Collateralized debt obligations

Diplomarbeit von Markus Lorenz

Abstract:

This work aims to give the reader a holistic introduction to Collateralized Debt Obligations (CDOs), an asset category which has recently experienced both popularity and criticism. Collateralized Debt Obligations represent a subset of asset-backed securities. As opposed to classical types of asset-backed-securities like mortgage-backed securities or credit card debt-backed securities, a Collateralized Debt Obligation is a vehicle transforming bank loans or commercial paper into tranches of traded securities.

While Collateralized Debt Obligations have been an established part of the U.S. fixed income market, it was only recently that academics showed interest in this asset category. From an asset pricing standpoint, CDOs represent a challenge as credit risk from a heterogeneous pool is passed through to tranches. Hence, asset pricing models have to account for expected defaults and default correlation on the one hand while incorporating the structural support the CDO is offering to the debt tranches on the other.

Also, regulatory agencies such as the Basel Committee on Banking Supervision have increasingly covered CDOs and their use in credit risk management, thus further stimulating interest in this asset category.

The report is mainly organized in three parts. The first part presents the basic ideas of Collateralized Debt Obligation as well as their structure and principal economics. Part II is the core of the report focusing on the aforementioned asset pricing problem and presenting various models to cope with it. Finally, the third part presents some of the multifaceted applications of Collateral Debt Obligations and concludes with an outlook for the product category. Here, special focus is laid on the European and German market as this is seen as a major area for growth.

Table of Contents:

Text Sample:

Chapter 5.1.1, Default Rates and Severity:

A starting point for assessing the credit quality of the collateral pool is the rating of its assets. However, such ratings may be unavailable for individual bank loans.

The difficulty of quantifying the credit quality of the commercial loans to be included in a CDO transaction has proved to be a major impediment to obtaining an acceptable rating for purely CLO transactions. In CBO transactions, which have structures that are virtually identical to those of CLOs, yet the underlying bonds collateralizing the CBO obligations usually have pre-existing credit ratings, which permit the rating agencies to analyze the proposed CBO structure by reference to the known ratings.

In contrast, commercial loans are not generally rated by the rating agencies, and the task of assessing the credit quality of a large portfolio on a loan-by-loan basis can be very cumbersome.

One approach to this problem is the attempt to concatenate the lender's internal credit rating system and loan underwriting criteria to the rating agency’s rating. While the initial installment of such a correlation pattern is time-consuming, the rating process on further transactions a single lender issues should be significantly simpler.

Another alternative to a loan-by-loan evaluation is the actuarial analysis on a portfolio. This actuarial approach requires extensive information about payment, delinquency, default and recovery characteristics over time for a fairly homogeneous product type, preferably originated or serviced by a single bank or company. This seems currently feasible with large European CDOs as the recent European trend is towards very large CDO transactions from a single lender as discussed in Herrmann and Tierney. In applying this approach, rating agencies apply their own historical data to the default rate and recovery rate factors in the following well-known equation:

Expected Loss = Default Probability Loss Given Default.

However, rating agencies still have to solve the problem that data on default probability and recovery rates are clustered according to the credit ratings and relative seniority, which again forces the evaluator to classify unrated debt in consonance with the established categories.

Pool Diversity:

As discussed in section 4.1.1 correlation is a major factor in evaluating CDOs as its ability to serve senior tranches is especially sensitive towards default correlation in the collateral pool.

Concentrations of loans with borrowers in the same or related industries or borrowers in the same geographic area are viewed as increasing the risk of a portfolio, and usually result in rating agencies downgrading the CDO tranches or demanding higher levels of credit enhancement. Generally speaking, the adverse impact of concentrations of loan types is decreased by diversification of the loan portfolio, which may be achieved by increasing the portfolio size or decreasing the average size of a loan.

All the rating agencies use cash flow models to stress test the collateral pool and its ability to service outstanding debt under a variety of economic, interest rate, default and loss scenarios.

In addition to those tests, Moody’s has established its diversity score based on the binominal expansion method which is discussed in section 4.1.2. Despite its lack of deep theoretical foundations it has become a standard metric in evaluating CDOs and is typically being referred to in the prospectus accompanying the initial offering.

Credit Enhancement Levels:

Credit enhancement levels represent a means of ensuring that mezzanine and senior tranches may be served from the cash flow derived from the collateral portfolio.

Based on the aforementioned analysis of credit quality, expected loss and pool diversity, the rating agencies set credit enhancement levels required for senior and mezzanine classes to achieve the desired rating. This is based on the level of losses that a pool can sustain while continuing to service rated debt issues. The lower the credit quality of the collateral and the higher the desired rating, the larger the required credit enhancement.

For example, for a diversified pool of B-rated collateral, the base case default rate for Moody’s is 31.8%. Assuming a conservative recovery rate of 30% the expected loss is 22.3%. But to obtain a target rating of Aa2, Moody’s analysis shows the rated class must be able to survive a 39% loss rate or 55.7% default rate on the collateral without default of the securities.