Korea's Two-Quadrillion-Won Megabet and the Limits of State-Backed Industrial Policy

When word emerged that Korea's government, together with Samsung Electronics and SK Hynix, would channel roughly two quadrillion won into three so-called megaprojects spanning semiconductors, artificial intelligence, and the infrastructure beneath them, the first thing that registered was the sheer implausibility of the figure. Two quadrillion won approaches eighty percent of the nation's annual output and dwarfs the yearly state budget several times over. The number is, of course, a cumulative tally stretched across roughly a decade, and the overwhelming share of it is private capital expenditure rather than government outlay. Yet because the state underwrites the foundation of that investment through subsidies, tax credits, power and water infrastructure, fast-tracked permitting, and policy finance, this is not merely a corporate decision. It is a national wager on the direction of an entire economy. And the trouble with wagers is that the larger they grow, the larger the bill society inherits when they go wrong.

The Hidden Trap in State-Underwritten Investment

Korea's brand of industrial policy differs both from the American model, where the state writes the checks directly, and from a laissez-faire posture that defers to market signals. Seoul does not shoulder most of the capital itself. Instead, it props up the lower edge of risk so that private investment can pencil out. The enormous power load, transmission grid, and industrial water flowing into the Yongin semiconductor cluster, along with compressed permitting timelines, are the archetype. The appeal is obvious: direct fiscal spending stays comparatively modest while vast pools of private capital are steered in a chosen direction. But that appeal is also the trap. The moment the state cushions the downside, the signal the market sends to corporate decision-makers is distorted. Profitability and demand forecasts should ordinarily govern the pace and scale of investment, yet when the state backstops the foundation, firms acquire an incentive to build beyond what the market alone would justify.

When this meets the hazard of picking the wrong winner, the danger multiplies. Few dispute that semiconductors and AI will anchor the coming decade. But the precise question—which chips, which AI infrastructure, at what scale and what moment—is not one the government can answer more accurately than the market. Whether HBM demand will keep climbing at its current slope, or at what point AI data-center buildout tips into glut, is something even the world's leading chipmakers cannot forecast with confidence. Once the state concentrates infrastructure around a particular technological path and a particular set of firms, the cost of reversing course when that bet proves wrong far exceeds what decentralized private adjustment would incur. What gets stranded is not only corporate capital but the grids and water systems the state laid down, and the political capital that justified the decision in the first place.

Concentration and Fiscal Rigidity, Two Converging Thresholds

The more structural danger is concentration. When two quadrillion won funnels into the narrow corridor of chips, AI, and their infrastructure, that capital has been drawn out of somewhere else. As financial resources, talent, policy attention, and the career paths of the most capable engineers crowd into a handful of industries, the economy's portfolio tilts perilously to one side. Korea already resembles a monoculture economy overexposed to semiconductor exports, and this megabet entrenches that imbalance structurally. When the chip cycle turns down or US–China technology friction rattles the supply chain, an undiversified economy has no buffer to absorb the shock.

Set against Taiwan and the United States, Korea's position comes into focus. Taiwan has effectively staked its national fate on a single company, TSMC, but that firm enjoys a clear business model in pure-play foundry and a proven moat in overwhelming market share. The United States dispenses subsidies directly through the CHIPS Act, yet the scale is held to a few percent of output and spread across multiple firms and regions. Korea's wager is as concentrated as Taiwan's and as large as America's, while its fiscal headroom is shrinking faster than either. Committing fiscal capacity to industrial policy during a period when aging demographics are structurally inflating welfare spending narrows the room available to respond when the next shock arrives.

The point, in the end, is not that the bet itself is mistaken. Concentrating national resources on semiconductors and AI may well be a rational choice given Korea's circumstances. The problem is that the bet has grown so large that no exit is visible if it misses. With three thresholds—the hazard of picking losers, capital concentration, and fiscal rigidity—approaching at once, what is needed is not to halt the wager but to design the diversification and exits that keep society standing if it fails. The real test of two quadrillion won lies not in where the money is poured, but in what remains when the judgment behind it turns out to be wrong.