Fitch Affirms Candide Financing Dutch RMBS Series
The Dutch RMBS transactions comprise loans originated by Bank of Scotland, Amsterdam Branch, which is a 100%-owned subsidiary of Lloyds Banking Group plc (A+/Stable/F1).
KEY RATING DRIVERS
Increasing Foreclosures, Arrears Trending Down
Over the 12 months prior to the last reporting dates, late stage arrears (loans with more than three monthly instalments overdue) decreased in Candide 2008 and Candide 2011. They increased in Candide 2008-2 by 9bps. Late arrears were reported at between 0.5% (Candide 2008-2) and 0.2% (Candide 2011) of the current portfolio balance, lower than the Dutch Prime RMBS Index (0.7%).
The general improvement of late arrears during the post-crisis period has been offset by the increased foreclosure stemming from house price recovery. Cumulative foreclosures were reported at between 2.6% of the original portfolio (Candide 2008 +46bps yoy) and 1.4% (Candide 2011 +38bps yoy) while realised losses spanned from 0.9% (Candide 2008-2 +2011bps yoy) and 0.3% (Candide 2011 +9bps yoy). On average, the Dutch prime RMBS rated by Fitch reported cumulative losses of 0.3% of the aggregate original balance.
Fitch believes the high original weighted average loan-to-value (LTV) and debt-to-income ratios, representing the borrowers' willingness to repay their mortgages and their affordability, are the main reasons for the worse than average asset performance, particularly for the deals originated in 2008. Loans with a current weighted average LTV above 100% represent between 60.3% (Candide 2008) and 24.7% (Candide 2011) of the portfolio, which explains higher than average losses.
Nevertheless, the credit enhancement (CE) available to the rated notes is sufficient to absorb Fitch's 'AAAsf' expected losses, with the exception of Candide 2008-2 class A notes where the marginal CE shortfall has led to their affirmation with Negative Outlook.
High Concentration of Interest Only (IO) Loans
IO loans represent between 74.8% (Candide 2011) and 64.4% (Candide 2008). These products are subject to a foreclosure frequency (FF) adjustment based on their current LTV/maturity profile. Higher LTV loans with shorter maturities have fewer refinancing opportunities and hence will attract a higher increase. In addition, where IO loans maturing within a three-year period make up more than 20% of the portfolio, Fitch assesses the sensitivity of the ratings assuming a higher FF for those loans. No rating action was required from the results of the sensitivity analysis.
Insurance Set-off Exposure
The intention of insurance policies is that the proceeds of the investments can be used to repay the mortgage loan in full or in part at maturity. In the event that the policy providers are no longer able to meet their obligations, for example as a result of insolvency, borrowers may seek to set-off the claim over the insurance provider against their mortgages, on the basis that the intention is for the loan to be repaid using the proceeds from the policy.
The risk that this set-off could be successfully exercised depends on whether the lender and insurance policy provider are the same legal entity, or whether the mortgage and insurance policies are offered as one product. If they are, Fitch assumes a set-off probability equal to 100%. This figure is reduced to 25% if the insurance provider and lender are different institutions or if the mortgage and insurance policy are not a joint product.
Currently, mortgages with an insurance policy attached represent between 29.8% (Candide 2008) and 15.3% (Candide 2008-2) of the total portfolios. The derived set-off exposure is accounted for in the analysis of the available credit enhancement.
Payment Interruption Risk Mitigated
Fitch believes the risk of an interruption in payments to noteholders due to the servicer's default is sufficiently mitigated by the well-developed third party servicing market in the Netherlands and the presence of a delegated sub-servicer. In addition, the notes across all transactions can withstand non-payment of at least six months thanks to appropriate liquidity sources, under stressed Euribor assumptions.
Application of PAF
In its analysis of Candide 2008-2 and Candide 2011, Fitch applied a Performance Adjustment Factor (PAF), which varies from its EMEA RMBS Rating Criteria. The PAF compares cumulative defaults and current delinquencies with the expected lifetime cumulative defaults of the transaction at specific points in time.
In Candide 2008-2, Fitch adjusted the PAF down to 1.0 from the calculated 1.85 in consideration of the fact that at the next forecasting point (in two months time), current and projected defaults are expected to be aligned.
For Candide 2011, the comparison between actual and expected foreclosures results in a PAF equal to 2. In consideration of the deal's standalone performance, relative to the rest of the series and to the market, we did not consider such a stressful adjustment appropriate. We calculated a more reasonable PAF equal to 1.45 comparing actual foreclosures across the series.
Deterioration in asset performance may result from economic factors, in particular the effect of increasing unemployment. A corresponding increase in new defaults and associated pressure on excess spread levels and the reserve fund could result in negative rating action.
Candide 2008 fixed-to-floating swap is structured so that the post swap margin available to the SPV is the sum of the weighted average margin on the floating rate loans (8.3% of the pool) plus the weighted average historical margin of fixed rate loans (91.7%) over the IRS rate. Historically, the excess spread generated by this structure has always been sufficient to provide for period losses. However, in a rising interest rate scenario, both the margin on variable rate loans and the margin paid by the swap on the fixed rate portfolio are assumed to compress. The reduced excess spread may lead to outstanding principal deficiency ledgers and negative rating actions.
USE OF THIRD-PARTY DUE DILIGENCE PURSUANT TO RULE 17G-10
Form ABS Due Diligence-15E was not provided to, or reviewed by, Fitch in relation to this rating action.
Fitch has checked the consistency and plausibility of the information it has received about the performance of the asset pools and the transactions. There were no findings that affected the rating analysis.
Fitch has not reviewed the results of any third party assessment of the asset portfolio information or conducted a review of origination files as part of its ongoing monitoring.
Applicable to Candide 2008 and 2008-2: Fitch did not undertake a review of the information provided about the underlying asset pools ahead of the transactions' initial closing. The subsequent performance of the transactions over the years is consistent with the agency's expectations given the operating environment and Fitch is therefore satisfied that the asset pool information relied upon for its initial rating analysis was adequately reliable.
Applicable to Candide 202011: Prior to the transaction closing, Fitch did not review the results of a third party assessment conducted on the asset portfolio information. Prior to the transaction closing, Fitch conducted a review of a small targeted sample of Bank of Scotland, Amsterdam Branch's origination files, which indicated that updates to loan details are not always completed on the system. These findings were considered in this analysis by applying a lender adjustment factor equal to 1.05 to the base default probability for this transaction.
Overall and together with the assumptions referred to above, Fitch's assessment of the information relied upon for the agency's rating analysis according to its applicable rating methodologies indicates that it is adequately reliable.
SOURCES OF INFORMATION
The information below was used in the analysis.
- Investor reports and Loan-by-loan data provided by Lloyds NL as at 31 August 2016 (Candide 2008 and Candide 202011) and as at 30 June 2016 (Candide 2008-2)
- Insurance Policy Providers Breakdown provided by Lloyds NL as at 26 and 27 October 2016.