Morgan Stanley has released a research report indicating that TSMC and Samsung Electronics are progressively consolidating and rationalizing their mature node process capacity. This strategic move aims to prioritize resource allocation towards advanced processes and packaging to meet AI-related demand. The firm believes that, over the medium to long term, this could create spillover demand opportunities for second-tier wafer foundries and specialized process manufacturers, as clients seek alternative manufacturing partners to ensure stable supply and cost-effective production.

The report suggests that an upturn cycle for mature node processes is imminent. Given that AI infrastructure construction may persist for another two years, some smartphone and PC clients are now beginning to express concerns about potential shortages in mature node capacity. Both SMIC and United Microelectronics Corporation (UMC) highlighted similar trends in their recent earnings calls. Consequently, Morgan Stanley does not anticipate inventory adjustments or order cuts for mature node wafer foundries in the near term.

Morgan Stanley has raised its target price for SMIC's H-shares from HK$70 to HK$85, maintaining an "Overweight" rating. The firm has increased its earnings per share (EPS) forecasts for SMIC for 2026, 2027, and 2028 by 11%, 14%, and 10%, respectively. This revision reflects strong second-quarter revenue and gross margin guidance, which the bank attributes to robust expansion in advanced process capacity and improvements in the blended average selling price (ASP), driven by an enhanced product mix and price increases on the BCD platform. Accordingly, Morgan Stanley has also raised its revenue forecasts for SMIC for 2026, 2027, and 2028 by 9%, 15%, and 15%, respectively, and increased its gross margin forecasts for 2026 and 2027 by 1.4 and 0.1 percentage points.

Morgan Stanley has also raised its target price for HUA HONG SEMI from HK$88 to HK$118, maintaining a "Market Perform" rating. While the firm has lowered its EPS forecast for the current year by 7%, it has raised its EPS forecasts for the next two years by 15% and 20%, respectively. This adjustment accounts for strong demand for AI-related energy management semiconductors and potential price increases on the BCD platform, an area where HUA HONG SEMI possesses specialized expertise. Even as consumer-facing products may face headwinds due to memory price increases, Morgan Stanley believes that demand for AI-related products can offset these negative impacts. The firm notes that HUA HONG SEMI's capacity and shipment growth momentum is encouraging, with Fab 9 currently undergoing expansion.