When will AI peak? Analysis of the inflection point in the US stock market AI cycle.
Recently, a deep review of the US stock market AI sector has revealed significant divergences among top Wall Street institutions regarding the timing and signals of the sector peaking. The market has shifted from narrative exuberance to a critical window of empirical profit.
Current institutional views present a stark contrast: Bank of America (BofA) has issued a clear warning that the AI sector will face an air pocket effect in 2026âcompanies continue to increase capital investment, but commercial returns lag severely, and the mismatch between funding and performance will trigger a sharp correction in stock prices; Goldman Sachs, on the other hand, estimates that capital expenditure in the AI sector could reach $700 billion by 2026, equal to the GDP share at the peak of the 1990s telecom bubble, believing that the industry still has room for expansion and is unlikely to peak in the short term.
Combining market data and industry logic, my core judgment is as follows:
Firstly, the overall US stock market AI sector is in an overvalued range and at the tail end of a bull market. Valuations have fully discounted future growth expectations, and long-term investors like Buffett are shifting to defensive allocations, signaling a contraction in market risk appetite.
Secondly, the life-and-death question of the AI sector lies in the realization of revenue at the application layer. Wall Street has entered a performance verification period, and the current slow commercialization and lower-than-expected conversion to paid services have become the sector's biggest soft spotâif the application layer cannot achieve scalable profitability, the valuation of the entire AI industry chain will face systemic reconstruction.
Thirdly, the growth rate of computing infrastructure has reached an inflection point, with the first half of 2026 being the peak of the acceleration in computing cluster deployment; although sales of large model tokens remain within a normal range, revenue has not shown significant growth on the books, making the industry's characteristic of 'heavy investment, light return' increasingly evident.
Fourthly, considering the cyclical position, industry rhythm, and external events, the AI sector is already in the late stages of a bull market, with continuous adjustment pressure accumulating.
Fifthly, combined with event-driven strategies, SpaceX is expected to go public in mid-2026, increasing the probability of capital diversion and market sentiment switching. The layout of the AI sector can complete a full exit before its IPO, locking in profits and avoiding uncertainty.
Research on the US stock market is still being deepened and improved. In the future, it will continue to adjust the judgment of the AI sector's inflection point by combining financial report data, capital expenditure rhythm, and liquidity changes, seeking the optimal balance between enthusiasm and rationality.
$BTC
{future}(BTCUSDT)
$ETH
{future}(ETHUSDT)
$BNB
{future}(BNBUSDT)