Is the US stock market currently in an "extreme" bubble?

Report Summary

● The reversal of the carry trade overseas in early August was a more extreme "stress test", but the core contradiction lies in the "micro-structural deterioration of the global Popular Trade represented by Japanese and American stocks"! Market participants are using language to tell you that "the micro-structure of certain global markets is currently quite fragile". In this article, we will focus on the key issues in the U.S. stock market.

● When the core asset bubbles of the "Fabulous 50" and the "Tech Bubble" burst, the micro-structure deteriorated significantly. Arbitrageurs engaging in convergent trading of "core assets" lead to the "deterioration of the micro-structure of the market" -> when the belief in "core assets" begins to waver and the relative profit advantage is difficult to maintain -> it often punctures the fragile micro-structure of the market -> position stampedes trigger severe liquidity shocks forming a negative feedback loop -> leading to the bursting of the "core asset bubble". The concentration of this round of U.S. stocks has reached historical highs, and whether the performance of the "tech leaders" can renew the "AI faith" is crucial to maintaining the high concentration of U.S. stocks.

● From a micro-structural perspective, this round is not the "ultimate bubble" of the dot-com era in 2000. Comparing this round with the dot-com bubble of 2000 using several key indicators: although most indicators have some similarities, this round can be relatively optimistic. (1) The degree of increase in the index level in this round is far less than that of the "dot-com bubble"; (2) The differentiation between tech and non-tech valuations in this round is far less than that of the "dot-com bubble", to some extent confirming that the differentiation between tech and non-tech stocks in this round is being driven by the profit advantage of tech stocks over non-tech stocks; (3) In terms of investor sentiment, the current market frenzy is significantly lower than that of the dot-com bubble period.

● From a meso-level perspective: behind the bursting of the dot-com bubble in 2000 was a significant slowdown in industry trends. Based on current data, the matching of stock prices and performance of AI in the U.S. stock market in this round is relatively good, but attention should be paid to the sustainability of AI performance. (1) The performance of internet companies in the late 1990s deteriorated, indicating a significant downturn in industry trends. The penetration rate of the internet industry declined significantly in the early 2000s, with the gradual shrinking of incremental market space indicating that high demand growth rates are unsustainable. Additionally, the overcapacity issue brought about by continuous investment expansion in the 1990s became evident, leading to a decline in capacity utilization, and both the primary and secondary markets of the internet industry encountered a cold winter. (2) Compared to the dot-com bubble around 2000, based on current data, the matching of stock prices and performance of AI in the U.S. stock market in this round is relatively good, but attention should be paid to the sustainability of AI performance. In 1999, the bubble in U.S. technology stocks accelerated, manifested by the extreme divergence in the "stock prices & performance" ratio between tech and non-tech stocks, laying the groundwork for the bubble burst.

Introduction: Comparing the dot-com bubble of 2000, how to gauge the risks of AI in this round?

The reversal of the carry trade overseas in early August was a more extreme "stress test", but the core contradiction lies in the "micro-structural deterioration of the global Popular Trade represented by Japanese and American stocks"! Mr. Market uses language to tell you that "the microstructure of global partial markets is currently very fragile". In this article, we will focus on the key issues of the US stock market:

Looking back at the history of the US stock market, during the bursting of the core asset bubbles of the "Nifty Fifty" and the "Dot-com Bubble", the microstructure also deteriorated significantly.

Arbitrageurs' convergence trading on "core assets" leads to the "deterioration of the microstructure" of the market -> when the belief in "core assets" begins to shake and the relative profit advantage is difficult to maintain -> it often punctures the fragile microstructure of the market -> position stampede triggers severe liquidity shocks forming a negative feedback loop -> leading to the bursting of the "core asset bubble".

The concentration of this round of US stocks has reached a historical high. The performance of the "tech leaders" in the future and whether they can renew the "AI faith" are key to maintaining the high concentration of US stocks.

From the perspective of the microstructure of the US stock market, we compare several indicators with the situation of the 2000 Dot-com Bubble: although most indicators have certain similarities, this round may not be an "extreme bubble", so we can be relatively optimistic.

(1) The degree of increase in the index level in this round is far less than that of the "Dot-com Bubble";

(2) The differentiation between tech and non-tech valuations in this round is far less than that of the "Dot-com Bubble", to some extent confirming that the differentiation between tech and non-tech stocks in this round is mainly driven by the profit advantage of tech stocks over non-tech stocks;

(3) In terms of investor sentiment, the current market enthusiasm is significantly lower than that during the Dot-com Bubble.

From a medium-term perspective: behind the bursting of the Dot-com Bubble in 2000, there was a significant slowdown in industry trends. From current data, the matching degree between the stock price and performance of AI stocks in the US stock market is relatively good, but attention should be paid to the sustainability of AI performance in the future. (1) At the end of the 1990s, the performance of internet companies in the US deteriorated, indicating a significant downturn in industry trends. The penetration rate of the US internet industry declined sharply in the early 2000s, indicating that the high-speed demand growth was unsustainable, and the overcapacity brought about by continuous investment expansion in the 1990s became apparent, leading to a decline in capacity utilization, and both the primary and secondary markets of the internet industry encountered a cold winter. (2) From current data, the matching degree between the stock price and performance of the AI sector in this round is relatively good. In 1999, the bubble in US technology stocks accelerated, manifested by the extreme divergence in the "stock price & performance" ratio between tech and non-tech stocks, laying the groundwork for the bursting of the bubble.

The microstructure deteriorated significantly during the bursting of the core asset bubbles twice

During the bursting of the core asset bubbles of the "Nifty Fifty" and the "Dot-com Bubble", the microstructure of the US stock market deteriorated significantly, and the fragility of the market was at a temporary high.

Arbitrageurs' convergence trading on "core assets" leads to the "deterioration of the microstructure" of the market -> when the belief in "core assets" begins to shake and the relative profit advantage is difficult to maintain -> it often punctures the fragile microstructure of the market -> position stampede triggers severe liquidity shocks forming a negative feedback loop -> leading to the bursting of the "core asset bubble". ** The concentration of this round of US stocks has reached a historical high. Whether the performance of "technology leaders" can renew the "AI faith" will be the key to maintaining the high concentration of US stocks. According to Wind data, as of the end of June 2024, the concentration of US stocks (Market Concentration = Market cap of the largest stock (TOP 5% Average) relative to the 75th percentile stock) hit a historical high, with a slight decline in July and August, still remaining at historical highs. Combining the framework analysis deduced from the historical two rounds of "core asset" market trends in US stocks, if the market's optimistic expectations for the relative profit advantage of AI technology stocks can be maintained, the trend of increasing concentration of US stocks may continue, otherwise, it may reverse.

Microstructure: This round is not the "ultimate bubble" of the 2000 dot-com era

Analyzing from the microstructure level of the US stock market, we compare several indicators to compare this round with the 2000 dot-com bubble: although most indicators have certain similarities, this round may not be the "ultimate bubble," and we can be relatively optimistic.

1. The degree of index-level increase in this round is far from the "dot-com bubble"

From the index perspective, as of August 28, 2024, the 1-year/2-year/5-year rolling CAGR of S&P 500 Information Technology and Nasdaq 100 are higher than the average level, but significantly lower than the levels during the "dot-com bubble" period.

2. The differentiation of valuations between technology and non-technology sectors in this round is far from the "dot-com bubble," and the market relies more on profit drivers

Looking at the ratio of S&P 500 Information Technology PE to S&P 500 PE, the differentiation of valuations between US technology and non-technology sectors in this round is far from the "dot-com bubble," to some extent, confirming that the differentiation between US technology and non-technology stocks in this round is more driven by the profit advantage of technology stocks over non-technology stocks.

In addition, the valuation levels of current US technology giants are still in a relatively reasonable range. As of August 28, 2024, except for Apple, the PE ratios of leading US technology stocks are generally lower than the +1STD level of the average of the past 5 years

3. From the perspective of investor sentiment, the current market frenzy is significantly lower than that of the dot-com bubble period

The Bullish indicator of the American Association of Individual Investors can to some extent reflect the enthusiasm of investors. In mid-July 2024, the Bullish indicator reached a cyclical high of 53%, higher than the historical average (38%), but significantly lower than the level during the dot-com bubble period. This to some extent reflects that investors during the internet bubble period were more enthusiastic.

Medium-term perspective: The matching degree of stock prices and performance of US AI companies is relatively good

(1) Behind the bursting of the dot-com bubble, there is a significant slowdown in industry trends

In the late 1990s, the performance of internet companies in the United States deteriorated, and behind the bursting of the bubble was a significant downturn in industry trends. Looking at the penetration rate of the internet in the United States (the proportion of internet users to the total population), the late 1990s was a golden period of rapid penetration rate increase. Towards the end of the 1990s and the beginning of the 2000s, there was a clear downturn in the penetration rate, indicating that the incremental market gradually shrank, implying that high-speed demand was unsustainable and the industry's prosperity marginally deteriorated. At the same time, after experiencing investment expansion in the 1990s, the issue of overcapacity in computer software and hardware gradually became apparent, and from 2000 onwards, the capacity utilization rate of the US computer industry rapidly declined. As a result, technology investment entered a trough, and fixed asset investment and venture capital in the US computer industry showed a significant decline after 2002.

(2) Currently, the matching degree of AI stock prices and performance is relatively good in the US, focusing on the sustainability of future performance

From the current data, the matching degree of stock prices and performance in the current round of US technology sector is relatively good, but attention should be paid to the sustainability of AI performance. During the dot-com bubble in the late 1990s, there was a significant acceleration of the bubble in 1999, during which the valuation and stock prices of tech stocks reached new highs and significantly deviated from the fundamentals. This not only manifested in the divergence between the trends of tech and non-tech stocks becoming extreme, but also in the mismatch between tech stocks and the overall market in terms of stock prices and performance. While the overall profit ratio between tech and non-tech sectors in the S&P 500 remained stable, the price ratio steeply increased, laying the groundwork for the bursting of the bubble Comparatively speaking, based on current data, the matching degree between stock prices and performance in the current round of the US technology sector is relatively good, but attention needs to be paid to the sustainability of AI performance.

Risk Warning

The US economy may enter a recession due to the Federal Reserve maintaining a high policy interest rate, leading to an unexpected rate cut by the Federal Reserve or an early end to balance sheet reduction; the issue of the US debt ceiling escalates, causing a sharp drop in US bond yields; escalation of the Russia-Ukraine situation triggers global inflation to rise again; accelerated transfer of savings from European and American banks leads to credit contraction beyond expectations, and so on