Almost all Japanese residential mortgage-backed securities (RMBS) appear built to withstand the relatively moderate domestic economic conditions we foresee. Significantly more will tumble down the rating scale if conditions deteriorate sharply beyond our expectations.

We estimate a 5 percentage point rise in defaults or a 10% decline in real estate prices would hurt only about 2% of classes we rate. More severe stress scenarios in which default rates and real estate prices deteriorate simultaneously would affect about 10% of classes we rate.

The macroeconomic environment for Japanese RMBS transactions changed significantly in 2022. Japanese consumer prices began to rise. As of January 2023, Japan's consumer price index (CPI) had climbed 4.2% (excluding fresh food items) from a year earlier. We expect it to remain volatile in the immediate future. A hike in consumer prices without higher wages may hurt obligors' abilities to repay loans.

In this analysis, we focus on two factors. First is the impact of rising consumer prices on default rates. Second, we examine fluctuations in real estate prices, which closely match the performance of RMBS transactions. We look at how deterioration in these factors affects our ratings on existing RMBS transactions.

An insufficient level of credit enhancement to absorb higher default rates and lower real estate prices may adversely affect our ratings on RMBS transactions. In this analysis, we found transactions with either of the following characteristics would risk downgrades:

• Transactions in pro rata payment in which credit enhancement does not increase over time; or
• Transactions in sequential payment that have a limited buildup of credit enhancement having been issued recently.

A Drop In Real Wages Hurts Obligors' Ability To Repay

We forecast inflation will ease to 2.8% in Japan in 2023, and we expect it to continue subsiding gradually to about 1.0%-2.0%. Some companies and industries have implemented or are considering higher pay rises than previously. However, generally, if pay rises are insufficient to cover higher consumer prices, obligors' loan repayment capabilities are likely to deteriorate. Consumer prices started rising in Japan in 2022, due in part to a drop in the value of the yen on top of elevated demand and soaring raw material costs as the pandemic eased. The CPI, excluding fresh food items, jumped 4.2% year on year by January 2023, its highest rise since 1981. Price hikes have been moderate in Japan by global comparison. However, the country has generally limited wage growth. Thus, individual incomes, which property owners depend on as obligors of housing loans, are declining. Real wages, which incorporate price fluctuations, fell 4.1% in January 2023 (provisional basis) from a year earlier.

Chart 1

Table 1

Central banks often raise policy interest rates when prices rise. Policy rates rose in the U.S. and Europe in and after 2022. In Japan, the Bank of Japan (BOJ) announced on Dec. 20, 2022, that it would raise the cap on 10-year government bond yields, the benchmark for long-term interest rates, to about 0.5% from about 0.25%. We have not observed any significant change in interest rates on housing loans. However, Japan's CPI rose to 3% during 1990-1993. In that period, the short-term prime rate in Japan, which influences floating-rate housing loans, climbed to 8%. Thus, if the short-term prime rate rises in the future, we believe it may adversely affect floating-rate housing loans.

Chart 2

Regarding repayment capabilities of obligors in RMBS transactions, we consider inflation a factor that elevates default rates. Higher consumer prices may weaken obligors' capabilities to repay loans if unaccompanied by wage growth. In addition, obligors' repayment burdens will expand further if they have floating-rate loans. Table 2 shows our calculation of a change in an obligor's debt-to-income (DTI) ratio given two factors: a fall in repayment ability due to higher consumer prices (calculated as a decline in annual income), and an increase in repayments due to higher interest rates. We found a 1.5 percentage point rise in interest rates on housing loans and a 10% decline in annual incomes would produce a 10 percentage point increase in DTI. This would translate to a 5 percentage point rise in the assumed cumulative default rate at the 'AAA' level, according to our calculation based on our criteria for rating RMBS, "Global Methodology And Assumptions: Assessing Pools Of Residential Loans," published Jan. 25, 2019.

Table 2

Official Land Prices Edge Up

Official land prices in Japan dropped amid the coronavirus pandemic in 2021, both nationwide and in the three major metropolitan areas. However, prices started rising in 2022. Nationwide land prices have trended upward moderately each year since 2017 except in 2021. Historically, official land prices swung hugely during the bubble economy in the late 1980s and the recessions that followed. Of note, nationwide land prices fell a sharp 8.7% in 1993.

We consider real estate prices a major factor in analyzing an obligors' credit quality. A decline in the collateral liquidation value of properties is likely to push up default rates, in our view. This is because it not only reduces the amount that can be collected on collateral properties in the event of default but also affects obligors' abilities to fully repay loans by selling collateral properties when having difficulty continuing repayments.

Chart 3

Assumptions

Defaults rise as consumer prices rise

We set stressed cumulative default rates by imposing an additional 5 percentage points on rates currently assumed for relevant rating categories. This levels with the increase in default rate we calculated above, assuming a rise in consumer prices. In addition, it is about four times the actual cumulative default rate of underlying loan receivables in all private-sector RMBS transactions we rate (include those redeemed fully), which is about 1.2% (weighted average based on amount outstanding).

Real estate prices drop

We stressed recoveries from collateral properties using 10% and 15% discounts on existing assumed recovery rates for relevant rating categories. We based these haircuts on the largest year-on-year declines in official land prices since 1975: about 9% for the national average, and about 15% for the three major metropolitan areas. In this analysis, a drop in collateral value not only affects our assumed loss severity rates (1-recovery rate), but also additionally pressures our assumed cumulative default rates above. Loan-to-value (LTV) ratios affect cumulative default rates. Therefore, a fall in the value of collateral properties will increase LTV ratios, triggering a further rise in cumulative default rates.

Stress Scenarios: Results Of Analysis

Impact of stress scenarios on default rates and loss severity rates

We established the following five scenarios based on the combination of a rise in default rate and a drop in collateral value. The table below shows the average cumulative default rate and loss severity rate at the 'AAA' level for rated RMBS transactions in each scenario. As noted, scenarios assuming a drop in collateral value also assume a rise in cumulative default rates, as well as in loss severity rates.

Prior to imposing stress (the no stress scenario), the cumulative default rate is 22.3% at the 'AAA' level. Leaving the scenario untouched, the cumulative default rate at the 'BBB' level calculated in accordance with our criteria for rating RMBS transactions is about 7.1%. Meanwhile, in scenario 5, the most severe scenario assuming an additional 5 percentage point uptick in default rate and a 15% drop in collateral value, the cumulative default rate is 7.2% higher than that in the no stress scenario. This indicates the level of stress under scenario 5 is equivalent to adding the cumulative default rate for the 'BBB' level to that for a current rating.

Table 3

Impact On Ratings

As a result of our analysis, we found a 5 percentage point rise in the cumulative default rate (scenario 1) and a 10% drop in collateral value (scenario 2) would have limited impact on our ratings. We would likely downgrade only about 2% of rated classes in both cases. Classes rated 'AAA', which make up most of the total number of classes we rate, indicated high resiliency against stress. Less than 1% would likely be downgraded and by one notch only. Other classes, albeit numbering only a few, would face more severe impact than classes rated 'AAA' both in percentage and range of downgrade. Scenario 3, which assumes a larger decline in collateral value, had considerably harsher impact than scenario 2. In this scenario, we would likely downgrade about 8% of 'AAA' rated classes by one notch.

As we applied more stress, we found potential for downgrades in about 10% of classes we rate in scenario 4 and in about 15% of classes we rate in scenario 5. Furthermore, in scenario 4 and scenario 5 about 4% and about 8% of 'AAA' rated classes, respectively, would face possible downgrades by two or more notches. Most classes rated 'A' or below would be downgraded by three or more notches.

Table 4

In typical RMBS transactions, occurrences of loan defaults and declines in value of collateral properties are key factors that adversely affect our ratings. Such negative factors can be mitigated using a structural mechanism by which credit enhancement rises as repayment of rated debt progresses. In particular, seasoned transactions with a sequential payment structure would benefit substantially from such a mechanism. In our view, classes with the following characteristics tend to be less resilient to negative factors:

• Classes in pro rata payment with no accumulation of credit enhancement over time;
• Classes issued relatively recently that have only seen a limited buildup of credit enhancement since closing;
• Classes suffering from reduced credit enhancement due to loan defaults; and
• Classes subject to recent rating actions and with little buffer between actual credit enhancement and that required for the revised ratings.

Our scenario analysis also found transactions with characteristics described above would likely be downgraded. Most classes with potential for downgrades are those issued recently, which have a limited buildup of credit enhancement. For example, in scenario 5, which assumes the highest levels of stress--a 5 percentage point increase in default and a 10% decline in collateral value--the cumulative default rate rises 7 percentage points and the loss severity rate rises about 11 percentage points (see table 2). Thus, the net loss rate (after accounting for recoveries from collateral properties) would rise about 5 percentage points. Meanwhile, a 5 percentage point buildup in credit enhancement would normally take about five years, although it may vary depending on transaction structure and performance, such as prepayment. Therefore, if the default rate rises or collateral value falls within a short time since a transaction closed, we believe this will have a relatively greater impact on our ratings on such transactions.

Rising consumer prices have adversely affected obligors' capacity to repay, in our view. However, we think it is unlikely that the performance of underlying assets in Japanese RMBS transactions we rate will deteriorate sharply. This is because while we forecast Japan's inflation rate will be 2.8% in 2023, we expect it to continue subsiding gradually to about 1.0%-2.0% in and after 2024. We also do not anticipate a major change in the unemployment rate. This report is intended to illustrate how Japanese RMBS will be affected if the macroeconomic environment deteriorates sharply beyond our expectations.

A rise in default rates and a drop in collateral value have a relatively large impact on RMBS transactions. In this scenario analysis, we looked at effects of these factors on our ratings on existing RMBS transactions. Our actual rating actions may differ from this analysis, as our rating analysis incorporates other factors, such as interest rate scenarios and prepayments.

Scope Of Analysis And Assumptions

• This analysis covers all RMBS transactions outstanding as of Nov. 30, 2022. We applied our criteria "Global Methodology And Assumptions: Assessing Pools Of Residential Loans," published Jan. 25, 2019.
• We excluded transactions, such as repackaged securitizations, for which creditworthiness is dependent solely on factors other than the creditworthiness of the asset pool, such as counterparty ratings.
• Base-case scenarios are based on our assumptions as of our most recent rating analysis on the classes. They do not reflect the updated status of repayments and performance.
• When real estate prices are declining, we calculate cumulative default rates assumed under stress scenarios by multiplying assumed cumulative default rates for current ratings by additional stress owing to a rise in LTV ratios. The additional stress level varies depending on the seasoning of each transaction after closing.

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.

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
• Economic Outlook Asia-Pacific Q2 2023: China Rebound Supports Growth, March 26, 2023
• Japan Structured Finance Outlook: Economic Growth To Shield Performance, Jan. 12, 2023
• Japan Private-Sector RMBS Performance Watch: In Stability? Dec. 5, 2022
• JHF RMBS Performance Watch November 2022: Toward Normality, Nov. 30, 2022