Three days ago SK Hynix priced its ADR at $149 in New York and raised $26.5 billion. In the largest foreign listing in US history. First day it closed up 13% at $168.49. Today Seoul dropped it another 15.4%. Kospi followed straight down 8.95% at the 7,000 level, triggering a circuit breaker for twenty minutes of silence.
Domestic pricing is sitting right now around $122.70 per ADR-equivalent. Almost a 28% gap versus Friday's close with no conversion mechanism to force arbitrage closing it. The disconnect just sits there waiting for one side to blink.
What ties this directly to the Strait of Hormuz rather than some generic macro risk-off flow is how concentrated Korea's energy exposure actually is. Roughly 70% of their crude and 20% of LNG comes from the Middle East, and more than 60% of Korean crude imports plus half its naphtha transited that chokepoint in 2025. That matters for fabs specifically since they run on cheap, reliable electricity and depend on petrochemical feedstocks that move through those same shipping lanes, and it shows up in FX pressure that tracks oil prices closely. Korea's vulnerability to a Hormuz disruption is tighter than a standard oil-importing economy's baseline.
The pricing gap also needs to be viewed against structural precedents instead of assuming idealized convergence over time. The closest analog is China A-share/H-share listing structures, where identical entities trade at persistently different valuations across segmented markets. That premium hasn't been narrowing. A 2026 arXiv study of 67 dual-listed A/H firms found that Shanghai-Hong Kong Stock Connect, the mechanism built specifically to narrow this gap, was associated with an 18.4% average increase in the premium instead. Semiconductor names in that same framework like SMIC and Hua Hong currently carry H-share discounts pushing near fifty percent, and that's the live number, not a historical footnote.
SK Hynix's ADR is three days old so we cannot project its exact convergence path yet. The mechanics for a fresh geopolitical shock hitting a leveraged retail market definitely differ from decades-deep structural separation. Still the existing analogs lean toward persistence over quick snaps, not the other way around. US-Iran tensions near Hormuz keep escalating, vessel traffic is sitting at five-week lows, and reporting offers zero signs of the de-escalation needed to trigger a snap-back recovery.
TSMC's revenue reflects orders placed months before they ship, across the entire chip ecosystem, not just memory. So a 68% June jump doesn't say much about today's sentiment, but it does say demand wasn't cracking when those orders were locked in. If SK Hynix's crash reflected a real break in AI chip demand rather than a positioning unwind, you'd expect to see it first in forward guidance or bookings, not in a memory stock's one-day move following an oil shock.
What specific catalysts or price levels would you want to see before calling this a structural discount rather than temporary panic pricing?