Morgan Stanley, Nomura split sharply on chip stocks

Résumé détaillé

Morgan Stanley and Nomura have issued sharply divergent forecasts for South Korean chipmakers Samsung Electronics and SK hynix, exposing a growing divide among global investment…

Morgan Stanley and Nomura have issued sharply divergent forecasts for South Korean chipmakers Samsung Electronics and SK hynix, exposing a growing divide among global investment banks over how long the artificial intelligence-driven memory chip boom will last.

Morgan Stanley slashed its price targets for the two companies, warning that additional production capacity and slower AI infrastructure investment could erode memory chip profitability after 2028. Nomura, by contrast, expects supply shortages to persist through at least 2028 and has maintained price targets nearly three times as high as Morgan Stanley’s for SK hynix.

In a report released Monday, Morgan Stanley cut its price target for SK hynix by 38 percent to 1.6 million won ($1,196) from 2.6 million won. It also lowered its target for Samsung Electronics by 19 percent to 309,000 won from 381,000 won, while reducing its target for Samsung’s preferred shares to 247,200 won from 304,800 won.

The report, titled “Memory Semiconductors — Key Issues After the U.S. Midterm Elections,” was written by Morgan Stanley analyst Shawn Kim, who has previously drawn attention for bearish calls on the semiconductor cycle.

Kim warned of deteriorating industry conditions in August 2021 and again raised concerns about a potential memory supply glut in September 2024, when he sharply reduced his price targets for chipmakers.

This time, Morgan Stanley’s central concern is that memory shortages will begin easing as new manufacturing facilities come online toward the end of 2027, potentially leading to excess capacity and weaker profit margins in subsequent years.

In an Oct. 20 report, the Japanese investment bank maintained its price targets of 670,000 won for Samsung Electronics and 4.7 million won for SK hynix, arguing that strong demand and supply constraints would sustain favorable conditions for memory producers through at least 2028.

The contrasting forecasts reflect disagreements over when bottlenecks in AI infrastructure…

Morgan…

Début de l'article, extrait de Pulse (Maeil Business). Pour lire la suite, ouvrez l'article chez le média.