Taiwan has thus become the world"s fifth-largest stock market, trailing only Hong Kong, Japan, China, and, much further ahead, the United States. South Korea follows closely in eighth place. Driven by SK Hynix and Samsung Electronics, it recently overtook the UK, with Canada potentially next in line.
This shift is all the more striking as these markets are overtaking Western economies with significantly greater economic weight. South Korea"s GDP is lower than Spain"s, while Taiwan's sits below that of Poland.
The speed of this movement is impressive. Taiwan and South Korea have become global equity heavyweights. This ascent is also built on high concentration. TSMC accounts for over 40% of Taiwan's market capitalization. In South Korea, SK Hynix and Samsung Electronics together command a comparable weight.
Market conditions have amplified this surge. TSMC holds a near-monopolistic position in advanced foundry services and remains indispensable to major American chip designers. In South Korea, SK Hynix and Samsung are benefiting from a bottleneck in memory. So far, their customers have shown little resistance to price hikes, resulting in a spectacular impact on margins and operating leverage. SK Hynix's net profit jumped 398% in Q1, while Samsung"s also leapt - up 475%.
However, this strength carries its own fragility through "national champion" risk. Saudi Arabia and Denmark offer two recent examples. Following the IPO of Aramco, Saudi Arabia became one of the world"s leading financial hubs. In Denmark, Novo Nordisk's prospects in obesity drugs significantly bolstered the national economy. Yet, falling oil prices and Novo Nordisk's challenges in the US market subsequently placed these two national indices among the worst performers of 2025.
Concentration looks like a strength, but as soon as pricing, competition, or expectations turn, it becomes a vulnerability. The key question is therefore whether the current rally in Taiwanese and South Korean markets represents a structural shift or a temporary peak.
TSMC appears to hold the most structural position. This is reflected in its valuation, with a 2027 P/E of 18x, roughly 3x higher than that of both the South Korean groups. The primary risk would be that a partial relocation of production outside Taiwan might reduce the efficiency of its model. However, the island has become so central to the AI revolution that it remains difficult to bypass. Nvidia, for instance, recently announced $150bn in annual investments in Taiwan.
Nevertheless, these companies all depend, directly or indirectly, on hyperscaler spending. Memory shortages are unlikely to disappear quickly, but suppliers remain exposed to AI returns on investment and the capital expenditure decisions of Big Tech. Customers could also pivot towards alternative memory architectures for their chips, especially if incentives to innovate remain strong for competitors.
These upheavals in the stockmarket hierarchy may be more durable than expected. TSMC, SK Hynix, and Samsung are not rising solely on a market narrative. They are rising because they own rare physical assets that are difficult to replicate and impossible to build quickly. The real question is not whether current margins will remain intact, as they likely won't. Rather, it is whether the sector's profitability floor has shifted. If HBM, agentic AI, and multi-year contracts make the memory cycle less brutal than in the past, current valuations may already be underestimating the entry into a new regime.
