S&P Global Ratings believes default rates of pools of receivables in Japanese RMBS transactions we rate correlate to some extent with Japan's real GDP, land prices, unemployment rate, and results of the Bank of Japan's Tankan survey of business sentiment. This study covers Japanese RMBS transactions backed by housing loans and condominium investment loans originated by private-sector lenders. We also recognize some correlation between default rates and the Nikkei average stock price index and the consumer price index (CPI).

Our analysis found relatively high correlation between the default rate and three factors. Real GDP (one year lead), the Urban Land Price Index of six largest city areas (year-on-year, one year lead, average of all categories), and the Tankan survey (one year lead, all size/all sectors/actual results) showed negative correlation with the 12-month moving average default rate, with correlation coefficients of -0.64, -0.66, -0.58, respectively. Meanwhile, the unemployment rate, had a positive correlation, with a correlation coefficient of +0.57. Borrowers of loans backing Japanese RMBS are individuals, and the macroeconomic environment surrounding them materially affects their ability to repay loans.

Our analysis covers pools of loan receivables underlying RMBS we rate, including fully repaid transactions. We took the default data of these transactions since 2003 as well as macroeconomic factors in the same period and examined their correlation to determine which macro factors affect the default rate. We also analyzed the default rate's sensitivity to changes in macroeconomic factors by measuring the volatility of the factors during the spread of coronavirus, when the default rate remained relatively volatile.

Correlation With Macroeconomic Factors

We calculated and analyzed correlation coefficients between the default rate of underlying assets in rated RMBS and key macroeconomic factors (see table 1). We also examined correlation coefficients among macroeconomic factors (see table 2).

Table 1

Table 2

Economic trend (real GDP, stock prices, Tankan survey)

The correlation coefficient between Japan's real GDP and the 12-month moving average default rate of the underlying loan pools was -0.64. The negative correlation is attributable to a fall in real GDP and a rise in the loan default rate during the economic slowdown from 2008 to 2010, as well as a moderate increase in real GDP and a fall in the loan default rate in and after 2010.

Real GDP is one of the macroeconomic factors showing the prevailing economic trend. In an economic slowdown, it affects obligors' creditworthiness through deterioration of business performance at obligors' places of employment and loss of jobs. However, a decline in real GDP does not immediately affect obligors' debt servicing capability, in our view; there is a lag before it affects obligors. Correlation between real GDP and the RMBS default rate is stronger when one-year lead GDP is adopted (see chart 1).

The Nikkei index and Tankan survey are other macroeconomic factors that indicate economic trends. We observed a certain level of correlation between real GDP (one-year lead) and the Nikkei index (year on year, one-year lead) and Tankan (business conditions sentiment diffusion index (DI), all enterprise/all industries/actual, one-year lead), with correlation coefficients of 0.46 and 0.51, respectively (table 2). Accordingly, the Nikkei index and Tankan both correlate negatively with the 12-month moving average default rate of underlying loans, with correlation coefficients of -0.55 and 0.58, respectively. However, we observed the opposite trend when the coronavirus spread in 2020. This is because continued global monetary easing pushed up the Nikkei index as the default rate of underlying loans went up similarly (chart 2).

Chart 1

Chart 2

Chart 3

Land prices

We have observed similar movements between the default rate of underlying loan pools and land prices. We measured a -0.66 correlation coefficient between the Urban Land Price Index of Japan's six largest city areas (average of all categories, year-on-year) and the 12-month moving average default rate. The correlation is particularly distinctive between 2008 and 2010 when a fall in land prices coincided with a rise of the default rate (see chart 4).

Collateral value is an important factor in examining housing loan applications. When land prices are high, obligors have options to either repay the residual amount of loans using proceeds of sales of properties or to refinance loans with lower interest rates to ease repayment burdens. As a result, obligors are more likely to evade default.

Since the global financial crisis in 2008 and 2009 wound down, Japan's land prices have been mostly stable without sudden fluctuations, which has contributed to stable performance of RMBS transactions, in our view.

Chart 4

Unemployment

We calculated the correlation coefficient between Japan's unemployment rate and the 12-month moving average default rate of underlying loans to be +0.57, which we believe indicates some correlation. Because borrowers of these loans are individuals, we expect a reduction in individual income due to unemployment would significantly impact the default rate of the underlying loans.

Japanese housing loan lenders generally focus on a borrowers' ability to repay, because recourse loans dominate Japan's housing loan market. To confirm a borrower's income is stable, lenders check not only their annual income but also their employment status, time in a job, and the amount of other loans they have, if any.

In our view, correlation between the unemployment rate and the default rate of the underlying loans points directly to the relationship between the performance of these loans and underwriting practices in the Japanese housing loan market. In addition, Japan has a low unemployment rate by global comparison because of its strict regulations on layoffs. We consider this one of the reasons for stable default performance of the underlying assets.

Chart 5

Inflation

Japan's CPI and the 12-month moving average default rate of the underlying loans had a correlation coefficient of -0.55.

Japan's CPI has not shown much volatility over the long term. However, prices have risen basically three times in the last 20 years. First, during the economic recovery that ended just before the 2008 financial crisis; second, in 2014 when the consumption tax rose; and third, currently, due to rising raw material costs and a weaker yen. Wages have not risen much when prices were on the rise. As a result, real wages fell in each of these three occurrences, weakening the ability of borrowers in general to repay their loans.

The CPI negatively correlated with the default rate of the underlying loans--that is, the default rate tended to fall as prices rose. Japan's CPI showed a similar movement to real GDP and the Nikkei index, with prices rising moderately in economic recovery phases. As noted above, real GDP and the Nikkei index negatively correlate with the default rate. This explains the same results for the CPI, in our view.

Chart 6

Interest rates

The yield on 10-year Japanese government bonds (JGBs) and the 12-month moving average default rate did not show an evident correlation, with a correlation coefficient of +0.28.

It is generally believed that housing loan interest rates are highly correlated with long-term interest rates for fixed-rate housing loans and policy interest rates for floating-rate housing loans. However, our analysis found no correlation between the yield on 10-year JGBs and the default rate of underlying loans of RMBS transactions. During the period of this survey, from 2003 to the present, Japan's policy rates have not changed materially. If interest rates on housing loans surge due to higher policy rates, repayment amounts on housing loans will increase materially. This may affect the default rate of the underlying loans.

Chart 7

Other macroeconomic factors

Comparison of macroeconomic factors other than those mentioned above with the default rate of underlying loans in RMBS transactions showed relatively high correlation between the real effective exchange rate and the 12-month moving average default rate, with a correlation coefficient of +0.62. The real effective exchange rate incorporates the inflation rate into the effective exchange rate, which is a measure of relative currency strength.

In the case of Japanese RMBS transactions, borrowers of underlying assets are usually Japanese citizens and permanent residents, and all repayments are made in Japanese yen. Therefore, exchange rates do not affect the performance of underlying assets. We believe the correlation coefficient is high due to the interaction of several macroeconomic factors including exchange rates, total value of trade, and inflation.

Chart 8

Sensitivity Analysis

As a result of our sensitivity analysis, a 10% fall in real GDP increased the annual default rate of the underlying loans by 0.55 ppt. A 1 ppt rise in Japan's unemployment rate pushed up the default rate by 0.68 ppt. In addition, a 50% fall in Nikkei index raised the default rate by 0.67 ppt.

Table 3

The default rate of underlying loans in Japan's RMBS transactions has remained stable over the long term. Even in the economic slowdown between 2008 and 2010, the rise in default rate was limited. However, when the coronavirus infection spread in 2020, the annualized default rate of underlying loans rose to 0.48% in September 2020, despite the rise being temporary. In this analysis, we studied the impact of macroeconomic factors on the default rate of loans backing RMBS transactions by analyzing changes in the default rate during the coronavirus pandemic since 2020 as well as fluctuations of macroeconomic factors before and after the spread of infection.

In this sensitivity analysis, we used as a reference point changes in the following macroeconomic indicators during the spread of coronavirus infection.

Table 4

In the sensitivity analysis, we calculated the impact of a change in a single macroeconomic factor on the default rate of underlying assets. In reality, several factors may change simultaneously in times of economic slowdown. This would cause a greater impact on the default rate.

Notes

In this report, figures include rating actions by S&P Global Ratings and S&P Global SF Japan Inc. (SPSF). SPSF is a registered credit rating agency under Japan's Financial Instruments and Exchange Act (FIEA) but is not registered as a Nationally Recognized Statistical Rating Organization (NRSRO) under U.S. Laws. Therefore, the credit ratings assigned by SPSF are Registered Credit Ratings under FIEA but are not Credit Ratings issued by an NRSRO under U.S. laws. The data in this study includes both public ratings and confidential ratings.

Related Criteria

• Global Methodology And Assumptions: Assessing Pools Of Residential Loans, Jan. 25, 2019

Related Research

• Default, Transition, and Recovery: 2022 Annual Japanese Structured Finance Default And Rating Transition Study, March 27, 2023
• Japan Structured Finance Outlook: Economic Growth To Shield Performance, Jan. 12, 2023