A product-by-product view of what CXMT's STAR Market IPO means for HBM, server DRAM, client DDR5, LPDDR, NAND, Samsung Electronics, SK Hynix and Micron.
A proprietary Monte Carlo and breadth analysis of why beating the pure KOSPI benchmark has been unusually difficult in 2026, when Samsung Electronics and SK Hynix drove a narrow mega-cap AI memory rally while the median Korean stock was negative.
A June 2026 macro event map covering jobs, CPI, PPI, ECB, BOJ, FOMC, BoE and the remaining gates: PMI, PCE, GDP, durable goods, Fed stress tests and Russell reconstitution. The conclusion is conditional buying in Korea AI memory, not chasing.
A follow-up on the June 2026 FOMC, arguing that the bigger shift is not an immediate rate-hike cycle but less forward guidance, a weaker Fed put, and more selective capital flowing toward real bottlenecks.
A follow-up on Samsung Electronics and SK hynix HBM4E 12-high samples, separating sample timing, specifications, customer lock-in, NVIDIA and AMD channels, and the 3Q-4Q 2026 proof window.
The use-of-proceeds plan in SpaceX's S-1 puts AI compute infrastructure expansion ahead of launch vehicles. The biggest capex and loss center is not the rocket but xAI. In other words, the ~$75bn raise is fuel that accelerates AI infrastructure investment, and that demand flows into GPUs, HBM, memory, and power. Today's reported Chey-Musk meeting (custom HBM for Tesla, memory for xAI data centers) is the first link tying that flow to Korean memory.
The official KOSDAQ premium 70-name list has not been released yet, but KRX Global criteria plus market cap, turnover, fundamentals and governance narrow the first 30 core candidates. The key is not simply market cap, but which KOSDAQ names passive money can actually own.
A macro follow-up on whether AI looks more like the 1996 productivity disinflation story or the 1999 expectations and CapEx cycle, and how that frames the June 2026 FOMC.