Relative profitability of the mortgage industry.
One can't argue with the evidence of consolidation on the servicing side of the mortgage banking business. The share of servicing rights held by the top 10 firms in the industry has risen from less than 10 percent in 1988 to more than 30 percent in 1997. The largest servicer in 1988 was Citicorp Mortgage, with a portfolio of $45.6 billion. The top servicer in 1997 was Norwest Mortgage, with a portfolio of $205.8 billion. Furthermore, as recently as 1990, Norwest didn't even have a servicing portfolio.
The consolidation on the origination side of the business is less clear. The entire landscape has been altered by a fracturing among origination channels between consumer-direct (e.g., phone and Internet applications), broker, correspondent and retail options. The origination process is becoming more like the automobile industry with many input suppliers (brokers), some module assemblers (correspondents), and a few final assembly firms (the top 25 mortgage lenders).
The firms normally recounted as the top 10 or top 25 mortgage lenders by market share typically report their origination volume based upon combined retail, correspondent and broker originations. There are literally thousands of brokers, and they report receiving more than 50 percent of all mortgage applications in the country. Thus, the issue of consolidation, on the origination side of the business, depends heavily on the definition of an origination.
This difference between events on the origination and servicing sides of the business is a result of a variety of factors, which will be discussed later. However, first let's look at a standard measure of the profitability of an industry, return on average equity (ROE), and compare year-over-year results for mortgage companies and then compare returns across financial firm types and industries. Ultimately, it is this measure of profitability that brings capital into or shifts it out of an industry.
The final returns for mortgage banking companies for 1997 are not yet available, so we have displayed the most recent 10-year period of industry returns. The dynamics of industry change are readily discernable in the mortgage company results. The FHA delinquency problems depressed returns in the late 1980s when mortgage banking companies were still heavily focused on government loans. The data shows their emergence as conventional lenders in 1989 and 1990 as the thrift crisis peaked and the secondary market agencies emerged to replace the thrifts. From 1991 through 1993 industry returns peaked as the secondary market became fully developed and interest rate declines generated huge volumes of refinancing activity.
In 1994 the Federal Reserve pushed interest rates up dramatically, killing off the refinance activity and revealing massive excess origination capacity. Industry financial losses worked in combination with the technology applications that had been developed to deal with the refinance volume of the 1991-93 period to ignite consolidation. The low returns led many firms to exit the industry through merger, sale or failure. Well-capitalized firms employed the technological lessons to achieve dramatic reductions in direct costs of servicing, and the chase for economies of scale was on in servicing. Thus, after the tough year of 1994, returns rebounded in 1995 and 1996 but not to the level of the early 1990s.
Also, while the data is not shown here, the intrayear dispersion of returns was higher in the late 1990s than in the earlier years. This means there were more big winners and more big losers. The forecast is for more industry exits and recombinations as these trends continue.
Average returns for mortgage bankers compare favorably with thrifts, banks and the auto industry but not to the secondary market agencies. Thrifts, for example, have seen average returns rebound from the crisis of the late 1980s but never to the pre-crisis level. However, their 1997 average ROE was 10.5 percent, so the trend in thrift profits continues its recent slight upward drift.
Banks' returns, while not as high as those of mortgage banking companies during the 10-year period, have been steadier and higher for the last five years. The average ROE for banks in 1997 was 14.7 percent, which continues the recent trend. The auto industry has experienced wide cyclical swings during the decade, and that around an average return not as high as mortgage bankers have achieved.
Throughout the entire period, the level and variability of returns to the secondary market agencies, Fannie Mae and Freddie Mac, has been steady with returns at very high levels and variability at very low levels. This is an effect of their duopsony market structure and the advantages it confers. A duopsony is a pair of firms operating in a market where they are the only two purchasers of an output. In this case, the status conferred by their being government sponsored (with private stockholders) gives them a cost advantage in capital markets with which private firms cannot compete. Thus, they are able to generate long-term risk-adjusted returns that exceed market norms.
A simple comparison of the level of risk-adjusted returns is presented in the table in the row titled "Coefficient of Variation" (CV). This number is a measure of the amount of risk incurred per [TABULAR DATA FOR FIGURE 1 OMITTED] unit of return and allows a comparison of return achieved relative to the level of risk taken across firm types. The higher the CV, the more risk taken per unit of return. The industry that has been most risky per unit of return in the last decade has been the auto industry, followed closely by the thrift industry, which was, of course, nearly completely restructured subsequent to the crisis of the 1980s.
Commercial banks earned a slightly lower average ROE during the decade than mortgage bankers, but mortgage bankers incurred nearly 60 percent more risk per unit of return than banks. However, both secondary market agencies earned substantially higher ROEs than either banks or mortgage banking companies over the decade and at less than one-third the risk per unit of return. The 1997 ROEs of the agencies are nearly identical to their 1996 figures.
What is expected for final ROE results for the mortgage banking industry for 1997 and in the future? Preliminary results show figures similar to but possibly slightly stronger than 1996. The numbers will probably not be as strong as they might have been expected to be, given the level of origination activity, the relatively healthy economy and strong home sales. This is because the consolidation of the industry has not yet reached equilibrium as represented by the continued wide dispersion of earnings among mortgage banking firms.
Some individual firms have achieved significant economies of scale and have strong and diversified origination strategies that achieved strong levels of profits in 1997. However, there continued to be firms that exited the industry because of low returns as well. Further, the duopsony status of the secondary market agencies allows them to continue to attract a disproportionate share of industry profits.
Douglas Duncan is senior economist at the Mortgage Bankers Association in Washington, D.C.
|Printer friendly Cite/link Email Feedback|
|Date:||Jun 1, 1998|
|Previous Article:||Branching out into subprime.|
|Next Article:||Training surveys.|