(Editor's notes: These criteria have been superseded by the article titled "Taiwan Ratings' Principles Of Credit Ratings," published on May 19, 2011)

2010/08/06

To accommodate the inherently dynamic and flexible nature of the global structured finance market, Taiwan Ratings Corp. employs a largely principles-based methodology for assigning and monitoring ratings on structured finance securities. Our approach helps us respond to the market's evolving needs and enhances the comparability of the ratings we assign by considering:

• The constantly changing and innovative nature of structured financings;
  
• The differences in the maturity and sophistication of markets, as well as particular sectors and/or participants within those markets;
  
• The differences and changes in regulatory and legal frameworks among countries and market sectors;
  
• The differences and changes in market conventions and industry practices;
  
• The cultural differences among markets that may influence behavior and performance;
  
• The unique considerations at the asset, originator, and transaction levels;
  
• The key characteristics of structures or asset classes that may not have been securitized before; and
  
• The interplay of the above factors, which may minimize the significance of each risk in determining default risk for a given transaction relative to another transaction.

Core Methodologies For Analyzing And Rating Securitization Transactions

Our analytical framework for structured finance securitization ratings comprises five key areas of analysis:

Credit quality of the securitized assets

For most asset-backed securities (ABS), residential mortgage-backed securities (RMBS), and collateralized debt obligations (CDOs) whose common thread is that they pool financial assets such as loans, receivables, or corporate debt, our analysis focuses on determining, under "worst-case" stress scenarios that are commensurate with the rating, the following:

• The portion of the original asset pool that will default or otherwise be lost over the life of the security;
  
• The portion of the defaulted or lost assets, if any, that can be recovered from exercising rights to the underlying collateral, security, or through other means; and
  
• The potential ultimate loss on the debt issue as a result of determining the two previous items.
  

To this end, Taiwan Ratings reviews information from both internal and external sources and employs a variety of methodologies and quantitative tools, such as default and cash flow models. Ultimately, we choose credit enhancement levels that can protect against these scenarios so that the likelihood investors will receive timely interest and ultimate principal repayments no later than the issue's legal final maturity date is commensurate with the ratings assigned.

For commercial mortgage-backed securities (CMBS), corporate and whole business securitizations, and future flows ABS, where debt repayment is predicated largely on future revenue generation (for example, rental income and commodity sales) and/or the underlying asset's market value, we stress-test assumptions about those factors that affect the valuation of the underlying asset or operating company and its ability to generate future cash flows. Ultimately, asset values and future cash flows coverage must be sufficient to ensure that the likelihood of timely interest and ultimate principal repayment on the rated debt by the issue's legal final maturity date is commensurate with the ratings assigned.

Legal and regulatory risks

For all types of securitization structures, we determine, under the applicable legal and regulatory regime, whether the securitized assets have been appropriately isolated from the bankruptcy or insolvency risk of any entities that participate in the transaction. Specifically, we seek assurance that such bankruptcy or insolvency would not impair the securitization structure's rights, including timely access, to the securitized assets and the cash proceeds those assets generate. Typically, we focus on the entity that originated and owned the assets before the securitization. Often, we look for a "true sale" of assets from the originator/seller to a "bankruptcy-remote" issuer (a special-purpose entity (SPE)). However, in some jurisdictions, we may look for true control over the assets through effective security rather than a true sale. We also review the enabling documentation for any SPEs that are part of the securitization structure to ensure that the likelihood such entities could become bankrupt or insolvent is sufficiently remote in the context of the rating requested and the relevant legal jurisdiction. Taiwan Ratings' analysts work with legal counsel to evaluate the securitization structure's legal integrity.

Payment structure and cash flow mechanics

Taiwan Ratings reviews relevant transaction documentation and cash flow modeling output, if appropriate, to understand the mechanisms through which the securitized assets allocate cash flows they generate to pay interest and principal payments to investors, as well as transaction expenses, such as servicing and trustee fees when due. We give particular attention to the documented payment priorities that determine the relative seniority of the various investors and transaction payment obligations across all rated classes of debt. We also analyze the "trigger events" and other performance or credit-driven events that are built into the structures, which dictate the use of excess cash and cause changes to the terms of investor payments. In addition, we review any potential cash flow leakages from the structure that may reduce the likelihood of meeting investor payments when due. For securitization structures that depend on third-party payment obligations, such as insurance policies, guarantees, bank credit and liquidity facilities, and derivatives instruments, we examine the payment mechanics behind these obligations to determine their compatibility with the rated securitization structure's payment terms.

Operational and administrative risks

For all new issuers, Taiwan Ratings reviews the entity that will be responsible for servicing or managing the securitized assets to determine whether it's competent and capable of managing the program post-closing. The servicer's or manager's duties include collecting asset payments, working with delinquent obligors, tracking cash receipts and disbursements, disposing of collateral, and providing timely and accurate investor reports. For actively managed CDOs, we review the asset manager's capabilities and track record. For most commercial asset, corporate or whole business, commercial real estate, and future flow securitizations, we also review the operational and administrative risks involved in the day-to-day management and administration of revenue-producing assets or properties.

Since we may rate the servicer of the securitized assets lower than the securitized issue, or not rate it at all, we consider scenarios where the original servicer may be unwilling or unable to perform its duties for the life of the transaction. As such, as part of our analysis we determine the likelihood of an alternative or backup servicer's availability and willingness to step into the servicing functions for the transaction if the original servicer resigns or is removed at any time during the life of the transaction. Several factors may influence our opinion, including the sufficiency of the servicing fee to attract a substitute servicer, the seniority of the servicing fee in the transaction's payment waterfall, the availability of alternative servicers in that sector or region, and the specific characteristics of the assets and servicing platform that may hinder an orderly transition of servicing functions to another party.

Counterparty risk

We analyze a transaction's exposure to counterparty risk by reviewing each third-party obligation in the context of the counterparty's willingness and ability to meet its obligations and the rating we plan to assign. Some examples of common third-party obligations are financial guarantees, bank liquidity or credit support facilities, letters of credit, and interest rate and currency swaps. Our counterparty criteria framework for securitized debt considers the nature of the link between the debt rating and the counterparty rating by recognizing that there is a continuum between a "weak-linked" analysis and a completely "de-linked" analysis. In a weak-linked analysis, the debt rating is directly tied to the counterparty rating. In the de-linked approach, structural features significantly reduce the likelihood of a lower counterparty rating negatively affecting the debt rating. Since the majority of recent securitization structures fall between the weak-linked and de-linked approaches, our analysis of the counterparty risk incorporates both the counterparty rating and structural features, such as rating downgrade triggers and collateralization and/or counterparty replacement remedies.

Sovereign risk may impact a structured rating in cases where the issuer, the assets, or the obligors behind the assets could be subject to government interference that could adversely impact payments on securitized debt. For example, such actions can take the form of governmental controls over the transfer and convertibility of foreign currency. Country risk and the likelihood of government interference are considered in the context of the securitization's structural features and the rating requested.

Supporting Guidelines To Aid Market Transparency And Expedite The Rating Process

To complement our principles-based approach, we periodically release operating guidelines for analyzing specific markets, asset classes, or structural features that provide practical examples of how we may employ the rating process in actual circumstances. These guidelines don't, however, preclude an issuer from proposing alternative approaches that we would independently assess to ensure a commensurate and comparable rating outcome relative to other securities we rate. In addition, they don't preclude the rating committee from determining the relative weighting of these factors within its assessment or determining that other factors need to be taken into account, given the specific set of circumstances under consideration at that time.

When we assign ratings to structured finance securities, we use a general framework and established guidelines, as well as various quantitative techniques and models, to enhance the rating committee's qualitative opinions. These qualitative judgments are an integral part of our rating process. Through this process, we can address and assess a transaction's country, collateral, and transaction-specific factors, as well as changes in the market and the environment, in our rating outcomes.

The Range Of And Meaning Behind Our Structured Finance Ratings

Taiwan Ratings' structured finance ratings generally address the timely payment of interest and the ultimate return of principal no later than the legal final maturity date, according to the transaction terms. Furthermore, our ratings only address the credit quality of structured finance securities. However, other risk factors may affect a security's yield or total return. For example, our ratings don't address prepayment or any reinvestment risk that may result if investors receive principal ahead of the scheduled or legal final maturity. Some securitization structures are more prone to this risk than standard corporate debt because the underlying assets (for example, mortgage loans) may be subject to partial or full prepayments, which, in turn, are typically passed through to investors. In addition, some structures contain "early amortization triggers" that result in accelerated principal repayment if certain adverse credit or performance events occur.

We base our ratings framework on the likelihood of default rather than expected loss or loss given default. In other words, our ratings at the rated instrument level don't incorporate any analysis or opinion on post-default recovery prospects. Therefore, two or more debt classes or tranches within the same securitization structure can have the same rating based on their default probabilities even though they may have different expected loss profiles due to differences in post-default recovery prospects or their relative priority in the capital structure.

While our ratings don't directly incorporate post-default recovery or expected loss analytics, we do use an expected loss approach to analyze securitizations at the asset pool level to determine credit enhancement amounts that we believe are appropriate for a given rating. For example, our CDO and RMBS default models generate assumptions for asset level default probability and loss given default, resulting in an expected loss determination that can become the basis for sizing minimum credit enhancement levels at different rating categories. However, we don't subsequently apply this expected loss approach at the rated debt level.

Our ratings represent a uniform measure of credit quality across all types of debt instruments. In other words, a 'twAAA' rated corporate bond should exhibit the same degree of credit quality as a 'twAAA' rated securitized debt issue. However, we recognize that the capital markets may not always price similarly rated debt types the same, all things being equal. This is also true when comparing different securitized issues. Such differences may be based on both credit and non-credit or market considerations, including perceived prepayment risks based on asset or structural characteristics; seller/servicer characteristics; the asset class' historical track record; the availability of historical performance data; and market liquidity considerations, including the depth of secondary markets in certain sectors or markets.

Empirical studies, such as transition and default reports, provide an opportunity for us to measure and compare the characteristics of default rates across all types of debt instruments, including structured finance. In particular, comparing structured finance and corporate default rates is challenging for several reasons, including:

• The structured debt default rate is measured at the issue rating level, while its corporate counterpart is measured based on the issuer's credit ratings;
  
• The market argues in favor of using a vintage-based default measure for the structured debt rather than the corporate's cohort approach;
  
• The structured finance defaults mainly come from the post-2001 period, while the corporate defaults cover a longer historical timeframe; and
  
• The recent default volatility and default correlation behavior of structured finance ratings may imply the need for a longer horizon to measure the "true" default rates.

Overall, recent empirical data suggests different default performance among structured finance and corporate ratings, and even among ABS, RMBS, CMBS, and CDO issues. However, the historical performance data must be viewed in light of the previously cited challenges regarding the limited amount of available structured finance default and transition data, and the differences between how we calculate structured and corporate default rates. These considerations call into question the statistical validity of any differences in performance between structured and corporate debt and among the different structured finance sectors based on the empirical data.

In addition, the differences in historical experience partly reflect the structured finance market's relatively short history and the differential impact of economic and other credit-related events on the default performance of a given debt type. As we generate additional historical default and transition data over time, we will continue to review and update our ratings criteria.