Reporter Cai Ding
After trading on Wednesday, February 21 (Eastern Time), NVIDIA, which received widespread attention, released its fourth quarter financial report for the fiscal year ended January 28, 2024. The current revenue and profit both increased significantly year-on-year, exceeding analyst expectations. Revenue and profit have reached record highs for three consecutive quarters. In addition, Nvidia's revenue guidance for the first quarter of fiscal year 2025 was $24 billion (± 2%), exceeding market expectations of $21.9 billion.
Some argue that analysts may have underestimated the driving force behind Nvidia's business growth, as to why it has consistently exceeded expectations.
During the earnings conference call, NVIDIA CEO Huang Renxun stated that generative artificial intelligence has reached the "tipping point". Nvidia CFO Kress said, "We estimate that about 40% of data center revenue in the past year came from artificial intelligence." The market demand for Nvidia's next-generation products far exceeds supply.
Driven by this, Nvidia's stock price surged by 10% after trading, with a 9.04% increase as of press release.
Dan Ives, Managing Director and Senior Stock Analyst of Wedbush, stated in a comment email to the Daily Economic News that Nvidia's latest financial report further consolidates the fact that the artificial intelligence revolution is just beginning rather than reaching its peak, which is the most important moment in the market and technology industry for many years.
The far exceeding fourth quarter report is almost just a matter of time for Nvidia to hit a market value of $2 trillion. However, the record valuations of AI related stocks, including Nvidia, are reminiscent of telecom stocks during the Internet foam. The Financial Times reported that the current NVIDIA was compared with Cisco, the leading telecom company during the Internet foam, warning investors and institutions to remember history.
What does Nvidia rely on to exceed analyst expectations?
According to the financial report, Nvidia achieved a revenue of $22.1 billion in the fourth quarter of the 2024 fiscal year, a 22% month on month increase and a significant 265% year-on-year increase, benefiting from the surge in demand for server AI chips. This was higher than analysts' expectations of $20.41 billion, and its quarterly revenue was even higher than the full year of 2021. During the reporting period, Nvidia achieved a net profit of $12.3 billion, a year-on-year increase of 769%, and an adjusted earnings per share of $5.16, higher than analyst expectations of $4.59. Revenue and profits have reached record highs for three consecutive quarters.
NVIDIA
Some argue that analysts may have underestimated the driving force behind Nvidia's business growth, especially its data center revenue generating capabilities, as to why its performance has consistently exceeded expectations.
The financial report shows that Nvidia's largest revenue source, the data center division, achieved a revenue of $18.4 billion in the fourth quarter, a year-on-year surge of 409%, higher than analysts' expectations of $17.21 billion. Nvidia stated that the growth in data center revenue is mainly due to the rebound in demand generated by the easing of global economic downturn concerns. Nvidia CFO Kress analyzed that "the revenue of data centers in the fourth quarter was mainly driven by generative artificial intelligence and its related training. We estimate that about 40% of data center revenue in the past year came from artificial intelligence."
Huang Renxun, founder and CEO of Nvidia, pointed out that accelerated computing and generative AI have reached the "tipping point", and the demand for companies, industries, and countries around the world is surging, indicating that Nvidia's products and solutions in these fields are warmly welcomed by the market.
"Nvidia's strong quarterly report and performance guidance exceeded market expectations, further consolidating the fact that the artificial intelligence revolution is just beginning rather than reaching its peak. This is the most important moment in the market and technology industry for many years," said Dan Ives, Managing Director and Senior Stock Analyst of Wedbush, in a comment email to a reporter from The Daily Economic News
Ives pointed out that in the past few weeks, companies including Microsoft, Alphabet, Meta, Amazon, and others have all indicated that their AI capital expenditures are accelerating.
He pointed out, "We believe that 60% to 70% of businesses will eventually move towards the path of artificial intelligence applications, and we estimate that AI spending will increase by $1 trillion in the next decade. Currently, we are only talking about generative AI in enterprises, and AI in the consumer sector led by companies such as Alphabet, Meta, Amazon, and Microsoft is also about to arrive."
Nvidia CFO Kress stated during the earnings conference call that the market demand for the company's next-generation products far exceeds supply, especially the new generation chip B100 that the company expects to ship later this year. He stated that "building and deploying AI solutions has touched almost every industry," and it is expected that the scale of data center infrastructure will double within five years.
The Daily Economic News reporter noticed that next month, Nvidia will hold a flagship event GTC conference for artificial intelligence developers. At that time, Nvidia may reveal the details of B100. Morgan Stanley analyst Joseph Moore stated in a report earlier this month that based on some early disclosures about the system, the B100 will be a "major advancement in state-of-the-art technology.".
For Nvidia, new products are crucial for coping with increasingly fierce competition and maintaining profitability. Vivek Arya, senior analyst at Bank of America, believes that Nvidia's upcoming B100 product will be priced at least 10% to 30% higher than the company's currently popular H100 system.
Nvidia now=Cisco before the Internet foam burst?
The Daily Economic News reporter noticed that under the strong demand for chips, Nvidia's stock price has risen by 40% this year, with a market value increase of over $400 billion, reaching $1.67 trillion. During the first 30 trading days ending February 20th, Nvidia's stock had an average daily trading volume of approximately $30 billion, surpassing Tesla, which had an average daily trading volume of $22 billion during the same period.
Nvidia is being aggressively pursued by funds in the secondary market, reflecting a strong interest in artificial intelligence and underlying computing chips in the market. However, the record valuations of AI related stocks, including Nvidia, are reminiscent of the boom and subsequent slump of telecom stocks during the Internet foam.
According to the Financial Times, at that time, hundreds of millions of investors around the world expected the Internet to change the world, and the demand for network hardware such as servers and related routers also increased sharply. Telecom companies and hardware suppliers became the biggest winners in the capital market. The stock prices of telecom equipment stocks such as Cisco peaked before the bursting of the Internet foam in 2000. Among them, Cisco's stock price surged more than 30 times in just a few years.
Comparison between Cisco's share price before and after the Internet foam burst and Nvidia's share price since 2020 (Financial Times)
However, the collapse of the telecommunications industry caused by the Internet speculative boom came earlier than expected - it only took four years for the entire industry to go from prosperity to depression, and it was much faster than the Internet to change human life. By 2002, two years after the Internet foam burst, the excess supply had led to the bankruptcy of more than 20 telecom groups, and the shares of other companies that did not go bankrupt had also suffered a sharp sell-off.
In the current era of artificial intelligence, computing power chips, as the underlying infrastructure of the entire industry, are essential products for all artificial intelligence software. As artificial intelligence models become increasingly powerful, the demand for high computing power chips is also increasing, and the continuous shortage of high-end computing power GPUs further increases the urgency of the entire industry.
However, there is controversy over how long the shortage of high-end computing power GPUs will last. Two years ago, due to a severe shortage of automotive chips, the global automotive industry almost came to a standstill. However, this crisis was alleviated in less than a year. Nowadays, the supply of automotive chips has not only been normalized, but many types of chips are in a state of oversupply.
The biggest risk of investing too much money in AI computing chips is the subsequent overcapacity.
Due to the prolonged downturn of the older generation of chips exceeding expectations, Samsung had to significantly reduce production last year to cope with the increasingly severe chip oversupply. In the fourth quarter of 2023, Japanese chip company Kioxia recorded a record loss of $1.7 billion. At the same time, global silicon crystal shipments in 2023 decreased by 14.3%, partly due to the cyclical downturn in the chip industry and a decrease in market demand for consumer electronics products.
Another issue is Moore's Law: high-end computing chips will soon be commercialized. Taking the old 40 nanometer chips used in household appliances as an example, these chips are no longer in short supply, but they were also scarce cutting-edge products when they were first launched in 2008. With the depreciation of capital equipment, the prices of such chips have experienced a significant decline in a short period of time.
The research and manufacturing speed of chips is accelerating every year, and software efficiency is also improving every year. The chip was upgraded from a 7-nanometer process to the advanced 5-nanometer process used in Nvidia's latest chips in just two years. This rapid technological advancement means that in the future, industry companies may spend much less on chips than they expected today.
The Financial Times reported that there is no doubt that there is a clear difference between the emergence of the Internet and the current prosperity of AI from the end of last century to the beginning of this century. For example, OpenAI's annual revenue has exceeded $2 billion, and it joined the fastest-growing technology startups in history just a few months after the launch of its generative artificial intelligence ChatGPT. Moreover, today's companies have more diversified profit methods than those more than 20 years ago. But the report also warns that the widespread application of artificial intelligence still has a long way to go, and we need to avoid repeating the overrated mistakes of the 1990s.
Nvidia's forward P/E ratio remains relatively low
The hot market for AI computing power GPUs proves that Nvidia may soon achieve a market value of $2 trillion.
About 8 months ago, shortly after Nvidia's 30th anniversary, it became the first chip company with a market value exceeding $1 trillion. Last week, thanks to the continued upward trend, Nvidia's market value surpassed that of Amazon and Google's parent company Alphabet. Currently, the company's market value is close to $1.8 trillion, making it the third largest listed company in the US stock market, second only to Microsoft and Apple.
Currently, apart from technology giants like Amazon and Google's parent company Alphabet, almost all artificial intelligence startups around the world are also competing for Nvidia's limited high-end computing power GPUs. This also explains why Nvidia, which relied mainly on selling gaming graphics cards five years ago, can now achieve such a high market value. It is reported that Amazon and Alphabet will invest billions of dollars in capital budgets annually for high-end computing GPUs.
However, FactSet's data shows that at a market value of $2 trillion, Nvidia's forward P/E ratio will be around 38 times, which is about 9% lower than the average level of the past three years. Among the 30 constituent stocks of the Philadelphia Semiconductor Index, only 7 currently have an expected P/E ratio higher than 38 times, including Nvidia's competitor AMD.
Nvidia's Forward P/E Ratio Changes Over the Past Three Years (Wall Street Journal)
"Bears will continue to debate the valuations of Nvidia, Microsoft, Palantir, MongoDB, and other pure artificial intelligence companies. Investors who have missed out on the biggest transformative technology stocks of the past decade, such as Amazon, Netflix, Meta, Apple, and Alphabet, will stubbornly adhere to the forward P/E valuation method in today's technology companies," Ives said.
Disclaimer: The content and data in this article are for reference only and do not constitute investment advice. Please verify before use. Based on this operation, the risk is borne by oneself.
