Korea's mortgage market now rations by queue: 79 lending curbs at 13 banks by August 12

The gap between two ways of counting the same tightening is where the mechanism sits. Collapse curbs that hit several loan products or sales channels simultaneously into single decisions and the tally falls to 47, against 21 in the same window last year ^[4]. Set that against the product-level count and you get 1.68 restrictions per underlying decision this year, up from 1.38 ^[15]. Each time a bank moved, it took more of the shelf with it.

Ninety percent of the year's measures, 71 of 79, arrived after the government said in April it would hold annual household loan growth near 1.5% ^[5][6][14]. The calendar is the part that reaches actual buyers. Last year the squeeze came late, with 30 of 63 measures taking effect after November ^[8]. This year 57 measures, 72% of the total, were packed into June and July ^[9], close to one a day across those two months ^[18], and inside the transaction season rather than after it. Housing carried the weight: the two mortgage categories together make up 58% of everything counted ^[16].

What changed for a rejected borrower is the missing second door. SC First Bank, Suhyup Bank, Kwangju Bank and iM Bank imposed no credit-loan restrictions at all last year; this year they blocked applications arriving through outside platforms and cut loan limits ^[10]. The channels that used to absorb overflow are running their own quotas.

Song's reading of the result is that a policy meant to prevent a year-end lending cliff instead produced a midsummer first-come, first-served rationing system, hitting owner-occupiers who needed balance-payment loans to complete a purchase ^[12]. He also says the authorities lifted the target from about 1.5% to 3% within four months, and calls it a policy failure in which the government conceded its own demand forecasts and design were wrong ^[13]. Both halves matter to anyone underwriting Korean housing risk. The 79 measures were calibrated to a ceiling that no longer applies, and they already exceed last year's full-year total of 63 by a quarter, with four and a half months left to run ^[17].

The pricing consequence follows from the rationing. A quota-managed book does not clear through rate, and critics in the account say banks adjusted criteria repeatedly to hit annual volume targets ^[19]. The variable that decides an application is therefore where a given bank stands against its own allocation on the day the file arrives, which is a number the applicant cannot see and the seller cannot price into a closing date. Cross-checking helps only a little: the Seoul Economic Daily's separate tally of 73 criteria changes lands near the 79, so the direction is not an artifact of one office's counting method ^[3].