원문 보기
Are we really facing the largest stock market bubble in American history driven by artificial intelligence? Well, according to Jeremy Grantham the answer is a resounding yes. Welcome to this analysis where we will systematically unpack, piece by piece, this audacious thesis. This is a very strong statement, we know, but precisely for this reason it deserves to be examined with absolute rigor, putting aside the sensational headlines and focusing exclusively on the macroeconomic data.
To do this, we will divide our study into six very quick steps. We'll start with Grantham's actual warning, then examine the current market euphoria, discuss multiple data, address the crucial issue of capex, explore the infamous Buffett indicator, and finally draw conclusions on price, value, and risk.
The goal: to separate hard data from assumptions. Let's get started. Section 1: Jeremy Grantham's warning . Let's immediately try to identify who is issuing this warning because, after all, the credibility of the source is everything. Do you know what the funny thing is ? Grantham is known for being a bubble spotter, but he himself openly admits that his real Achilles heel is timing.
There is a famous quote of his that perfectly sums up the situation: "The market can remain irrational longer than the customer can remain patient." This basically means that you can have correctly diagnosed a blister, but guessing exactly when it will burst, well, that could take years. So why are major financial institutions hanging on his every word?
The reason is on this timeline. He predicted the collapse of the Japanese market in 1989, anticipated the bursting of the dot-com bubble in 2000, and was right on target with the subprime mortgage crisis of 2007. That's why when he talks today about a possible 70% collapse in AI-related stocks, the entire financial world stops and listens.
Let's move on to section 2: euphoria and extreme narratives. What is the first major symptom that Grantham sees in the market today? Pure euphoria. A prime example he often uses concerns SpaceX. There is a narrative circulating in the markets today that SpaceX's potential market could be worth the equivalent of a quarter of the entire global GDP, a quarter.
It is such a massive and, if you like, disproportionate assumption that Grantham immediately catalogues it as the classic textbook sign of a market that has reached peak speculative activity. But there is a very important distinction to make here. Be careful not to confuse the value of a technology with its stock price.
Think about the internet revolution in 2000. The technology was great, it changed the world, and yet giants like Amazon saw their shares fall by 90% at the beginning, simply because prices had gone out of control. Today with AI we are seeing a similar dynamic: the technology is real, it is powerful, but valuations are based on promises of future profits that are nothing short of monumental and have not yet been realized.
Which brings us straight to section 3: the multiples debate. And here things get interesting because the raw data seems to work against the bubble theory. Let's look at these numbers together because they are precisely the ones that undermine Grantham's alarm. Looking at the mid-2026 estimates, the forward P/E, or expected price-to-earnings ratio, for the S&P 500 stands at around 22.1.
That's actually lower than the record of 24 reached in 2000 during the dot-com peak. The historical P/E is also hovering around 25, not far from its long- term average. In short, the numbers don't exactly scream panic. What's more, both giants like BlackRock and the Federal Reserve itself point out a glaring difference compared to 2000.
During the internet bubble, shares of companies that didn't generate a single dollar of profit were bought at crazy prices. Today, no, AI leaders generate insane profits and have enormous, real, and absolutely tangible cash flows. But, and there is a big but, let's get into section 4, the CAPEX issue.
This is where the huge asterisk to this whole thing comes in. Goldman Sachs has come up with a thought-provoking figure. Currently, the capital expenditures, or CAPEX, required to build the physical infrastructure of AI are absorbing a whopping 75% of the cash flow of these large tech companies. A level of spending that, curiously, almost eerily mirrors that of the late 1990s.
Building huge data centers and buying chips continuously has its risks. First, there is a growing reliance on corporate debt. Second, there's the raw math. It will be very difficult for all these companies to generate long-term returns capable of justifying such expenditures. So, if those famous future profits fail to materialize, those reassuring P/E multiples we saw earlier could turn out to be just a big, dangerous illusion.
Here we are at point 5, the Buffett Indicator. If we have been arguing so far , brace yourselves because this is by far the strongest and most undeniable macroeconomic argument in favor of Grantham's thesis. How does this indicator work? It's very simple. It takes the total value of the entire U.S. stock market and divides it by gross domestic product.
End. No debate about future earnings or interest rates , it simply measures how large the financial market is relative to the real economy that, theoretically, should support it. Now, look at the progression on this table, it's literally mind-boggling. During the famous bubble of 2000, the indicator was at 140%.
In the post-pandemic euphoria of 2021, it soared 200% and today it reached an all-time high of between 230 and 236%. Figures that completely overshadow what two decades ago already seemed pure madness. Of course, to be intellectually honest, we must say that today large American companies make much more money abroad, which somewhat inflates their value relative to domestic GDP alone, but even taking this into account, the gigantic proportional leap from 2000 to today remains an objective fact.
It is a clear and undeniable disconnect between market prices and the real economy, the very heart of potential systemic risk. We now come to our final section: price, value, and risk. It's time to wrap up the discussion and connect the dots. We must remember that the financial industry is not at all united behind Grantham's alarm; in fact, many large institutions disagree at all .
BlackRock, for example, shows real and solid profitability in the sector. Morgan Stanley predicts that productivity will soon increase in many other sectors, not just tech, and Goldman Sachs considers this cycle of investment in physical infrastructure for AI, estimated at $7.6 trillion, absolutely justified and structural, not just a frenzy.
Ultimately, this journey through the data leaves us with a single, powerful question. We know that artificial intelligence technology is real, it 's not smoke and mirrors, but looking at this incredible price level, with Buffett pegged at unprecedented levels and colossal investments to be repaid, the real question is: is the price the market is paying today truly consistent with the profits this technology will be able to generate tomorrow?
A crucial question to understand the real direction of our economy. Thank you for following this analysis.
Comments 0