The market is witnessing a puzzling divergence: robust tech company earnings and aggressive capital expenditure plans are failing to fuel corresponding stock price gains. This disconnect is sowing confusion among investors. The narrative around a potential end to the AI boom has evolved from a single storyline to a complex web of multiple scenarios, significantly heightening market uncertainty.
According to a report by Nomura Securities strategist Naka Matsuzawa on July 17, the cost-benefit balance of AI investment has become ambiguous. Market participants are now grappling with several potential "AI boom endgame scenarios," leaving them in a state of doubt and caution. Concurrently, a clear divergence is emerging within the tech sector: semiconductor stocks have suffered significant declines, the so-called MAG7 group has largely traded sideways, while software stocks have shown relative strength. This indicates a quiet but significant internal repricing of capital across subsectors.
Over the next two weeks, U.S. hyperscale cloud providers are scheduled to report earnings and unveil their capital expenditure plans. Matsuzawa notes that even if these figures are strong, the complex backdrop of multiple scenarios makes it difficult to predict whether the market will respond positively. The bond market has not yet seen a surge in bets on interest rate cuts in anticipation of an AI investment slowdown, suggesting the market has not fully priced in a potential end to the AI boom.
Complex Endgame: Multiple Scenarios Cloud the AI Outlook
Previously, market assessments of a potential turning point in the AI investment cycle centered on three main scenarios. First, that overheated AI investment could squeeze the cash flows of hyperscalers, leading to a subsequent investment slowdown. Second, that persistently high memory prices could escalate investment costs, triggering a contraction in AI spending. Third, that rising raw material costs could exacerbate inflationary pressures, prompting central banks to adopt a more hawkish stance.
However, the Nomura report indicates the market is now alert to a potential fourth scenario: high memory prices spurring an overheated semiconductor investment cycle, which in turn leads to a future decline in memory prices. In essence, the cost-benefit calculus of AI for the broader economy has become unclear. While semiconductor stocks previously rode a wave of surging memory prices, that trend appears to have reached an inflection point.
The coexistence of these four scenarios makes it increasingly difficult to discern the overall economic cost-benefit balance of AI. Matsuzawa believes it is precisely this multi-path uncertainty that has plunged market participants into a state of skepticism and watchful waiting.
Sector Divergence Deepens: Strong Fundamentals Fail to Impress
A central market paradox is the failure of healthy tech profits and ambitious investment plans to translate into upward momentum for stock prices. This disconnect has been particularly pronounced recently.
Looking within the sector, semiconductors and the MAG7 have shown weakness, while software stocks have held relatively firm. In the broader U.S. market, defensive sectors and consumer-related stocks have performed better, while technology, capital goods, and banking stocks have faced pressure. The VIX index has risen to 16.7, while volatility measures in the bond and foreign exchange markets continue to decline. This indicates the current uncertainty is primarily concentrated in the equity market.
Matsuzawa points out that even if hyperscalers report stellar earnings and capital expenditure plans in the coming weeks, predicting the market's reaction across subsectors is complicated by the overlay of multiple scenarios. Strong data may not necessarily lead to a broad-based rally.
Bond Market Holds Firm: No Pricing for an AI Slowdown Yet
A key signal to watch is the bond market. If the market truly began pricing in the end of the AI boom, one would expect to see bond buying as investors position for future rate cuts. This dynamic, however, has not yet materialized.
According to the Nomura report, the U.S. Treasury yield curve is experiencing a bear flattening. The 10-year real yield has rebounded to 2.31%, while the 10-year breakeven inflation rate has continued to fall to 2.23%. Market expectations for Federal Reserve rate hikes have intensified, with pricing for a 3-basis-point increase at the July meeting, a cumulative 14 basis points by September, and 27 basis points by December. The 2-year forward OIS rate, a proxy for the terminal rate, has risen to 3.84%.
This suggests the bond market's current pricing logic remains dominated by expectations for further rate hikes, not a pivot towards anticipating future cuts. Based on this, Matsuzawa concludes that the market's pricing of an AI boom conclusion is incomplete. Investor skepticism is still in a phase of development rather than having crystallized into a firm consensus.
Comments