Source: Yaya Hong Kong Stock Circle In the year and a half since GPT was introduced, the entire US stock market has revolved around AI and has experienced a bull market. Nvidia is the undisputed leader of the US stock market and has also experienced the greatest increase in market cap. While many people think that Nvidia's performance is overvalued and that its short-term gains are unrealistic and perhaps even bubble-like, closer tracking of the company's valuation shows that it is a stock truly driven by performance, and even if the AI bubble were to burst, Nvidia would not be the company that is hurt the most. It is indeed a technology monopolist and the company that has earned the most real profits from this boom.
In contrast, AMD is the second largest competitor in the AI chip industry and has been developing new computing chips in hope of taking a share from Nvidia. Almost all downstream companies in the industry chain are willing to support AMD. However, the truth is that despite the fact that AMD's stock price has increased by 100%, its revenue growth has been quite poor. Its revenue after the guidance for the second half of 2024 is only 25.2 billion, which is only a 6% increase compared to 2022. The profit margin is not rising but instead is falling. This revenue growth rate shows that AMD has played the role of a pseudo-AI company in recent years. A 6% increase in revenue has resulted in a doubling of the stock price.$NVIDIA (NVDA.US)$It can be said that if we compare the trend of AMD with Nvidia in the past two years, AMD has followed the downturn instead of the upswing. Moreover, its upswing amplitude is smaller, and its pullback amplitude is larger. Currently, AMD's high-end decline has exceeded 44%.
Many other companies within Nvidia's industry chain, such as in the server, optical module, and chip sectors, have seen similarly impressive increases in their stock prices. However, based on the performance numbers, there is little controversy.
During the AI boom that began in early 2023, the reasonability of a company's stock price could be understood by comparing its cumulative increase in stock price, expected revenue, and profit growth. As can be seen, Nvidia's stock price has risen by an astonishing 577% from 2023 to the present day. The revenue in 2022 was $14.8 billion, and the profit was $4.3 billion. Looking ahead to the farthest performance prediction for Nvidia, the revenue for the first half of fiscal year 2025 is expected to be $54 billion, with an annual growth rate of 323% compared to 2022. Due to the hot demand for AI chips, Nvidia's profit margin is currently much higher than in 2022, with close to $14.8 billion in profit in a single quarter. Therefore, the profit growth rate is more than 10 times higher. From the perspective of PE, Nvidia's stock price is currently lower than at the end of 2022, while from the perspective of PS, it is slightly higher. However, if Nvidia continues to decline, it could soon reach a new low in valuation. Regardless, Nvidia's several years of several-fold increases will ultimately have a significant increase in value.
One way to judge a company's performance during the AI boom is to look at their hardware performance.
From the PE perspective, Nvidia is currently lower than its valuation at the end of 2022, while from the PS perspective, it is slightly higher. However, if Nvidia continues to decline, it could soon reach a new low in valuation.
As can be seen, in recent years, Super Micro Computer has been rising faster than Nvidia and has become the number one competitor in this AI bull market.
However, despite the fact that its business model is not ideal, Super Micro Computer's performance shows a clear upward trend. In the past two years, Super Micro Computer has increased by 521%. Of course, at its peak, its increase was even higher than that of Nvidia, but its recent drop in stock price was due to a significant drop in profit margin in its financial report. The company provides a revenue guidance of $26-30 billion in fiscal year 2025, with a median of $28 billion. Its revenue was $5.196 billion in 2022, an increase of 338% compared to 2022. SMCI's estimated profit rate for 2025 compared to 2024 is expected to decline, but it is still at 8%, which translates to a profit of $2.2 billion, an increase of 785% compared to 2022.
Therefore, it is easy to see that Nvidia's performance increase is even better than its stock price increase. Concerns about the low return on investment in AI applications and the delayed shipment of Nvidia's new chips have recently caused the company to retreat by 30%, making it appear that its valuation is more reasonable. Even if it continues to decline, Nvidia's PE will hit a 5-year low. Therefore, no matter what happens, Nvidia's several years of several-fold increases will ultimately have a significant increase in value.
The danger lies in the companies that are not closely related to AI but have followed the Nasdaq market all the way up, such as many A shares, and a lot of them have returned to their original state this year. There are also many U.S. stocks that have returned to their original state, but there are still many that remain high up. Perhaps we should carefully assess the performance and determine which companies are truly pseudo-AI.
$Advanced Micro Devices (AMD.US)$ If we compare the trend of AMD with Nvidia in the past two years, AMD has followed the downturn instead of the upswing. Moreover, its upswing amplitude is smaller, and its pullback amplitude is larger. Currently, AMD's high-end decline has exceeded 44%. AMD has once again hurt Nvidia, encouraging investors to pursue the undervalued stock of the former.
As can be seen, Supermicro Computer is the number one little brother that has risen even more fiercely than Nvidia in this AI bull market. $Super Micro Computer (SMCI.US)$
Although this company's business model is not ideal, functioning primarily as a data center assembler, it faces a significant amount of competition and has never had a good profit margin. However, its performance trend is undeniable. In the past two years, Supermicro Computer has increased by 521%. At its peak, its increase was even higher than that of Nvidia. However, its recent drop in stock price was due to a significant drop in profit margin in its financial report.
However, it is worth noting that the gap between the increase in its stock price and the increase in revenue is still considerable. The company provides a revenue guidance of $26-30 billion in fiscal year 2025, with a median of $28 billion. Its revenue was $5.196 billion in 2022, an increase of 338% compared to 2022. SMCI's estimated profit rate for 2025 compared to 2024 is expected to decline, but it is still at 8%, which translates to a profit of $2.2 billion, up 785% compared to 2022.
According to the proportion of these two data, SMCI's stock price increase has been digested after a round of adjustment, and it can't be regarded as pseudo AI. Because orders are backlogged, profits have a certain predictability. The probability of major errors in actual performance in the next 25 years will not be very high. Product structure, the operating income of 10-30 billion yuan products are 401/1288/60 million yuan respectively.
As a company that sells shovels, the only question about their market is whether the demand is stable or not. There is a possibility that once everyone gives up investing in AI, orders will drop sharply. But this is a demand problem. The supply side has not made any mistakes. Similar to cyclical stocks and vaccine stocks, we cannot question the real profits they have earned.
However, as can be seen from AMD, there are also a large number of pseudo-AI companies in the hardware industry. They are more likely to use the concept of AI to drive stock prices soaring by sharing the expected benefits of existing beneficiaries. Representative stocks with high increase rates include Applied Optoelectronics (AAOI.US), and other semiconductor companies such as Broadcom. Both of them have performance completely incompatible with the increase in stock price. Which ones will really benefit from the chain effect and which ones are fake will naturally be differentiated. One example is Taiwan Semiconductor, which has a monopoly-like competition pattern, making it easier to receive profits from Nvidia. While the expectations of Broadcom and AMD are very similar and there is a chance of repeating the same mistakes. $Applied Optoelectronics (AAOI.US)$Please use your Futubull account to access the feature.$Arista Networks (ANET.US)$Please use your Futubull account to access the feature.$Vertiv Holdings (VRT.US)$ and other semiconductor companies.$Broadcom (AVGO.US)$,$Taiwan Semiconductor (TSM.US)$,$Applied Materials (AMAT.US)$They all have performance completely incompatible with the increase in stock price.
It is reasonable for the leading companies to become leading companies, but it is not logical for companies waiting for profit overflow to share. Once they do not move forward as planned, they become pseudo-AI, starting to fall instead of rising. Even if you are optimistic about the AI and believe that the prosperity will continue, you must understand that there are better and worse choices. In the hardware industry, which ones will really benefit from the chain effect and which ones are fake will naturally be differentiated. For example, as a semiconductor company, if Broadcom's expectation is similar to AMD's, it may repeat the same mistakes.
However, there are a few things to note. First, currently the U.S. stock market is facing a big decline, and the valuations of companies such as Nvidia that have already earned profits have already adjusted. However, the valuations of those that rely on forward-looking valuations are still to be realized, because the rationalization of their valuations still depends on the performance of the next few quarters.
Second, their profitability is a feature of the lagging release after the leading profits. If companies like Nvidia keep rising and their performance keeps improving, it's okay for them to catch up later. However, if they are faced with a great collapse in the AI industry and the demand for Nvidia declines, their performance is still soaring, is it to look at the premonitory performance of Nvidia or their financial reports? Their stock prices and valuations are waiting for performance to catalyze them, and in this situation, even good performance cannot catalyze the stock prices, which is a drawback.
Therefore, from the performance of the hardware industry, it is reasonable for the leading companies to become leading companies, and the logic of companies waiting for profit overflow is not that good. Once they do not move forward as planned, they become pseudo-AI, starting to fall instead of rising. Even if you are optimistic about the AI and believe that the prosperity will continue, you must understand that there are better and worse choices.
The profit trend of software is unclear.
The discussion on the AI bubble has been very extensive. The most discussed issue basically comes from the mismatch between input and output. There are data indicating that, assuming Nvidia's annualized revenue is 90 billion, the input of computing facilities and other equipment accounts for half, so it needs to build data centers with an input of 180 billion. If calculated based on the software industry's profit margin of 50%, AI needs to drive 360 billion in revenue to make sense.
However, the current situation is that the revenue growth of the software industry is far from this estimation.
Of course, this 360 billion is not necessarily an increase in revenue. If AI saves half of the manpower and produces the same efficiency, its impact on enterprises is mainly cost-related. With half the manpower cost, this part of the profit is released.
For global software companies, releasing 180 billion in incremental profits through profit margin optimization after large investments have been made in AI each year is an acceptable number. However, this excludes the profit growth generated by the development of enterprises themselves.
Currently, the profit improvement of global software application companies has not yet reached this level.
Microsoft, which holds OpenAI, is also the world's largest software company. From financial data, its revenue growth and profit margin have increased simultaneously. Specifically, Microsoft's profit in fiscal year 2024 has increased by about 21% compared to fiscal year 2022, with an increase of about 15.4 billion, and its cumulative revenue has increased by 23%. However, the forecast revenue next quarter is the same as the current revenue. This has lowered the expected revenue growth rate for 2025 to about 15%. So even if we calculate the revenue growth rate of 2025, Microsoft's revenue growth in these three years is only about 41%. $Microsoft (MSFT.US)$But Microsoft's net profit margin is declining, so there is reason to believe that its three-year cumulative profit growth rate until 2025 will not exceed 41%.
At present, Microsoft's stock price has also increased by 70% in two years, and both AI PC and Copilot have been launched for a long time. In theory, they should have driven some growth. But the growth slope has not changed much. The three-year growth rate is not as good as the two-year stock price increase, which shows that Microsoft's performance is not compatible with its stock price.
$Alphabet-A (GOOGL.US)$ / $Alphabet-C (GOOG.US)$Please use your Futubull account to access the feature.$Meta Platforms (META.US)$ As a competitor in the face of Microsoft's large-scale model arms race, the product has a lower degree of integration but a higher increase. However, in comparison with performance, the performance is indeed slightly better than that of Microsoft. Google and meta's income in the past two years has increased by 37% and 33%, respectively. However, meta's profit in 2022 is very low due to excessive investment in the metaverse two years ago. In the past two years, profit has increased by nearly 160%. However, these two companies have not seen their stock prices rise significantly, regardless of income or profit.
Of course, the software industry's performance benefit from AI is certainly slower than that of hardware, so everyone has given more optimistic expectations for the future. However, it should also be noted that Microsoft and Google's latest fiscal year expected profit margins are not much improved compared to 2022, and both capitalization and cost accounting will increase computing power expenses. The growth of enterprise profits has a correction effect on the business (such as meta's cut in VR business) and natural growth accompanied by the economy. Even without AI, growth is reasonable, and the profit release brought by cost optimization through AI should theoretically improve profit margins, but this is not the case for now. This makes the goal of generating an additional profit increase of 160 billion even further away.
In addition, Microsoft is the company that combines AI applications the fastest and has the fastest decline in profit margin performance and revenue growth rate. Meta and Google are still exploring. The good results they have now are due to low base numbers and the superiority of their own business models, and have little to do with AI. So, is the result of the determination of AI going down continuous growth, rather than Microsoft? This is a question worth considering.
Other software service companies can be said to be a fake AI or not belonging to the AI sector, because the prosperity of AI started with GPT and Wensheng Diagram, and most software services have no logical business growth in this area, perhaps only cloud computing services have gained some prosperity by following Microsoft's AI infrastructure. But in terms of performance, most of them have a large increase in growth compared with income. From...to... They all have a significant increase in valuation. $Salesforce (CRM.US)$Please use your Futubull account to access the feature.$SAP SE (SAP.US)$ Saying that AI will promote software prosperity, most of them have a decreasing growth rate since 2022. AI has not changed the downward curve of the entire industry's growth rate. There are some improvements in performance growth rate, such as... this, which has a large increase, but looking back, it has a PS of more than 10 and a revenue growth rate of 30%. The entire software sector is now stuck at 30% growth rate, which is worrying. $ServiceNow (NOW.US)$Please use your Futubull account to access the feature.$Palantir (PLTR.US)$ In this environment, a bunch of software service companies whose growth rate is declining rapidly have already fallen back to their original form and plummeted. For example, star stocks such as...have been expelled from the AI sector and have become one of the worst-performing sectors in the US stock market all year round, which is the destiny of fake AI.
The development of AI will greatly promote productivity, but it is all talk without quantitative analysis. If it does not drive performance growth and profit improvement, or invests 100 billion to produce 10 billion, it is all empty. From the perspective of current performance and stock prices, the market has dismissed some AI stocks, but there are still many stocks in the hardware and software sectors that have not small bubbles due to the AI concept.$Palantir (PLTR.US)$and $ServiceNow (NOW.US)$ In the future, no matter whether the prosperity continues or doubts increase, there are several logics that need to be paid attention to. The logic of interest deferral and external overflow itself has greater risks. Secondly, for some leading Internet companies, AI is not necessarily a promoting factor for subsequent performance. Perhaps when the market can rationally regard AI and not think that it is a completely positive business promoting factor, but a factor of coexistence of good and bad, it will be a correct attitude. According to the degree of AI and the performance related to AI, it is obviously wrong to become the first evaluation index of the company's value.
$Cloudflare (NET.US)$,$Snowflake (SNOW.US)$The destiny of fake AI. Most software service companies whose growth rate is declining rapidly have already fallen back to their original form and plummeted, such as... and other star stocks, which are one of the worst-performing sectors in the US stock market throughout the year, have also been expelled from the AI sector.
III. Conclusion
Editor / jayden