DBRS Morningstar Publishes Final European RMBS Insight: Spanish Addendum

DBRS Morningstar finalised its “European RMBS Insight: Spanish Addendum” (the Methodology).

This Methodology presents the criteria for which Spanish residential mortgage-backed securities (RMBS) ratings, and, where relevant, Spanish covered bonds ratings and Spanish nonperforming loan (NPL) transactions are assigned and/or monitored.

The Methodology supersedes the prior version published on 6 July 2021 and is effective as of 26 April 2022.

DBRS Morningstar updated its Spanish loans scoring approach (LSA), which DBRS Morningstar has recalibrated using an updated modelling sample that includes approximately 120,675 Spanish mortgages sourced from both internal and external datasets. The resulting LSA consists of 22 model parameters from 15 variables. Some of the variables employed in the LSA are new or have substantially changed compared with the LSA in the previous version of the Methodology.

Similar to the LSA in the previous version of the Methodology, which included two qualitative variables (the Spanish Underwriting Score and the Adverse Performance), the updated LSA also has two judgmental variables: the Portfolio Score and the Restructure Quality (the latter only for portfolios of mortgage loans that contain restructured loans).

The Portfolio Score is an ordinal variable with three values: high, neutral, and low. Factors considered for setting the Portfolio Score of a pool of mortgage loans include the historical performance of the vintages securitised in the pool, portfolio characteristics that affect the creditworthiness and that are not captured by the other variables of the LSA, as well as the accuracy and completeness of the loan-by-loan tape provided for the analysis.

A Restructured Score variable is applied to loans that have undergone a restructuring since the origination date. These are normally assumed to be riskier and are more likely to default. The Restructured Score will be a function of the following factors: (1) the quality of the restructuring, (2) the payment due, (3) the current interest rate, and the (4) past borrower performance (if the loans has been ever in arrears in the past two years). Of these factors, the quality of the restructuring arrangement is a judgmental variable (Restructure Quality) that has three values—high, neutral, or low—where high is the best and low is the worst. This would depend on the length of the time the loans in the portfolio have been performing since restructuring, the actual payment received compared with the payment due over the past two years, and redefault rates on the loans post-restructuring.

The updated LSA includes a variable that takes into account the seasoning of each loan. Loans that are 48 months seasoned will receive a benefit adjustment with the maximum benefit up to 120 months. After a loan reaches 120 months seasoning, no further benefit is applied. Repayment loans benefit from both the seasoning adjustment and the decrease in the loan-to-value (LTV) adjustment as their age increases. At the same time, the updated LSA also envisages a higher probability of default (PD) adjustment for mortgages granted between 2004 and 2008 because of the more aggressive lending policies under which those vintages were originated.

The updated LSA introduces a lower PD adjustment for loans that are fixed-rate loans for life. Fixed-rate loans have become more popular in the Spanish mortgage market since 2015 and are considered to be less risky compared with floating-rate loans. In addition, there is a new variable to account for higher PDs of second-lien loans.

Under the updated Spanish LSA, DBRS Morningstar now calculates the margin for each loan in the portfolio by subtracting the 12-month Euribor rate as of the last reset date to the interest rate of the loan.

The updated LSA also introduces specific PD adjustments for employed, protected (civil servants), self-employed, and unemployed borrowers. For borrowers who do not report income, the variable missing income will lead to a higher PD adjustment.

Furthermore, the total income in place under the previous version of the Methodology has been replaced by the combined loan-to-indexed income ratio (CLTI), which is a better predictor of the loan's affordability and eventual likelihood of default. A higher total income is generally correlated with low CLTI values at the time of origination, which is an indicator of sound origination policies. DBRS Morningstar will index the income using disposable income on a regional level from Eurostat.

The updated LSA continues to include variables that are part of the previous version of the Methodology where the coefficient has changed due to the recalibration of the model. These variables include PD adjustments for (1) loans granted to foreign borrowers, (2) loans granted with the purpose of purchasing a residential property, (3) current indexed LTV, (4) credit lines, and (5) government-subsidised loans (Vivienda de Protección Oficial).

In addition, the updated LSA will no longer include the following variables: (1) a total income between EUR 1 and EUR 100,000, (2) a current balance when lower than EUR 10,000, (3) a loan maturity longer than 30 years, (4) a property type that is not a house, flat, condominium, or multifamily unit, and (5) adverse performance.

DBRS Morningstar also updated the Spanish Delinquency Migration Matrix (DMM) using the same dataset employed to revise the Spanish LSA by computing the average roll rates observed in the loan-level data from 2013 to Q4 2017. The Spanish DMM is now based on 13 risk segments.

DBRS Morningstar also updated its house price indexation and market value decline rates to reflect data through Q4 2020 and change the ceiling correlation to 30% from 24%.

The distressed sale assumptions (DSD) remain unchanged, but DBRS Morningstar intends to apply lender-specific DSD assumptions where data for such specific lender is available.

The updated methodology does not change the loss given default floors, which remain the same as in the previous version of the Methodology.

DBRS Morningstar deems these updates to be material. DBRS Morningstar currently rates 84 classes of notes across 40 Spanish RMBS transactions. No downgrades are expected as a result of these changes. The update could have a potential positive rating impact on a small number of transactions. No rating impact is expected on Spanish covered bonds, structured credit, and NPL securitisation transactions.

No comments were received during the request for comment (RFC) period for the Spanish addendum.

All comments received during the RFC period have been published to the DBRS Morningstar website, except in cases where confidentiality is requested by the respondent.

Notes:
A description of how DBRS Morningstar considers ESG factors within the DBRS Morningstar analytical framework can be found in the DBRS Morningstar Criteria: Approach to Environmental, Social, and Governance Risk Factors in Credit Ratings at https://www.dbrsmorningstar.com/research/373262.

The DBRS Morningstar Sovereign group releases baseline macroeconomic scenarios for rated sovereigns. DBRS Morningstar analysis considered impacts consistent with the baseline scenarios as set forth in the following report: https://www.dbrsmorningstar.com/research/384482/baseline-macroeconomic-scenarios-application-to-credit-ratings.

DBRS Morningstar methodologies are publicly available on its website www.dbrsmorningstar.com under Methodologies & Criteria.

For more information on this methodology or on this industry, visit www.dbrsmorningstar.com or contact us at [email protected].