Xinhua Finance Shanghai, June 27 (Ge Jiaming) In the field of AI hardware, computing power, storage and network are the three most critical elements. Nvidia dominates the market in terms of computing power, with a market capitalization of more than $3 trillion. In terms of storage, Micron's stock price has also been pushed to new highs driven by the increasing application of AI in PCs and mobile phones, as well as the demand for HBM (High Bandwidth Memory) for AI servers.
Yesterday, Micron handed over a good "report card", the third fiscal quarter performance and the fourth fiscal quarter guidance were higher than market expectations, but still could not meet investors' expectations, which also caused the company's stock price to fall nearly 8% after hours.
At the same time, Nvidia held a shareholder meeting, but there was no new incremental information, and Nvidia's stock price continued to fall. With NVIDIA's market capitalization exceeding $3 trillion, further verification of performance is undoubtedly needed to go further.
Andrew Jackson, head of Japan equity strategy at Ortus Advisors Pte in Singapore, believes that the market's unrealistic illusions about the performance of tech giants have also weighed on Micron's stock even after earnings reports beat expectations.
According to Jackson, Wall Street investment banks are well aware that U.S. technology stocks are overhyped, and too many investors are pursuing short-term gains and gradually abandoning long-term investment strategies in the real value of stocks.
Vincent Lin, a prime broker (PB) analyst at Goldman Sachs, said in a recent report that as technology stocks continue to climb to all-time highs, hedge funds have begun to choose to reduce their holdings over the past month, while retail investors continue to flock to the most crowded trades such as long tech stocks.
Tech giants are struggling to meet market expectations
The reincarnation of market sentiment cannot escape even if it is a "leading AI stock".
Xinhua Finance has repeatedly analyzed that NVIDIA's short-term performance depends on the giants' capital expenditure plans and the allocation of sovereign computing power, and in the long run, it depends on whether there are enough data centers and power to support it.
Just as Nvidia topped the U.S. stock market capitalization list, Rosenblatt Securities analyst Hans Mosesmann raised the price target on Nvidia stock to $200 from $140, saying it still has 50% upside.
Analysts told Xinhua Finance that Nvidia's stock price has risen by about 700% since OpenAI launched ChatGPT at the end of November 2022, so those who have already made a profit of 7 times will start to consider whether the 50% upside is still worth holding onto.
Analysts believe that NVIDIA has built a complete ecosystem, relying on the binding of hardware to the "moat" of CUDA, making full use of the giants' investment in AI, continuously accelerating the iteration speed of GPUs, and expanding its advantages in the "AI arms race".
But now there are two major disadvantages for Nvidia, first, the update speed of AI large language models is slowing down, which may lead to a decrease in the demand for computing power from giants.
In April this year, researchers from Google Research and Johns Hopkins University conducted a study on the efficiency of AI models in image generation tasks, showing that when computing resources are limited, smaller models may outperform larger models at the same sampling cost, which may mean that small AI models on the device side can meet the needs of most users, and to a certain extent, it may affect the needs of GPUs.
On the other hand, the deployment of computing power also needs to rely on energy, and the limit of power also means that there is a bottleneck in computing power.
NVIDIA's GPU supply problem will gradually be resolved over time, and there is already a backlog of GPUs and idle in some data centers, so with the development of computing infrastructure, the focus has become more central - the power issue. Even if there is enough electricity, there is a need to worry about the transportation capacity of the electricity.
Since Nvidia announced its fiscal first-quarter results on May 22, more than one-third of Nvidia's management has chosen to reduce its holdings while its stock price continues to rise, hitting a multi-year high. Therefore, it is still unknown how Nvidia will create new stimulus points for the market and push the market value further up.
It can also be seen from the stock price trend after Micron's earnings report that the technology giant seems to be unable to meet the market's expectations for its performance.
Micron achieved total revenue of $6.811 billion in the third quarter of fiscal 2024, up 81.5% year-over-year and slightly better than the consensus of $6.661 billion. The revenue recovery accelerated, mainly driven by the growth of the core businesses DRAM and NAND. Micron achieved a net profit of $332 million in the third quarter of fiscal 2024, continuing to remain profitable.
For the fourth quarter of fiscal 2024, Micron expects operating income of $7.4 billion to $7.8 billion (up 89.5% year-over-year), largely in line with market expectations ($7.57 billion), and gross margin (GAAP) of 32.5% to 34.5%, which continues to improve sequentially, in line with market expectations of 33.5%. Micron's shares fell despite earnings reports that were in line with expectations.
Analysts believe that although Micron's performance continues to improve, the market's expectations for Micron have been reflected in the stock price, Micron has risen by 67% so far this year, so Micron did not give the market stronger expectations, which has affected the market's confidence in it to a certain extent.
Hedge funds are reducing their holdings and exiting the market
The debate about the "AI bubble" has not stopped this year, and there are even huge disagreements within Bank of America and Goldman Sachs, but analysts generally believe that even the bubble will last for a long time.
According to the June Global Fund Manager Survey released by Bank of America strategist Michael Hartnett, long the "Big Seven" (Apple, Microsoft, Google, Amazon, Nvidia, Tesla and Meta) has been the most crowded trade for the 15th consecutive month, with 69% of fund managers choosing to continue to go long.
However, Goldman Sachs analysts said in a report that current data shows that hedge funds have begun to choose to reduce their holdings over the past month, selling large net stocks in the TMT sector (technology, media and communications), especially semiconductors and semiconductor equipment companies, including Nvidia.
The size of hedge fund sell-offs in the TMT sector in June hit the highest level on record with Goldman Sachs, while retail investors continued to pour into long-tech stocks, among others.
Vanda Research, which focuses on the observation of retail fund flows, said in its latest market report that the breadth of overall retail fund flows is rapidly shrinking, and funds are still flowing into the tech giants, continuing to chase Nvidia through 2x long Nvidia's ETF.
According to Vanda Research, beneath the calm of the market, a huge "asset redistribution" is taking place. Institutional investors are selling large amounts of tech stocks to retail investors. Retail investors will stop adding to large-cap tech stocks only if the sectors that retail investors prefer to "prefer" such as small-cap stocks, crypto concept stocks, or retail huddle stocks can consistently outperform the broader market.
Editor: Tan Rui
Statement: Xinhua Finance is a national financial information platform undertaken by Xinhua News Agency. In any case, the information published on this platform does not constitute investment advice. If you have any questions, please contact customer service: 400-6123115
![Galaxy L7 intelligent technology makes travel better[fig]](data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=)