The Fate of NVIDIA

Key Points

First question: From the perspective of industrial evolution, how to understand the stage of the AI wave represented by NVIDIA?

As early as May 2023, in a heavyweight in-depth special topic released by the Guotou Strategy team, we proposed, "From a longer-term perspective, we may now be at the starting point of a new cycle of technological innovation and economic paradigm revolution. AI is not a flash in the pan and will be the primary theme of technology growth investment in the next 3-5 years."

Currently, NVIDIA in the US stock market and its counterparts have become the core of technology growth investment in A-shares and even globally. Based on evolutionary economics and Carlotapeles's "Technological Revolutions and Financial Capital," there are four stages of technological waves:

• Eruption Stage
• Frenzy Stage (the era of finance, the separation of financial capital's book value from the real value of production capital)
• Turning Point (period of reflection and adjustment of development path, liquidation and acceptance)
• Synergy Stage (the era of production, the real golden age, with production capital at the helm, book value coupled with real value)
• Mature Stage. We believe that AI is likely at the turning point of the technology introduction to deployment phase of this technological revolution, and will subsequently enter the synergy stage, which will be the true golden age of AI, the key to pricing the AI wave.

Regarding the huge controversy surrounding the current AI wave, including NVIDIA, the market invariably draws parallels with the dot-com bubble more than 20 years ago. In fact, when we look back at the spectacular dot-com bubble at the end of the 1990s to the beginning of the 21st century, we will find that the bursting of the Nasdaq bubble in 2000-2002 was based on the concentration of technology in society consuming a large amount of resources, but failing to act on the economy, leading to early signs of economic decline, which is the turning point emphasized by Carlotapeles.

In fact, after experiencing the turning point, Carlotapeles' prophecy of the technology internet industry entering the synergy stage and opening the golden age was eventually validated.

Specifically, at the turning point stage, due to the imbalance between the scale of demand and the potential of supply on one hand, and the deviation between book value and real value on the other hand, when the economic collapse comes, sometimes recession or depression will bring financial capital back to realityIt is worth noting that although the turning point experienced by the Nasdaq from 2000 to 2002 brought the profit expectations of the technology internet in the 90s back to a realistic level, the technology itself did not stop advancing during this period. Instead, it continued to explore opportunities to integrate with various production activities. As a result, in 2003, the combination of technology internet and e-commerce gave birth to Amazon, while the combination with mobile terminals gave birth to Apple. The continuous integration of technology internet triggered waves of massive production movements and consumption revolutions. After 2003, financial capital once again cooperated with production capital, and internet technology entered a collaborative stage.

Subsequently, after 2010, it gradually entered a mature stage, where the technological paradigm of technology internet dominated, permeating its logic into various activities, from business to administration to education. Higher levels of new productivity and quality became widespread throughout the economy, and society could feel the economic growth progressing at a good pace.

Second question: How do macro constraints and industry trends affect the pricing of NVIDIA?

When we realize that the current AI wave is also at a turning point, there are concerns about a repeat of the collapse of the tech bubble, which is more intuitively reflected in whether NVIDIA's stock price is analogous to Cisco in 2000.

Regarding this issue, a widely held view is based on the fact that the Federal Reserve, led by Greenspan, continued to raise interest rates in the second half of 1999, leading to the bursting of the massive bubble in the Nasdaq and the subsequent significant loss of social wealth, triggering a sharp economic downturn in the United States from 2000 to 2002. Coupled with the severe delay in rate cuts until 2001, this process was further exacerbated. Observers of this view believe that the stock market crash precedes the economic downturn, making it easy to conclude that the current pricing of the AI wave will mirror that of the tech bubble in the year. After all, this round of Nasdaq continued to rise for 2 years under the pressure of the Federal Reserve's interest rate hikes, NVIDIA's stock price is at a historical high, and the landing of applications on the terminal side seems to have not clearly exacerbated the significant risks of industry trends, giving the pricing of the AI wave a sense of being built on air.

Once the high-tech stock prices fall from their peak and trigger a Nasdaq crash, it is easy to create a situation similar to the bursting of the tech bubble in 2000However, based on our careful review and the records of Carlota Perez, the reason why the dot-com bubble burst in 2000 was more due to the macro constraints brought about by the delayed impact of technology on the U.S. economy, indicating signs of a recession. In 1999, the U.S. economy already showed signs of weakness, with the core CPI remaining below market expectations in the second half of the year, public utility capacity utilization continuously declining, and both S&P 500 index earnings and revenue growth rates declining throughout 1998-1999. In other words, the signs of an economic downturn came first, followed by the Nasdaq collapse.

Therefore, behind the bursting of the dot-com bubble, it was more about the overall risks in the U.S. stock market rather than a complete denial of industry trends.

From the perspective of "economic downturn signs come first, stock market decline comes later," it is necessary for us to analyze whether the U.S. economy will experience a "soft landing" or a "hard landing" next. If it is a "hard landing," it will have a trend-negative impact on industry trends; if it is a "soft landing," then after the macro constraints are lifted, U.S. stocks will resume their rise, and technology companies with performance support will further lead the way.

By comparing the U.S. stock market trends with the current AI wave, it is not difficult to conclude that in a round of technological wave pricing, the two core factors are macro constraints and industry trends. Regarding which of these two core pricing factors is more important, we must emphatically state that "industry trends are paramount."

We can also see that in 2010, the U.S. stock market experienced severe fluctuations due to macro constraints, but in the following 3 years, investing in Apple was unavoidable because Apple was the largest technological industry trend at that time; in 2020, due to the pandemic, the U.S. stock market experienced severe fluctuations under macro constraints, but in the following 3 years, investing in Tesla was unavoidable because Tesla was the technological industry trend at that time.

Therefore, the viewpoint that industry trends are paramount implies that often we need to first understand the long-term outlook of industry trends in order to correctly address the pricing impact of macro constraints.

At the same time, it is worth noting that besides macro constraints and industry trends, in the process of technological growth and industry evolution, the focus of pricing at different stages changes, and the investment pace follows the following four stages:

1. In the stage of explosive products, buy giants (for example, buy Microsoft when the operating system appeared in 90-95; buy Tesla when Model 3 appeared in 12-13; buy Microsoft when ChatGPT appeared in early 2023, etc.);

2. Giants start massive capital expenditure to buy infrastructure (for example, buying Cisco from 95-2000; buying charging piles and grid equipment from 18-19; buying NVIDIA in 2023);

3. The industrial chain is formed, completing the purchase of key links in the industrial chain from 0 to 1 (for example, buying Amazon and Apple after 2003; buying lithium batteries in 2020);

4. Buy the supply-demand gap in the process of 1-100 (for example, the supply-demand gap of new energy vehicles in the second half of 2020: lithium and automotive electronics)

Third question: In the next 2-3 years, will the application landing of AI be important for pricing NVIDIA? From the perspective of industry trends, referencing the stock price trend of Cisco, what is the core of NVIDIA's true pricing?

Furthermore, when we compare Cisco in the past with the current NVIDIA in the technology sector, the two indeed play similar roles in the industrial chain.

From a macro-constraint perspective, the current NVIDIA is indeed going through a turning point process similar to Cisco. The apparent difference between the two is that Cisco had a PE valuation of over 200 times at its peak, while the current NVIDIA is only valued at 60 times, indicating that NVIDIA's profit guidance is much stronger than Cisco's current fundamental expectations. Therefore, it seems that NVIDIA's current bubble pricing is much smaller than Cisco's.

However, taking a longer-term view, we believe that the fundamental difference between the two lies in the industry competitive landscape.

In fact, after 2003, although Cisco's stock price showed a rebound in profit growth, it continued to fluctuate downward. During the subsequent process where many tech giants such as Microsoft, Apple, and Amazon saw their stock prices recover or even reach new highs, Cisco was submerged in that wave of the tech internet boom. The reason behind this was that Cisco had a strong competitor as early as 1998 - Juniper, which once held a 15% market share; after 2003, Huawei fully entered the U.S. market, bringing significant challenges to Cisco.

Although, based on the perspective of industry evolution, there is indeed a transition in pricing center towards key links in the industry chain such as Amazon (chain) and Apple (chain), as well as the supply-demand gap, gradually moving away from infrastructure. However, the reflection on Cisco's pricing cannot be limited to the turning point stage of 2000-2002, but also needs to consider why Cisco fell into obscurity after 2003.**

Obviously, its lonely fate comes from the collapse of the industry's competitive landscape, falling in the fierce competition of the technology internet, which means that in the process of fierce competition in the technology wave, the industry's competitive landscape is the ultimate source of excess returns.

So, will the application landing of AI in the next 2-3 years be important for the pricing of NVIDIA? Objectively speaking, from the perspective of industry trend pricing, this is not the core. Taking a slightly longer view, the industry's competitive landscape of NVIDIA is the primary focus of its valuation support, and it is the core pricing content of the industry trend.

It can be seen that in the 1990s, most of Cisco's upward phases did not have a clear application of technology internet that triggered a large-scale production movement and consumption revolution, but the stock price continued to rise throughout the 1990s decade. The reason behind this is that Cisco's strong revenue and profit growth in most moments of the 1990s, fundamentally, this revenue and profit growth benefited more from the high-speed development of the U.S. economy in the 1990s and the industrial policy support for information technology investment.

Therefore, the impact of AI application landing on the pricing of NVIDIA in the next 2-3 years is nothing more than concerns about downstream companies reducing their investment in AI chips, which is actually more related to the performance of the U.S. macroeconomy, not an industry trend issue.

In other words: For the judgment of NVIDIA's investment value, we emphasize that based on whether the application end can land in the short term of 2-3 years, it can be a technical issue or an economic issue, but not an industry trend issue. In this process, it is only necessary to clarify that technology itself is still progressing continuously and has not stopped extensive contact with production, so the industry trend has not ended. In fact, the turning point of technology and the landing of the industry are mostly retrospective, and it is difficult to predict ahead.

The fourth question: How to predict the fate of NVIDIA as a great company? How to price it at the current stage?

Here, we have reason to believe that although NVIDIA is going through the same turning point in its fate, its stock price cannot be simply compared to Cisco after 2000. Based on this, we recommend the following decision-making logic framework, which we believe is worth considering:1. First of all, based on the industry competitive landscape as the logical first entry point, if it is believed that there are no barriers for AI chips, or if NVIDIA will face strong competitors in the next 2-3 years, then its investment value will be significantly weakened;

2. Secondly, if it is believed that the future competitive landscape of the AI chip industry will remain stable, with NVIDIA's competitive advantage in its field far ahead, then we can discuss the pricing impact of macro constraints on NVIDIA in the US economy.

• If there is a hard landing in the US economy, downstream will face financial constraints and reduce investment in AI chips, which will obviously have a negative impact on NVIDIA;

• If it is a soft landing, then waiting for the economy to stabilize and rebound will lift the macro constraints, leading NVIDIA to lead Nasdaq to resume its rise.

Before the macro constraints are lifted, that is, before confirming that the US economy is officially stabilizing and rebounding, NVIDIA's stock price is more likely to show fluctuations with bottom support (more similar to the form of Apple's stock price in 2010), but it also means that NVIDIA's best excess phase has come to an end.

Main Content

1. Deep retrospective analysis of the four pricing stages of US tech stocks: How did it come about?

Firstly, based on the historical details of the US at that time, we analyze the characteristics of the market and divide the tech stock bubble into four stages: incubation, outbreak, bubbling, and bubble burst, combining index gains and historical events to deeply review each stage.

In the first stage, the incubation stage, the core sign was the invention and popularization of web browsers. The information technology sector as a whole did not show significant excess returns, but strong individual stocks such as Microsoft and Cisco saw significant price increases.

In the second stage, the rapid outbreak stage, marked by Clinton's "Information Superhighway Plan" in September 1993, investment and growth in the information technology industry accelerated significantly. Based on the optimistic background of the US macroeconomy of "low inflation, low unemployment, high growth," the information technology industry developed rapidly with policy support. During this period, various applications flourished, and internet companies like Amazon, Yahoo, Google, and Apple rose rapidly.

In the third stage, the bubbling stage, the Southeast Asian economic crisis erupted, the US dollar strengthened relatively, and under the pricing background of "East falling, West rising," the US tech sector became a safe haven for global funds, further exacerbating the valuation bubble of tech stocks. As the tech bubble approached its peak, the information technology sector showed significant excess returns compared to other sectors, with trends diverging from other sectors.

In the fourth stage, the bubble burst stage, a series of events and market changes led to the bursting of the tech stock bubble and subsequent long-term adjustments. The Nasdaq index peaked in March 2000 ahead of fundamentals, and the "Microsoft antitrust" case judgment landed in April of the same year, marking the formal burst of the tech stock bubbleWhen the U.S. economy entered a recession period, with declining corporate profits and multiple interest rate hikes by the Federal Reserve, the stock prices of technology companies plummeted significantly. Some internet companies that went public with a large amount of financing based on previous hype but lacked fundamental support quickly declined in the tightening external environment. The bubble of overvalued technology stocks burst, leading the U.S. internet industry into a long period of adjustment.

Based on the above division, we categorize individual stocks and key events into four stages, deeply reviewing the evolution process of the U.S. stock market's tech sector pricing in four stages.

1.1. First Stage of the U.S. Stock Market Tech Bubble: 1988-1994, incubation stage, information technology sector's excess returns not significant

In the early 1990s, the U.S. technology industry completed an important transition from the electronic era to the computer era. During the incubation period of the technology stock computer era, the overall excess returns of the information technology sector were not significant, but strong individual stocks emerged within the sector. Stocks of companies like Cisco, Oracle, Andrew, and Microsoft surged, while companies like Apple, IBM, and Texas Instruments performed relatively poorly during this stage.

1.2. Second Stage of the U.S. Stock Market Tech Bubble: The U.S. information technology investment boom started in September 1993 with the "Information Superhighway Plan," and after 1994, the share of the information technology industry in GDP significantly increased

In September 1993, President Clinton officially launched the "National Information Infrastructure" project for the new century, known as the "Information Superhighway" strategy. The Clinton administration made the information industry a key driver of economic development, especially focusing on improving and popularizing the Internet. With the "Information Superhighway Plan" as a symbol, the U.S. internet industry entered a new phase of explosive growth, with the share of the information technology industry in GDP significantly increasing.

In August 1995, Netscape went public, with its stock price soaring from $25 to $75 on the first day, marking the gradual beginning of the internet industry boom.

In December 1996, amidst the continued prosperity of global stock markets, Alan Greenspan delivered a speech on "irrational exuberance," warning of stock and bond bubbles. Despite a brief dip, the three major U.S. indices continued to rise, with the Nasdaq rising by 23% for the year.

In 1997, the Asian financial crisis erupted, negatively impacting the U.S. economy, leading to a certain pullback in U.S. stocks. In contrast to the Asian economy, the three major U.S. indices still maintained gains of over 20% for the yearAfter the current crisis, the Nasdaq index rebounded first in the profit cycle stimulated by loose policies.

Against the backdrop of a strong US dollar, overseas capital hedged by purchasing US dollar assets, with technology stocks leading the way. In 1998, the Nasdaq rose by 40% for the whole year, showing significant returns compared to the Dow Jones Industrial Average and the S&P 500 Index. By the end of the year, Microsoft became the largest company in the United States by market value.

1.3. The third stage of the US stock market dot-com bubble: the Southeast Asian financial crisis + the significant development of US mutual funds as important incremental funds; after 1998, funds poured into the technology sector, and the bubble approached its peak

In 1997, the Southeast Asian economic crisis broke out, and the US dollar strengthened relatively. Against the backdrop of "East down, West up" pricing, the significant development of mutual funds at the same time as the Southeast Asian crisis and the global capital flowing back into US stocks became important incremental funds. From 1995 to 1999, the year-on-year growth rate of US mutual funds exceeded 20%, reaching a cyclical high point in 1999. US technology stocks became a "safe haven" globally at that time, exacerbating the bubble pricing of technology stocks.

To address the "Y2K bug" issue of cross-century date processing, many technology companies increased equipment investment. The fear of the "Y2K bug" became an important speculative logic in the Internet bubble and also one of the factors that later weighed on the performance of Internet companies.

In early 1999, Microsoft was found to have used its monopoly power over personal computer operating systems to hinder competition and innovation, marking the beginning of the "Microsoft antitrust case." In September of the same year, Microsoft CEO Steve Ballmer pointed out that tech companies, including Microsoft, were severely overvalued by the market. By the end of the year, Yahoo was included in the S&P 500 Index, triggering a frenzy of buying with daily gains exceeding 20%.

During the peak of the dot-com bubble, the information technology sector showed significantly higher excess returns compared to other sectors. The trend of the "new economy" represented by the Nasdaq index and the "traditional economy" represented by the Dow Jones Industrial Average diverged. By the end of 1999, tech companies such as Microsoft, Cisco, Lucent, Intel, IBM, and AOL became the top 10 companies by market value in the S&P 500 Index

1.4. The Fourth Stage of the U.S. Tech Bubble: Marked by the Microsoft Antitrust Case, coupled with the U.S. economy entering a recession, the tech bubble burst, followed by a long period of correction

In the late stage of the tech bubble, stock prices peaked ahead of fundamentals, with the "Microsoft antitrust" case marking the official burst of the tech bubble. On March 10, 2000, the Nasdaq index hit its peak during the bubble period at 5048 points before starting to decline. In April 2000, Microsoft was found to have violated antitrust laws, triggering market panic and capital outflow from the tech sector. In June 2000, Microsoft was ordered by a judge to split into two companies (this decision was overturned in June, avoiding the fate of being split).

Around 2000, U.S. corporate earnings began to decline or grow at a slower pace, making it difficult for the tech stocks' weak expectations to continue supporting the optimistic outlook of high valuations. In the first quarter of 1999, PC sales in the U.S. market remained sluggish, with expectations of declining profits for companies like Dell and IBM. In early September 2000, tech companies such as Dell and Apple, along with large companies in traditional industries, issued profit warnings. Overall, U.S. corporate earnings were on the decline.

Internet companies that went public and raised a large amount of funding through IPOs and marketing tactics during the previous investment frenzy quickly declined after the bubble burst in a tightening external environment, as their investment logic was gradually refuted. For example, Pets.com, a company selling cat litter, invested heavily in advertising but ultimately closed down in September 2000 due to the inability to turn a profit, less than a year after its official listing. As the first listed website company to go bankrupt, Pets.com was a microcosm of the bursting of the internet bubble.

Due to the overall decline in U.S. corporate earnings and multiple consecutive rate hikes by the Federal Reserve, U.S. stocks re-entered a downtrend in September 2000. After peaking in March 2000, the Nasdaq index fell by 39.3% for the year. The tech stock bubble of high valuations was burst by weak performance and tightening policies, leading to a long period of reshuffling in the internet industry.

2. What was the macro and micro environment like during the burst of the U.S. tech bubble? Signs of a recession in the second half of 1999

2.1. Macro Perspective: U.S. economy weakening, Federal Reserve policy continuously tightening

The reason for the burst of the tech bubble lies in the delayed demonstration of technology's driving force for productivity. In fact, the U.S. economy showed signs of a recession in the second half of 1999, with core CPI data consistently below expectations, particularly evident in the utilization rate of public utility capacity in the U.S. From 1995 to 2000, the proportion of equipment investment to GDP increased, reaching nearly 90%, but did not significantly boost the utilization rate of public utility capacity

In the 1990s, the US economy was in a state of "low inflation, low unemployment, high growth", with GDP growth mostly above 4%, inflation at a relatively low level showing a decreasing trend, and the unemployment rate and fiscal deficit continuously decreasing since 1992, with the government's fiscal deficit turning into a surplus from 1998 to 2000. This trend was particularly favorable for equity assets, especially in the information technology sector.

The trend of the US economy's "low inflation, low unemployment, high growth" gradually peaked around 1999 and 2000. After 2000, the characteristics of economic recession in the US became more apparent, with the peak of US stock technology profit growth rate occurring in 1999Q3. Fundamental data such as S&P 500 revenue, ROE, and earnings all weakened after 1998. Therefore, the bursting of the dot-com bubble was more based on the overall risks of US stocks rather than a complete denial of industry trends.

Revisiting the policy of the Federal Reserve at that time, Greenspan's response to the 1997-1998 financial crisis was textbook-worthy. However, the interest rate hike cycle restarted in the second half of 1999 failed to promptly contain the dot-com bubble. Due to CPI being lower than expected in the second half of 1999, market expectations for a Fed rate hike weakened, leading to a continued frenzy in the US stock market.

In August 1999, Greenspan pointed out that interest rate policy should pay attention to stock and other asset prices. Before the Nasdaq index reached its peak, the Fed had raised interest rates four times, with the benchmark interest rate rising from 4.75% to 5.75%However, the market in the bubble period did not react much to the first three rate hikes. After the fourth rate hike, the tightening of liquidity had a significant downward impact on the Nasdaq index. After the bursting of the internet bubble in 2000, due to the failure to cut interest rates in time, the economic downturn in the United States worsened further post the tech bubble.

2.2. Micro Clues: During the tech bubble period, the profit growth of tech stocks came from a large number of mergers and acquisitions

From 1998 to 2000, the amount of IT issuance accounted for over 25% of the overall US stock market, reaching over 50% in 2000, exacerbating the bubble pricing at its peak. Against this backdrop, a large amount of merger and acquisition financing entered the internet industry. As a result, internet industry valuations further increased, leading to a significant increase in financing scale and IPO numbers, exacerbating the capital bubble of tech stocks lacking sufficient fundamental support post the tech bubble.

Objectively speaking, it was the signs of an impending recession in the US economy that led to the bursting of the tech bubble, rather than the bursting of the tech bubble causing a recession in the US economy. Other macro and micro explanations for the bursting of the bubble: 1. Measures against Microsoft's antitrust actions; 2. Greenspan's rate hikes leading to the Nasdaq crash, we tend to view these as triggers rather than the root cause. In fact, the true reasons for the collapse of the tech bubble in 2000 still remain highly debated in the academic community.

From the perspective of "economic recession signs precede stock market declines," it is necessary to analyze whether the US economy will experience a "hard landing" or a "soft landing" in the future: if it is a "hard landing," it will have a trend negative impact on industry trends; if it is a "soft landing," it is expected to resume an upward trend after macro constraints are lifted.

3. From a micro-enterprise perspective, what are the characteristics after entering the bubble period? Valuations overextend performance growth for more than three years

3.1. In-depth review of representative stocks during the tech bubble period

Before Q3 of 1999, the pricing of core tech leaders in the US stock market could be supported by fundamentals, but thereafter, a serious deviation between fundamentals and stock prices began to emerge. The market's enthusiasm for investing in internet companies surged, and tech stock valuations gradually detached from the fundamentals of companies, with PE growth rates exceeding changes in earnings per share.

The market's blind optimism about the "capital story" and profit expectations of internet companies led to a large number of companies without fundamental support successfully going public and being sought after. **From the perspective of fundamental pricing anchors, profit growth rate sensitivity is highest for single-quarter pricing, while revenue growth corresponds to pricing sustainability. During the bubble pricing stage, the market often shifts from PE valuation to PS valuation

After the burst of the bubble in 2000, Cisco, despite some performance recovery, did not achieve significant excess pricing. In the subsequent process where many tech giants such as Microsoft, Apple, and Amazon saw their stock prices recover and even reach new highs, Cisco was submerged in that wave of the tech internet, with the core reason being the deterioration of the router competition landscape.

In fact, Cisco had a strong competitor, Jusiper, as early as 1998, which once held a 15% market share. After 2003, Huawei entered the US market comprehensively, causing a significant impact on Cisco.

The reflection on Cisco's pricing should not be limited to the turning point from 2000 to 2002, but should also look at why Cisco declined after 2003. Obviously, its decline was due to the deterioration of the industry competition landscape, losing ground in the fierce competition of the tech internet. In the tech wave, the industry competition landscape is the ultimate source of excess returns.

In contrast, Microsoft, as an operating system, still possesses the scarcity barrier and can continue to be priced upwards after the fluctuations in 2000. From the perspective of industry trends, the consideration of the core valuation of hardware pricing as infrastructure lies in whether the subsequent industry competition landscape deteriorates.

3.2. Analysis of Overdraft Situation of A-share Prosperous Companies

In the A-share market, if the valuation years of prosperous companies exceed three years, caution is needed. If it exceeds five years, it is often overvalued. For example, CATL reached an overdraft of nearly three years during the peak period in 2020-2021, and currently has fallen back to within one year; companies like WuXi AppTec, Mindray Medical, and Huichuan Technology all overdrafted for more than three years at their stock price peaks.

It is difficult to judge the peak point of the bursting of the bubble of representative companies in the technology sector based on logic, it is more based on trading formation. From the perspective of individual stock pricing, it may not make sense logically but the stock price breaks first. At the peak position, the stock price PE ratio overdrafts the performance growth rate for more than 3 years, with Cisco reaching 8 years. From the historical pricing experience of A-shares, overdrafting the performance growth rate for 3 years is an important threshold for industry trend stocks, and 5 years is the maximum level.

3.3. Analysis of NVIDIA's Overdraft Years of Valuation

Looking at the overdraft years of NVIDIA's stock price, the current number of overdraft years is less than three, and the degree of overdraft is less than representative stocks like Cisco and Microsoft during the peak period of the technology bubble. Therefore, we believe that NVIDIA's stock price is still some distance away from the bubble peak.

4. Observations on the Mid-term Industrial Theory of the Bubble Pricing in the U.S. Stock Market: Experiencing a Turning Point Pricing under Macro Constraints

4.1. Mid-term Perspective: Based on the analysis of technological revolution and capital theory, the current AI is gradually entering a turning point pricing after the frenzy

Looking back at the spectacular dot-com bubble at the end of the 1990s to the beginning of the 21st century, it can be seen that the Nasdaq bubble from 2000 to 2002 was based on the aggregation of technology attracting a large amount of social resources, but it failed to impact the economy, leading to the economy showing signs of decline before collapsing. This is what Carlotapeles emphasized: the turning point.

The imbalance before the turning point lies between the scale of demand and the potential of supply, while the imbalance after the turning point is the deviation between the nominal value and the real value. The arrival of economic collapse, sometimes in the form of recession or depression, will bring financial capital back to reality.

From the perspective of industrial theory, comparing the history of the dot-com bubble and the current situation of AI, we can make judgments and deductions about the trends of different industrial stages in the technology sector.

With the slowing growth of AI investments, concerns about the resurgence of the dot-com bubble have emerged. Comparing the similar scenes of the decline in U.S. capacity utilization, the slowdown in economic growth, and the financial capital's obsession with technology clustering from 1999 to early 2000, the current pricing of AI is also at a turning point after the frenzy.

The reason for the formation of the turning point is that AI has not led to a large-scale productivity explosion, and the economy has entered a depression, thereby forming macro constraints on pricing. The dual constraints of industry and macro have compressed the upside of the technology sector. Analogous to the Internet entering the collaborative stage after 2003, the key to the recovery of technology lies in when AI can be widely integrated with production activities.

The essence of thematic investment in technology growth industries is to grasp the "technology-economy" paradigm, and evolutionary economics is an important theoretical basis. Referring to Carlotapeles' book "Technological Revolution and Financial Capital," in this round of AI investment frenzy, we need to address two core issues: 1. Whether to increase positions: how capital markets price technology; 2. Who is the next Google: which companies will be eliminated by technology, and which companies will emerge victorious?

According to Carlotapeles' theory, we can divide the period of the U.S. dot-com bubble into two periods (introductory period and development period, with turning points marking the transition of the period) and four stages, reviewing and analyzing from the perspective of industrial theory. In each period, the roles and interactions between industrial capital and financial capital are different:

1. Introductory period, including the outbreak period and the frenzy period. (1) Outbreak period (1990-1993): The era of technology, financial capital begins to fall in love with production capital. Correspondingly, within the U.S. technology sector, strong individual stocks emerged, and financial capital, including venture capital, became an important support for many start-up technology companies during this period.（2）Euphoria Period (1994-2000): The era of finance, where financial capital's book value is separated from the real value of production capital. The peak of the dot-com bubble saw a valuation frenzy in the technology sector in the capital market, with a significant divergence from actual corporate profits and macro trends.

2. Turning Point (2000-2002): A period of reflection and adjustment, a time of liquidation and acceptance. The imbalance before the turning point lies between the scale of demand and the potential supply, while the imbalance after the turning point is the deviation between face value and real value. The onset of economic collapse, sometimes in the form of recession or depression, brings financial capital back to reality.

During this period, the U.S. economy continued to decline, and the Microsoft antitrust case was settled. The technology stock bubble, under the dual constraints of macro and micro factors, quickly burst after reaching its peak, with once fervent profit expectations falling back to a reasonable range.

3. Development Period: Includes the synergy period and the mature period. (1) Synergy Period (2003-2010): The era of production, a true golden age where production capital takes the lead and the book value is coupled with real value. During this period, financial capital provided strong support to leading companies within the sector, with Amazon establishing an e-commerce ecosystem and Apple Inc. leading the technology sector by popularizing Apple computers.

(2) Mature Period (after 2010): An era of questioning and complacency, where financial capital seeks new technologies, regions, and sectors for investment. Correspondingly, during the era of mobile internet, technology stocks generally benefited, with overall stock prices trending higher. The technological paradigm of tech internet dominates, permeating its logic into various activities, from business to administration to education, leading to higher levels of new productivity and quality becoming widespread throughout the economy, allowing society to feel economic growth progressing at a good pace.

From the perspective of evolutionary economics, we believe that AI is currently at the turning point of the technology revolution from the stage of technology introduction to deployment, and will subsequently enter the synergy stage, which will be the true golden age of AI.

In line with the theory of Carlota Perez, we propose the four stages of technology pricing in the tech cycle and analyze the core pricing and allocation strategies:

1. Giant Investment Stage: Buying giants when hits appear. For example, in 1990-1995, the emergence of the Microsoft operating system led to buying Microsoft; in 2012-2013, the appearance of Model 3 led to buying Tesla; in early 2023, the emergence of ChatGPT led to buying Microsoft.

2. Infrastructure Investment Stage: Giants start a period of massive capital expenditure on infrastructure, benefiting companies primarily engaged in infrastructure. For example, from 1994 to 2000, Cisco, a provider of hardware equipment, led the way with its stock price rising more than thirty times in six years.

3. Formation of the Industry Chain, Speculation on Key Links in the Industry Chain: The pricing focus shifts, with the most significant benefits accruing to key links in the industry chain. For example, after the dot-com bubble burst, Amazon successfully turned losses into profits with the establishment of an e-commerce ecosystem, while Apple achieved the popularization of Apple computers, resulting in better performance than Microsoft and Cisco, becoming an unavoidable link in U.S. stock investments4. 1-100 Supply and Demand Gap in Trading: The pricing of infrastructure hardware reaches its peak when giants start massive capital expenditures. As the industry evolves to the 1-100 stage, the pricing focus transitions to the application end supply and demand gap. At this point, it is crucial to assess the industry competitive landscape risks in the hardware sector (such as Cisco in the US stock market after 2000, and charging piles in the A-share market after 2020).

4.1.1. Introduction Period - Four Stages of Expansion Period: Outbreak Stage - Frenzy Stage - Coordination Stage - Maturity Stage

According to the interaction between production capital and financial capital, the transition period from the introduction period to the expansion period can be divided into four stages:

Outbreak Stage (era of technology, financial capital starts dating production capital) - Frenzy Stage (era of finance, separation of financial capital's book value and production capital's real value) - Turning Point (period of reflection and adjustment of development path, liquidation and acceptance) - Coordination Stage (era of production, the real golden age, production capital takes the helm, book value and real value are coupled) - Maturity Stage (era of questioning and complacency, financial capital seeks new technologies, regions, and sectors for investment).

4.1.2. Outbreak Stage: The beginning of technological revolution, the romance between financial capital and technological revolution

In the outbreak stage, the wave of technological revolution begins. During this stage, the two paradigms of old and new coexist for a long time, core countries start experiencing real economic and social dilemmas, while some catching-up countries are reaching their peak of glory. The hallmark of this stage is the emergence of more cost-effective new products that attract consumers and competitive entrepreneurs. They gradually combine new ideas and successful actions to make it the best practice, which is the new technological-economic paradigmThe financial and banking industry quickly modernized its operations. Each development and dissemination of technological revolutions tends to stimulate innovation in the financial sector, benefiting from the momentum generated by financial innovation. For example, during the steam and railway era, very few individual enterprises could afford the massive investment required for railway development. However, the innovation of joint-stock companies brought an effective way to centralize capital and diversify risks.

4.1.3. Frenzy Stage: Self-satisfied financial capital dominates the casino; social differentiation and wealth gap widen, eventually making the bubble unsustainable; infrastructure investment expands to a sufficient scope, completing phase tasks

The frenzy stage is a tumultuous period where financial capital takes off on its own. The explosive growth in productivity affects more and more activities, triggering a restructuring process in the production sector. With the acquisition of new infrastructure, the restructuring process is further strengthened. These infrastructures have covered a sufficiently large scope and exhibit clear externalities.

Everyone benefiting from this opportunity believes that the world is heading towards an incredible stage. The success of the successful has generated enormous wealth, which is concentrated in a few economic entities. These entities, in turn, aim to continue increasing this wealth at the speed at which they created it. As the financiers' focus shifts to the task of making money from money, the relationship between the inflated monetary economy and the rebuilt real economy becomes increasingly tense, with financial capital moving further away from its role as a supporter of creating real wealth.

For those pursuing wealth accumulation, the higher profits in the financial sector make them reluctant to engage in production activities unless they are in areas related to the most active technologies that can generate high profits and attract more capital inflows.

Frenzied development will polarize society, increase the wealth gap, making it difficult to be accepted at the societal level. Due to the imbalance in the growth of both sides, the economy becomes unsustainable. One is the imbalance between the demand side and the potential supply side. An investment-intensive economy concentrates income in the upper echelons, becoming an obstacle to the growth of any specific product production and ensuring the overall economic scale for all.

The other is the mismatch between book value and actual value. Therefore, this structurally unstable system cannot guarantee sustained growth. Economic collapse is followed by economic recession, sometimes economic depression, bringing financial capital back to reality. This process, due to social pressure, leads to institutional restructuring.

In this tense atmosphere, many social innovations will emerge, gradually taking shape in the initiation stage, possibly bringing new regulatory measures in the financial sector and creating an environment conducive to growth. This restructuring is crucial, usually occurring at the turning point after the initiation stage, followed by institutions and society choosing to follow suit, leading to a paradigm shift into a golden age

4.1.4. Turning Point: Reflecting on and Adjusting Development Paths

The decline after the frenzy and the collapse that triggers the decline are consequences of an unsustainable structural arrangement. There are three structural tensions that make it impossible for the frenzy process to continue indefinitely, which exist between real wealth and book wealth; the existing demand scale and potential supply scale of core products in technological revolutions; and the excluded groups in society and those who have reaped the benefits of the bubble.

The turning point represents a necessary fundamental change, transitioning the economy from a frenzy shaped by financial standards to a collaborative approach, relying on gradually increasing production capacity.

The turning point is neither a specific event nor a specific stage. It is a process of change that occurs in a specific environment and can last from several months to several years. For decisions regarding social systems, it is a crucial crossroads.

Profit expectations that become confused and unrealistic during the frenzy phase must be brought back to the norm. This means establishing full control over financial capital and setting up a system framework that favors the real economy over the book economy. Only when the collapse evaporates a large amount of earned profits and when the decline indicates that the casino cannot be revived, can financial capital accept this control.

4.1.5. Collaborative Stage: Re-coupling of Production Capital with Financial Capital

The collaborative stage is the first half of the paradigm deployment period, which can be called the true "golden age." Its form is manifested in: the basic externalities that contribute to technological revolutions, especially infrastructure, have already been formed in the frenzy phase, as have the foundational investments made in industrial sectors that serve as growth engines.

The conditions required for vibrant expansion and economies of scale are in place. Under the appropriate framework, growth will tend to stabilize, and society can feel economic growth progressing at a good pace.

The new paradigm now dominates, with its logic permeating all activities, from business to administration to education. Higher levels of new productivity and quality are becoming widespread throughout the economy. People view technology as a positive force, and the same goes for finance, as it now truly supports production capital.

New rules are established in the financial sector, typically including a new framework for banking operations and monetary activities. People have also established game rules that restrict business, employment relationships, and other aspects, as well as institutional innovations at the international levelThis is a period of orderliness and disciplined action, as well as a period of continuous investment in production capacity, increasing employment, and expanding markets. Production is the keyword of this stage, and actual growth in production becomes the basic source of wealth.

4.1.6. Mature Stage: Old industries monopolize, new technologies begin to emerge

In the mature stage, many original core industries in the mainstream paradigm show signs of depletion, with only a few new industries maintaining high growth rates. The latest technological systems and products have short lifecycles, as accumulated experience leads to rapid learning and demand saturation curves, gradually revealing the limitations of the paradigm.

Financial capital joins hands with enterprises to try to boost their threatened profits. Because in the previous synergistic stage, financial capital and production capital established profitable cooperative habits, years of successful experience continue to support the strategic direction of production capital.

Strengthening control over the market is one of the early solutions that most dominant enterprises find when facing difficulties, achieving this goal through various means such as mergers, acquisitions, and squeezing out smaller competitors from the market. The driving force behind monopolization is a market reaction to the beginning of growth contraction.

Another trend is that enterprises in mature industries try to expand outward from existing investment areas (both in terms of sectors and geography). Financial capital begins to support investments in peripheral areas, selling to new customers in distant places, and shifting production to cheaper locations. Early overseas investment opportunities accompany the driving force of market expansion towards peripheral regions by mature industries.

4.2. Introduction of the Four Stages of Technology Pricing: Buying Giants with Explosive Products - Giants Start Massive Capital Expenditure to Buy Infrastructure - Industry Chain Formation to Buy Key Links in the Industry Chain - Buying Supply and Demand Gaps from 1 to 100

In accordance with the theory of Carlota Perez, we propose the four stages of technology pricing from the perspective of a complete technological cycle, using important events from different technological periods as examples, and analyze the core pricing and allocation strategies respectively:

1. Giant Investment Stage: Buying Giants with Explosive Products. For example, from 1990 to 1995, the emergence of the Microsoft operating system, buy Microsoft; from 2012 to 2013, the appearance of the Model 3, buy Tesla; at the beginning of 2023, the emergence of ChatGPT, buy Microsoft.

2. Infrastructure Investment Stage: Giants start a period of massive capital expenditure to purchase infrastructure, benefiting companies primarily engaged in infrastructure. For example, from 1994 to 2000, Cisco, a provider of hardware equipment, led the way with its stock price rising more than thirty times in six years; from 2018 to 2019, buy companies related to charging piles and electric grid equipment;In 2023, buy NVIDIA, a chip provider.

3. Formation of the industrial chain, speculation on key links of the industrial chain: The shift in pricing focus occurs, with the most significant benefits coming from key links in the industrial chain. For example, after the dot-com bubble burst, Amazon successfully turned losses into profits with the establishment of an e-commerce ecosystem, while Apple popularized Apple computers, resulting in better growth than Microsoft and Cisco, becoming an unavoidable link in the U.S. stock market investment.

4. Speculation on the 1-100 supply-demand gap: The pricing of hardware infrastructure reaches its peak when giants initiate massive capital expenditures. As the industry evolves to the 1-100 stage, the pricing focus transitions to the supply-demand gap at the application end. At this point, it is important to focus on evaluating industry competition risks in the hardware field (such as Cisco in the U.S. stock market after 2000, and A-share charging piles after 2020).

5. Risk assessment of NVIDIA's pricing bubble from multiple perspectives: More similar to Apple's fluctuations in 2010 than Cisco in 2000

Taking a longer view, the fundamental difference between Cisco and NVIDIA lies in the industry competition landscape within the industry trend. In fact, after 2003, despite the rebound in profit growth, Cisco's stock price continued to fluctuate downward. During the subsequent process where many tech giants such as Microsoft, Apple, and Amazon saw their stock prices recover and even reach new highs, Cisco was submerged in the tide of the tech internet, due to the gradually deteriorating competitive landscape of Cisco.

As early as 1998, Cisco faced a strong competitor - Juniper, which once held a 15% market share. After 2003, Huawei entered the U.S. market comprehensively, posing a significant challenge to Cisco.

While there is indeed a transition of pricing focus towards key links in the industry chain such as Amazon and Apple, and a shift towards supply-demand gaps away from infrastructure based on industry evolution, reflecting on Cisco's pricing requires a deeper look at why Cisco declined after 2003. Clearly, this was due to the worsening competitive landscape, losing ground in the fierce competition of the tech internet, indicating that the industry competition landscape is the ultimate source of excess returns in the tech wave.

Although NVIDIA is going through a similar turning point in its fate, its stock price cannot be simply compared to Cisco after 2000. When evaluating the investment value of NVIDIA, it is emphasized that technology itself is still advancing and has not stopped widespread contact with production. Taking a slightly longer view, the industry competition landscape of NVIDIA is the primary lever supporting its valuation.

Based on this, we recommend the following decision-making logic framework:

Firstly, based on the industry competition landscape as the logical first entry point, if it is believed that AI chips have no barriers, or that NVIDIA will face strong competitors in the next 2-3 years, then similar to Cisco's fate in 2000, NVIDIA's investment value will be significantly weakened.**

Secondly, if it is believed that the future competitive landscape of the AI chip industry will remain stable, NVIDIA will maintain a stable competitive position. Then, we can discuss the pricing impact of macro constraints on NVIDIA in the context of the U.S. economy.

If the U.S. economy experiences a hard landing in the future, downstream sectors will face financial constraints and reduce investment in AI chips, which will obviously have a negative impact on NVIDIA, similar to the decline in Microsoft's stock price in 2000. If it is a soft landing, waiting for the economy to stabilize and recover to lift the macro constraints, it is expected that NVIDIA will lead Nasdaq to resume its upward trend.

A soft landing can be further divided into two scenarios for discussion: In the first scenario, AI technology drives a consumption revolution, rapidly advancing the overall social productivity, completing the upgrade process of the industry chain from 0 to 1. The investment focus will shift to key links in the industry chain, so we should look for companies in the ecological application end, such as Amazon and Apple in the past.

In the second scenario, AI technology is still in the infrastructure investment stage, and there is no clear landing and outbreak in the application end. In this case, we believe that NVIDIA will continue to lead Nasdaq.

The current risk of bubble burst is more based on the macro constraints brought about by the signs of U.S. economic recession (the so-called turning point by Carlotapeles), rather than a denial of industry trends.

From the perspective of macro constraints, unless the U.S. economy enters a "hard landing" in the future, it is difficult to form a trend-negative impact on Nasdaq. From the perspective of industry trends, we believe that there was a severe volatility in the U.S. stock market in 2010, but in the following 3 years, the capital market could not avoid Apple; similarly, there was a severe volatility in the U.S. stock market in 2020, but in the following 3 years, Tesla could not be avoided.

Similarly, we believe that although there may be short-term fluctuations due to macro constraints, NVIDIA will still be the core investment in the U.S. technology sector for the next 3 years. Before the macro constraints are lifted, that is, before confirming that the U.S. economy officially stabilizes and recovers, NVIDIA's stock price is more likely to show fluctuations with bottom support in the short term (more similar to the form of Apple's stock price in 2010), but it also means that NVIDIA's best excess phase has come to an end.

By comparing the U.S. stock market trends with the current AI wave, it is not difficult to conclude that in a round of technology wave pricing process, the two core factors are macro constraints and industry trends. Which of these two core pricing factors is more important, we must emphatically state that "industry trends are paramount."

In fact, after the technology market experienced a turning point, Carlotapeles' prophecy in 2002 that the technology internet industry trend was about to enter a collaborative stage and usher in a golden age was verifiedSimilarly, we can also see that: In 2010, the US stock market experienced severe fluctuations due to macro constraints, but in the following 3 years, US stock investments could not avoid Apple, as Apple was the largest tech industry trend at the time; In 2020, due to the epidemic, the US stock market experienced severe fluctuations under macro constraints, but in the following 3 years, Tesla could not be avoided, as Tesla was the tech industry trend at the time.

Based on Bridgewater's research evaluation in the first half of the year, compared to Cisco, the current valuation around NVIDIA is not considered high, and the overall pricing bubble risk in Nasdaq is relatively small. Cisco's static P/E ratio reached around 200 times in early 2000, even considering the forward P/E ratio based on performance growth two years later, it exceeded 100 times. While NVIDIA's current P/E ratio is 73 times, the forward P/E ratio in two years is still less than 30 times.

Currently, Wall Street is confident in NVIDIA's performance realization in the next year. From an industry theory perspective, the biggest risk of valuation trend decline comes from the deterioration of the AI chip competitive landscape. We believe that NVIDIA's pricing form will be more similar to Apple in 2010 with strong performance support during the volatility of the US stock market, rather than Cisco after a long period of stock price decline since April 2000.

Therefore, the short-term rebound of NVIDIA's stock price needs to wait for the release of macro constraints, and the medium-term upward movement requires its deep involvement in the process of Industry 1-100.

5.1. Looking at the next three years, NVIDIA's performance forecast is very optimistic

Looking at the next three years, the current performance forecast for NVIDIA remains very optimistic, with the year-on-year growth rate expected to remain at a high level. Over the past five years, NVIDIA's revenue and EPS have maintained positive growth in most periods, with the year-on-year growth rate reaching a remarkable peak in 2023.

5.2. How to identify the relative market value of tech stock bubbles: The current expansion of the Nasdaq index relative to the S&P 500 is lower than the tech bubble stage, and has not exceeded the 2021 high point

From the perspective of market value ratio, the degree of tech stock bubble at the current stage is not significant. Before 1999, the market value ratio of the Nasdaq index and the S&P 500 index remained around 30%, and at the peak of the tech bubble, this ratio rapidly expanded to 60%, doublingSince 2010, this ratio has risen to around 40%, and currently it has risen to around 65%, lower than the peak in 2021.

5.3. How to identify the differentiation of tech stock bubble through yield: There is no significant differentiation in the yield between the Nasdaq Index and the S&P 500

During most periods, the difference in the 120-day rolling yield between the Nasdaq and the S&P 500 does not exceed 20%. However, during the tech bubble period, this difference once exceeded 60%, indicating a significant differentiation of tech stocks compared to other US stocks. This phenomenon is currently difficult to observe in the market.

5.4. How to identify the measurement index of tech stock bubble: Bridgewater Fund's Six-Dimensional Stock Market Bubble Index

Bridgewater Fund's Dalio defines a bubble market in six dimensions and constructs corresponding indicators to describe the degree of market bubble, ultimately integrated into an Equity Market Bubble Gauge. According to this index, the overall degree of market bubble in the US stock market is at the median level (52%).

5.5. Overall observation of stock market bubble and evaluation of individual stock profit valuation support: The US stock market currently shows signs of bubble, but it is not completely bubbled; The price of NVIDIA compared to Cisco has completely different profit valuation matches

Specifically, the performance of the six dimensions measuring the overall bubble of the US stock market is as follows, with the US tech stock M7 more inclined towards a bubble compared to the overall stock market: 1. Stock price percentile (currently 73%); 2. Profit growth rate required to achieve current stock yield (currently 73%); 3. Investor market entry heat (currently 55%); 4. Investor optimism (currently 58%); 5. Scale of leveraged financing (currently 23%); 6. Long-term investments such as corporate capital expenditures (currently 38%).

Further comparing Cisco during the tech bubble phase and NVIDIA currently, behind similar stock price trends are completely different profit matches. Cisco's static P/E ratio reached 180 times in early 2000, even considering the forward P/E ratio based on performance growth two years later exceeded 100 times. While NVIDIA's current P/E ratio is 73 times, the forward P/E ratio two years later is less than 30 times, indicating that NVIDIA's current high valuation is built on strong performance growth and profit capabilities.

Author: Lin Rongxiong (S1450520010001), Peng Jingtao, Source: Lin Rongxiong Strategy Salon, Original Title: "The Fate of NVIDIA (Text Version)"