Korea's banks found a loan with no credit risk in it, and the guarantee funds are paying

The 85 to 90 percent norm at the Korea Credit Guarantee Fund and the Korea Technology Finance Corporation is not a rounding convention. That residual slice is the reason a bank reads the borrower's books, and both institutions say it exists precisely to keep loss risk on the lender and hold down bad guarantees ^[6]. Take the slice to zero and the asset changes character: the guarantor repays the whole principal on default ^[2], so the bank is left underwriting nothing and earning a spread on quasi-public credit. Kim Seok-ki of the Korea Institute of Finance puts it plainly, saying banks have little incentive to assess business viability themselves under the current system ^[14].

The funds' own numbers show what that costs before the new vintages are even tested. The combined guarantee balance grew 2.9 percent from 2023 to last year ^[16], while combined repayments made on borrowers' behalf rose 41.7 percent to 4.0039 trillion won, the highest in ten years ^[9]. Payouts expanded roughly fourteen times faster than the book they came from ^[20]. Measured against the balance, the loss rate went from about 3.15 percent to about 4.34 percent, a rise of 1.19 percentage points ^[17]. None of that damage comes from the full guarantees now being signed: the productive finance agreements date from April ^[3] and the Hana programme from the 23rd of this month ^[1], so the deterioration is happening on a book written mostly under partial cover.

Seoul has already ruled on the instrument once. After regional credit guarantee foundations watched their subrogation ratio climb from 1.01 percent in 2021 to 5.66 percent last year, the government moved in June to ban full guarantees there in principle, allowing them only where policy needs require ^[10][11]. The national funds are going the other way, with 100 percent cover appearing repeatedly inside special-contribution agreements framed as policy support ^[7].

On stated program size, the full-guarantee push is still modest: the six agreements named by Sedaily total about 2.86 trillion won ^[18], near 3.1 percent of the 92.3 trillion won outstanding ^[19]. The trend line matters more than the stock. Guaranteed loans already made up 18.2 percent of the 42.6578 trillion won in small and mid-sized enterprise lending the five major banks booked as productive finance ^[12], and an official at one commercial bank defends the mechanism on the grounds that startups without collateral cannot borrow on their own credit at all ^[13]. That defence covers the guarantee. It does not explain why the last ten to fifteen points of it had to go, on an instrument the paper describes as a crisis tool ^[15].

There is a fuse on this. The 1.4 trillion won of productive finance cover runs at 100 percent for the first three years only ^[3], and the 610 billion won regional growth tranche with IBK is on the same three-year clock ^[4]. In year four those loans revert to partial cover, and the banks inherit a share of credit files they never built.