M AY 2 0 1 7
METHODOLOGY
European RMBS Insight:
Spanish Addendum
European RMBS Insight: Spanish Addendum
DBRS.COM
2
Contact Information
Table of Contents
Sebastian Hoepfner
Vice President
Global Structured Finance
+44 20 7855 6663
[email protected]
Executive Summary
3
Spanish Mortgage Scoring Model
3
Spanish Segmentation and Delinquency Migration Matrices
7
Vito Natale
Head of EU RMBS & CBs
Global Structured Finance
+44 20 7855 6649
[email protected]
Spanish Correlation
8
Spanish House Price Model
9
Rehanna Sameja
Vice President, EU RMBS
Global Structured Finance
+44 20 7855 6677
[email protected]
Spanish Cash Flow Assumptions
10
Kali Sirugudi
Vice President, EU RMBS
Global Structured Finance
+44 20 7855 6609
[email protected]
Asim Zaman
Assistant Vice President, EU RMBS
Global Structured Finance
+44 20 7855 6626
[email protected]
Belen Bulnes Meneses
Senior Financial Analyst, EU RMBS
Global Structured Finance
+44 20 7855 6699
[email protected]
Juan Jose Mendez
Senior Financial Analyst
Global Structured Finance
+44 (020) 3356 1550
[email protected]
Christian Aufsatz
Managing Director
Global Structured Finance
+44 20 7855 6664
[email protected]
Claire Mezzanotte
Group Managing Director
Global Structured Finance
+1 (212) 806 3272
[email protected]
DBRS is a full-service credit rating agency
established in 1976. Spanning North
America, Europe and Asia, DBRS is
respected for its independent, third-party
evaluations of corporate and government
issues. DBRS’s extensive coverage of
securitizations and structured finance
transactions solidifies our standing as a
leading provider of comprehensive, in-depth
credit analysis.
All DBRS ratings and research are available
in hard-copy format and electronically on
Bloomberg and at DBRS.com, our lead
delivery tool for organized, web-based,
up-to-the-minute information. We remain
committed to continuously refining our
expertise in the analysis of credit quality and
are dedicated to maintaining objective and
credible opinions within the global financial
marketplace.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
3
Executive Summary
This report is the Spanish Addendum to the DBRS European RMBS Insight Methodology (published 17 May 2016). The report
outlines the country-specific aspects of the methodology to estimate defaults and losses for Spanish RMBS and Covered Bonds. This
addendum should be read in parallel with the DBRS European RMBS Insight Methodology. The addendum includes an overview of:
1. Data used for building the Spanish Mortgage Scoring Model and Spanish Delinquency Migration Matrices
2. Spanish Mortgage Scoring Model
3. Spanish Dynamic Delinquency Migration Matrices
4. Spanish Correlation Assumption
5. Spanish House Price Model
6. Spanish Distressed Sale Discount
7. Spanish Foreclosure Costs
8. Spanish CPR Assumptions
9. Spanish Cash Flow Assumption
Data
The Spanish Mortgage Scoring Model (Spanish MSM) and Spanish Dynamic Delinquency Migration Matrices (DMM) were
constructed using Spanish loan-level data from the European Data Warehouse (EDW).
Home price data from Instituto Nacional de Estadística (INE) and TINSA was used to generate a base house price forecast and
market value declines (MVDs). Additionally, the monthly Spanish Harmonised Consumer Price Index from Eurostat was used to
calculate real house prices.
Spanish Mortgage Scoring Model
Spanish MSM Modelling Sample
In the first step for the estimation of portfolio defaults and portfolio losses for Spanish mortgages, loans are scored dynamically
over a forecast horizon in the Spanish MSM. The technology used to build the Spanish MSM is logistic regression. To build
a score in this framework, a loan’s risk characteristics at a point in time (the “as-of” date) and a binary outcome variable that
represents the loan’s performance over a period of time subsequent to the as-of date are needed.
A set of data cleaning rules were applied to the universe of Spanish loans in the EDW to identify a dataset to develop the Spanish
MSM. The dataset included loans with the following characteristics:
• For each reporting date, there was a 12-month performance window of data available subsequent to that reporting date;
• The account is not in arrears at the as-of date;
• The current balance is greater than zero;
• The current loan to value (LTV) is less than 125%; and,
• The current interest rate margin is between 0% and 7%.
Given the universe of loans, the data on each loan is assembled with the risk characteristics at the as-of date, and the binary
outcome variable calculated from its subsequent performance. The outcome variable is determined based on the loan’s
performance over the subsequent 12 months from the as-of date. A loan is considered “bad” if during the 12-month period it
either i) entered into 90+ days arrears, ii) was foreclosed, iii) was repurchased, or iv) prepaid from a delinquency status of two
months or more. Additionally, this status persisted for at least three or more observations during that time window.
The modelling sample for the Spanish MSM included approximately 84,000 loans of which 12,000 were considered “bad” and
72,000 were considered “good”. Loans with multiple as-of dates were subject to entering the modelling sample only once. “Bad”
loans did not represent a very high proportion of the population and, as a result, were over-sampled to increase their presence
in the modelling sample.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
4
Spanish MSM Parameters
The Spanish MSM was built with objective variables (loan and borrower characteristics) and judgmental variables (Underwriting
Score and Adverse Performance). The Spanish MSM consists of 19 model parameters from 15 variables (Figure 1). Figure 1 also
reports the odds ratio for each parameter. The odds ratio measures the effect of each parameter in the Spanish MSM.
Figure 1 – Spanish MSM Parameters
Parameter
Odds Ratio
Variable Change
Variable Type
0.91
6 Years to 8 Years
Continuous
1%-40%
1.16
30% to 40%
Continuous
40%-80%
1.20
70% to 80%
Continuous
80%-125%
1.56
110%-120%
Continuous
0.0%-0.5%
0.30
0.3% to 0.5%
Continuous
0.5%-0.9%
1.47
0.7% to 0.9%
Continuous
0.9%-1.2%
1.57
1.0% to 1.2%
Continuous
Borrower Not Employed
1.79
Binary
Maturity > 30 years
1.15
Binary
Foreign Borrowers
1.46
Binary
Property Type, Not a House, Flat, Condo or Multi-Family Unit
1.20
Binary
Credit Line
1.53
Binary
VPO Guarantee
0.20
Binary
Purchase Loan
0.70
Binary
Loan Missing Income
0.88
Binary
Total Income Between EUR 1-100,000
0.96
60k to 80k
Continuous
Current Balance < EUR 10,000
1.22
10K to 5K
Continuous
Loan Age Between 6-15 Years
Indexed LTV
Margin
Underwriting Score
1
0.66
Binary
2
0.90
Binary
3
1.00
Binary
4
1.19
Binary
5
1.62
Binary
6
1.79
Binary
Adverse Performance (for Underwriting Scores 1-3 only)
1.20
Binary
Given the nature of the scoring model, the most direct way to measure the effect of a variable is by examining the odds ratios. Take,
for example, the binary variable Foreign Borrower; the odds ratio comparing Foreign Borrower with non-Foreign Borrower is:
(P[Foreign Borrower]/(1-P[Foreign Borrower]))/(P[non-Foreign Borrower]/(1-P[non-Foreign Borrower]))
Here, P[Foreign Borrower] is the probability of a Foreign Borrower becoming ‘bad’ in the 12-month time horizon. For logistic
regression, the odds ratio constructed on the values of one explanatory variable does not depend on the values of any of the other
explanatory variables. Variables included in the Spanish MSM are either binary, where the loan includes the relevant parameter
or does not, or continuous, where the odds ratio represents a specific change in the variable. For binary variables, the odds ratio is
calculated for each value as being a loan characteristic versus not being a loan characteristic. For continuous variables, the odds
ratio can be calculated for a specific change in the variable. Furthermore, for the continuous variables Indexed LTV and Margin the
change in odds ratio is not constant across the spectrum. As can be seen, the odds ratio for a 10% Indexed LTV change is greater
from 110% to 120% versus 70% to 80%.
Spanish Underwriting Score and Adverse Performance
The Spanish MSM includes two judgmental variables to assess the credit risk of Spanish residential portfolios: Underwriting Score
and Adverse Performance. Underwriting Score is a scaled scoring classification with values of 1 to 6 where 1 is considered the least
risky and 6 considered the highest risk. Adverse Performance is a binary variable which is reserved for a Spanish Underwriting
Score between 1 and 3.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
5
Spanish Underwriting Scores are assigned to each loan in a portfolio based on multiple factors including current underwriting
guidelines; changes in underwriting criteria; product types; and, length, consistency and relative quality of historical performance.
As seen in Figure 1, 3 is the base Underwriting Score where if a value of ‘3’ is assigned there is no risk adjustment to the loan
score in the Spanish MSM (Odds Ratio = 1.00). As a starting point, Spanish Underwriting Scores of ‘3’ or lower are reserved for
loans where DBRS observes established origination practices accompanied by historical data which is consistent with market
performance given the underlying product types. DBRS would typically assign an Underwriting Score of ‘3’ where a stable
relationship is observed between the originator’s strategy and the performance is deemed to be at least average. This typically
entails at least five years of performance history with average performance relative to the market in terms of 90 days arrears,
repossessions, recoveries or other metrics DBRS considers relevant. A score lower than ‘3’ would be assigned to loans where
above-average performance data is observed for a substantial period of time.
Spanish Underwriting Scores of ‘4’ and higher are assigned to loans where DBRS finds it difficult to establish a consistent
relationship between an originator’s strategy and historical performance. This may be determined in situations where (1) the
originator provides insufficient historical performance data, (2) performance data has underperformed relative to the market,
or (3) DBRS believes the origination or servicing strategy for the respective loans to be riskier than market standards. Although
there have not been new lending platforms established in the Spanish market, there have been mergers and acquisitions which
have led to rebranding of previous lenders. These rebranded platforms may have defined updated underwriting criteria which are
consistent with market standards; however, the performance of a portfolio under the redefined criteria may be considered riskier
than a lender with an established track record leading to a higher Underwriting Score. Spanish Underwriting Scores between ‘4’
and ‘6’ may also be assigned in situations where an established lender introduces one or several new products, DBRS observes
that expanding underwriting criteria or historical performance has deviated significantly compared with the market average.
High Underwriting Scores may also be assigned to portfolios consisting of reperforming loans (e.g. previous loan modifications or
payment holidays) or more severe restructurings (e.g. debt consolidation) as a result of loss mitigation strategies.
Although each loan is assigned a Spanish Underwriting Score, in practice the score is generally assigned at the portfolio level. In
circumstances where DBRS determines that a pool consists of sub-portfolios which would entail multiple scores being assigned,
the sub-portfolios would be defined with a different respective Underwriting Score assigned to each.
Generally, DBRS would expect the assigned Underwriting Score to remain constant over time. However, given the ongoing
evolution of the Spanish market, the Underwriting Score may change following acquisitions, servicing transfers or changes to
servicing strategies.
The Spanish Underwriting Score was constructed based on an assessment of the originating entities for loans in the EDW. Two
qualitative variables were considered based on an analysis of the Spanish mortgage banking sector:
• Type – three classifications based on a bank’s market share as of 2011
• Strategy – two classifications based on a bank’s being a going concern as of 2012
“Type” consists of three classifications – National, Regional and Local. Banco Bilbao Vizcaya Argentaria, S.A. and Banco
Santander S.A., the two largest issuing entities in the EDW, are classified as National. Banks with a market share between 1% and
10% are classified as Regional and all other originators are classified as Local. “Strategy” consists of two classifications: Risky and
Average. Risky originators are those which received financial support from the Fondo de Reestructuración Ordenada Bancaria
(FROB) or were acquired by another entity.1 Each of the qualitative variables were assessed in the logistic regression individually.
The combination of Type and Strategy results in six possible Spanish Underwriting Scores (Figure 2).
Figure 2 – Spanish Underwriting Score Scale
Score
Type
Strategy
Odds Ratio
1
Local
Average
0.66
2
National
Average
0.90
3
Regional
Average
1.00
4
Local
Risky
1.19
5
National
Risky
1.62
6
Regional
Risky
1.79
1. Entities receiving support from FROB or acquired by another entity as detailed in the IMF report “Spain: Financial Stability Assessment”, June 2012.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
6
The Spanish Underwriting Score is driven by the evolution of the banking sector through the financial crisis and how each of
these two variables potentially had an impact on the performance of loans in the data set. Annual Spanish mortgage origination
grew to a peak of almost EUR 300 billion in 2007 from EUR 142 billion in 2003 (INE). Volumes dropped on average 20-35% per
year before bottoming out at just EUR 37 billion in 2013. Regional savings banks represented almost 50% of total volumes in 2003
and gained market share reaching approximately 57% in 2007. Savings bank originations dropped off at an even faster rate before
disappearing in 2014, as a result of mergers and acquisitions.
Spanish home purchase non-performing loans (NPLs) increased to a peak of 6.10% at the beginning of 2014 from 0.37% at the
beginning of 2006 (Bank of Spain), driven by increased unemployment, falling home prices and a pull-back in mortgage lending.
All Spanish originators were affected by increasing residential NPLs although the severity of the NPL impact differed across
institutions. The increase in NPLs at the originator level was most likely affected by each of a bank’s Type and Strategy. Local banks
(mainly cooperative or specialised banks in the data sample) performed better than the overall market as their originations did not
grow at the same rate as the market or other types of originators gaining market share. Banks with aggressive strategies expanded
underwriting guidelines to chase market share and suffered from ambitious and growing exposure to the housing market2..
Figure 3 – Spanish NPLs, Mortgage Originations and Unemployment
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
Home Loan NPLs
Quarterly Change Mortgages
06-2015
06-2014
06-2013
06-2012
06-2011
06-2010
06-2009
06-2008
06-2007
06-2006
06-2005
06-2004
06-2003
06-2002
06-2001
06-2000
06-1999
06-1998
-5.0%
Unemployment
Source: Bank of Spain and INE.
The Spanish Underwriting Score for transactions issued prior to 2012 is assigned based on the classification of the bank within the
relevant type and strategy. For transactions issued after 2012, DBRS assesses the origination guidelines, post-2012 performance
of the issuer and vintage distribution of the portfolio to assign the Spanish Underwriting Score.
The analysis of transaction performance for the development of the Spanish Underwriting Score also showed that despite
on average better underwriting the actual performance of transactions within one originator can differ. This performance
differentiation was observed for the Spanish Underwriting Score between 1 and 3. The Adverse Performance variable is applied
to portfolios where a Spanish Underwriting Score between 1 and 3 is applied, and portfolio performance may be expected to
deviate from the average.
Fixed rate interest rate and interest margin
The Spanish mortgage market has seen an increased origination volume in fixed-interest rate mortgage loans since 2015 and
throughout 2016. DBRS will estimate interest rate margins for fixed-rate loans that lack specific information in reference to
a margin. The estimated margin will be assessed in the context of DBRS MSM margin parameters. Typically DBRS will use
the five-year Spanish government bond or swap rate as a reference rate and calculate the estimated margin as the difference
between the interest rate as of the loan’s origination date and the reference rate.
2. Ostensibly, this overview is a simplification of what happened in the Spanish banking and housing sector; however, this approach yielded a reasonable level
of explanation of asset performance.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
7
Spanish Segmentation and Delinquency Migration Matrices
The Spanish MSM was applied to approximately 825,000 loans in the EDW to score each loan and assess the distribution of
scores. 16 risk segments were identified for the purpose of calculating DMMs with a corresponding DMM for each risk segment.
Figure 4 below shows cumulative percentiles for the distributions of scores within each risk segment. Additionally, Figure 4
shows cumulative percentiles for the distributions of scores within each risk segment as well as the low and high Spanish MSM
score for each of the 16 risk segments.
Figure 4 – Spanish MSM Score Distribution
Segment
Cumulative Percentile
Low
High
1
5%
0.0%
3.3%
2
10%
3.3%
4.3%
3
15%
4.3%
5.3%
4
20%
5.3%
6.4%
5
30%
6.4%
8.7%
6
40%
8.7%
11.0%
7
50%
11.0%
13.5%
8
60%
13.5%
16.5%
9
70%
16.5%
20.5%
10
80%
20.5%
26.0%
11
85%
26.0%
30.0%
12
90%
30.0%
35.3%
13
92%
35.3%
38.3%
14
95%
38.3%
42.0%
15
97%
42.0%
50.5%
16
100%
50.5%
100.0%
DMMs are calculated for each risk segment by computing the average roll rates observed in the loan-level data between Q1 2013
and Q2 2015. For Spain, the roll rates are quarterly and measure the periodic transition rates between the six states described in
the DBRS European RMBS Insight Methodology.3
The risk segmentation exhibits separation in the default risk across the distribution of loan scores. This can be observed in two
ways. First, in Figure 4 the loan scores increase within each risk segment when moving from lower to higher risk segments
(left to right). Second, when iterating the DMM within each risk segment over time, higher cumulative default rates are
calculated for higher risk segments. Figure 5 shows the cumulative defaults forecasted per risk segment over 50 iterations for a
non-amortising pool4 of current loans in period zero.
Figure 5 – Cumulative Default Expectation per Spanish MSM Risk Segment
30%
25.8% 26.3%
Cumulative Defaults
25%
22.2%
20%
14.9%
15%
9.9%
10%
5%
3.8%
2.3% 2.3% 2.6% 3.0%
1.1% 1.3% 1.9%
4.9%
11.8%
6.7%
0%
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Risk Segment
3. The six states are DQ0, DQ1, DQ2, DQ3 Default, and Redeemed.
4. Assumed 0% CPR.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
8
Looking at Figure 4 and Figure 5 together, for a portfolio of current non-amortising Spanish loans the following could be expected:
• High quality portfolio where all of the loans have an initial Spanish MSM score less than 3.3%, risk segment 0, the expected
default rate is approximately 1.1%.
• Average quality portfolio where all of the loans have an initial Spanish MSM score between 13.5% and 16.5%, risk segment 7,
the expected default rate would be approximately 3.0%.
• Low quality portfolio where all of the loans have an initial Spanish MSM score greater than 50.5%, risk segment 15, the
expected default rate would be approximately 26.3%.
Figure 5 illustrates a simplified estimate of potential cumulative default rates for each of the 16 different segments. In this example
the respective DMM for each risk segment is applied to a portfolio of performing non-amortising loans for 12.5 years. Ultimately,
forecasted portfolio default rates will be a function of the underlying loan characteristics and each loan’s amortisation profile.
The loan amortisation profile determines the exposure at default (EAD) to calculate the periodic loan-level defaults. For Spain,
the EAD for a given period is equal to the outstanding loan balance 18 months5 prior to the loan defaulting which is reflective of
market practice of recognising defaults.
Spanish Correlation
The correlation applied to each portfolio for estimated rating scenario default rates is a function of the expected default rate
of the portfolio. Correlations assigned to each portfolio range from 12% to 24% with higher default portfolios being assigned
lower correlations and lower default portfolios being assigned higher correlations (Figure 6). The scale of correlations for each
expected portfolio default rate is based on the Basel III framework.6
Figure 6 – Correlation Versus Expected Portfolio PDR
25%
Correlation
20%
15%
10%
9.75%
9.25%
8.75%
8.25%
7.75%
7.25%
6.75%
6.25%
5.75%
5.25%
4.75%
4.25%
3.75%
3.25%
2.75%
2.25%
1.75%
1.25%
0.75%
0%
0.25%
5%
Expected Portfolio Default Rate
DBRS estimated a correlation across the Spanish risk segments between 10% and 12% over a period of two years. Although
relatively low correlations were estimated for the lower risk segments, higher correlations are applied to lower default portfolios
due to potential exposure to future idiosyncratic risks not reflected in the available data.
5. Defined as ‘k’ in the Base Case Default Rates section of the DBRS European RMBS Insight Methodology.
6. Asset correlation assumption per Basel III – BIS, Basel III: A global regulatory framework for more resilient banks and banking system, December 2010, pg. 39.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
9
Spanish House Price Model
The Spanish House Price7 model generates MVDs for each rating scenario. The Spanish MVDs are estimated at the national level
and for the 19 autonomous regions. There are two publicly available data sources which are commonly used to analyse Spanish
home prices: INE and TINSA. Each data source has advantages and disadvantages for analysing Spanish home prices. Data from
INE is quarterly covering the period from Q1 2007 to Q2 2015. Additionally, the INE data is reported for the 19 autonomous
regions which is advantageous to map to data in the EDW template. Data from TINSA is monthly covering the period from
January 2001 to October 2015. TINSA data does report sub-indexes; however, these sub-indexes are not easily mapped to loanlevel data. Given the shorter time period for the INE data, DBRS analysed the time periods where data was available for both
datasets to model each INE series as a function of the TINSA data to backfit missing INE values to 2001. Real home prices are
calculated using Consumer Price Index (CPI) data with June 2003 as the base year. MVDs for each of the autonomous regions
and the national level are shown below in Figure 7. MVDs are applied to the updated property value to discount the sales price
of a property to calculate periodic losses. This allowed a more granular data set over a longer period of time.
Real home prices are calculated using harmonised CPI data with June 2003 as the base year. MVDs for each of the autonomous
regions and the national level estimated using the Spanish Home Price Model are shown below in Figure 7. MVDs are applied to
the updated property value to discount the sales price of a property in the periodic loss calculation.
Figure 7 – Spanish MVDs
Region
AAA
AA
A
BBB
BB
B
Volatile
Overheated
Andalucía
-33.0%
-30.4%
-25.5%
-21.6%
-17.5%
-12.6%
Yes
Yes
Aragón
-16.3%
-13.4%
-9.8%
-6.2%
-2.5%
0.0%
Yes
No
Asturias
-44.6%
-40.8%
-35.0%
-30.0%
-24.7%
-19.4%
Yes
Yes
Balearic Islands
-38.5%
-35.3%
-29.9%
-25.2%
-20.3%
-15.7%
Yes
Yes
Canary Islands
-43.0%
-39.9%
-34.2%
-29.6%
-24.7%
-18.5%
Yes
Yes
Cantabria
-41.7%
-38.2%
-32.3%
-26.8%
-21.3%
-16.7%
Yes
Yes
Castile-La Mancha
-47.4%
-44.3%
-38.5%
-33.3%
-27.7%
-22.0%
Yes
Yes
Castile and León
-36.3%
-32.9%
-27.6%
-22.6%
-17.8%
-13.7%
Yes
Yes
Cataluña
-17.4%
-14.5%
-10.4%
-6.6%
-2.8%
0.0%
Yes
No
Ceuta
-25.0%
-20.8%
-15.0%
-10.5%
-5.8%
-2.4%
No
No
Communidad Valencia
-31.7%
-29.1%
-24.1%
-19.9%
-15.7%
-11.8%
Yes
Yes
Extremadura
-43.2%
-40.0%
-34.2%
-29.5%
-24.7%
-18.2%
Yes
Yes
Galicia
-49.6%
-46.0%
-40.3%
-35.4%
-30.0%
-22.0%
Yes
Yes
Madrid
-16.2%
-13.5%
-9.5%
-6.1%
-2.6%
0.0%
Yes
No
Melilla
-36.4%
-33.6%
-28.5%
-24.6%
-20.2%
-14.5%
Yes
Yes
Murcia
-45.4%
-42.0%
-36.1%
-31.2%
-26.2%
-19.6%
Yes
Yes
Navarra
-16.7%
-14.2%
-10.4%
-6.5%
-2.7%
0.0%
Yes
No
Pais Basque
-33.2%
-29.9%
-24.1%
-18.4%
-13.0%
-9.1%
Yes
Yes
Rioja
-34.6%
-30.5%
-24.6%
-18.7%
-13.1%
-9.1%
Yes
Yes
National
-31.1%
-28.0%
-23.2%
-18.6%
-13.7%
-10.2%
Yes
Yes
Figure 7 also includes the market status for the purposes of the MVD simulation where markets are considered overheated if the
real house price index (June 2003 = 100) has increased above 150 in the most recent run-up and has yet to drop below 85. Volatile
indicates if a market has ever been overheated.
Spanish Distressed Sale Discount
The Spanish Distressed Sale Discount (DSD) is 40%. This is based on analysis of over 6500 repossessed properties which were
sold between 2007 and 2015. Over 95% of the sales occurred after 2010. DSDs are applied to the expected property value after
applying the MVD and meant to address a property sale in a liquidation scenario.
Spanish Foreclosure Costs
DBRS estimates Spanish foreclosure costs to be 8.0% of the outstanding loan balance at the time of default and a fixed cost of
EUR 5,000. DBRS’s estimates of the variable and fixed costs are based on consultation with external counsel.
7. See European RMBS Insight Methodology, Appendix 3 for further details.
Structured Finance: European RMBS
May 2017
European RMBS Insight: Spanish Addendum
DBRS.COM
10
DBRS has implemented minimum loss given default (LGD) rates for certain rating scenarios in Spain. DBRS recognises in
periods of high economic stress it may be difficult or even impossible to find a market clearing price for residential real estate
resulting in potential losses which may be greater than those implied by stressed liquidation values. The minimum LGDs are set
for the AAA, AA and “A” rating scenarios at 25%, 20% and 15%, respectively.
Spanish Constant Prepayment Rate (CPR) Assumptions
Between 2004 and 2007, CPRs averaged approximately 6.5%. CPRs steadily declined until 2012 in tandem with the drop in house
prices and decrease in mortgage lending (Figure 8). Spanish CPRs have been averaging around 2% since 2012. Mortgage rates
have continued to drift lower. However, the decline in mortgage rates is not incentivising borrowers to refinance given that most
mortgages tend to be floating rate indexed to 12-month Euribor. As Euribor trends downward, borrowers’ monthly payments
reset lower. As monthly payments remain low, the incentive to refinance is absent without substantial equity in the property.
House prices have begun to recover (up 6.5% between Q1 2014 and Q3 2015) but not to a level which would be expected to have
a significant impact on CPRs. DBRS expects the current observed trend in CPRs to continue and assumes a 2% CPR for the
Spanish MSM.
35.0
12.0%
30.0
10.0%
25.0
8.0%
20.0
6.0%
15.0
4.0%
10.0
Mortgage Volumes (left)
Avg Mortgage Rate (right)
2015-07
2015-01
2014-07
2014-01
2013-07
2013-01
2012-07
2012-01
2011-07
2011-01
2010-07
2010-01
2009-07
2009-01
2008-07
2008-01
2007-07
2007-01
2006-07
2006-01
0.0%
2005-07
0.0
2005-01
2.0%
2004-07
5.0
2004-01
Monthly Lending (billions EUR)
Figure 8: Spanish CPRs, Mortgage Rates and Lending Volumes
3 Mo CPR (right)
Source: Band of Spain, Intex and INE.
Spanish Cash Flow Assumptions
Prepayments
Four prepayment stresses are generally run for the DBRS cash flow stresses. The typical stresses for Spain are 0% - Very Slow,
5% - Slow, 10% - Mid and 20% - Fast. DBRS will determine in its analysis which stresses should be the most relevant.
Recovery Timing
The Spanish recovery timing for cash flow analysis is 48 months. This is based on further analysis of the repossessed loans used
to estimate the Spanish DSD. The recovery timing is the time between the first period a loan stops contributing principal and
interest payments to the collections and receipt of the recoveries on a loan.
Structured Finance: European RMBS
May 2017
The DBRS group of companies consists of DBRS, Inc. (Delaware, U.S.)(NRSRO, DRO affiliate); DBRS Limited (Ontario, Canada)(DRO, NRSRO affiliate);
DBRS Ratings Limited (England and Wales)(CRA, DRO affiliate); and DBRS Ratings México, Institución Calificadora de Valores S.A. de C.V. (Mexico)(CRA,
NRSRO affiliate, DRO affiliate). Please note that DBRS Ratings Limited is not an NRSRO and ratings assigned by it are non-NRSRO ratings. For more information
on regulatory registrations, recognitions and approvals, please see: http://www.dbrs.com/research/225752/highlights.pdf.
© 2017, DBRS. All rights reserved. The information upon which DBRS ratings and reports are based is obtained by DBRS from sources DBRS believes to be reliable. DBRS does
not audit the information it receives in connection with the rating process, and it does not and cannot independently verify that information in every instance. The extent of any factual
investigation or independent verification depends on facts and circumstances. DBRS ratings, reports and any other information provided by DBRS are provided “as is” and without
representation or warranty of any kind. DBRS hereby disclaims any representation or warranty, express or implied, as to the accuracy, timeliness, completeness, merchantability,
fitness for any particular purpose or non-infringement of any of such information. In no event shall DBRS or its directors, officers, employees, independent contractors, agents and
representatives (collectively, DBRS Representatives) be liable (1) for any inaccuracy, delay, loss of data, interruption in service, error or omission or for any damages resulting therefrom,
or (2) for any direct, indirect, incidental, special, compensatory or consequential damages arising from any use of ratings and rating reports or arising from any error (negligent or
otherwise) or other circumstance or contingency within or outside the control of DBRS or any DBRS Representative, in connection with or related to obtaining, collecting, compiling,
analyzing, interpreting, communicating, publishing or delivering any such information. Ratings and other opinions issued by DBRS are, and must be construed solely as, statements of
opinion and not statements of fact as to credit worthiness or recommendations to purchase, sell or hold any securities. A report providing a DBRS rating is neither a prospectus nor a
substitute for the information assembled, verified and presented to investors by the issuer and its agents in connection with the sale of the securities. DBRS receives compensation for its
rating activities from issuers, insurers, guarantors and/or underwriters of debt securities for assigning ratings and from subscribers to its website. DBRS is not responsible for the content
or operation of third party websites accessed through hypertext or other computer links and DBRS shall have no liability to any person or entity for the use of such third party websites.
This publication may not be reproduced, retransmitted or distributed in any form without the prior written consent of DBRS. ALL DBRS RATINGS ARE SUBJECT TO DISCLAIMERS
AND CERTAIN LIMITATIONS. PLEASE READ THESE DISCLAIMERS AND LIMITATIONS AT http://www.dbrs.com/about/disclaimer. ADDITIONAL INFORMATION REGARDING
DBRS RATINGS, INCLUDING DEFINITIONS, POLICIES AND METHODOLOGIES, ARE AVAILABLE ON http://www.dbrs.com.
DBRS Ratings Limited
| 20 Fenchurch Street, 31st Floor, London, EC3M 3BY United Kingdom
|
TEL +44 (0)20 7855 6600
|
www.dbrs.com