Sixth in a series

Loan servicing is one of the least understood aspects of mortgage banking, but one that has wreaked havoc with borrowers and at times posed significant financial volatility for banks and other firms engaged in this activity. 

A major contributing factor to the unprecedented problems in servicing nonperforming loans during the mortgage meltdown is the simple flat pricing structure for compensating servicers for their activities.  A fiercely competitive market during the housing boom, coupled with this pricing model, led to vast underinvestment in systems and personnel capable of handling the onslaught of nonperforming loans. Servicers were caught flat-footed when the crisis hit.

The failure to sufficiently price for the costs and risks inherent in the servicing process, under both benign and stressful environments, is an important lesson for servicers as it was for mortgage originators.  The Federal Housing Finance Agency is considering adjustments to the servicing compensation model. The best alternatives are those that offer a risk-based approach to the problem.

Such an approach is likely to improve the borrower's experience with the servicer, reduce the volatility of a traditionally difficult asset to hedge and open the door for other market participants to compete for the servicing asset – thereby addressing the concerns over excessive concentration in this activity by a few large players.

Under today's pricing structure for Fannie Mae or Freddie Mac loans, servicers retain a minimum servicing fee of 25 basis points on the outstanding principal remaining of loans being serviced, regardless of the mortgages' risk profile.  During normal periods when defaults are relatively low, the primary activities of mortgage servicers include processing payments and other related administrative functions, which are characterized by a high degree of automation.  Servicers can dramatically improve operational efficiencies by scaling up their operations and leveraging technology.  The fees earned by servicers more than compensate them for the costs borne during these periods.

However, the economic pressures to improve efficiency reduce investment in the predominately manual process associated with servicing nonperforming loans. The limitations of the flat-fee servicing model are laid bare once loans begin defaulting. Dunning late payers, tracking down unresponsive borrowers, negotiating loan modifications: these aren't jobs for a machine, or inexperienced temps.

In addition, the servicer's compensation, an interest-only strip off the mortgage interest payments from the borrower, poses significant risks.  These assets, referred to as mortgage servicing rights, are created from cash flows associated with the servicing process including the servicing fee, ancillary income, float interest income and other revenue sources.  MSR valuations are notoriously unstable, due to the risk that borrowers will pay off their mortgages sooner than expected, the fair-value accounting treatment and the dearth of market pricing information for this asset.  And hedging the interest rate risk of such assets poses its own unique challenges given the complex relationship between the hedge position and underlying MSR asset and the difficulty of establishing an “effective hedge” under accounting rules.

One alternative under discussion is to pay servicers a reduced flat servicing fee (say, 20 basis points rather than 25), with the difference stored in a reserve account that may be tapped to defray the costs associated with delinquent loans. Another idea is a fee-for-service model that pays the servicer a set number of dollars per loan, rather than a percentage of unpaid principal, in addition to providing for extra compensation for nonperforming loans.  

Stan Kurland, the founder, chairman and CEO of PennyMac – a company formed in the depths of the crisis, now a leading a nonbank mortgage company – believes capital to support servicing rights is critical to a better-functioning market.  His solution would retain the 25 basis point flat fee structure but allow servicers to pledge a strip of that – up to 18 basis points – as collateral for financing. This would open the business to more nonbank competitors, since they'd have a new way to finance operations. The amount that could be stripped off would depend on the risk profile of the asset as well the quality of the servicer. 

Stan's plan would also give servicers an incentive to invest in the high-touch capabilities needed for problem loans. If a servicer did a poor job of resolving troubled loans, the investor in the loans could bring in a specialty servicer – and make the original servicer foot the bill. It could pay the cost directly, or in the form of reduced cash flow to the financed strip (in which case the servicer would have to cover the difference for its lenders).

This solution elegantly addresses the bifurcated nature of the business, in which performing loans can be handled in assembly-line fashion and nonperformers require TLC. Such a straightforward and executable concept could critically change the mortgage industry and attract more private capital into the servicing market by allowing other participants to efficiently invest and leverage MSRs and thus compete on a level playing field with banks for servicing.

Next: The Qualified Residential Mortgage rule and risk retention.

Clifford V. Rossi is the Executive-in-Residence and Tyser Teaching Fellow at the Robert H. Smith School of Business at the University of Maryland.