SK Hynix Target Price Raised to 870,000 KRW, A Memory Semiconductor Boom Unseen in 30 Years
Editor
Upon hearing the news that KB Securities has raised SK Hynix’s target stock price from 730,000 KRW to 870,000 KRW, a 19% increase, it struck me how significantly the landscape of the memory semiconductor market is changing. Particularly noteworthy is the phrase “the first memory semiconductor boom in 30 years since the internet expansion era of 1995,” which is based on actual market data rather than merely optimistic forecasts from a securities firm.
KB Securities’ Head of Research, Kim Dong-won, has analyzed that current fourth-quarter DRAM demand exceeds supply by more than threefold, highlighting how tight the memory market is. Personally, the most intriguing aspect is the forecast that this extreme supply shortage will persist for two years. Typically, in the semiconductor industry, when supply shortages occur, efforts are made to quickly increase capacity, but that is not easily achievable at present.
Particularly convincing is the analysis that SK Hynix’s Yongin semiconductor cluster is scheduled to commence operations in the first half of 2028, meaning supply expansion will be limited over the next two years. This is likely to lead directly to rising memory prices. In fact, KB Securities projects that SK Hynix will achieve an impressive operating profit growth rate of 84% this year and 89% next year, which is truly remarkable.
To understand the structural changes in the memory semiconductor market, one cannot overlook the impact of the ongoing AI boom. SK Hynix’s dominant position in the High Bandwidth Memory (HBM) market is particularly impressive. Even with new entrants to the HBM market next year, SK Hynix is expected to maintain a market share of 60-65%, due to competitors’ HBM4 redesign issues, illustrating the high technical barriers to entry.
Restructuring of the Supplier-Dominant Memory Market
The analysis that the DRAM market will be restructured to favor suppliers by 2027 is particularly noteworthy. In recent years, the memory semiconductor market has suffered from price declines due to oversupply, but now the situation is completely reversed. The explosive increase in demand for AI server memory is creating a market on a different scale from traditional PC or smartphone memory.
With Samsung Electronics and SK Hynix dominating the global memory market, SK Hynix shows overwhelming superiority, especially in the HBM sector. SK Hynix supplies a significant portion of the HBM used in NVIDIA’s AI chips, which is more of a technological partnership than a mere product sale. SK Hynix is also leading in the development of next-generation products like HBM4, suggesting that this technological advantage is likely to continue for some time.
Although U.S. companies like Micron Technology are entering the HBM market, they still find it challenging to match SK Hynix’s technological prowess and production scale. HBM requires much more complex 3D stacking technology than general memory, and significant expertise is needed to improve yield. These technical barriers further solidify SK Hynix’s dominant position.
The situation is similar in the general DRAM market. The surge in demand for server memory is increasing the need for high-performance products different from traditional PC memory. The expansion of data centers, the increase in cloud services, and the expansion of servers to handle AI workloads all contribute to increased memory demand. Moreover, server memory typically yields much higher margins than consumer-grade memory.
Financial Outlook and Market Expectations
KB Securities’ financial outlook for SK Hynix is truly astonishing. They forecast next year’s operating profit and net profit to increase by 89% and 73% year-on-year, reaching 81 trillion KRW and 70 trillion KRW, respectively, marking an all-time high. Even for the fourth quarter of this year, sales and operating profit are expected to surge by 42% and 87% year-on-year, reaching 28.1 trillion KRW and 15.1 trillion KRW, respectively, raising questions about the sustainability of this growth trend.
The backdrop for such forecasts is the continuous rise in memory prices. The semiconductor industry typically exhibits cyclical characteristics, alternating between booms and busts, but the emergence of AI as a new demand driver is creating a different pattern from the traditional cycle. The increasing proportion of high-value products like HBM is significantly improving overall profitability.
The target price of 870,000 KRW is also quite significant, indicating substantial upside potential from the current stock price level. While securities firms’ target prices are not always accurate, this case appears to be based on concrete data and market analysis, lending it a degree of credibility. Particularly, the medium- to long-term outlook until 2027 suggests that this is not just a short-term recommendation but an analysis of structural changes.
However, despite such optimistic forecasts, there are several risk factors. The AI boom could cool faster than expected, and Chinese memory companies might quickly catch up in terms of technology. Additionally, a global economic slowdown could lead to a general decline in IT demand. Nonetheless, the growth drivers currently seem stronger than these risks.
Personally, I think KB Securities’ analysis captures the structural changes in the memory semiconductor industry well. It is not merely predicting a stock price increase but analyzing the paradigm shift in the industry and evaluating the company’s positioning accordingly, which is persuasive. The expression “the first boom in 30 years” aptly illustrates the uniqueness of the current situation. Compared to the internet expansion era of 1995, the current AI revolution seems to have a similarly significant impact on memory demand.
This article was written after reading an article from Hankyung Korea Market, with additional personal opinions and analysis.
Disclaimer: This blog is not a news outlet, and the content reflects the author’s personal views. Investors are responsible for their own investment decisions, and the author assumes no responsibility for investment losses based on this article.
