How to Be Right and Lose Everything

Core argument: being right about a macro trend (AI-driven chip demand) doesn’t protect you from being wiped out by market mechanics — leverage risk is independent of prediction accuracy.
South Korea’s KOSPI is simultaneously one of the world’s best-performing and worst-crashing markets in 2026 — down roughly 27% from its June peak despite genuinely profitable AI-era companies (Samsung Electronics, SK Hynix) underpinning the index.
Over 1.2 million retail accounts were hit with margin calls; hundreds of thousands of Korean retail investors were wiped out by leveraged single-stock ETFs that created mechanical feedback loops on the way down.
Uses Victor Haghani’s “biased-coin experiment” to illustrate how even a demonstrably favorable bet can ruin someone who sizes it wrong.
Takeaway: correct macro calls are necessary but not sufficient — position sizing and leverage, not conviction, determine whether being right pays off or wipes you out.