정책 이슈

To summarize the $350 billion package created by Korea and the US in 2025 in one sentence, the US is using tariffs on Korean products as a weapon, telling Korea to 'bring cash if you want to maintain market access.' In exchange for lowering tariffs on automobiles and parts from 25% to 15%, Korea promised $200 billion in cash for strategic industries and $150 billion for shipbuilding cooperation. The deadline for selecting projects is January 2029, when Trump's term ends, and the cash portion is to be deposited in installments with an annual limit of $20 billion.

Calling this structure 'investment cooperation' is the language of diplomatic documents. Looking at the content, a more accurate term would be 'conditional remittance' secured by market access. The logic presented by the US is simple: since Korea has a surplus from exports to the US, either a portion of that surplus should be collected in the form of tariffs, or it should be invested in projects designated by the US. The outward appearance of this binary choice is negotiation. Internally, if you refuse, it's tariffs; if you accept, it's cash.

This choice is close to an imposition, as revealed by three facts.

First, Korea does not have the final authority to choose what to buy. According to the MOU of November 14, 2025, the $200 billion projects are recommended by an investment committee chaired by the US Secretary of Commerce and chosen by the US President. The consultative committee, chaired by the Korean Minister of Industry, can provide opinions in advance. The phrase 'only commercially reasonable investments are recommended' is included. However, it is clear who interprets the threshold for recommendation and who gives the final approval. The party paying the money and the party choosing the goods are different.

Second, it is not an equal joint venture in terms of profit distribution. Before repayment of principal and interest, it's 50-50 between Korea and the US. After repayment, it's 1 for Korea and 9 for the US. There is a safeguard that allows for adjustment of the ratio if recovery seems difficult within 20 years. Nevertheless, the basic design is clear. The high-risk initial phase is split evenly, while the profitable later phase is taken by the US. In September 2026, the US went a step further. Reports emerged that the US was weakening the umbrella-type SPV (a structure where losses from one project are covered by profits from another) that Korea proposed, and trying to change it to project-specific profit and loss. If that happens, 'commercial reasonableness' will remain just a phrase, and the holes from failing projects will remain in Korea's account.

Third, the projects favored by the US and those that Korea can handle are at odds. As of September 2026, the first project being discussed, the Texas Encinal Gas Combined Cycle, has a total project cost of approximately $22.3 billion and a capacity of about 6.3 GW. There is a justification for AI data center power demand. However, the project cost increased from $19.8 billion to $22.3 billion, after the US requested $25 billion. The eight large nuclear power plants in the US, which are being discussed as the second project, cost approximately $15 billion each, totaling about $120 billion. The US is talking about Westinghouse AP1000 and a minority stake (5-10%), while Korea is talking about at least two KEPCO APR-1400s and a 15% stake and board membership. Alaska LNG, a key project for Trump, is considered by Korea to have low commercial viability. Nevertheless, it repeatedly comes up at the negotiating table. Although there are reports that the Louisiana LNG terminal, Midwest carbon capture, and spent nuclear fuel recycling have been removed from consideration, even with the remaining three projects, the $200 billion limit is tight.

여기에 선집행 압박이 겹친다. 미국은 연 200억 달러 한도의 상당액을 2026년 말까지 먼저 보내라고 한다. 한국은 올해 소액, 내년부터 본집행을 상정했다. 이재명 대통령이 9월 18일 “동의하기 어려운 부분이 있다”고 한 지점이 바로 회수·배분·손실이다. 같은 주 미 국무부는 외교장관 회담 자료에서 신속한 이행을 앞세우고 반도체 협력을 거론했다. 협상이 늘어지면 관세 카드가 다시 보인다. MOU에도 납입을 안 하면 이자를 미국이 가져가고, 관세가 오를 수 있다고 적혀 있다.

The funding story also aligns with this picture. The government says it "will not touch foreign exchange reserves." That's true. Official procurement comes from interest and dividends earned by the Bank of Korea and KIC from managing foreign currency assets, and the shortfall is covered by government-guaranteed dollar bonds from the Korea-US Strategic Investment Corporation (KUIC). The corporation's capital of 2 trillion won is not principal but an operational and guarantee buffer. However, even if the principal is not reduced, the profits that would have increased the original holdings will go to US projects. What Korea is putting forth is not "idle money" but the future increase in public foreign currency assets. This is a different account from the National Pension Service or corporate retained earnings.

The $150 billion for shipbuilding is a bit different. It's not a lump-sum cash payment, but rather direct investment by Korean companies in the US, guarantees, and ship financing. HD Hyundai Heavy Industries, Samsung Heavy Industries, Hanwha Ocean, and the Export-Import Bank of Korea, Korea Development Bank, Korea Trade Insurance Corporation, and Korea Ocean Business Corporation formed a consultative body in June 2026. Hanwha's Philadelphia shipyard is symbolically mentioned. This axis offers Korean companies the opportunity to secure work. The government has also explained that the profits from the shipbuilding sector will accrue to Korean companies. Therefore, covering the entire package with the single word 'extortion' is an exaggeration. However, for the $200 billion cash axis, that expression is not an exaggeration in many respects. The pressure for selection rights, allocation, advance execution, and transfer of bad assets is concentrated there.

So why is it difficult for Korea to refuse? Economic figures alone cannot explain it. This is because the political and military structure gives weight to the tariff card.

The ROK-US alliance is not the backdrop for a trade agreement, but the foundation of the negotiations. The joint fact sheet from the same period included a 3.5% increase in defense spending as a percentage of GDP, the purchase of $25 billion worth of US weapons by 2030, and expanded support for US Forces Korea. The US included its support for Korea's peaceful uranium enrichment and spent nuclear fuel reprocessing procedures, and approval for the construction of nuclear-powered attack submarines in the same document. Trade and security are on the same table. If tariffs are rejected, automobile and parts exports will be hit first. Following that are heavier items such as alliance management, wartime operational control, extended deterrence, and deterrence against North Korea. The Korean government's repeated statement that 'we won't do it if there's no commercial rationality' means that while it understands this foundation, it cannot lift the entire foundation.

Trump-style deals make this foundation even more explicit. They view alliances not as a communal obligation, but as a contract where the bill is due immediately. Japan's promise of a larger package also put pressure on the Korean negotiating team. The comparison that 'other countries have paid, so if Korea doesn't, tariffs will return' is at play. Domestic politics are the same. If tariffs rise, export-oriented large corporations and their partners, employment, and growth rates are seen first. The long-term failures of the investment package are seen later. The election cycle makes it difficult to overcome the latter.

So the question, 'Should we succumb to this extortion?' is not a moral issue, but a question of who pays the price of refusal. The price of refusal is immediate. The 15% tariff could revert to 25%, semiconductor treatment could falter, and the door to shipbuilding cooperation could narrow. The clauses regarding nuclear submarines and enrichment in the alliance document could also be politically reinterpreted. The price of acceptance is dispersed: a limit of $20 billion per year, over 10 years, foreign exchange operating profits, public bonds, and the failure of individual projects. It's less noticeable. That's why nations often lose in such deals.

To make it a choice rather than submission, the criterion is whether Korea adheres to the mechanisms it has already secured in documents in actual contracts. This includes the explicit mention of the $20 billion annual limit, the quantification of commercial rationality—whether the revenue returning to Korea covers the principal and interest—the maintenance of an umbrella-type SPV, Korean project managers, the participation of Korean companies in equipment and EPC, the exclusion of projects with no visible recovery like Alaska, and the refusal of advance execution. If this list is omitted, the 15% tariff will be received, but the $200 billion will become seed money for US domestic political projects with a low probability of recovery. This is also the point where President Lee Jae-myung put the brakes on the working agreement. It is true that putting the brakes on increases tariff risks. If the brakes are not applied, only the form of the risk changes.

The minimum sovereignty Korea can exercise in this deal is not to say 'no' entirely, but to insist on not doing certain projects. The current structure does not have the power to overturn tariffs and alliances all at once. However, it is not the destiny of an alliance to pour public foreign exchange profits into every project chosen by the US president. An alliance is a transaction of stationing and deterrence, not a contract to cover the losses of Texas power plants and Alaska pipelines with Korean finances.

In summary, the US intends to recoup the benefits of South Korean exports through either tariffs or cash. The form of this recoupment has taken the shape of MOUs, special laws, and public corporations. South Korea, facing both tariffs and security concerns, finds it difficult to close the door. However, accepting commercially unviable projects after opening the door is not an alliance, but rather a transfer. The issue for September 2026 is already on that borderline. Whether or not there is capitulation will be determined not by the large sum of 350 billion, but by the recoupment clauses in the first and second contracts to be signed in the future.