General Structured Finance Rating Methodology Structured Finance

Transcription

1 General Rating Methodology 28 August 2017 Contacts Guillaume Jolivet Managing Director Martin Hartmann Senior Analyst Carlos Terré Managing Director Sebastian Dietzsch Associate Director

2 Table of Contents 1. Introduction Areas of application Summary Rating definitions Methodology Collateral risk analysis Collateral pool characteristics: non-granular or variable composition Collateral asset types and characteristics Collateral default and loss-distribution analysis Collateral market value risk Sensitivity analysis, long-term assessment of senior protection buffers Counterparty risk analysis Materiality of financial and operational risk exposures Asset management quality Structure analysis Legal risks Structural and enhancement features and cash flow analysis Country and industry risks Representation and warranties Monitoring Appendix I. Scope s idealised expected loss and default probability tables Appendix II. Technical note on the expected loss framework Appendix III. Technical note on timely payment Appendix IV. Technical note on general asset recovery analysis August / 19

3 1. Introduction This document is an update of Scope Ratings AG s (Scope) General Methodology published on 31 August Aside from minor editorial changes and clarifications, this update does not propose any material changes to our rating approach. These clarifications or changes do not affect Scope s existing ratings of structured finance instruments or covered bonds. The main adjustments and clarifications are: A revision of ratings definition (Section 4) to reflect those are now available directly on our website and a clarification applicable to unfunded instruments; An update of Figures 1 and 2 and the summary of collateral asset types in structured finance; A clarification and substantial editing of sections 5.1.3, and without modification of their substance; and The addition of a section highlighting market value risk requiring specific consideration. Scope generally uses a fundamental bottom-up analysis to capture the rating impact of different asset, portfolio or structure characteristics, taking into account the context of origination and relevant jurisdictions. Scope avoids one-sizefits-all assumptions and analyses securitisations in a European context, applying its knowledge of the local market. European markets exhibit laws, features and a history that differ from those of other markets. European markets are typically more controlled and regulated, and benefit from safeguards or economic stabilisers that protect against certain shocks. This is reflected not only in the way Scope looks at key input assumptions, but also in how it applies stress to these assumptions. Scope s rating approach incorporates the historical performance of European structured finance transactions without a bias towards North American markets. Scope complements the analysis of a historical performance data series with forward-looking views on economic cycles and generally incorporates macroeconomic factors into its base case assumptions. Scope s forward-looking approach consists of assessing how structural factors may affect European credit markets, for example, regarding the evolution of loan-underwriting standards, demographics, wage structures, consumption patterns, or changes in European regulations or supervision. Scope analyses counterparty risks by building on post-crisis realities, including the new regulatory and supervisory framework for banks, such as bail-in and stronger prudential metrics, and the resulting limited likelihood for banks to default in the short term. In addition, Scope does not mechanistically limit the maximum rating achievable by a securitisation based on the sovereign credit quality of the country of the issuer or of the securitised assets. 28 August / 19

4 2. Areas of application This methodology applies to all types of structured finance instruments, in particular to debt instruments issued by a special-purpose vehicle and exposed to the performance of real or financial assets. The securitised asset s economic risk can be transferred to the special-purpose vehicle/issuer either by transferring the securitised assets legal ownership (a true sale) or through a synthetic transfer of risks via credit derivatives. Structured finance instruments issued by a special-purpose vehicle are generally non- or limited-recourse debts. Their repayment is driven primarily by the performance of the underlying collateral and the priorities of payments in the transaction. The collateral of structured finance instruments includes a wide range of financial or real assets related to different sectors such as real estate, consumer credit, SMEs, corporates, project finance and infrastructure. This general methodology ensures consistency when we rate structured finance instruments. It applies primarily to European structured finance instruments but may also apply to non-european transactions when appropriate. When rating structured finance instruments from a certain asset class, Scope might refer to methodologies specific to those assets. Scope might also expand on the tailored analytical approach for a given transaction in its rating report. 3. Summary This methodology lays down the key principles we apply to the rating and monitoring of all types of structured finance instruments. A structured finance instrument is a transaction in which the credit risk associated with an underlying exposure or a pool of exposures may be securitised. The payments in the transaction depend primarily on the performance of the underlying exposure or pool of exposures. If the liability structure of the transaction is tranched, the seniority of tranches may drive either the distribution of payments or the allocation of losses during the life of the transaction. Scope s rating methodology relies on three analytical building blocks, which include an analysis of: The underlying collateral; The risks related to key parties to the transaction; and The transaction s structure. An important component of the analysis is the calculation of an expected loss figure associated with the rated instrument. Scope benchmarks the expected loss outcome over the life of the instrument against its idealised expected loss table (see Appendix I. Scope s idealised expected loss and default probability tables). The result of the quantitative analysis is Scope s anchor for determining the rating, and is supplemented by its qualitative analysis of unquantifiable risks. Scope believes the analysis of qualitative factors is crucial for assigning a rating. The qualitative elements identified when analysing the structure of the transaction (counterparty risks, quality of origination, management, and incentives) are incorporated into the rating process. Scope takes the view that a qualitative assessment is essential to differentiate transactions whose assets require intensive care, dynamic management or an active work-out. Qualitative considerations can limit a rating or link it to the performance of a counterparty. It may even lead to Scope not assigning a rating. 28 August / 19

5 4. Rating definitions Scope s structured finance credit ratings constitute a forward-looking opinion on relative credit risks. A rating reflects the expected loss associated with payments contractually promised by an instrument on a particular payment date or by its legal maturity. It factors in both the likelihood of a default on such payments and the loss severity expected upon default. While the expected loss approach forms the cornerstone of Scope s analysis, the agency closely assesses the frequency of default and may limit the rating if an instrument has a low expected loss and a high default frequency. Ratings assigned to unfunded instruments such as credit default swaps measure the expected loss for the credit protection seller, i.e. the risk to make a payment with respect to a credit event under the transaction s terms. Ratings assigned to such unfunded instruments do not address potential losses resulting from the early termination of the transaction nor any market risk associated with the transaction. For more details, refer to the technical notes on the expected loss framework and timely payment under Appendix II and Appendix III. Coupon deferral: Some structured finance instruments contractually allow for interest deferral and the accrual of deferred interest. In general, prolonged periods of interest deferral occur when the credit performance of the underlying collateral is worse than expected, and will be reflected in a downward adjustment of the instrument s rating. Scope believes that an investment grade rating is incompatible with an instrument that allows the likely or discretionary deferral of interest payments over long periods, i.e. deferral periods exceeding the shorter of either i) one year, or ii) two interest payment dates. An instrument which structurally allows interest deferral will only be assigned an A+ or above if the deferral likelihood remains commensurate with the timely payment standards highlighted in Appendix III. The default of a structured finance instrument includes one of the following events: i) a missed payment of interest or principal, which is incurred under the instrument s terms and conditions; ii) the instrument s restructuring or repurchase to avoid a payment failure, which ultimately leads to an economic loss for the debt investor; iii) an event of default (EoD) under the instrument's terms that leads to an enforcement of the security. Scope applies the SF suffix to structured finance instruments in line with Regulation No of the European Parliament and the European Council. Such instruments include asset-backed securities (ABS), mortgage-backed securities (MBS) and collateralised debt obligations (CDOs). Scope does not apply the suffix to instruments such as covered bonds and non-tranched asset securitisations. 28 August / 19

6 5. Methodology To analyse the credit risk of a structured finance instrument, Scope assesses i) the collateral underlying the related transaction, in particular the default distribution, loss distribution, and associated cash flows from the collateral (collateral risk); ii) the parties that affect the performance of the transaction (operational risk and counterparty credit risk); and iii) the features of the transaction s structure including integrity of the transaction s structure. Scope s rating methodology for structured finance instruments relies on three analytical blocks shown in Figure 1: Figure 1. Rating Collateral risk analysis Quantitative and qualitative analysis of the collateral characteristics Portfolio characteristics Loss distribution Sensitivity analysis Scope estimates an instrument s expected loss using a quantitative approach, supplemented by qualitative considerations for each of the three analytical areas shown in Figure 1. Scope benchmarks the expected loss outcome over the life of the instrument with the idealised expected loss tables in Appendix I. Scope uses the quantitative results as an anchor for determining the rating, although qualitative considerations may override them. Scope assumes an investor has a passive role in a transaction when considering the transaction s optional features. This means that when an investor can actively alter the transaction s features, Scope will generally not consider its impact on the rating, but will monitor the effect on the transaction should an investor actively exercise an option. In addition, even when an investor actively makes it possible to modify a transaction mechanism, Scope assumes the investor might not proactively take this option Collateral risk analysis Scope focuses firstly on the credit quality of the underlying assets or pool of assets (collateral). Scope evaluates the characteristics of assets in the collateral pool to understand default patterns, loss severity upon default and dependency structures. The collateral risk assessment involves both a qualitative and quantitative analysis at various rating levels to estimate the collateral s default and loss distribution under different scenarios. Scope s preferred approach to analyse collateral risk is to access and analyse loan-by-loan data on collateral. From a qualitative standpoint, Scope analyses the context of the origination or the sourcing of the asset. This approach therefore assesses the product type of the assets as well as the underwriting process, incentives, strategy and standards of the originator. In order to calibrate model inputs, Scope relies on historical data related to the collateral assets when available, 28 August / 19

7 supplemented by qualitative considerations on the collateral. In some instances, input calibration may be based on loss characteristics experienced across structured finance transactions exposed to comparable collateral. Shorter historical data sets or poor-quality data regarding the collateral pool usually result in more conservative modelling assumptions to reflect greater uncertainty. In some instances, the lack of sufficient quantitative or qualitative data may even make it impossible to assign a rating Collateral pool characteristics: non-granular or variable composition Transactions with collateral which is non-granular or has high concentrations require additional analysis. For instance, assets composing collateral can range from a handful of heterogeneous assets in CMBS or CDO transactions to severalthousand relatively homogeneous assets in ABS, RMBS or even SME ABS transactions. For SME ABS, the collateral may comprise numerous assets, of which only a few represent a substantial portion of the total balance. Considering this aspect is important to avoid reliance on averages. Scope measures the equivalent effective number of exposures the inverse of the Herfindahl Index to assess portfolio concentration under different criteria depending on the asset class. This diversity metric may be applied to measure concentration in among others obligors, industries or regions. Depending on the asset class, a detailed analysis of individual assets may be required to analyse the collateral pool risk. For this purpose, Scope may produce a credit rating or an internal assessment of some or all of the assets in the pool. Should some assets in the pool already benefit from third-party ratings such as from a regulated bank or credit rating institution, Scope may incorporate the ratings into its analysis, but reserves the right to adjust them. Expression (1) shows the diversity index that measures obligor concentration: (1) Scope focuses on the actual collateral s characteristics when the portfolio of assets is static, i.e. when assets cannot be removed from or replaced in the portfolio. By contrast, when collateral can be replenished, liquidated or actively traded, a hypothetical portfolio that follows asset-eligibility criteria covenants is used for the analysis. The agency does not generally analyse a worst-case portfolio based solely on the transaction covenants, but does consider characteristics of the initial portfolio and its stickiness. For this analysis, Scope evaluates the manager s ability and incentives to perform their duties and its potential to add value to the transaction. This is particularly important for managed transactions with covenants that are weak, limited or very different from the actual characteristics of the invested collateral. 28 August / 19

8 Collateral asset types and characteristics The collateral behind structured finance instruments includes a large variety of assets, such as loans, receivables or bonds. These may be secured or unsecured and can also be real assets. Obligors can be consumers, corporates or even public entities, sovereigns or sub-sovereigns. Summary of asset types and characteristics of core structured finance asset classes Deal types ABCP Auto ABS Underlying assets Commercial discount credits or credit advances Auto loans or auto leases Typical characteristics of the asset types Asset analysis Risk horizon Granularity Homogeneity Focus Rating assessments Short-term Granular Homogeneous Medium-term Granular Homogeneous Originator loan book Securitisation portfolio No No CMBS Commercial mortgages Typically, long-term Non-granular Heterogeneous Loan by loan Yes Consumer ABS Consumer loans Medium-term Granular Homogeneous Securitisation portfolio No CRE loans Commercial real estate loans Long-term Non-granular Heterogeneous Loan by loan Yes Credit cards Credit card balances Short-term Granular Homogeneous Originator loan book No Credit-linked notes/repackag ing Any financial assets Typically, medium-term Typically, single asset N/A Pass-through rating/asset per asset Yes LL CLOs Corporate leveraged loan Medium-term Non-granular Relatively Homogeneous Loan by loan Yes PF CLOs RMBS SME ABS Trade receivables Other/Esoteric Project finance debt Residential mortgages Loans to small and medium-sized enterprises Commercial credit Real assets, funds shares, credit default swaps, other Long-term Non-granular Heterogeneous Loan by loan Yes Long-term Granular Homogeneous Typically, medium-term Short-term Short- to longterm Granular Typically, granular Typically, nongranular Mixed Homogeneous Loan by loan or securitisation portfolio Loan by loan or securitisation portfolio Originator loan book No Possible No Heterogeneous Bespoke Possible Transactions may be exposed to single assets with different levels of concentration. Scope therefore assesses the credit risk of large exposures individually and will assess them differently depending on single asset concentration levels. This is because large exposures within a securitised portfolio may pose significant idiosyncratic risks to the rated instrument. 28 August / 19

9 Analytical approach to assess and monitor different levels of single asset risk Top obligor concentration (% of portfolio balance) Credit quality derived from: Less than 2% Mapping of external credit risk measures available to Scope 1 Less than 5% Less than 10% Mapping an individual outcome validated by Scope s analysts Credit estimate or similar assessments by Scope or its affiliates 10% or more Public or private ratings by Scope Consumer ABS and RMBS consumer credit and residential mortgage loans Consumer credit includes auto loans, lines of credit, consumer loans and credit card debt. The collateral pools backing consumer credit ABS or RMBS transactions usually show limited heterogeneity and often contain a significant amount of loans, for which a portfolio analysis may be best suited. In general, the analysis comprises an assessment of the pool s characteristics, the quality of the asset pool s servicing, the asset originator s lending standards and Scope s forwardlooking expectations on performance based on macroeconomic forecasts. SMEs loan ABS Depending on the granularity of the securitised pool, Scope may either carry out a loan-by-loan analysis or model an ideal portfolio. Scope may assess the credit quality of the pool by: examining individual credit ratings and internal assessments; calibrating historical data; and incorporating information based on the internal rating system of the loans seller. The analysis also includes Scope s macroeconomic view on the relevant SME market. Details on these procedures can be found in Scope s SME CLO Rating Methodology. CMBS and commercial real estate loans Underlying collateral for CMBS transactions and securitisations of commercial real estate loans are non-granular and most often exhibit a high level of heterogeneity. When rating CMBS transactions, Scope relies on an internal assessment of each underlying commercial real estate loan. This is achieved by reviewing in detail the economics of the commercial real estate assets and by assessing the loan-servicing abilities. In addition, Scope evaluates the asset s tenant structure, rentroll and economic environment. CLO leveraged loans Leveraged loans are usually broadly syndicated instruments that exhibit homogenous characteristics. However, leveraged loan portfolios are usually non-granular and require a loan-by-loan analysis to assess the credit risk of the portfolio. As a result, Scope relies on ratings or internal assessments for each underlying loan, or on monitored ratings from other regulated credit rating agencies, when available. Credit-linked notes and asset repackaging Credit-linked notes (CLN) can be used to repackage a variety of assets under a different format or to create hedging instruments. In some instances, these instruments provide a strict pass-through of the cash flows and risk characteristics from the securitised asset. In some cases, it may add significant risks from key counterparties or modify the underlying asset s payment characteristics (e.g. payment maturity profile, currency, coupon basis). The analysis of the legal structure and counterparty risks is key to such transactions. Scope analyses collateral using the expected loss approach including elements described in Appendix I, Appendix II and Appendix III. In the case of a single-asset repackaging with limited legal and counterparty risks, the credit risk of the instrument is likely to be in line with that of the repackaged asset. By contrast, CLNs used as a hedging instrument could be exposed to credit risk from all counterparties in the structure, increasing the note s likelihood of default. Project finance loans Project finance (PF) assets are usually very heterogeneous since the asset class covers financing for infrastructure, transportation, energy and real estate. PF often relates to projects for public needs. When analysing structured finance instruments backed by PF loans, Scope generally prepares a rating or a credit assessment of each underlying PF loan in the collateral pool. The assessment incorporates a detailed view of the economics of each project, the project phase and the liability-servicing abilities, including seniority and credit enhancement. Scope also considers the off-takers and guarantee providers, which often play a significant role with respect to going-concern operations. 1 Internal rating models of the originator or public ratings from ECAI. Scope may use those credit measures and adjust them as necessary. 28 August / 19

10 Collateral default and loss-distribution analysis In most cases, Scope analyses collateral risk by projecting future losses and deriving cash flows associated with the securitised assets performance. A central aspect of this analysis includes the selection of an appropriate modelling process to capture the collateral s characteristics. Such processes measure the loss distribution of the collateral to best mimic the collateral assets behaviour over time. To adequately capture the risks of different collateral pools, Scope applies a framework that provides a consistent projection across asset classes and collateral exhibiting various granularities and homogeneities. The probability of each possible default rate is produced using either a Monte Carlo simulation or a standard parametric distribution. Concentrated collateral pools with limited diversification, such as CMBS, CDO and certain ABS, call for using a default distribution that reflects detailed and specific assumptions for each asset making up the collateral. To analyse such portfolios, Scope produces a non-parametric distribution of losses based on a market-standard, Monte Carlo simulation, typically implementing a Gaussian copula dependency framework. Many simulations allow results to converge with a level of volatility commensurate with the target rating s expected loss. For homogenous and granular collateral pools, Scope generally analyses pools according to standard, parametric probability distributions laws such as the inverse Gaussian, using a large homogeneous portfolio approximation approach. This approach limits the number of required inputs to define the default distribution as: i) a measure of mean default probability, and ii) a variance (or correlation parameter). These inputs can be calibrated based on historical data and adjusted for the qualitative assessment of the securitised assets. If possible, Scope also considers the performance of other European structured finance transactions exposed to similar collateral. The determination of each input s characteristics is generally subject to sensitivity analysis, based on information provided by the originator of the assets. Public historical data and proprietary data, as well as market studies by reputable providers or academic research, are also used. This approach notably applies to retail mortgage loan pools, consumer credit or granular pools of SME loans. To analyse the cash flow generated by the collateral, once the default pattern of the collateral pool has been determined, certain assumptions are made on: the default timing affecting the pool; the recovery amount and recovery timing upon default; the prepayment pattern of the pool; and other elements driving the cash flows of the transaction. Default timing, recovery patterns, prepayment profile and amortisation schedule and asset yield assumptions depend significantly on the characteristics of assets in the collateral pool. Scope may analyse the sensitivity of these assumptions Collateral market value risk Securitisation instruments may be exposed to collateral market value risk. For example, when the cash flows used to repay the instrument were generated from a sale of all or part of the securitised assets. Under such circumstances, the instrument will be subject to the price volatility of the sold assets. This risk will typically depend on characteristics that affect the asset s market liquidity, duration and currency. Large exposure to market value risk is uncommon for European securitisation transactions. Therefore, Scope assesses this risk using a transaction-specific approach to reflect the characteristics of the assets and their respective markets Sensitivity analysis, long-term assessment of senior protection buffers To supplement the quantitative approach described above and the model input calibration, Scope performs sensitivity analyses on the rated notes. Such a test includes selecting the variables that mainly drive the performance of the collateral pool and assessing which assumptions on these variables would cause a loss or a change in the instrument s rating. Afterwards the rating committee decides whether such potentially extreme assumptions correspond to a scenario whose likelihood remains consistent with the instrument s rating. Scope may also test whether different stress levels in the collateral pool might shift the rated instrument from investment grade to non-investment grade, or vice versa. For asset classes with sufficient data, Scope looks at long-term performance information from the market and takes a forward-looking view of the economic cycle. It incorporates market information into the base case assumptions such as macroeconomic factors correlated to defaults in a relevant asset class. For example, GDP and unemployment rates can be used to infer SME default assumptions. Scope s forward-looking approach may incorporate structural factors of European markets that may impact credit performance. When possible, we use a long-term distribution of portfolio defaults based on an average through-the-cycle performance reference to calculate expected loss under an AAASF stress, and the point-in-time distribution of portfolio defaults to calculate expected loss under a BSF stress. Under all other rating stress scenarios, from BBSF to AASF, Scope blends the results obtained from the point-in-time and long-term defaults. As the rating target becomes higher, this blending process assigns more weight to long-term distribution and less to the point-in-time distribution. 28 August / 19

11 With this approach, Scope prevents undue volatility on high investment grade ratings over the economic cycle. We can thus say that this methodology is not pro-cyclical. The stabilisation effect on sub-investment-grade ratings gradually becomes less noticeable, and the BSF ratings are driven only by the point-in-time distribution Counterparty risk analysis Materiality of financial and operational risk exposures Scope evaluates links between the risks of the rated instrument and various parties to the transactions. Scope assesses the materiality of a risk exposure to an external counterparty as excessive, material and immaterial. The agency differentiates between financial risk exposure and operational risk exposure. It also assesses how effective available measures are at mitigating or reducing risk exposure to counterparties in the specific context of the transactions. Scope s detailed approach to assessing counterparty risk in structured finance transactions is explained in Scope s Rating Methodology for Counterparty Risk in Transactions Asset management quality Qualitative factors play a crucial role in the analysis of structured finance transactions whose assets require intensive care, dynamic management or active work-out necessitates. Over the past few years, several CMBS transactions in Europe have demonstrated the importance of special servicers in maximising recovery proceeds for investors. Similarly, the active role of CLO managers helped to not only preserve the portfolio par value of several transactions during the recent credit crisis, but also accelerate the recovery of those transactions by seizing investment opportunities to reconstruct notional par. Scope s methodology emphasises qualitative credit judgment based on objective components to form its rating opinion. For transactions that involve active management of the collateral pool, i.e. to source, work out, add, exchange or remove assets, Scope examines the potential risks related to the asset manager s performance. The impact and importance given by Scope to this analysis depends greatly on the level of discretion left to the manager or servicer, and the effect of this on maximising/preserving/destroying the collateral pool s value. When Scope holds a positive opinion on the asset manager, Scope may regard additional management flexibility or a longer reinvestment period as credit-positive for the transaction. Scope s analysis of the asset manager includes a review of the manager s company, skills, expertise, processes, performance and track record, focusing on: The manager s economic incentives within the structure, e.g. the manager s remuneration, interests in the transaction s performance, and how or to what extent its interests are aligned with those of debt investors; The importance of the securitised asset segment within the manager s overall development strategy; and Standard of care and general liability as well as reputational risk for the manager. When available, we may rely on Scope s Asset Management Rating analysis. An Asset Management Rating does not measure credit risk, but rather the quality of management companies acting as service providers for the issuer Structure analysis In addition to assessing collateral risks and additional counterparty risks, Scope evaluates the structural characteristics of the rated transaction. Scope analyses the characteristics of the issuer, the structural aspects of the rated instrument and other transaction-specific risk drivers not captured in the previous steps of the analysis. Key structural elements reviewed by Scope in relation to a rated transaction include: (i) legal risks and tax considerations, (ii) structural enhancement features, (iii) systemic risks, and (iv) other transaction- or sector-specific risks Legal risks The analysis of the legal framework aims to assess the legal integrity of the structure and identify any legal issues or weaknesses that could affect the performance of the transaction. This part of the analysis considers tax aspects associated with the collateral that may affect cash flows within the transaction. To assess the integrity of the structure, it is important to evaluate the likelihood that the issuer could default for reasons unrelated to collateral or counterparty risks. In some instances, even when collateral and counterparty performance are satisfactory, such defaults could lead to a liquidation of the collateral, and expose the rated instrument to market value losses. A key part of this analysis includes a review of the issuer s bankruptcy remoteness. It is impossible to remove the risk of issuer bankruptcy altogether. However, issuer can generally be protected through standard securitisation features specific to the issuer s nature, and activity and relationships with parties to the transactions. Scope evaluates the strengths of protective elements implemented in the rated transaction. These elements include the legal nature of the issuer, the restriction of its activity, its ownership structure and its limited liabilities. Scope also reviews the limited-recourse and non- 28 August / 19

12 petition provisions included in the transaction contracts which prevent other contractual parties from causing the issuer to default. This analysis allows Scope to form an opinion on the issuer s insolvency risks. In many securitisation transactions, the true sale of such collateral to the issuer by the seller which is generally also the originator is a key mechanism to isolate the risks of the securitised collateral and its originator. For a large majority of structured finance types, Scope assesses the legal robustness of the true sale to evaluate the risk of collateral claw-back and consolidation on seller s balance sheet should the seller default shortly after the sale of the collateral. In addition, Scope may examine whether the obligors of the securitised assets and the originator of those assets could become subject to cash flow set-offs upon default of the originator (set-off risk). If the obligor of an asset holds a cash deposit account at the originator of the asset, such a set-off may allow the obligor to give up its deposit against a reduction of the originator s claim. This set-off risk may therefore result in lower cash flows generated by the collateral to the benefit of the issuer and available to repay the rated instrument. Scope generally assesses the risks related to unclear or broad definitions of the legal documentation, for example, pertaining to key transaction mechanisms such as the definitions of transaction default and termination events. Scope considers opinions provided by third-party experts on the likelihood of an issuer s tax liability affecting a transaction s cash flow as well as the issuer s ability to pay principal and interest on the rated instrument. To rate a transaction, Scope would need access to the tax analysis of the transaction. For more about legal risks on structured finance transactions, please review Scope s analytical considerations in Legal Risks in (14 January 2014) Structural and enhancement features and cash flow analysis Structural features can contribute to the improved or weakened performance of the transaction from the perspective of a rated debt. Given the wide variety of asset classes in the structured finance landscape, it is impractical to compile a full list of every structural item reviewed during the analysis. However, key structural features generally include: i) the priorities of interest and principal payments under the rated notes; ii) payment frequency of the instrument; iii) enhancement features such as excess spread, cash reserves or liquidity buffers; iv) mismatch of cash flows between the underlying collateral and the issuer s financial obligations; v) coverage of the issuer s ordinary and extraordinary expenses; vi) guarantees or hedging mechanisms; vii) covenants, performance triggers or other protective mechanisms; and viii) call, early redemption, asset substitution or new issuance features. Scope analyses the efficiency and impact of such features and generally includes the most important in the quantitative analysis of the transaction. Scope combines quantitative methods based on simulation or standard distribution laws (see collateral risk analysis section) with a cash flow analysis that reflects key cash flow characteristics of the transaction. Such characteristics include interest and principal priority of payments, sources of credit enhancements, liquidity lines, reserve accounts or costs borne by the structure over the entire term of the rated instrument. In each scenario, we calculate a loss for the rated instrument to produce the expected loss associated with the rated instrument. 2 The objective of the cash flow analysis is to replicate as closely as possible the structure of the rated transaction and reflect structural and enhancement features. As a result, the analysis makes it possible to measure how, when and to what extent the cash flows generated from the collateral assets cover costs and liabilities borne by the structure. Many assumptions for cash flow analysis are based on the legal documents related to the rated issuance. Most constant parameters relevant to income and expense assumptions are typically derived from contractual terms governing the structure. However, when parameters are not contractually specified, or include provisions for variable components, these are incorporated into the analysis based on Scope s qualitative assessment. For structured finance transactions with a simple structure and where allocation of cash flows does not drive the rating outcome, Scope may derive an expected loss directly by modelling the loss distribution of the collateral pool and allocating losses to the rated instrument in each scenario, instead of computing the expected loss through a fully-fledged cash flow allocation. For instruments exposed to non-tranched portfolios and not subject to material cash flow enhancements, the expected loss of the instrument may equal a simple weighted average of the expected loss of each asset securing the repayment of the instrument Country and industry risks In addition to the analysis of collateral and counterparty risks, Scope carries out a qualitative, forward-looking evaluation of the trends affecting the country and the industry sector the transaction is exposed to. It considers macroeconomic, environmental, sovereign or industry risk factors that may impact the performance of the rated instrument. 2 The expected loss for the structured finance instrument is therefore the sum-product of i) the probability of occurrence of a given scenario associated with a given asset performance; and ii) the loss derived from the cash flow model and specific to the transaction in each scenario. 28 August / 19

13 Scope does not see a valid analytical reason for systematically capping structured finance ratings by sovereign ratings. As a result, Scope does not limit the maximum rating achievable by a securitisation based on the sovereign credit quality of the country of the issuer or of the securitised assets. However, Scope believes that credit ratings allow investors to consistently compare credit risks between different instruments and securitisation types across different locations. As a result, credit ratings must adequately and consistently reflect the credit risks of an instrument, including risks arising from an exposure to a country with weak economic fundamentals. As a result, Scope includes a measure of country risk in its analysis. Scope assesses convertibility risk (risk of eurozone exit) and the risk of institutional meltdown in the context of the tenor of each rated transaction. Scope takes macroeconomic factors into account for its ratings. A material exposure of a transaction to a financially weak domestic sovereign is viewed as a material credit risk that will negatively impact the ratings of the transaction. However, Scope believes the credit assessment of a sovereign is not an adequate anchor for applying a rating cap or a ceiling, particularly in eurozone countries Representation and warranties Scope considers the strength and expected impact of representation and warranties made by the parties to the transaction, including those made by the originator of the assets included in the collateral pool. In some instances, Scope may have to rely on audits on the pool to supplement representations and warranties Monitoring The monitoring process of the rated instrument starts immediately after a rating has been assigned. Scope continuously monitors the credit risk and performance of the collateral, as well as the key counterparties to the transaction. As a result, Scope may adjust the rating if the instrument s performance materially differed from expectations. Scope typically monitors structured finance transactions based on performance reports produced by the management company, the collateral agent, or the trustee in the transaction, as well as information from the originator or transaction sponsors. If information provided by issuer or its agent is irrelevant, of insufficient quality, or inappropriately delayed, Scope will have to consider the impact of such deficiencies on the rating, and may even be forced to withdraw the rating as insufficient information makes monitoring impossible. The ratings are monitored continuously and reviewed at least once a year or earlier if warranted by events. The monitoring process typically involves frequent high-level checks and a detailed annual review. Since Scope s ratings aim to provide a long-term view based on the maturity of the rated instrument, a temporary performance dip may not necessarily be a reason to downgrade the rating of an instrument. Scope aims to avoid rating pro-cyclically and, where possible, seeks to anticipate the effect of cyclical trends on particular structured finance asset classes. This translates into ratings that are forward-looking rather than reactive. Scope may therefore include any change in opinion regarding the outlook for the collateral asset s credit cycle in its re-assessment of the collateral s credit quality. 28 August / 19

14 Appendix I. Scope s idealised expected loss and default probability tables 3 Figure 2. Maximum expected loss for each rating level and expected life of the instrument Scope Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10 AAA 0.001% 0.003% 0.008% 0.015% 0.025% 0.038% 0.055% 0.076% 0.101% 0.130% AA % 0.005% 0.012% 0.023% 0.039% 0.059% 0.084% 0.115% 0.151% 0.193% AA 0.004% 0.013% 0.028% 0.050% 0.079% 0.115% 0.158% 0.208% 0.265% 0.329% AA % 0.016% 0.037% 0.068% 0.108% 0.157% 0.215% 0.281% 0.356% 0.439% A % 0.028% 0.064% 0.114% 0.175% 0.248% 0.331% 0.422% 0.523% 0.632% A 0.021% 0.054% 0.106% 0.172% 0.252% 0.344% 0.447% 0.560% 0.683% 0.815% A % 0.082% 0.160% 0.260% 0.377% 0.509% 0.653% 0.810% 0.978% 1.156% BBB % 0.170% 0.306% 0.462% 0.635% 0.823% 1.025% 1.240% 1.465% 1.702% BBB 0.106% 0.287% 0.499% 0.733% 0.987% 1.258% 1.543% 1.842% 2.153% 2.475% BBB % 0.533% 0.923% 1.334% 1.758% 2.192% 2.634% 3.083% 3.538% 3.998% BB % 1.142% 1.713% 2.280% 2.841% 3.398% 3.950% 4.499% 5.044% 5.586% BB 0.889% 1.778% 2.668% 3.526% 4.354% 5.154% 5.929% 6.683% 7.417% 8.133% BB % 2.541% 3.812% 5.014% 6.147% 7.220% 8.241% 9.217% % % B % 4.604% 6.280% 7.691% 8.952% % % % % % B 2.971% 5.941% 8.032% 9.746% % % % % % % B % 8.971% % % % % % % % % CCC % % % % % % % % % % CC % % % % % % % % % % C % % % % % % % % % % Figure 3. Maximum default probability for each rating level and expected life of the instrument Scope Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10 AAA 0.003% 0.007% 0.015% 0.029% 0.049% 0.076% 0.110% 0.151% 0.201% 0.260% AA % 0.010% 0.025% 0.047% 0.078% 0.118% 0.169% 0.230% 0.302% 0.386% AA 0.007% 0.025% 0.056% 0.101% 0.159% 0.231% 0.316% 0.416% 0.530% 0.657% AA % 0.032% 0.073% 0.135% 0.216% 0.314% 0.430% 0.563% 0.712% 0.878% A % 0.056% 0.128% 0.227% 0.351% 0.496% 0.661% 0.845% 1.046% 1.264% A 0.041% 0.107% 0.211% 0.345% 0.505% 0.689% 0.894% 1.120% 1.366% 1.630% A % 0.164% 0.321% 0.520% 0.754% 1.018% 1.307% 1.620% 1.955% 2.311% BBB % 0.341% 0.612% 0.924% 1.270% 1.646% 2.050% 2.479% 2.931% 3.404% BBB 0.211% 0.574% 0.998% 1.467% 1.975% 2.516% 3.087% 3.684% 4.306% 4.950% BBB % 1.066% 1.846% 2.667% 3.516% 4.384% 5.269% 6.167% 7.076% 7.996% BB % 2.284% 3.426% 4.559% 5.682% 6.795% 7.900% 8.998% % % BB 1.778% 3.557% 5.335% 7.053% 8.709% % % % % % BB % 5.082% 7.624% % % % % % % % B % 9.208% % % % % % % % % B 5.941% % % % % % % % % % B % % % % % % % % % % CCC % % % % % % % % % % CC % % % % % % % % % % C % % % % % % % % % % 3 Scope relies on idealised expected loss to benchmark quantitative results produced in structured finance transactions, covered bonds or Project Finance. Scope assumes a 50% loss given default to derive the default probabilities associated with a specific rating over a given time horizon. Scope's assumptions reflected in this table may be modified. The table is provided purely as an accommodation by Scope, and such an accommodation creates no obligations of any kind on the part of Scope other than those parts of the regulation applicable to European credit rating agencies. 28 August / 19

15 Appendix II. Technical note on the expected loss framework Scope analyses the probability-weighted average loss, the expected loss, and the probability-weighted average life (WAL), i.e. the expected WAL, for each rated tranche. The rated structure is tested for every possible portfolio default rate from 0% (no defaults) to 100% (the entire portfolio defaults). Scope compares the expected loss and the expected WAL calculated for a given tranche to Scope s idealised expected loss table as an indicator of a rating for a given instrument. The probability of each possible default rate is taken either from an idealised distribution such as the inverse Gaussian distribution, or from the probability distribution produced by a Monte Carlo simulation. Losses in the structure generally result from the application of rating-specific recovery assumptions and from costs of carry. The probabilities are used to weight the losses obtained for each rated tranche under every default rate scenario. This is shown in expression (1). Figure 4. Diagram of the structure analysis and cash flow model implementing the expected loss framework Cash-flow model Structure, Documents Tranche-specific: expected loss; expected WAL; missed payment probability Idealised expected loss table Portfolio default rate distribution Asset analysis Amortisation profile, default timing Recovery rates and timing, cure rate Interest rates, prepayments and other Tranche-specific: Default rating indication (timely payment see Appendix III) Tranche-specific: Model-driven rating indication Tranche-specific: Model-driven expected loss rating indication (expected loss) The loss of a tranche under a given default rate scenario i, Li, is the difference between the par value of the tranche and the present value of all principal and interest cash flows for the investor, discounted at the promised rate of the tranche being considered as seen in expressions (2) and (3). Similarly, the probabilities are used to weight the different WALs resulting in the cash flow model for each rated tranche under every default rate scenario from 0% to 100%. This is shown in expression (4). For consistency, the WAL of a given default rate scenario i is derived by considering all principal and interest cash flows for the investor. (1) (2) (3) (4) (5) Figure 5 illustrates the losses on each level of a three-tranche structure for all portfolio default rates. The loss rates are expressed as a percentage of the tranche notional at closing. In this example, it is notable how class C benefits from excess spread that is not trapped by the transaction until the first assets are classified as defaulted. This allows class C s maximum losses to be lower than the maximum possible for class B. The probability-weighted loss for class B would be, of course, far smaller than that of class C. 28 August / 19

16 Probability Tranche loss General Rating Methodology Figure 5. Sample portfolio distribution and corresponding losses in a three-tranche structure 0.7% 0.6% 0.5% 100% 90% 80% 70% 0.4% 0.3% 0.2% 0.1% 60% 50% 40% 30% 20% 10% 0.0% 0% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100% Portoflio default rate Normal Inverse Loss A (Timely) [RHS] Loss B (Timely) [RHS] Loss C (Timely) [RHS] 28 August / 19