Improved bond valuation methodology benefits mortgage bankers and correspondents

Larry Barnett, co-founder and principal, BlackBox Logic LLC

As published in Scotsman Guide's Residential Edition, May 2012.

Much has been written about the cause of the meltdown in the mortgage-backed securities (MBS) market. Some argue that bond investors and portfolio analysts need enhanced looks at downstream loan information to improve bond valuation methodologies. Related to this discussion is the need for an improved feedback loop between the initial underwriting criterion and the subsequent performance of the underlying mortgages. Mortgage bankers and correspondent lenders — whose companies may be servicing these loans — should keep themselves educated and up-to-date about proper bond valuation analysis, as this ultimately will affect their companies’ bottom lines.

It’s now become commonplace to replace the data at the deal level — data that initially was the foundation for bond valuations — with an in-depth analysis that begins with reviewing loan-level attributes. By incorporating detailed data on individual mortgages, more accurate pricing will evolve for securities without liquid markets.

This accuracy, however, will be contingent on correlating the initial and updated views of underwriting attributes that describe borrowers’ credit profiles and property characteristics with mortgage delinquencies and foreclosures.

Mortgage professionals should know that the core of this bond valuation analysis concerns the calculation of individual scheduled and unscheduled mortgage payments made by a given borrower into a residential MBS trust. As transparency improves, the aggregation of information on unscheduled cash flows related to voluntary mortgage prepayments, delinquencies and loss severities caused by the sale of distressed assets — and the timing of those losses — is enhanced, as well. This in turn leads to more accurate estimates of future cash flows, improving bond valuations and enhancing the feedback loop between the initial mortgage underwriting and loan performance.

A recent advancement in the calculation of these future unscheduled cash flows comes from progress being made in determining the amount of principal and interest that have been advanced to the trust by the mortgage servicer. Thus, identifying loans modified by the servicer in an attempt to keep struggling borrowers current on their payments is a key component in the analysis of servicing advances and their subsequent reimbursements.

The estimation of servicer advances at the loan level can make bond valuations more accurate, while also improving models that tie the initial underwriting to downstream performance. Mortgage bankers and correspondent lenders who know how to properly evaluate — and use — their businesses’ data can add lift to even the most junior and senior bonds in a given mortgage trust.

Analyze and evaluate

Servicer advances have become a significant tool in predicting future losses on individual mortgages and unscheduled cash flows coming into trusts. These cash flows can be used by mortgage originators to take a longer and more in-depth look at servicer performance, as well as the effectiveness of mortgage underwriting.

Of course, any analysis of these factors should be in addition to more traditional methods of predicting loss severities and calculating future cash flows. This analysis can be vital, as servicers are obligated to advance timely principal and interest into a given trust for borrowers who are unable or unwilling to make their monthly payments.

In general, however, the servicer continues to advance until they consider the fact that proceeds from the eventual sale of a distressed asset will not cover the advanced monthly principal and interest. In other words, stopped advances are a window into loss estimations made by the servicers themselves.

The counterbalancing number also is of interest to investors, namely the amount of reimbursements that the servicer receives from the trust to pay themselves back for these cash outlays. This number is taken off the top and is not remitted to the bond investors when the property is sold as a real estate owned (REO) property or modified to provide troubled borrowers some relief.

Fortunately, with a fair amount of data manipulation, a user of loan-level data sets can identify the loans that were advanced, giving them the ability to incorporate data into relevant models to predict cash flows.

Page: 1 2 Next