Our bi-weekly Opinion provides you with latest updates and analysis on major capital market and financial investment industry issues.
Structural Reform of Korea’s Household Debt Management
Publication date Jul. 22, 2025
Summary
The Korean government currently manages household debt by setting annual debt growth targets in line with nominal GDP growth projections. However, this framework suffers from two structural limitations. First, the high level of uncertainty in nominal GDP forecasts often leads to frequent revisions, weakening policy consistency and eroding market predictability. Second, the management approach based on short-term projections results in a procyclical policy stance—overly tightening credit supply during downturns and enabling excessive leverage in periods of expansion—thereby amplifying business cycle volatility. To mitigate these risks, it is worth considering a revised framework that anchors household debt management to medium-term trend nominal GDP growth. Notably, the current policy mix, comprising macro-level aggregate debt caps and micro-level debt service ratio (DSR) rules, has shown limited effectiveness in stabilizing household debt levels. As a complementary policy tool, a progressive macroprudential levy tied to loan size could be adopted to encourage borrowers to internalize the externalities generated by excessive leverage.
Household debt management is a key policy priority for ensuring financial and macroeconomic stability. As of the end of 2024, Korea’s household debt-to-GDP ratio stood at over 90%, ranking fifth among major economies according to the Bank for International Settlements (BIS). Although the ratio has declined gradually since the third quarter of 2021, it remains elevated, highlighting the need for a more systematic and sustainable policy framework. In response, the Korean government has sought to align household debt growth with nominal GDP growth. However, this approach has been increasingly challenged by recent shifts in the economic environment and rising uncertainty in economic forecasts. This article analyzes the structural limitations of Korea’s current household debt management strategy and outlines policy reforms aimed at mitigating cyclical volatility. In particular, it examines the growing procyclicality of household debt and the rising uncertainty over nominal GDP growth projections. Drawing on this analysis, this article proposes a revised medium-term household debt management framework and recommends complementary regulatory tools to internalize the externalities associated with excessive household borrowing.
Current household debt management framework and its limitations
The government currently anchors annual household debt growth to projected nominal GDP growth, in an effort to stabilize the household debt-to-GDP ratio. According to the Financial Services Commission (FSC)’s “2025 Household Debt Management Plan” (released on February 27, 2025), household debt growth for the year will be capped at 3.8%, in line with the official nominal GDP forecast. The underlying rationale behind this strategy is to ensure debt sustainability by preventing household debt from outpacing economic growth.
However, the current framework that links household debt growth to annual nominal GDP projections exhibits structural weaknesses. Most notably, setting policy targets in line with early-year forecasts may not remain valid over the course of the year. Economic projections are constantly revised, and in an environment of heightened uncertainty, the divergence between initial estimates and actual outcomes could widen considerably. Nominal GDP growth reflects the sum of real GDP growth and the GDP deflator, both of which are particularly subject to forecast uncertainty. As shown in Figure 1, forecast errors for real GDP growth have increased markedly since 2018, driven in part by the growing influence of non-economic shocks such as geopolitical tensions and public health crises.1) Traditional economic models, primarily built on interrelations among economic variables, often fail to accurately capture the spillover effects of exogenous disruptions, such as pandemics, wars, and natural disasters. Moreover, the increasing complexity of structural shifts, ranging from global supply chain restructuring and intensifying technological competition to climate change mitigation, has reshaped the global economic landscape, further diminishing the accuracy of traditional forecasting models.
Forecasting the GDP deflator—the second component of nominal GDP growth—also poses significant challenges. In the absence of benchmark projections, the GDP deflator is often estimated based on consumer price index (CPI) forecasts. However, this method has clear limitations due to the persistent divergence between the two indicators. The GDP deflator reflects price changes across all domestically produced goods and services, including consumer goods, exports, and capital goods. In contrast, the CPI measures price changes in a more limited scope of goods and services consumed primarily by households.2) Although both serve as inflation indicators, differences in statistical coverage result in a weak correlation between them, as illustrated in Figure 2. Notably, the gap between the two indicators reached 1.8 percentage points in 2024.
Since nominal GDP growth forecasts are susceptible to uncertainty in both real GDP growth and the GDP deflator, a policy framework that depends on annual nominal GDP projections may lead to frequent revisions of target levels and the resulting readjustments in household debt management policy. This, in turn, undermines policy consistency and reduces predictability for market participants. For financial institutions, in particular, the predictability of policy direction is essential in shaping lending portfolios and risk management strategies. Frequent policy shifts can complicate strategic planning. For households, abrupt tightening of lending regulations may restrict timely access to credit for home purchases or business activities. Both cases can compromise financial market efficiency and distort resource allocation.
A more fundamental limitation of the current framework lies in its procyclical policy stance, which risks amplifying business cycle volatility. When nominal GDP growth is expected to decline during economic downturns, household debt growth targets must also be adjusted downward, potentially further deepening the contraction in credit supply. This dynamic can exacerbate the economic slowdown and delay economic recovery. In contrast, in periods of economic expansion, stronger nominal GDP forecasts raise the permissible pace of household debt growth, fueling a spike in asset prices and excessive leverage. Over the long run, such a policy stance may undermine financial stability and heighten adjustment pressures in future business cycles.
Introducing a medium-term framework for household debt management
A medium-term approach to household debt management shares similarities with central banks’ monetary policy frameworks. Rather than reacting to short-term price fluctuations, most central banks typically pursue price stability over the medium term, recognizing that excessive responsiveness to short-term shocks could escalate business cycle volatility. Household debt management should adopt a similar strategy, prioritizing medium- to long-term financial stability and balanced economic growth.
To improve the existing framework, policymakers should consider replacing annual nominal GDP growth forecasts—with their high volatility as illustrated in Figure 3—with a medium-term estimate of nominal GDP growth trends as the basis for setting household debt targets. Trend nominal GDP growth can be derived using statistical techniques that filter out temporary shocks and seasonal effects, thereby offering a more accurate reflection of the economy’s underlying growth capacity. One viable option would be to set a three-year average target for household debt growth at a level slightly below the trend growth rate (e.g., 80–90% of trend nominal GDP growth).
A key advantage of adopting a medium-term management approach is its potential to dampen cyclical volatility. During downturns, when actual growth tends to fall below the trend growth rate, more accommodative credit conditions are permitted, thereby contributing to mitigating the economic slowdown. Conversely, during periods of expansion, when actual growth exceeds the trend growth rate, a tighter debt management policy would be implemented to help reduce the risk of overheating. In this way, this mechanism would function as an automatic stabilizer that adjusts the policy stance in line with macroeconomic conditions. Moreover, a medium-term management framework would enhance predictability for both households and financial institutions. Even if short-term projections fluctuate throughout the year, a stable medium-term policy stance would provide greater consistency in household debt management.
The need for a macroprudential stability levy on household lending
The current policy framework, combining macro-level debt caps with micro-level, borrower-specific debt service ratio (DSR) requirements, has proven insufficient to contain household debt risks, necessitating additional policy tools. As income inequality widens, DSR-based regulation may exacerbate disparities in credit access, potentially deepening structural divergence in housing prices between the Seoul metropolitan area and other regions. Microdata from the Survey of Household Finances and Living Conditions reveal that, as of 2024, the top 10% of income earners accounted for 28.7% of total mortgage credit. The recent surge in housing prices and mortgage lending in the metropolitan area can be attributed to high-income borrowers who expanded their lending before tighter third-phase stressed DSR rules were imposed, buoyed by strong expectations of rising housing prices.
As such, rational borrowing decisions made by individuals do not necessarily translate into the broader financial stability since household debt generates externalities.3) Just as one person’s noise may disrupt an entire neighborhood, a single borrower’s loan increases aggregate debt levels, reducing available credit for the wider population and generating negative externalities. Furthermore, during downturns, central banks’ concerns over excessive household debt could delay interest rate cuts, indirectly harming financially vulnerable borrowers with high debt burdens.4) These externalities are amplified when household lending is increasingly concentrated among high-income borrowers, as a result of DSR-based regulation. To address these risks, individually rational borrowing decisions should be aligned with socially optimal debt levels. This requires a policy mechanism that incentivizes borrowers to internalize the broader effects of their lending.
In response to overheating in the real estate market and a surge in household debt, the Korean government introduced a set of policy measures on June 27, including a limit on maximum mortgage loan amounts in the Seoul metropolitan area and designated regulated zones.5) While these measures may be effective as an emergency intervention aimed at curbing large-scale borrowing, they rely on a uniform, volume-based loan caps that potentially restrict credit access for genuine end-users. To address the externalities of excessive borrowing more systematically, financial authorities could consider implementing a progressive macroprudential stability levy. This price-based regulatory tool would impose a gradually increasing charge on household loans once the loan amount exceeds a defined threshold, regardless of whether the borrowing is for home purchase or lease financing (jeonse). The levy is expected to minimize market distortion by ensuring access to credit for end-users. Given that a small number of large-value loans account for a significant share of total household lending,6) this approach would enable more precise targeting by imposing the levy on high-value borrowers while leaving the majority of small-scale borrowers unaffected. In doing so, it would enhance both social equity and the effectiveness of macro-level aggregate debt management. A well-designed levy could complement the current nominal GDP-based debt cap and borrower-level DSR rules and better reflect borrower behavior, laying the groundwork for a more integrated macroprudential policy framework.
1) Quarterly GDP growth forecast errors, measured by the Root Mean Square Error (RMSE), where a higher RMSE indicates greater inaccuracy, increased from 0.14 during 2012–2017 to 0.35 during 2018–2024, representing a 2.5-fold rise.
2) In highly export-dependent economies such as Korea, forecast errors may be even larger. For instance, rising prices of high-bandwidth memory (HBM), driven by increased AI investment, can affect the GDP deflator while having no impact on the consumer price index (CPI).
3) Bianchi, J., 2011, Overborrowing and Systemic Externalities in the Business Cycle, American Economic Review vol. 101, no. 7, pp. 3400–3426.
4) In August 2024, despite slowing inflation and weak domestic demand, the Bank of Korea (BOK)’s Monetary Policy Board kept the base interest rate unchanged, citing concerns over rising housing prices and growing household debt. As a result, the anticipated rate cut was deferred to the following policy meeting in October. In August 2024 alone, household debt in the financial sector rose by KRW 9.8 trillion, providing empirical support for the BOK’s concerns.
5) Financial Services Commission, June 27, 2025, Authorities Introduce Measures to Strengthen Household Debt Management Centered on Seoul Metropolitan Area, Press release.
6) According to microdata from the 2024 Survey of Household Finances and Living Conditions, households with mortgage loans of KRW 600 million or more comprised just 3.0% of all indebted households but accounted for 19.2% of total mortgage debt. In contrast, those with mortgage loans under KRW 300 million represented 88.2% of all indebted households, yet held only 59.2% of total mortgage debt.
Current household debt management framework and its limitations
The government currently anchors annual household debt growth to projected nominal GDP growth, in an effort to stabilize the household debt-to-GDP ratio. According to the Financial Services Commission (FSC)’s “2025 Household Debt Management Plan” (released on February 27, 2025), household debt growth for the year will be capped at 3.8%, in line with the official nominal GDP forecast. The underlying rationale behind this strategy is to ensure debt sustainability by preventing household debt from outpacing economic growth.
However, the current framework that links household debt growth to annual nominal GDP projections exhibits structural weaknesses. Most notably, setting policy targets in line with early-year forecasts may not remain valid over the course of the year. Economic projections are constantly revised, and in an environment of heightened uncertainty, the divergence between initial estimates and actual outcomes could widen considerably. Nominal GDP growth reflects the sum of real GDP growth and the GDP deflator, both of which are particularly subject to forecast uncertainty. As shown in Figure 1, forecast errors for real GDP growth have increased markedly since 2018, driven in part by the growing influence of non-economic shocks such as geopolitical tensions and public health crises.1) Traditional economic models, primarily built on interrelations among economic variables, often fail to accurately capture the spillover effects of exogenous disruptions, such as pandemics, wars, and natural disasters. Moreover, the increasing complexity of structural shifts, ranging from global supply chain restructuring and intensifying technological competition to climate change mitigation, has reshaped the global economic landscape, further diminishing the accuracy of traditional forecasting models.
Since nominal GDP growth forecasts are susceptible to uncertainty in both real GDP growth and the GDP deflator, a policy framework that depends on annual nominal GDP projections may lead to frequent revisions of target levels and the resulting readjustments in household debt management policy. This, in turn, undermines policy consistency and reduces predictability for market participants. For financial institutions, in particular, the predictability of policy direction is essential in shaping lending portfolios and risk management strategies. Frequent policy shifts can complicate strategic planning. For households, abrupt tightening of lending regulations may restrict timely access to credit for home purchases or business activities. Both cases can compromise financial market efficiency and distort resource allocation.
A more fundamental limitation of the current framework lies in its procyclical policy stance, which risks amplifying business cycle volatility. When nominal GDP growth is expected to decline during economic downturns, household debt growth targets must also be adjusted downward, potentially further deepening the contraction in credit supply. This dynamic can exacerbate the economic slowdown and delay economic recovery. In contrast, in periods of economic expansion, stronger nominal GDP forecasts raise the permissible pace of household debt growth, fueling a spike in asset prices and excessive leverage. Over the long run, such a policy stance may undermine financial stability and heighten adjustment pressures in future business cycles.
Introducing a medium-term framework for household debt management
A medium-term approach to household debt management shares similarities with central banks’ monetary policy frameworks. Rather than reacting to short-term price fluctuations, most central banks typically pursue price stability over the medium term, recognizing that excessive responsiveness to short-term shocks could escalate business cycle volatility. Household debt management should adopt a similar strategy, prioritizing medium- to long-term financial stability and balanced economic growth.
To improve the existing framework, policymakers should consider replacing annual nominal GDP growth forecasts—with their high volatility as illustrated in Figure 3—with a medium-term estimate of nominal GDP growth trends as the basis for setting household debt targets. Trend nominal GDP growth can be derived using statistical techniques that filter out temporary shocks and seasonal effects, thereby offering a more accurate reflection of the economy’s underlying growth capacity. One viable option would be to set a three-year average target for household debt growth at a level slightly below the trend growth rate (e.g., 80–90% of trend nominal GDP growth).
The need for a macroprudential stability levy on household lending
The current policy framework, combining macro-level debt caps with micro-level, borrower-specific debt service ratio (DSR) requirements, has proven insufficient to contain household debt risks, necessitating additional policy tools. As income inequality widens, DSR-based regulation may exacerbate disparities in credit access, potentially deepening structural divergence in housing prices between the Seoul metropolitan area and other regions. Microdata from the Survey of Household Finances and Living Conditions reveal that, as of 2024, the top 10% of income earners accounted for 28.7% of total mortgage credit. The recent surge in housing prices and mortgage lending in the metropolitan area can be attributed to high-income borrowers who expanded their lending before tighter third-phase stressed DSR rules were imposed, buoyed by strong expectations of rising housing prices.
As such, rational borrowing decisions made by individuals do not necessarily translate into the broader financial stability since household debt generates externalities.3) Just as one person’s noise may disrupt an entire neighborhood, a single borrower’s loan increases aggregate debt levels, reducing available credit for the wider population and generating negative externalities. Furthermore, during downturns, central banks’ concerns over excessive household debt could delay interest rate cuts, indirectly harming financially vulnerable borrowers with high debt burdens.4) These externalities are amplified when household lending is increasingly concentrated among high-income borrowers, as a result of DSR-based regulation. To address these risks, individually rational borrowing decisions should be aligned with socially optimal debt levels. This requires a policy mechanism that incentivizes borrowers to internalize the broader effects of their lending.
In response to overheating in the real estate market and a surge in household debt, the Korean government introduced a set of policy measures on June 27, including a limit on maximum mortgage loan amounts in the Seoul metropolitan area and designated regulated zones.5) While these measures may be effective as an emergency intervention aimed at curbing large-scale borrowing, they rely on a uniform, volume-based loan caps that potentially restrict credit access for genuine end-users. To address the externalities of excessive borrowing more systematically, financial authorities could consider implementing a progressive macroprudential stability levy. This price-based regulatory tool would impose a gradually increasing charge on household loans once the loan amount exceeds a defined threshold, regardless of whether the borrowing is for home purchase or lease financing (jeonse). The levy is expected to minimize market distortion by ensuring access to credit for end-users. Given that a small number of large-value loans account for a significant share of total household lending,6) this approach would enable more precise targeting by imposing the levy on high-value borrowers while leaving the majority of small-scale borrowers unaffected. In doing so, it would enhance both social equity and the effectiveness of macro-level aggregate debt management. A well-designed levy could complement the current nominal GDP-based debt cap and borrower-level DSR rules and better reflect borrower behavior, laying the groundwork for a more integrated macroprudential policy framework.
1) Quarterly GDP growth forecast errors, measured by the Root Mean Square Error (RMSE), where a higher RMSE indicates greater inaccuracy, increased from 0.14 during 2012–2017 to 0.35 during 2018–2024, representing a 2.5-fold rise.
2) In highly export-dependent economies such as Korea, forecast errors may be even larger. For instance, rising prices of high-bandwidth memory (HBM), driven by increased AI investment, can affect the GDP deflator while having no impact on the consumer price index (CPI).
3) Bianchi, J., 2011, Overborrowing and Systemic Externalities in the Business Cycle, American Economic Review vol. 101, no. 7, pp. 3400–3426.
4) In August 2024, despite slowing inflation and weak domestic demand, the Bank of Korea (BOK)’s Monetary Policy Board kept the base interest rate unchanged, citing concerns over rising housing prices and growing household debt. As a result, the anticipated rate cut was deferred to the following policy meeting in October. In August 2024 alone, household debt in the financial sector rose by KRW 9.8 trillion, providing empirical support for the BOK’s concerns.
5) Financial Services Commission, June 27, 2025, Authorities Introduce Measures to Strengthen Household Debt Management Centered on Seoul Metropolitan Area, Press release.
6) According to microdata from the 2024 Survey of Household Finances and Living Conditions, households with mortgage loans of KRW 600 million or more comprised just 3.0% of all indebted households but accounted for 19.2% of total mortgage debt. In contrast, those with mortgage loans under KRW 300 million represented 88.2% of all indebted households, yet held only 59.2% of total mortgage debt.