[Editorial] The Risks of 'One-Legged' AI Circular Financing: Prepare for the Worst-Case Scenario
- Input
- 2026-11-17 19:15:08
- Updated
- 2026-11-17 19:15:08

Circular financing is a structure in which money invested in AI companies flows back to investors through purchases of products and services. Money circulates as long as investment and purchases continue, but the system is vulnerable because it could collapse in a chain reaction if one link breaks.
Warning signs emerged at NVIDIA, a key source of funding for circular financing. NVIDIA's market capitalization fell by $1.7 trillion (about 2,278 trillion won) amid the disgraceful departure of CEO Jensen Huang. Swiss investment bank UBS estimated that if NVIDIA reduces its financial support, $50 billion or more (about 67 trillion won) could flow out of the market in a short period.
One scenario suggests that if returns on AI investment fall short of expectations, U.S. stock prices could drop 20% this year and global gross domestic product (GDP) in 2028 could be 1.1% below the previous forecast. Since AI has supported global growth and trade, the impact of a correction would inevitably be substantial. Last month, the IMF raised its forecast for global economic growth this year to 3.1%, factoring in AI investment despite the energy crisis triggered by the Iran war.
This is no time to treat this as someone else's problem. Just look at how the 1.0% growth our country recorded in the third quarter depended on semiconductors and government consumption. An AI correction could expose the structural vulnerability of exports, tax revenue and the stock market all coming under pressure at once. Earlier this month, concerns about AI regulation stemming from the U.S. midterm elections triggered a circuit breaker on the KOSPI. The KOSPI also showed signs of unease on the 17th, closing at 4,968 and falling below the 5,000 mark.
Some argue that such concerns are excessive. Georgieva described recent volatility as being at an "orderly level." But she also said that "short-term disappointment could lead to a disorderly correction across asset classes."
A market axiom is that building buffers in calm times costs the least. Financial authorities should conduct stress tests (a method of assessing the ability to withstand losses under extreme conditions) this year, based on the worst-case scenario, and publish the results. The first areas to examine are semiconductor exports and tax revenue, as well as the financial sector's exposure to AI and big tech risks. In its budget deliberations, the National Assembly should screen out spending that relies on temporary tax revenue from the semiconductor boom. Companies, too, should retain profits from the boom as financial buffers and a source of funding for future investment. An economy standing on one leg cannot stay upright for long.