The Korean economy is being showered with a semiconductor-driven windfall. Buoyed by strong semiconductor exports, the cumulative trade surplus through May has reached 102.1 billion dollars. The largest annual trade surplus on record was 95.2 billion dollars in 2017, but the all-time high was surpassed in just five months. The current account surplus as a share of gross domestic product (GDP), a gauge of foreign currency earning capacity, also came in at 9.4% in the first quarter, the highest on record.

It is the second such windfall in the history of the Korean stock market. The first appeared during the mid-to-late 1980s three-lows boom. The three lows then stemmed squarely from shifts in the global environment. Low interest rates, low oil prices, and a weak won were all products of exogenous changes. The US federal funds rate, which had reached 20% in 1981, fell to 5.8% by 1986, allowing the Korean economy, burdened by external debt from the heavy and chemical industry drive of the 1970s, to catch its breath. The price of Dubai crude also plunged from around 35 dollars per barrel in 1981 to the eight-dollar range in 1986. On top of that, after the 1985 Plaza Accord the yen strengthened, and the won per 100 yen exchange rate rose from 290 in 1985 to 590 in 1988.

With external conditions improving, Korea saw GDP growth exceed 10% for three consecutive years from 1986 to 1988, and the current account moved into surplus for the first time. The dollars flowing in on the back of exports swelled the money supply. From 1987 to 1988, growth in the M2 money supply exceeded 30%, and consumer price inflation topped 8%. Abundant liquidity flowed into asset markets. The KOSPI jumped 666% in four years, and land prices nationwide nearly doubled. Home prices and jeonse deposits in Seoul and the greater capital area also surged.

Selective boom, costs borne by society

Yet at the time, the fruits of the boom spread relatively widely. As the bargaining power of labor unions strengthened during democratization, wages rose sharply, and employment and income expanded across manufacturing. It was also the period that laid the foundation for the middle class of today. The government also moved to expand supply, advancing a 2 million-home construction plan to counter soaring home prices. There were side effects, but policy measures were taken to spread the gains of the boom across society.

The current semiconductor boom looks different. Aggregate macro indicators are better than ever, but the benefits are concentrated among a few. Semiconductor firms and their employees, and investors who hold their shares, enjoy the largest gains. One can extend the scope to power equipment makers and some equipment companies benefiting from expanded investment in artificial intelligence (AI), but across industry the range of beneficiaries is narrower than expected. By contrast, many manufacturing segments still face weak demand and competitive pressure from China. Service industries and the self-employed also find it hard to feel the semiconductor boom.

The problem is that society as a whole bears the costs of the boom. A boom inevitably stokes prices. For those whose incomes rise in tandem, inflation may be a bearable burden, but for workers with flat wages, pensioners, and the self-employed, it means a decline in real income. Interest rates are no different. US big tech companies are raising funds for AI data center construction by issuing corporate bonds and equity on an astronomical scale. When global demand for funds surges, the price of capital, interest rates, comes under upward pressure. Market rates are already trending upward globally. With the government, households, and firms all carrying record levels of debt, rising rates impose a heavier burden on those who do not share in the boom. As of 2025, 43% of domestically listed manufacturers are not generating enough operating profit to cover interest. In short, the semiconductor boom is not a boom for every company.

The spread of AI is also a new variable. Technological innovation to date has tended to increase jobs, but AI is likely to boost productivity while replacing some roles. In that sense, this boom is far more selective than in the past: the semiconductor industry may grow while the labor market as a whole does not.

Imbalances in the housing market make the situation even more bleak. When home prices soared in the late 1980s amid excess liquidity from the three-lows boom, the government played a strong card with a 2 million-home construction plan. At the time there was physical and geographic space, such as Bundang and Ilsan, where large-scale housing could be built. Announcing large new towns signaled that people could someday own a home, serving as a buffer that stabilized housing market sentiment.

Policies needed to spread the gains more widely

How about now. The greater Seoul area is already saturated, with an absolute shortage of physical space to develop large residential sites. Urban reconstruction and redevelopment are slow due to complex interests and construction costs that have surged in recent years. Funds held by the lucky few are concentrated in core areas of Seoul and the areas around semiconductor clusters, but the physical capacity to absorb them with housing supply is woefully insufficient compared with the past. As a result, the asset price increases driven by this boom are highly likely to be far more prolonged and entrenched than in the 1980s. A small group with assets becomes richer without effort, while a marginalized majority is pushed downward as the housing ladder breaks, deepening disparities.

Policy direction is crucial to ease these imbalances. What is needed now are policies that spread the gains of the boom widely. Rather than blending the increased tax revenues, such as corporate and income taxes that naturally rise during a semiconductor boom, into general finances, a fiscal rule for boom times should be established that prioritizes allocating them to funding industrial transition and strengthening the social safety net.

Safety nets are also needed for vulnerable groups and marginal firms taking a direct hit from high interest rates and high inflation. However, rather than mechanically extending maturities for listed manufacturers that cannot even pay interest and thereby inflating distress, restructuring programs are needed to support preemptive debt workouts, downsizing, and business transitions. To increase housing supply, it is worth bringing even the use of greenbelts, a gift handed down by prior generations, into the public debate.

The task after the semiconductor boom is to protect those whom the windfall does not reach. If the three-lows boom of the 1980s created a middle class, this boom could, if mishandled, further widen gaps between asset holders and non-asset holders, between semiconductor workers and non-workers, and between high- and low-credit borrowers.