On October 13, 2022, the S&P 500 hit a stage bottom of 3,491 points, starting a bull run that lasted more than two years. On February 19, 2024, it hit a record high of 6,147 points, and has been falling since then; the performance of the other two major US stock indices (Nasdaq and Dow Jones) is similar. Considering that the decline of US stocks from the peak has reached 10% (Nasdaq is a little more), the whole market is concerned about one question: Is this the end of the bull run of the past two years and four months? Or is it another "fake fall" like the ones from July to October 2023, April 2024, and July 2024?
History has proven that predicting the top is as thankless as predicting the bottom. I have several friends who have rich experience in investing in US stocks. They all thought that "everything is in place" between April and July 2024 and sold all their holdings, including Nvidia and Apple. I am afraid that no one can "accurately" predict the market in units of one or three months, not even the most advanced quantitative funds. However, we can at least discuss some longer-term and more fundamental issues, such as the following two questions:
There is a company with an operating income growth rate of about 5% and a net profit growth rate of about 10%, but its static price-earnings ratio is as high as 34 times. Would you be willing to buy it?
There is another company with an operating income growth rate of about 12% and a net profit growth rate of 15-18%, but its static price-earnings ratio is as high as 31 times. Would you be willing to buy it?
There is another company with an operating income growth rate of about 15% and a net profit growth rate of 30-35%, but its static price-earnings ratio is as high as 45 times. Are you willing to buy it?
The above three companies are: Apple, Microsoft, and Netflix. The first two are the first and third largest companies in the US stock market, and the third is also a large and important technology company and a leader in the entertainment industry. Their P/E/G are obviously much higher than 1.0 times; to make a DCF model for them, very tolerant conditions (such as a very high perpetual growth rate and a very low capital cost) are required to draw the conclusion that they are undervalued. This is a microcosm of the current situation of US stock technology giants after a 28-month bull market.
Please note that when discussing the valuation of US stocks, we are often easily led by the computing power industry chain and fall into a vicious circle such as "Is Nvidia a cyclical or growth company?" and "How to value Broadcom?". The problem is that the computing power industry chain is only a small part of the US stock market, although it has already produced three trillion-dollar companies (Nvidia, TSMC, and Broadcom). Since the valuation of the computing power industry chain is generally recognized as "metaphysics", why not put it aside and take a look at other sectors. At least in the Internet industry that I am concerned about, all high-quality companies are not cheap. From a horizontal comparison perspective, some companies are slightly cheaper, such as Google and Meta, but we cannot say that they are "very cheap."
Looking back over the past two years, it is not difficult to find that the bull market in the US stock market is actually supported by two factors that are not completely related to each other: AI and the economy.
Generative AI has significantly boosted the computing power industry chain, and at the same time pushed up almost all major Internet companies. For some companies, such as Nvidia, TSMC and Microsoft, it has brought tangible revenue and profit promotion; on a larger scale, it has brought confidence.
The US economy is much stronger than expected, and the recession that investors expected has never come. From July to October 2023, the market experienced a long and uneasy adjustment, and then the consensus expectation became that there would be no recession in 2024. Both US consumption and employment data are very strong, and inflation has been controlled to a certain extent.
Starting from the second half of 2024, a third factor is added, namely the expectation of the Fed's interest rate cuts - this will also be the first "normal" interest rate cut cycle that has not been disturbed by major external events since 2001. In this way, the capital market has ushered in an extremely rare "trinity" of positive factors: the overall macroeconomic strength, the booming new technological revolution, and the Fed is still cutting interest rates! You can hardly imagine a better situation than this!
Even the frequent international geopolitical tensions in the past two or three years are also delivering ammunition to the US stock market: to a certain extent, the US stock market, like gold, plays the role of a geopolitical "safe haven". If you are nervous about the international situation, but not particularly nervous, you may transfer your positions from emerging markets and other developed markets to US stocks. So we have seen again and again that no matter who wins or loses, the beneficiaries are all US stocks when there is a slight disturbance in any corner of the world. At first glance, it seems unfair, but it is actually logical.
The current situation is very different from three months ago. First, the Fed's rate cut rhythm is likely to be interrupted, or even return to the track of rate hikes, because inflation is back. Most American consumers list inflation as their biggest economic concern, and the trade frictions provoked by the United States will only exacerbate rather than ease inflation. In the first half of 2023, the US stock market reversed under the pressure of the Fed's rate hike, but the market valuation level at that time was only more than half of the current level. If the expected rate cut turns into a rate hike, it will obviously be particularly unfavorable for those technology giants whose valuations are at historical highs. Secondly, when international geopolitical tensions reach a certain point and break through a certain critical point, US stocks, like other risky assets, are not immune. Just like in the initial stage of a major flu, people with relatively good health may be able to survive on their own and stand out from the crowd; but as the flu continues to evolve, even the healthiest people will get sick, and the only difference is the severity of the disease. After all, U.S. stocks are not gold or U.S. bonds, and will inevitably be affected by geopolitical risks. Have we reached that critical point now? It is unknown, but it is certain that we are getting closer and closer to the critical point. Again, the AI arms race may have progressed to the point where investors feel that it is unprofitable in the short term. In 2025, the total capital expenditure (Capex) plan of U.S. technology giants exceeded $300 billion, of which Amazon was $100 billion, Google was $75 billion, and Meta was $60-65 billion - the biggest driving force behind it is undoubtedly the purchase of AI computing power. The rising capital expenditure not only brings heavy depreciation cost pressure, but more importantly, squeezes free cash flow. You know, an important driving force of the U.S. stock bull market is the continuous dividend repurchase of listed companies. The dividend repurchase amount of technology giants in a single quarter can often reach tens of billions of dollars; spending tens of billions more on computing power means returning tens of billions less to shareholders. And 2025 is by no means the peak of computing power construction, and crazier days are still to come. This is certainly good news for the computing power industry chain, but again: the computing power industry chain is only a small part of the US stock market. As for whether the US economy will enter a recession, it has become a less important issue. Economic cycles exist objectively. There is no economy in the world that only expands and does not decline. Developed economies such as the United States are especially no exception. There was no recession in 2023, nor in 2024, and there will probably be no recession in the first half of 2025. However, assuming that the Federal Reserve returns to the track of raising interest rates, and the White House refuses to implement fiscal expansion under the banner of "saving", it means a double tightening of fiscal and monetary policies. How strong must the fundamentals of the US economy be to withstand this double tightening (as well as international trade and geopolitical pressures) without any damage? I can hardly imagine this possibility. I emphasize again: in any case, predicting the bottom or top of the market is a thankless act. However, common sense tells us that the capital market is like a pendulum, always swinging between extreme optimism and extreme pessimism, and it is an objective law that things will turn around when they reach their extremes. Common sense also tells us that no matter how high-quality an asset is, it has a reasonable valuation. If it deviates too far from the valuation center, it will inevitably produce the pressure of mean reversion. Mean reversion may occur in one month, three months or one year, but it will always occur in the long run. Whether it is time for mean reversion or not, there is one thing I am sure of: generative AI is a real technological revolution, and its long-term impact may be greater than the Internet itself, reaching a height comparable to electricity or even drilling wood to make fire. There is no contradiction between something changing everything in the long run and being overvalued in the short run. Because human society advances in waves, if we only know how to make linear extrapolations, we will make mistakes that are sometimes overly optimistic and sometimes overly pessimistic. Precisely because of my confidence in generative AI, even if I think that the valuation of US stocks is seriously overpriced, I am still optimistic in the long run - it is a great thing for the valuation of high-quality assets to be adjusted back to below the mean, and real value investors should welcome this. 
Joy
