An analysis has found that households that purchased homes with aggressive borrowing are particularly vulnerable to interest-rate shocks. The top 10% of households whose debt-service burden relative to income increased most sharply during the home-purchase process would see their delinquency rate climb by 0.81 percentage points if the base rate rose 1 percentage point.
According to the BOK issue note “Assessing Household Debt Risks Using a Household Database” released by the Bank of Korea on the 7th, the interest-rate sensitivity of so-called “young-kkeul” (soul-gathering) high-leverage home-purchasing households was markedly higher than that of ordinary households. The study is notable as the first attempt to comprehensively track debt, income, and spending at the household level rather than the individual level.
High-Leverage Households: Delinquency Risk Spreads Across the Entire Family
The Bank of Korea built a household database tracking the debt, assets, income, and spending of 2.06 million households on a monthly basis using KCB credit information. The rationale is that the income and debt of other household members, such as spouses, must be considered together to properly assess actual repayment capacity.
Researchers identified households that had no housing assets between 2023 and 2025, purchased a home, and took out a new mortgage. They then classified the top 10% of households by the increase in debt-service burden relative to income after the home purchase as high-leverage home-purchasing households.
Under a scenario of a 1 percentage point rate increase, the delinquency rate for these households rose by 0.81 percentage points after 12 months. This was larger than the increase for existing homeowners (0.52 percentage points) or households that purchased homes without excessive borrowing (0.64 percentage points). It also exceeded the average for all homeowners (0.56 percentage points).
Jang Hoon, head of the Bank of Korea’s Monetary and Financial Research Division, explained: “High-leverage home-purchasing households have only about 30% of their members in the first and second income quintiles. Despite the majority being middle- and high-income earners, they suffer impacts from rate hikes nearly identical to those experienced by low-income households.” He added: “We focused on these households because there is a need to examine the risks associated with so-called ‘young-kkeul’ borrowing.”
The possibility of credit-risk transmission among family members was also confirmed. In high-leverage home-purchasing households with two or more indebted members, the probability that another member would also become delinquent on a separate loan within 12 months after one member defaulted was 8.8%. This is 1.9 times the rate for existing homeowners (4.6%) and 3.4 percentage points higher than for ordinary home-purchasing households (5.4%).
When the scope was broadened to all indebted households, the impact of rate increases was relatively limited. Under a typical scenario of a 0.25 percentage point rate hike, the delinquency rate for indebted households was estimated to rise by 0.27 percentage points from the existing 3.35%.
However, disparities across income groups were pronounced. The delinquency rate for the first income quintile rose by 0.48 percentage points from 5.45% before the rate hike to 12 months after, while the second quintile climbed 0.40 percentage points from 4.38%. Self-employed and corporate-representative households also rose by 0.32 percentage points from 4.47%, showing greater sensitivity to interest rates than other occupational groups.
Jang assessed: “When rates rose by 0.25 percentage points, the overall increase in delinquency rates was not excessively large compared to existing trends. We believe the situation can generally continue on a stable path.” He clarified, however, that the 1 percentage point rate increase scenario is a stress test assuming a crisis situation, not a future rate forecast.
11.1% of Indebted Households Face Consumption Constraints from Debt-Service Burdens
The study also confirmed that household debt is suppressing consumption. The Bank of Korea’s analysis of the relationship between household card spending and total debt-service ratio (DSR) found that consumption begins to decline with statistical significance once debt-service payments exceed 46% of income. At lower DSR levels, consumption increases as borrowing provides additional funds, but beyond a certain threshold, the burden of principal and interest payments grows and consumption contracts.
As of 2025, an estimated 11.1% of indebted households exceeded this threshold. This means more than one in ten households with debt are in a situation where they must cut consumption while using nearly half their income for debt-service payments.
The problem is that the consumption capacity of low-income households continues to deteriorate. While the proportion of all indebted households exceeding a 46% DSR fell from 12.3% in 2024 to 11.1% last year, the first income quintile actually saw an increase from 13.1% to 14.5% over the same period. Compared to 11.4% in 2021, this represents a 3.1 percentage point increase over four years.
Based on this analysis, the Bank of Korea also offered policy implications. During periods of rising rates, it should closely monitor the debt-repayment burden and financial conditions of low-income households and, if necessary, consider targeted policy support to reduce the repayment burden on vulnerable households. The central bank also emphasized the need to manage high-leverage home-purchasing households to prevent their distress from spreading to other household members and sectors, while consistently operating monetary and macroprudential policies to curb expectations of rising home prices.
This study is significant in that it extends existing individual-level household debt analysis to the household level. The Bank of Korea explained that the household database, built on data submitted by financial institutions and credit information, offers high data accuracy and can complement the Household Finance and Welfare Survey, which is based on interview surveys and produced annually.