THE MORTGAGE SERVICING BUSINESS has historically been a very cyclical enterprise. In good years, when rates are increasing (e.g., 2000), we have been able to eke out earnings of $94 per loan per year.
The underlying profit dynamics of servicing have generally improved over the last four years. Servicing fee income has grown 27 percent, due in large part to an ever-increasing average loan balance. This increase in loan balance has also driven up borrower payments of both principal and interest as well as taxes and insurance. Escrow income, accordingly, has grown dramatically. Additionally, it appears that cross-sell efforts are finally starting to show some results, with ancillary income almost doubling over the last several years. These gains are expected to continue.
Offsetting this $151 per loan increase in income, however, is a $10 increase in direct operating expenses. This increase seemed to be pretty much across the board, including items such as salaries, benefits and the ubiquitous "other" category. The largest negative impact to the profit and loss (P&L), however, relates directly to the prepay frenzy we recently experienced. Interest losses on mortgagebacked securities (MBS) pools and impairment hits (i.e., amortization/hedging) overwhelmed the improvement in net operating income. These last two items should be greatly curtailed in 2005 and beyond, resulting in a return to profitability for this business line for the near-term future.
During this expected period of normalcy, the industry has a tremendous opportunity to mitigate the risk of impairment hits in the next cycle of declining interest rates.
1. Convert to an option-adjusted valuation methodology for booking mortgage servicing rights (MSRs): The 8 percent mortgages that we booked at six times service fee in 2000 were clearly not worth a six multiple in hindsight. Such a multiple implies an average expected life approximating eight years. Most of these MSRs are now gone after only three. The write-off of the lost five years' worth of cash flows is reflected in the amortization/hedging cost of $476 per loan and $358 per loan (2003 and 2004, respectively) shown in Figure 1.
Accordingly, the prepayment option cost must be better reflected in current MSR multiples. Traditional optionadjusted spread (OAS) technology is not needed. While OAS has great benefits in and of itself, it is not necessary to calculate a reasonably accurate assessment of the option risk imbedded in a portfolio. A fairly simple expected return analysis (i.e., averaging the value of a portfolio at today's interest rates as well as up/down 50, 100 and 200 basis points) would provide such a value.
IMAGE TABLE 2Figure 1 Servicing Profitability
Now is the time to start incorporating this option cost into MSR valuations. Since we are seemingly near the nadir of the rate cycle, this option cost is at its cyclical low. Thus, the pain of converting to this methodology would be fairly minimal today. As rates increase, however, this option cost would increase dramatically. The industry must be steadfast in its resolve to reflect this increasing option cost as rates rise in order to avoid another debacle similar to 2003.
2. Re-examine yield requirements on MSRs: The discount rate (i.e., yield requirement) is one of the most important valuation assumptions, second only to prepay speeds. Yet most servicing analysts still use basic rubrics for estimating this rate of return. For instance, many today use a flat 8.5 percent for conventionals, 10 percent on governments and rr percent on adjustable-rate mortgages (ARMs), across the board. Yet most would agree that these yields do not accurately reflect the idiosyncratic risks that exist within a servicing portfolio.
Investment analysts in other industries would use a "risk-free" rate such as the Treasury as a base, match up their investment's expected maturity to this Treasury yield curve and add an increment to that rate to reflect the various risks inherent in their investment vehicle that are not reflected in the Treasury. Servicing risk increments might include: maturity, marketing, inflation, operational, and credit and liquidity risks, among others. Going through this mental process rationalizes your discount rate. Additionally, pricing off a point in the yield curve that approximates the expected life of the portfolio in question would result in a lower yield requirement on high-coupon portfolios (shorter expected life) versus lowcoupon portfolios. This decreased yield requirement on higher-coupon portfolios would mitigate to some extent the negative impact of valuing them on an option-adjusted basis.
3. Implement fair value accounting: The current exposure draft proposing to amend FAS 140 would allow servicers the option to mark-to-market MSRs at each reporting period. If adopted, this will allow participants to take greater advantage of the natural hedge inherent in the mortgage business. When interest rates drop, there will be write-downs on the servicing portfolio (minimized, it's hoped, through the implementation of items 1 and 2), but increased income from originations will partially offset them. When rates rise and origination income plummets, we will now be able to book the gain in the value of the MSRs.
Additionally, for banks and others with more complex balance sheets, gains/losses incurred in their fixedincome investment portfolios as well as their held-for-sale loan portfolios will be offset by losses/gains on their servicing portfolios. By taking advantage of this natural counterbalance, the need for expensive, risky and intellectually intensive hedging exercises will be minimized.
Servicing profits should be attractive and stable for the next couple of years. While it might be tempting to maximize them by rationalizing too high MSR values, strategically it will result in the same modest profit/bust cycle that we have experienced over the last decade. By more accurately valuing and managing the option costs imbedded in servicing, impairment losses should be minimized the next time interest rates start to drop. This is key to building longterm normalcy into the mortgage servicing business.
SIDEBARThe underlying profit dynamics of servicing have generally improved over the last four years.
-THOMAS J. HEALY, CMB
HanoverTrade Inc.
AUTHOR_AFFILIATIONThomas J. Healy, CMB, is a senior vice president with HanoverTrade Inc., Edison, New jersey. He can be reached at tom.healy@hanovertrade.com.