim highs in the near term, which is actually causing "inflation" driven by the need for safe-haven assets.

In fact, panic selling has already appeared in the market, triggering a stampede.

Analysis in the institutional dark pool channel has revealed that institutional traders dumped orders worth hundreds of millions of dollars at the close, while selling a call option worth $1.17 billion in the options channel, speculating that this was due to stop-loss measures forced by risk management.

Top Wall Street strategists have already indicated that the long positions in the S&P 500 are worth $52 billion, with 88% of them in a loss-making state. If the market continues to turn negative, the speed at which they are sold to cut losses will be rapid, and the quantity will be larger, potentially triggering a larger-scale plunge.

Apart from the semiconductor industry, the recent performance disclosures of other tech giants have also brought significant pressure to the market.

For example, Tesla, which is about to disclose its performance, has recently experienced a global decline in sales and a large number of employee optimizations, intensifying market concerns about its performance.

Previously, Apple, the market value leader in the U.S. stock market, is also facing similar challenges, with recent data showing a significant decline in global iPhone sales in recent months, coupled with ongoing EU sanctions and global geopolitical influences, leading to market concerns about its performance expectations.

These are all heavyweight giants in the U.S. stock market, so it is conceivable how much of a short-term impact their collective "turnaround" in performance will have on the entire U.S. stock market!

So, returning to the semiconductor sector, this significant retreat can be understood.

But even so, this round of sell-off is still only "killing valuation, not logic."

The biggest problem for Nvidia is that the market's expectations for it are too high, leading to a significant increase in valuation over the past two years, which needs to be digested.

But the AI boom will not stop, even though Google, Intel, and AMD are also preparing to launch their own AI chips to grab a share of the market. However, the demand for its orders is still so large in the short term that it is difficult to digest.

Nvidia's orders are already scheduled into next year, especially with the introduction of the more powerful GB200, institutions estimate that its sales next year could reach as high as $90-140 billion (60-70,000 units).

Recently, it was reported that Microsoft plans to have 1.8 million AI chips by the end of 2024, which means that Microsoft plans to double the number of GPUs it has this year. Insiders revealed that from the current fiscal year to the 2027 fiscal year, Microsoft is expected to spend about $100 billion on GPUs and data centers.

Although it is not known how much of Microsoft will adopt Nvidia's GPUs, at least it reflects the capital expenditure of the giants on AI chips, which will definitely increase significantly in the future.

Looking at the entire AI industry chain in this context, its prosperity logic will not change.

For example, in the center of the recent market downturn, there is AMD. For the entire 2024 fiscal year, AMD expects its revenue to reach $14.3-14.7 billion, consistent with institutional forecasts. This scale has doubled compared to the 2023 fiscal year, showing strong growth.

According to institutional forecasts, AMD's adjusted earnings per share for the 2024 fiscal year are $21.95, even at a 30 times PE, the valuation will exceed $60 billion.

Compared to the current valuation level, there is still a lot of room for upward movement.

So, if this round of downturn in U.S. tech giants under the resonance of multiple negative factors leads to a larger decline, it could be a sharp drop. But there is a high probability that after this round, the AI chip semiconductor industry chain will likely see a very rare "golden pit".

Of course, it seems that the current decline is not large enough, so patience is still needed.

03 What about the A-share AI concept?

So far, the vast majority of the domestic AI industry chain is dominated by the several major AI giants in the U.S. stock market, and even some are "imported" in terms of underlying technology and AI models.

So, if the performance of these giants does not meet expectations in the short term and their stock prices plummet, it will inevitably have an impact domestically.

After all, the domestic market sentiment is not high, and confidence has weakened again, so the possibility of being led by the nose should also be considered.

Objectively speaking, in terms of the quality of A-share AI concept stocks, there are not many that truly have confidence. Because commercialization is still in the exploratory stage, even if there is revenue, it will not have a significant impact on overall revenue. Most are just concepts without performance. So, it is likely just an emotional impact at the trading level.

However, the domestic government is vigorously promoting the independent development of the AI industry, with no shortage of policies and funds, so the AI industry wave will definitely continue to develop rapidly. Although AI software assets will be under pressure due to the need to continue exploring business models, some hard assets such as data storage, CPO, liquid cooling, servers, and other hardware support will still receive performance support.

Recently, the performance of the domestic photoresist industry chain has been quite good, and the stock prices of CPO giants have also begun to rise.

Of course, at this point, you absolutely cannot rush to bottom fish, let alone leverage. Otherwise, before the counterattack arrives, you may have already exhausted your ammunition and fallen before dawn. (End of the full text)