The Federal Reserve Board published a FEDS Note by staff economist Geng Li examining whether major general purpose technologies can lead to overinvestment. The note argues that a large technology breakthrough such as artificial intelligence can produce a sustained investment boom and higher productivity, but that the same process can also end in excess capital and overcapacity even when investors behave rationally and account for that risk. The note also states that the views are those of the author, not necessarily those of the Federal Reserve Board or its staff. The analysis uses an embodied technology model in which installing new capital helps diffuse a new technology across the capital stock, so marginal returns stay elevated during the buildout phase. Because investors do not know the true size of the technology shock at the outset, they learn by investing and updating beliefs as new output data arrive. Under some assumptions about those prior beliefs, that learning process can cause investment to accelerate and overshoot the level needed to fully embody the new technology, leaving marginal returns below market interest rates and creating a capital overhang. The note points to the 1990s IT boom and bust as a historical example, and says the recent rise in U.S. intellectual property and equipment investment has been comparable in pace to the 1990s episode, with the share of GDP reaching 11.5 percent in 2000 and, by the first quarter of 2026, standing about 0.8 percentage points above its 2024 level and only slightly below the 2000 peak. A large part of that recent investment surge is described as AI-driven. The note concludes that another sustained AI-led investment boom is possible, that such a boom could also end in overinvestment, and that this risk does not necessarily imply investment should be preemptively restrained.
Federal Reserve Board2026-07-06
Federal Reserve Board publishes staff note on AI driven investment booms and overinvestment risk
The Federal Reserve Board published a staff FEDS Note arguing that major technologies such as AI can trigger long investment booms but can also end in overinvestment and excess capacity. The note says this can happen in a fully rational setting when investors learn about the scale of a technology shock through continued investment. It links the framework to the 1990s IT boom and to the recent AI-driven rise in U.S. IP and equipment investment.