INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
Jiseob Kim, Fellow
1. Issues The rise in household debt has accelerated since August 2014 and higher interest rates and an economic slowdown have been projected on growing uncertainties at home and abroad, aggravating concerns over households’ financial stability. Household credit has risen at an increasing rate (above 10%) on eased household loan regulations, widening the gap between the growths in household debt and income. Additionally, long-term interest rates have risen on looming risk factors both in Korea and abroad, including political uncertainties in the US, and the mounting downward pressure will likely weigh down on households’ financial stability.
Household Debt and Income Growth Rate (%)
15 Household credit (YoY change)
Nominal GNI (YoY) 10
5
0 I
II
III
2010
IV
I
II
III
2011
IV
I
II
III
2012
IV
I
II
III
IV
I
2013
II
III
2014
IV
I
II
III
2015
IV
I
II
2016
Note: Blue-shaded region indicates post-mortgage loan deregulation. Source: Bank of Korea.
1
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
This paper analyzes the trajectory of household debt since the easing of household loan regulations in August 2014 and examines changes in households’ financial stability as a result of the macroeconomic impact. For this, the changes in household debt before and after the easing of LTV and DTI regulations are compared and analyzed using the Survey of Household Finances and Living Conditions.
2. Characteristics of Changes in Household Debt To closely examine the traits of the recent rise in household debt, this paper used Statistics Korea’s 2012-2015 Survey of Household Finances and Living Conditions (balanced panel data). The 2015 survey includes data on household assets and debt as of 31 March 2015 and income and expenditure for 2014; enabling the analysis of households’ financial conditions before and after deregulation.
The share of households with more debt than a year ago is decreasing while their average increase in debt is rising, hinting at a climb in the over-indebted. The share of households with more debt than a year ago receded to 29.0% in early 2015 from 35.5% in early 2013 while their total debt increase rose an annual average of 44.70 million won from 36.40 million won. ╺ Accordingly, households with growing (shrinking) debt had an average debt of 66.00 million won (94.00 million won) in early 2014 and 110.00 million won (64.00 million won) in early 2015.
Rising
aggregate
household
debt
along
with
a
shrinking
share
of
households with increased borrowing suggest that there is a rise in the proportion of the over-indebted.
2
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
Share of Households with Increasing/Decreasing Debt (%)
50
Average Debt Increase/Decrease of Households with Increasing/Decreasing Debt
Households with decreasing debt 40
Households with increasing debt
(10,000won)
8,000 6,000
Average debt decrease of households with decreasing debt Average debt increase of households with increasing debt
4,000 30 2,000 20 0 10
-2,000
0
-4,000 2012~13
2013~14
2014~15
2012~13
2013~14
2014~15
Source: 2012-2015 balanced panel in the Survey of Household Finances and Living Conditions, Statistics Korea.
With the increase in household debt largely driven by real estate purchases and the downward turn in household loans for business operations, an increasing flow of capital in the real estate market may be a possibility. According to the survey, approx. 55% of the 2014-2015 increase in household debt was for residential homes and real estate purchases, marking a rise of 23% in 2013-2014. ╺ Specifically, 41 trillion won from the 74 trillion won increase in total household debt in 2014 was spent on homes and realty. ╺ This may be a consequence of the boom in the real estate market caused by low interest rates and deregulation of mortgage loans. * Household debt in this analysis is defined by the sum of mortgage and credit loans as the survey includes data on the amount by purpose for mortgage and credit loans only.
Loans for realty purchases account for more than half of the total household debt while those for business operations are on a moderate decline. ╺ Share of household loans for housing and realty (%): (‘13) 52.1 → (‘14) 50.6 → (’15) 51.0
3
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
╺ Share of household loans for business operations (%): (‘13) 28.5 → (‘14) 27.9 → (‘15) 26.2
Changes in Household Debt by Purpose
Total Household Debt by Purpose (%)
(%) 100
80
14.8 60
14.2 55.3 8.1 7.6
40
20
0 2014~15
For home and realty purchases For jeonse or monthly rent deposits
2013
2014
2015
For others For living expenses For business operations (incl. farming funds)
Source: 2012-2015 balanced panel in the Survey of Household Finances and Living Conditions, Statistics Korea.
Meanwhile, households’ financial stability remains mostly intact as the 30s-40s cohorts and the upper 20% income stratum lead the recent rise in household debt. Contrary to two years ago, when household borrowing was observed evenly across all age groups, it is currently concentrated in specific age groups. ╺ About 80% of the increase in household debt in 2014-2015 was attributed to households with householders in their 30s-40s.
Of the increase in 2014-2015, the share of households in the upper 20% income stratum rose significantly to 51% while that of lower income households in the first and second income quintiles declined. The fact that the recent increase in household debt is driven by stable income earners in their 30s-40s and by high-income households implies that the quality of household debt has not deteriorated substantially since the easing of real estate finance regulations.
4
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
The following presents the analysis on the recent changes in the debt service ratio (DSR) and loan-to-value (LTV) ratio to assess households’ financial stability following the deregulation of the real estate finance market.
Changes in Household Debt by Age Group
Changes in Household Debt by Income Quintile (%)
50
2012-13
2012-13 2013-14
40
2014-15
(%)
60 50 40
2013-14 2014-15
30 30
20
20
10
10 0
0
-10
-10 30s
40s
50s
60s
70s
1 quintile
2 quintile
3 4 quintile quintile
5 quintile
Source: 2012-2015 balanced panel in the Survey of Household Finances and Living Conditions, Statistics Korea.
3. Changes in Households’ Financial Stability on Eased Real Estate Finance Regulations The DSR remains at a similar level following the alleviation of LTV and DTI regulations in August 2014, implying that there will no sudden increases in households’ repayment burden. Despite the larger increment in household debt as a result of the eased real estate finance regulations, the DSR in 2015 (21.2%) advanced only slightly from 2013 (20.4%) on increasing nominal household income and share of long-term loans and a declining loan rate. ╺ Nominal household income growth (%): (‘13) 2.1 → (‘14) 3.4 → (‘15) 1.6 ╺ Household loan rate (balance, %): (‘13) 4.54 → (‘14) 4.02 → (‘15) 3.35
5
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
╺ Weighted average remaining maturity of banks’ mortgage loans (year): (‘12) 14.5 → (‘13) 15.4 → (‘14) 16.8 → (‘15) 17.7 → (Mar. ‘16) 17.5
According to an analysis on the impact of increasing nominal household income and loan duration and declining loan rates on the DSR, the DSR would have risen by an average 2.9%p if conditions in 2013 continued onto 2015 (see appendix for more details about calculation). ╺ The change in the DSR is attributed to the impact from increasing nominal household income (55%) and loan duration (30%) and declining interest rates (15%).
Yearly DSR Distribution
DSR Simulation
Note: The DSR simulation shows the distribution of a hypothetical DSR under the assumption that the conditions in household income, loan rate and duration in 2015 are as in 2013, Source: 2013-2015 data from the 2012-2015 balanced panel in the Survey of Household Finances and Living Conditions, Statistics Korea.
However, if the current internal and external risk factors materialize and deal a blow to income and interest rates, households are expected to face a much heavier burden of principal and interest payments. ╺ As global economic risks, such as the political unrest in the US and economic slowdown in China, resurface, the Korean economy is increasingly likely to experience weaker growth (income growth). Additionally, long-term interest rates are on a rapid upward trajectory, which is expected to weigh down the burden of principal and interest payments.
6
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
╺ Under the assumption that a shock has caused household income to drop by 5% and interest rates to increase by 1.0%p at the current rate of household debt growth, the average principal and interest, as of 2015, would climb 14%, from 11.4 million won to 13 million won. ╺ The average DSR of households in this case would increase from 21.2% to 25.5%.
The above suggests that a shock to household income and the loan rate may rapidly undermine households’ financial stability. ╺ In particular, the DSR in this paper does not include the principal paid under collective and interest-only bullet loans and the impact from the rapidly mounting
household
households’
burden
debt of
(125
principal
trillion and
won)
interest
in
2015-2016,
payments
may
therefore, have
been
underestimated. * The survey includes questions on the total principal and interest payment during the previous year. As collective and interest-only bullet loans do not require a principal payment before maturity, survey respondents may have answered “none” to the question on paid principal.
Changes in Short- and Long-term Interest Rates (%)
2.0 5-year government bond rate
3-year government bond rate
1.8
Call rate 1.6
1.4
1.2
1.0 6
7
8
9
10
11
2016
Source: Bank of Korea
Meanwhile, the LTV ratio inched up following the deregulation of mortgage loans, but the possibility of a relatively sharp increase in response to an income or house price shock cannot be ruled out.
7
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
The share of households with a LTV ratio of above 60% increased to 6.5% in 2015 from 5.2% in 2013 on eased mortgage loan regulations, mainly led by households in the upper 50% income stratum.
LTV Ratio Distribution by Income Quintile ▎Upper 50%
▎Bottom 50%
Source: 2014-2015 data from the 2012-2015 balanced panel in the Survey of Household Finances and Living Conditions, Statistics Korea.
By age group, the increase in the LTV ratio was mainly led by those in their 30s and 40s with relatively high income, meaning that is no severe deterioration in the quality of household debt.
LTV Ratio Distribution by Age Group ▎Under 60 yrs
▎Over 60 yrs
Source: 2014-2015 data from the 2012-2015 balanced panel in the Survey of Household Finances and Living Conditions, Statistics Korea.
8
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
If housing prices were to tumble to the 2013 level, the share of households with a LTV ratio of above 60% would rise from the current 6.5% to 7.2%. ╺ A drop in housing prices to the March 2013 level (3.4% drop based on Kookmin Bank's housing price index), based on the LTV ratio in March 2015, would increase the share of households with a LTV ratio of above 60% by 0.7%p.
On the other hand, a 5% drop in housing prices resulting from the aforementioned risks would raise the share of marginal households with a LTV ratio of above 60% to 10.2%. ╺ Based on the increased inflow of capital in the real estate market, the shock of falling
housing
prices
could
significantly
boost
the
number
of
marginal
households in terms of the LTV ratio. * This scenario assumes that the mortgage growth rate remains at 1.3% and housing prices have fallen 5%. ╺ As the materialization of the aforementioned internal and external risks cannot be ruled out in the midst of a prolonged economic slowdown, efforts are urgently needed to preemptively manage and supervise marginal households.
LTV Ratio Simulation ▎Housing Prices Return to the 2013 Level
▎Housing Prices Drop 5%
Source: 2014-2015 data from the 2012-15 balanced panel in the Survey of Household Finances and Living Conditions (2012-2015), Statistics Korea.
9
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
Besides, it was found that households with high LTV ratios use their loans for business operations, debt repayment and living expenses, rather than asset accumulation. Two groups were formed－those directly affected by the eased LTV ratio regulations and those who are not－and the characteristics of the debt increase for each group were analyzed. ╺ Group A: Households that have actively utilized the increase in the borrowing limit on eased LTV regulations, marking a LTV ratio of above 0%, below 60% in March 2014 but above 60% in March 2015. ╺ Group B: Households that have not been directly affected by the eased LTV regulations, marking a LTV ratio of above 0%, below 60% in both 2014 and 2015 but a higher LTV ratio in 2015 than in 2014. ╺ Accordingly, Group A can be understood as those whose debt soared extremely high after deregulation, compared to Group B.
The share of households with a sharp increase in the LTV ratio following deregulation (Group A) is not very high; their loan expansion was mainly for business operations, debt repayment and living expenses. ╺ Of households with a LTV ratio of above 0 in 2014, only 4.4% were classified as Group A while 13.6% were classified as Group B. ╺ Given that households in Group A used a large portion of the loans on business operations, debt repayment and living expenses, all of which do not include response assets, they may be vulnerable to unexpected shocks.
Share of Households with Rising LTV
Average Changes in Household Debt by Purpose (%)
20
6,000
(10,000won)
5,000 4,000
15
3,000 2,000 10
1,000 0
5
-1,000
0 Group A
10
Group B
Group A Group B For others Debt repayment For living expenses For business operations (incl. farming funds) For jeonse or monthly rent deposits For home and realty purchases
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
4. Summary and Policy Implications Although the deregulation of the real estate finance market has not diminished households’ financial stability to any significant degree, a shock such as a rise in the interest rate may weigh down heavily on households in the short-run. Despite the rapid increase in the aggregate household debt, the rise was mainly led by high-income households with relatively stable income and response assets. As such, households’ financial stability has not been significantly affected. However, due to the recent rapid rise in interest rates, households’ financial stability may turn negative in the short-run as a result of unexpected shocks. ╺ If household income is negatively affected or interest rates rise, the DSR and LTV ratio will swing upward in the short-term and households’ general economic activities will be restricted. ╺ Households with a sharp increase in the LTV ratio post-deregulation may be particularly vulnerable to external shocks as the majority of their loans are used for business operations, debt repayment and living expenses.
Preemptive measures are needed to enhance marginal households’ financial stability in preparation for internal and external shocks as mortgage regulations are strengthened to curb the increase in household loans. Aggregate household debt has hovered far above income growth since the easing of regulations on DTI and LTV ratios. As such, the regulations should be reversed to previous levels and collective loans should be placed under much stricter supervision to curb the increase in household loans. ╺ Substantial repayments are currently not required for collective loans, meaning that a much heavier repayment burden could weigh down on households later.
At the same time, efforts are also needed to prevent additional increases in the share of households with extremely high DTI ratios or rapidly increasing LTV ratio.
11
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
╺ Households with high DTI and LTV ratios are most likely high-risk borrowers (over-indebted persons, multiple debtors). Therefore, efforts are needed to induce a systematic payment of the principal and interest and to apply tighter screening measures for additional loans. ╺ Considering that households with variable rate loans are exposed to shocks to both income and interest rates, consistent efforts are needed to assist them in transitioning to fixed rate loans.
12
INSIDabcdef_:MS_0001MS_0001
Characteristics and Implications of Rising Household Debt: Pre-and Post-deregulation of Mortgage Loans
▎Appendix ▎ DSR Simulation
This section presents the calculation of the impact of a declining loan rate and prolonged loan duration on the DSR. The Survey of Household Finances and Living Conditions lacks data on the respective interest rate, duration and repayment type of each household loan. As such, it is assumed that all financial debts are contracted on identical loan terms (fixed rate, amortization). ╺ Each household is assumed to amortize their debt by paying the same amount of principal and interest every year.
* Here, denotes the total debt held by household at time ; represents the principal and interest that household has to pay evenly every year; represents the loan rate at time and the loan duration at time .
The average of actual and experimental DSR are set equally by adjusting the loan maturity ( ) after substituting the Bank of Korea’s household loan rate and households’ financial debt balance. ╺ The household loan rate (balance, ) provided by the Bank of Korea and households’ financial debt balance ( ) from the Survey have been substituted. ╺ has been adjusted so that the average DSR calculated using the Survey data is set equal to the average experimental DSR calculated using the adjusted . * With the , an experimental principal and interest payment ( ) is determined, and based on this, households’ experimental DSR can be calculated.
According to the analysis, the loan maturity has increased slightly since 2013. ╺ Based on the calculation, the maturity ( ) was 6 years and 1 month in 2015, 5 years and 9 months in 2014 and 5 year and 8 months in 2013.
Lastly, the experimental DSR is calculated by substituting with the 2013 household income, loan rate and duration in the 2015 financial debt.
13
