This morning's coffee was my usual favorite Georgia Black canned coffee, but it made me think for a little while.
While looking through some documents I had on hand, I came across some figures showing how much return on investment is expected in AI in the future. 。
According to JPMorgan's estimates, in order to achieve a return of 101 TP3T from AI investments by 2030, companies as a whole would need to generate $650 billion in revenue annually, or approximately 100 trillion yen. 。
This would mean that each iPhone user would pay over 60,000 yen per year to the AI. I wanted to take a moment to organize my thoughts about how to interpret these numbers before work began.
The ever-growing expectations and the creeping "predicted collapse"
Currently, the stock market is being driven by seven tech giants known as the "Magnificent Seven," reaching levels not seen since the dot-com bubble. 。
Investors are making huge bets on the future brought about by AI, but bank executives and the International Monetary Fund (IMF) are already sounding the alarm, warning that its market capitalization is "astronomical." 。
Unlike the subprime mortgage crisis, this current boom is primarily driven by equity financing, so the risk of an immediate and widespread paralysis of the financial system is considered low. 。
However, that doesn't mean we can be optimistic. In fact, some say that if the market were to crash, it would be one of the most "predicted collapses" in history. 。
The core issue lies in "the wallets of American consumers."
What I was most interested in while reading this article was the impact on "U.S. household assets." What distinguishes this AI boom from previous recessions is that 211 TP3T of U.S. household assets are held in stocks. 。
This figure approaches that of the peak of the dot-com bubble, and it is said that nearly half of the assets that have increased in the past year are related to AI. 。
In other words, if stock prices fall, American consumers will be the first to tighten their purse strings. This perspective suggests that this could be the trigger that drags the global economy into recession. 。
Three changes you should remember
When the bubble corrects, our business environment may undergo the following changes:
-
The end of America's "exceptional growth"
An estimate suggests that if a decline similar to that of the dot-com bubble occurs, US household net worth could shrink by 81 TP3T and GDP could shrink by 1.61 TP3T. The "America's dominance" that has driven the global economy until now may be shaken. 。 -
The dollar's appeal as a safe haven is declining.
Normally, the dollar is bought when the economy is weak, but with the US growth outlook being revised downwards, there is no guarantee that the dollar will remain the safe-haven asset it once was. In fact, the dollar has been on a downward trend since the beginning of 2025. 。 -
Intensification of friction due to overproduction in China
If U.S. consumer spending slows down, exports from China to the U.S. will decrease. These goods, having nowhere else to go, may flow into other markets, potentially accelerating a global "oversupply of goods" and protectionist movements. 。
These scenarios do not in any way mean that "AI technology itself will become worthless." 。
As history shows, even the most groundbreaking technologies that could change the world are often disappointed once the initial high expectations are dashed. 。
Up until now, we have focused solely on "what AI will change," but from now on, we may also need a more objective perspective on "how expectations for AI have altered the structure of the economy."
Rather than simply chasing market figures, being aware of the fragility of the real economy behind them will likely become an important aspect of "business acumen" going forward.
Well, it's almost time to start work.
Today, I'll try to approach my work with a slightly forward-looking perspective. I'll finish my warm coffee and head to my desk.
