Skip NavigationMarketsBusinessInvestingTechPolitics & PolicyVideoWatchlistInvesting Club
PRO
LivestreamMenuRecent weakness in Asian technology stocks does not mark the end of the artificial intelligence investment cycle, J.P. Morgan said in a note Wednesday, adding that investors have become overly concerned about the sustainability of AI spending despite little evidence of a fundamental slowdown. The Wall Street bank said the recent 25%-30% correction in Asian technology stocks and the Philadelphia Semiconductor Index marks the third major drawdown since the AI driven upcycle began in late-2022. “Stepping away from the share price moves, we do not see any fundamental indicators that signal meaningful weakness in the next 6-12 months,” it wrote in a report published Wednesday. J.P. Morgan said the key foundation of the current cycle remains intact, with frontier AI models continuing to improve every few months while demand for AI inference remains strong across both proprietary and open-source models. It also pointed to improving profitability across the AI ecosystem as agentic AI gains traction. While investors have increasingly questioned whether hyperscalers can sustain their aggressive AI spending, the bank said it does not expect cloud providers to scale back investment. “We do not anticipate any of the hyperscalers stepping back on AI compute investments in 2027,” the bank said, adding they are likely to tap equity and debt markets to finance further AI infrastructure expansion. The market is already pricing in a downturn that is unlikely to occur, it said . “The reverse,” with broader earnings upgrades and continued increases in AI-related capital expenditure is likely, the analysts said. Among the factors that could improve investor sentiment, the bank cited broader adoption of generative AI and agentic workflows by software companies, wider deployment across industries including financial services and healthcare, and continued advances by leading AI labs toward recursive self-improvement. Looking across the semiconductor supply chain, JPMorgan said semiconductor equipment manufacturers appear “the best-positioned sub-sector” over the next 12 months as wafer fab equipment spending accelerates. The bank also expects packaging and testing to see sharp growth as 2.5D packaging, which integrates multiple chips side by side, goes mainstream and the 3D packaging investment cycle begins at TSMC, while identifying IC substrates as the “most promising” among components. 3D packaging is a technology in which multiple chips are stacked vertically to boost speed and power efficiency. Memory remains a more nuanced story. Although memory supply-demand fundamentals remain solid, JPMorgan said supply is expected to run well short of demand for the next two to three years. It said said recent moves by Nvidia and AMD to lower memory content in future AI products have weakened the prevailing market narrative that “AI-driven memory demand is price-inelastic and this time is different.” “Market narrative on Memory is problematic, even though the fundamentals are sound,” J.P. Morgan analysts said, adding that while the sector could rebound over the next six months, it does not expect memory stocks to reclaim the highs reached in May. “As efficiency becomes paramount, interconnect could become the next key bottleneck,” the bank wrote. Over the next 18 to 24 months, however, it expects power availability to overtake chips as the main constraint on AI compute infrastructure as semiconductor capacity ramps up.Read More
About The Author
