SK Hynix: Can It Double Again This Cycle?

TL;DR

• Aletheia Capital has raised its price target for SK Hynix to approximately $3,500, significantly above the target range of roughly $2,000 to $2,520 set by many mainstream institutions.
• At the core of this valuation divergence is whether the market is willing to believe that the HBM shortage, DRAM price increases, and free cash flow improvements can extend into 2027.
• Related Tickers: SK Hynix, Samsung Electronics, Micron Technology, Nvidia Supply Chain.

In a report released today, Aletheia Capital raised its price target for SK Hynix to approximately $3,500, pushing it well above the target range of mainstream institutions.

Aletheia Capital is an independent research and investment advisory firm headquartered in Hong Kong, serving institutional investors and covering sectors including Asian tech hardware. In contrast, publicly reported price targets from SK Securities are around $2,000, and those from Mirae Asset and KB Securities are around $2,520.

What makes the $3,500 price target truly aggressive is not just that it's more optimistic than mainstream institutions, but that it requires the market to believe three things will happen simultaneously: a continued shortage of HBM (high-bandwidth memory for AI chips), further price increases for conventional DRAM, and AI server demand sustaining memory industry tailwinds and free cash flow through 2027.

The market already acknowledges that SK Hynix deserves a revaluation; the disagreement lies in how far that revaluation can go. Most mainstream institutions still apply a cyclical industry discount, while the $3,500 target pulls out an optimistic tail scenario on top of the revaluation.

The Divergence Lies in the 2027 Profit Base

The $3,500 target is most easily misinterpreted as a simple valuation issue: as long as SK Hynix is given a 10x multiple on its 2027 profits or free cash flow, the stock price can continue to rise. The difficulty lies not in the multiple itself, but in how much the company can actually earn and how much cash it can retain in 2027.

Memory companies' profits are highly volatile. During an upcycle, prices rise, inventories clear, and margins expand rapidly. During a downcycle, new capacity comes online, customers cut orders, prices fall, and profits can quickly reverse course. This is also why the market has historically assigned lower valuation multiples to memory companies.

Even with SK Hynix's current strong profitability, the 12-month forward P/E ratio mentioned in public reports remains in single digits. The market isn't blind to AI; it's worried that this rally will ultimately be priced based on cyclical peaks.

The aggressive price target attributed to Aletheia challenges this cyclical discount. Based on public accounts, it bets that sustained AI hardware demand will drive HBM and DRAM prices higher, and that SK Hynix's free cash flow in 2027 will significantly exceed current consensus estimates, allowing for repricing on a higher base.

The problem is that the $3,500 target requires multiple variables to align favorably simultaneously: HBM prices remain strong, conventional DRAM prices aren't crushed by new capacity, SK Hynix maintains its leading market share, capital expenditures don't consume too much cash, and the market is willing to give a cyclical stock a not-insignificant multiple. If any link underperforms, the target price transforms from a structural revaluation into an extrapolation of strong cyclical conditions.

HBM Transmits Shortage to Conventional Memory

HBM can change SK Hynix's pricing logic because it is not just a minor upgrade to conventional memory, but a core component alongside AI accelerators. No matter how fast an AI chip computes, overall performance will be bottlenecked if data cannot be fed in quickly enough. HBM's role is to provide a higher-bandwidth data pathway for GPUs or AI accelerators.

Retail investors can think of it this way: the GPU is the engine, and HBM is the high-speed fuel delivery system. The more powerful the engine, the greater the demands on the fuel system. In the past, the market traded AI hardware by first looking at Nvidia GPUs. Now, the market is increasingly realizing that whether GPUs can be shipped and AI servers can be built also depends on HBM supply.

Furthermore, HBM supply cannot be rapidly scaled just by slightly modifying conventional DRAM production lines. It requires more complex stacking, packaging, and customer qualification, and it consumes more wafer area and advanced packaging resources. Producing an equivalent capacity of HBM typically occupies more production resources than conventional DRAM.

This effect spills over into conventional memory. As manufacturers shift more resources to HBM production, supply in the DRAM segments used for conventional servers, PCs, and mobile phones tightens, potentially pushing up average DRAM selling prices.

This is the core mechanism that makes the ~$3,500 price target plausible. If HBM were just a fast-growing small product, it would only boost a portion of SK Hynix's revenue. But if HBM simultaneously constrains conventional DRAM supply and elevates the entire memory pricing curve, it becomes an amplifier for the company's overall profit margins and cash flow.

However, an HBM shortage can only prolong the cycle, not eliminate it. Samsung and Micron are both catching up, and SK Hynix itself will also expand capacity. New fabs and packaging capabilities will eventually be reflected on the supply side. The core of the debate is not whether the shortage exists, but how long it can persist and how long prices can remain strong.

SK Hynix Captures the Most Direct Supply Chain Premium

SK Hynix has become the focal point of this revaluation not just because it is a memory company, but because it has moved the fastest in HBM. According to data from Counterpoint cited by Reuters, SK Hynix held approximately 58% of the global HBM market share in Q1 2026, compared to about 21% for both Samsung and Micron. Reuters has also identified it as a key supplier in Nvidia's HBM supply chain.

This leadership is highly valuable in the semiconductor supply chain. AI chipmakers choose HBM based not only on price but also on performance, yield, stability, and qualification progress. The earlier a company passes customer qualification, the easier it enters the next product collaboration window. The earlier it secures orders, the better its position in capacity planning and price negotiations.

This is why the visibility of supply and demand for 2026 has garnered attention. According to a 2025 Reuters report, SK Hynix had completed discussions with key customers regarding its HBM supply for 2026. Numerous industry reports also suggest that the HBM shortage could extend into 2027. For investors, this means the company's 2026 performance is at least partially supported by tangible factors, not just narrative.

SK Hynix's benefit is not limited to HBM revenue itself. Because HBM consumes more capacity, squeezing conventional DRAM supply, its traditional memory business may also benefit from price increases. AI demand first enters the financial statements through HBM, then influences the entire DRAM pricing landscape through capacity reallocation.

This explains why institutional price targets are constantly being revised upwards. Even without accepting the extreme $3,500 scenario, the target range of approximately $2,000 to $2,520 indicates that mainstream institutions are already recalculating SK Hynix's profit elasticity for 2026-2027. The difference is that most still apply a discount appropriate for cyclical industries, without directly extrapolating the post-2027 tightness as a new normal.

A Double Requires Three Conditions to Materialize

The rumored $3,500 price target from Aletheia essentially bets that demand will remain strong, supply will remain tight, and cash flow will continue to exceed expectations. Over the past two years, massive purchases of GPUs by cloud providers and AI companies triggered explosive demand for HBM. Next, the market needs to see whether inference, enterprise AI, and custom ASICs can further expand memory consumption, extending demand beyond just training clusters.

On the supply side, conditions also cannot loosen too quickly. Tightness in 2026 is relatively easy to understand due to lags in capacity, packaging, and customer qualification. By 2027, new capacity and products from Samsung, SK Hynix, and Micron will gradually enter the market. If new supply comes online faster than expected, the pace of HBM price increases may slow, and conventional DRAM will face renewed pressure.

Ultimately, it comes down to cash flow. During an upcycle in the memory industry, companies tend to increase capital expenditure for expansion, process upgrades, and advanced packaging investment. Profit growth does not necessarily all stay on the books. If SK Hynix needs even larger investments to maintain its lead, the free cash flow base that the $3,500 target relies on would be eroded.

Therefore, this price target is better viewed as an optimistic scenario rather than a proven market consensus. The year 2027 will be the true observation window: as long as HBM prices, average DRAM prices, supply dynamics, and free cash flow remain favorable, the market will believe AI is raising the industry's profit floor. If prices soften first, supply materializes ahead of schedule, or cash flow is consumed by capital expenditure, the ~$3,500 target could shift from a revaluation anchor to a sentiment peak.