Margin Debt, Leverage, and the South Korean Stock Market Collapse as a Preview for the US

South Korea's stock market recently underwent a dramatic boom-and-bust cycle, described as 'the most concentrated major stock market in the world.' Borrowing heavily, 14 million retail investors (one in four adults) placed their savings and leveraged positions mostly into Samsung and SKINX, which together accounted for over 56% of the Korean market (Cost be index). The preceding run-up was fueled by AI-driven global demand for memory chips, with Microsoft, Google, Amazon, and Meta projected to spend $725 billion on AI infrastructure in 2026. However, in just 21 days, following a sell-off in US chip stocks led by Micron, the Korean market dropped 25%, triggering 1.2 million margin call thresholds, liquidating over 3 trillion won in investments and wiping out 320,000 accounts.

This scenario is considered a preview for the US market due to similarities in high margin debt levels: in June 2026, US margin debt reached 4.5% of GDP, the highest ever recorded—surpassing previous peaks preceding crises in 1968, 1972, 1987, 2000, and 2007. The official statistics likely understate true leverage, omitting leveraged ETFs, options, portfolio margin, and private credit. In the US, the S&P 500's concentration is historically high, with its top 10 companies representing 36% and most making the same AI bet.

The fundamental risk is that AI spending (capital expenditures by Microsoft, Google, Amazon, Meta) holds up the entire pyramid: unprofitable AI startups (OpenAI, Anthropic) drive hyperscaler spending, supporting chip earnings, which underpin stock market valuations. Jim Chainos notes ROI on AI investments is 'going toward 10%' from 40% a year and a half ago, warning that unsustainable spending will prompt a CEO to ask, 'does it make sense for our company to pay this much for AI?' When a major (hyperscaler) cuts CapEx, the ensuing earnings collapse in suppliers (like Nvidia and Micron) could trigger a US doom loop of margin calls and forced selling.

Two scenarios are considered: (1) Korea remains isolated, with AI spending and markets recovering and six–12 months continued growth; (2) Korea is the first domino, and the US—due to its unprecedented leverage—faces a similar unwinding if CapEx is cut. The speaker draws a parallel to the Titanic disaster, quoting a tweet: absolute confidence preceded sudden catastrophe. The final recommendation is caution around leverage and debt amid current valuations, noting that while AI may deliver tremendous future value, 'investors are paying future prices in the present time as though these companies have already accomplished all the things they promised.'