SK Securities predicted that Samsung Electronics and SK hynix could rise to 400,000 won and 2 million won, respectively. The firm said the judgment is based on a surge in average selling prices (ASP) for memory and the easing of the semiconductor industry's cyclical characteristics as long-term supply contracts expand.
Han Dong-hee, a SK Securities researcher, said, "As memory prices rise more than expected, we raised our earnings outlook for 2026 and 2027 and also lifted our target prices."
Han projected Samsung Electronics' operating profit at 327 trillion won in 2026 and 417 trillion won in 2027. SK hynix is also expected to post operating profit of 253 trillion won and 328 trillion won, respectively, over the same period.
The shift away from a cyclical industry through long-term memory supply contracts is a factor that boosts appeal. Long-term contracts can secure a certain level of volume and prices even during downcycles, which can ease earnings volatility.
However, Han noted that the market is considering only the drawbacks of long-term supply contracts. Han said, "The view that products sell for less than expected during an upcycle is short term," adding, "It is also necessary to consider the midterm view that products can be sold for more than expected during a downcycle."
Han also saw long-term supply contracts helping to mitigate the risk of oversupply from future capacity expansions. Han explained, "Operating leverage is not permanent, and given strong demand for artificial intelligence (AI), we are in a phase that requires capacity additions," adding, "Long-term supply contracts will hedge the oversupply risk from expansions."
Han also projected a structure in which long-term supply contracts and a market exposed to spot conditions run in parallel. Prices formed in the spot-exposed market will serve as the upper and lower bounds for long-term contract prices, maintaining the market's price discovery function, the analysis said.
Despite macroeconomic uncertainties such as the Middle East war, Han assessed that the AI industry's growth will continue. Han said, "Macroeconomic concerns are rising, but the justification for AI investment itself is unlikely to be undermined, and it remains true that memory is the key bottleneck," adding, "Structural improvement in profit-generating power and the easing of cyclical characteristics, along with stronger shareholder returns based on these, will underpin a revaluation of the semiconductor sector."