New Software Can Provide Risk Models

Article excerpt

The securitization of financial assets represents an important new tool in the financial markets of today. The basic premise is a simple one: Net value may be created by redistributing risk, and producing custom, liquid instruments from nooliquid assets.

This technique is now increasingly being applied to the transformation of the nonliquid portions of a corporation's balance sheet into liquid, saleable securities, a potentially effective means of raising capital. Today, the asset securitization market is over $500 billion and growing rapidly.

Since asset securitization involves the redistribution of risk, we must consider the different forms of risk, which may be shifted. A nonexhaustive list of risks might include:

* Maturity-mismatch.

* Prepayment Volatility.

* Interest Rate.

* Credit/default.

Each of these problems has a corresponding securitization technique as a solution.

Maturity Mismatch

This form of risk typically occurs when there is a substantial disparity between the maturity of an organization's financial assets and its liabilities, which might lead to significant value gaps if the shape of the yield curve were to change.'

This is also the most common problem with basic mortgage securities, in which principal payments may occur more or less steadily over a period of up to 30 years, rather than in a single bullet payment as in the case of Treasury or corporate bonds.

An instrument with such a long and ill-defined payment period is far less attractive than an instrument with a sharply concentrated maturity.

These problems are generally solved by tranching the cashflows of the security by maturity class, allowing one tranche to receive the earliest principal payments, others to receive the intermediate principal payments, and the remainder to receive the latest payments. This results in the production of bonds of short, intermediate, and long maturity.

For mortgages, a crucial element of risk is the risk of prepayment, namely the uncertainty regarding when and if individual homeowners will prepay their mortgage loans, effectively resulting in the calling of the bonds out at par.

Under certain circumstances, this might result in a tranche of a mortgage-backed security with a stated maturity of 20 years having an actual coverage life of just six months, with all the entailing disruption of investment strategy and maturity mismatch problems.

The solution to this problem is the assignment of certain tranches to act as "buffers" for other tranches, protecting them from a considerable degree of prepayment uncertainty by absorbing unexpectedly high or low prepayment rates.

Ideally, the increase in market price of the buffered tranches (usually called PACS, TACS, or VADMS) is somewhat greater than the decrease in market price of the buffering tranches, which become especially volatile with regard to prepayments (these are usually called support classes).

Large classes of investors desire to hold floating-rate assets, which are substantially protected against changes in interest rates' Since most assets used in asset securitization have fixed coupons, the use of such assets to produce floating-rate bonds is only possible if inverse-floating rate bonds are also created simultaneously.

This is typically done in CMO/Remic structures, in which certain bond classes are protected against interest rate risk by being assigned floating-rate coupons, while other classes are correspondingly made especially vulnerable to interest rate risk by functioning as inverse floaters.

Credit/Default Risk

This is one of the most crucial aspects of asset securitization, since broad classes of assets are - subject to considerable credit risk, whether they are mortgage whole loans, consumer loans, or commercial loans.

Unstructured pools of unrated assets are relatively unattractive investments, since they fall into no clear investment category. …