Has the AI bull market ended? The market is filled with doubts.

Technology companies are showing strong profitability and have proactive capital expenditure plans, but this has not been able to drive stock prices higher -- this deviation signal is causing confusion among investors. The path to the end of AI prosperity has evolved from a single narrative to multiple intertwined scenarios, leading to a significant increase in market uncertainty. According to Winds Trading Station, Nomura Securities strategist Naka Matsuzawa pointed out in a report released on July 17 that the cost-benefit balance of AI investments has become blurred, and market participants are facing various potential "AI prosperity end scenarios" and are currently in a state of doubt and observation. At the same time, there is a clear differentiation within the technology sector: semiconductor stocks are falling significantly, MAG7 is holding steady, and software stocks are counter-trend strength, showing that funds are quietly repricing within the sector. In the next two weeks, U.S. hyperscalers will sequentially release financial reports and disclose capital expenditure plans. Matsuzawa pointed out that even if the related data shows strong performance, it is difficult to predict whether the market will respond positively given the complexity of the above scenarios. The bond market has not yet seen a rate cut bet triggered by expectations of a slowdown in AI investments, indicating that the pricing of the end of AI prosperity has not yet been completed. With multiple scenarios overlapping, the path to the end of AI prosperity becomes increasingly complex. Previously, the market mainly assessed potential inflection points in the AI investment cycle based on three scenarios: first, AI investment overheating compresses cash flows of hyperscalers, leading to a slowdown in investment; second, high memory prices raise investment costs, causing AI expenditure to shrink; third, rising raw material costs exacerbate inflationary pressures, prompting central banks to shift to a hawkish stance. However, according to the Nomura report, the market is currently wary of the emergence of a fourth scenario -- high memory prices leading to overheating in semiconductor investments, which then leads to a fall in memory prices. In short, the cost-benefit balance that artificial intelligence brings to the broader economy has become blurred, semiconductor stocks have continued to rise in the past due to the soaring memory prices, but this trend seems to have reached a turning point. The coexistence of four scenarios makes it increasingly difficult to clarify the cost-benefit balance that AI brings to the overall economy. Matsuzawa believes that it is this uncertainty in the multiple paths that is causing market participants to be filled with doubts and observations. Increasing differentiation within the technology sector exacerbates the difficulty of strong performance gaining market recognition. One core contradiction in the current market is that while technology companies are profitable and have proactive investment plans, this has not translated into upward momentum in stock prices. This deviation has been particularly prominent recently. Looking inside the sector, both semiconductor and MAG7 trends are weak, while software stocks are relatively strong. Overall in the U.S. stock market, defensive sectors and consumer-related stocks have shown strength, while technology, capital goods, and bank stocks are generally under pressure. The VIX index has risen slightly to 16.7, while volatility indicators in the bond and foreign exchange markets continue to decline, showing that the current uncertainty is mainly concentrated in the stock market. Matsuzawa pointed out that even if hyperscalers announce impressive financial reports and capital expenditure plans in the next two weeks, the market reactions in various sub-sectors are difficult to predict simplistically under the interference of the above multiple scenarios -- strong data may not necessarily bring about a universal boost across sectors. The bond market has not yet priced in the end of AI, and expectations of a rate cut remain distant. One important signal to watch for is: if the market truly begins to price in the end of AI prosperity, there should be bond buying -- investors will position themselves in advance for future rate cut expectations. However, this situation has not yet occurred. According to the Nomura report, the U.S. bond yield curve is showing a bearish flattening trend, with the 10-year real yield rebounding to 2.31% and the 10-year breakeven inflation rate continuing to decline to 2.23%. Market expectations for the Fed's July meeting to raise rates by 3 basis points, 14 basis points cumulative rate hikes in September meeting, and 27 basis points cumulative rate hikes in December meeting have increased, with the 2-year forward OIS rate (terminal rate proxy benchmark) rising to 3.84%. This indicates that the current pricing logic in the bond market is still dominated by expectations of rate hikes, rather than a shift towards betting on future rate cuts. Based on this, Matsuzawa judges that the pricing of the end of AI prosperity in the market is still incomplete, investors' doubts are still fermenting, rather than forming a clear consensus. This article is reprinted from "Wall Street News", GMTEight Editor: Zhang Jinliang.