Perfecting the prepayment hedge.
Although accounting rules have not been the primary factor driving the search for a practical prepayment hedge, they have helped fan the flames of desire for a better-designed product. Devising a way to protect the economic value of servicing from prepayments goes well beyond a mere game played for the accountants on high.
Whirling dervishes have nothing on the mortgage banking industry. During the last five years, rapid change has left the industry spinning faster and faster until one might think it would either fly into pieces or collapse.
Wall Street whizzes have set sail into these winds of change with the creation of financial products for mortgage bankers tossed about in the storm. Indeed, it was the acquirers of servicing portfolios and some originators, too, that issued the SOS to Wall Street, which has come up with financial instruments to hedge so-called runoff or prepayment risks. In the process a new financial market has been created. At the same time, the hedges have already added a new dimension to the conduct of business among the major players, with an undefined potential to reshape the rest of the industry. The creators of these hedges include investment banking firms such as Goldman Sachs, Salomon Brothers, J. P. Morgan Securities and Merrill Lynch, along with First Interstate Bank of California in Los Angeles.
These hedges have helped accelerate the restructuring that is going on by reducing the fears of firms acquiring servicing portfolios, especially from the huge reservoir at the Resolution Trust Corporation, but also from busted mortgage bankers. So far, only about $100 billion, or 7 percent, of the estimated $1.5 trillion servicing universe in the United States is covered by hedges, according to William "Hal" Hinkle, director of mortgage banking at Goldman Sachs, the firm he says has the "lion's share" of the market. This is not exactly a stampede to hedges. Yet, the sellers of hedges see the potential as considerable, about 100 mortgage banking firms in the nation have servicing portfolios of at least $1 billion, according to First Interstate's senior vice president, Brian Cosgrove.
The industry's interest in hedges seems to follow the interest-rate cycle. A few nervous mortgage bankers rushed out to hedge servicing portfolios during the nightmare runoffs of 1986, when prepayments hit an all-time high of 30 percent on some portfolios. Then, early this year, a second wave of buyers took the plunge into hedging, as mortgage servicers experienced another troubling wave of early prepayments. This time around, the innovators on Wall Street seem to be gleaning a better harvest from worried executives. "It's a hot topic," confesses Austin Tilghman, a vice president at J. P. Morgan Securities.
At times it seems that balance sheet concerns are driving the industry to buy more and more hedges. Yet, according to some on Wall Street and in the mortgage banking industry, it is the desire of senior management to protect economic value that is really the primary driving force. Concern about purely accounting losses are secondary. As Tilghman points out, senior management is concerned that in a situation where they bought servicing for $20 million, and it all pays off at once, they have to take a $20 million loss. "This is not an accounting game--this is real value at risk. While the accounting concerns have added credibility to the products, people would have bought them anyhow," Tilghman says.
Accounting concerns, however, get a lot of attention by potential hedge buyers. This is because changes in accounting rules during the past decade have gradually increased the risks to the balance sheet of holding mortgage servicing. When interest rates rose beginning in 1987, it pushed down the volume of originations and left the industry with excess capacity. This pushed down profit margins to the point that the changes in accounting standards began to create a demand for hedges, Hinkle says.
The accounting straw that broke the mortgage banking camel's back was the introduction of the Financial Accounting Standards Board's (FASB) Statement No. 91 in 1987, according to Hinkle. This guideline proposed that loan origination fees could not be booked as income in the current period because they were deemed to be integral to the servicing process. This "funny" rule, as one major mortgage banker calls it, turned a financial asset into a short-term accounting liability. In retrospect, this change looks to some like an unintended fiat that forever split the mortgage banking kingdom into two distinct businesses: originations and servicing.
The new, big, corporate players that moved into the market were particularly vulnerable to the impact of the ruling. To compensate, they had to sell off big chunks of the mortgages they originated and then buy originations from other firms to build up their portfolio. Walter C. "Terry" Klein, chairman of Sears Mortgage Corporation, Riverwoods, Illinois, which moved into the business in a big way in 1986, explains the impact of the accounting change. If you had no servicing portfolio, and originated $1 billion at a loss of 50 basis points, the new rule required you to report a loss of $500 million. However, if you sold $400 million of the newly originated pool of mortgages for 150 basis points, you would bring in $600 million. If you subtracted the $500 million charge-off in origination costs from the $600 million gain from selling servicing, you'd be able to report a pre-tax profit of $100 million. "Thus, the accounting rules force you to sell what you originate and buy servicing from other originators in order to avoid a write-off," Klein says. (Selling can be avoided if a firm keeps a servicing portfolio five to six times the annual volume of originations; this will then yield enough profits to cover the balance-sheet losses, Klein says.)
The genesis of the servicing hedge
Before they asked Wall Street for better solutions, some mortgage bankers had already begun to hedge their servicing portfolios as early as 1985. First they bought long Treasuries or Ginnie Maes, and then they tested interest-rate swaps. Richard Stuckey, managing director of derivatives at Salomon Brothers, says that his firm was providing interest rate swaps as early 1986, along with caps and floors on LIBOR, both put and call options. Then in 1987, Goldman Sachs, after being approached by clients to provide better hedges, developed the first products designed specifically to hedge mortgage servicing. They were called interest-only (IOs) and principal-only (POs) strips, and Hinkle says they were developed in conjunction with Fannie Mae. This makes Goldman Sachs "the father of hedging mortgage servicing," Hinkle says.
Mortgage bankers held the POs and leveraged POs on their balance sheets as assets to provide an income stream to cover larger-than-expected losses. Compared to most of the hedges that were to follow, POs now seem very expensive for mortgage bankers. POs had an accounting problem, too. While they could be written up or down, servicing could not be written up when prepayments slowed. Thus, on the balance sheet the PO hedge was not a hedge, and accountants began to raise troubling questions for senior management.
The hapless POs had sailed into a fog. To gain hedge accounting, a hedge must satisfy FASB Statement No. 80, which says the performance of a hedge should have a high degree of correlation with the performance of the asset that is being hedged. After some mortgage bankers bought POs, their accountants began to tell them that POs and similar products lacked a sufficient level of correlation to meet Standard No. 80, according to Clark McGranery, vice president of fixed income market strategy at J. P. Morgan Securities. To be fair, the level of correlation could be sustained by purchasing additional POs if prepayments ran off too fast. That, however, demanded an active management of the hedge that was too demanding for most mortgage bankers.
Mortgage bankers were also beginning to balk at the cost of the first generation of hedges and the liabilities that came with them. "The POs were eating into profits," explains Tilghman, because their yield was low, 5 to 7 percent, while the yield on servicing was generally 15 to 17 percent." At the same time, the early interest-rate swap contracts carried heavy basis risk--that is, the hedge's performance did not always match, closely enough, the performance of the servicing portfolio. Finally, innovators sought to devise hedges that would reduce the level of the basis risk that made the original hedges unattractive.
With so much talent looking into innovative products on Wall Street, it came somewhat as a surprise that the first product in the second generation of hedges came from the West Coast. In the fall of 1989, according to industry sources, First Interstate Bank shocked Wall Street with its cash-flow hedges--the first hedges to be off the balance sheet, portfolio-specific and meet the requirements of hedge accounting. Whether it realized it or not, First Interstate had fired the first shot in the hedging revolution.
Launching a new generation
According to Cosgrove, thrifts and mortgage bankers were coming to First Interstate to try to devise ways to protect excess capitalized servicing they were putting on the balance sheet as an asset. This concern about servicing originated in-house was light years from the merger activity and purchased servicing frenzy that preoccupied Wall Street. In November 1989, First Interstate announced by an advertising tombstone the sale to Home Savings of America, Irwindale, California, of the first option-type prepayment cap that would meet hedge accounting treatment with a lump-sum, up-front payment that would protect the mortgage servicer for five years. As the name implies, First Interstate's cashflow hedge supplies the cash flow that is lost when prepayments accelerate beyond a predetermined level. Within days of the Home Savings deal, "Wall Street firms were calling us to find out about it," says Gary Spehar, vice president in First Interstate's mortgage-backed sales and trading group. By January 1990, J. P., Morgan Securities, Salomon Brothers and Goldman Sachs were all offering a raft of new generation hedges. The, the following summer Merrill Lynch joined the fray.
What had First Interstate wrought? First and foremost, according to Spehar, they had taken the worry out of managing the prepayment hedge. No longer would it be necessary to add POs if prepayments accelerated too much. Secondly, it protected the asset of excess capitalized servicing while it also protected the economic value of the servicing portfolio. While many institutions could absorb the losses of fast prepayments, they were beginning to realize that regulatory and accounting-standards pressures were gradually forcing them to develop more and more sophisticated analyses of the risk of prepayment. As mortgage bankers began to realize that the servicing portfolio had within it "imbedded options" with which they would have to take into account if they were doing their analysis properly, they decided it was more efficient to turn to a firm with expertise in this area, rather than develop it on their own, Spehar explains.
To improve the performance correlation between hedge and prepayments, for example. First Interstate and the Wall Street innovators have developed reference portfolios that come as close as possible to the characteristics of the hedged servicing portfolios. Tilghman says the reference portfolio is a real portfolio that is nearly identical to the hedged portfolio's weighted-average coupon, weighted-average maturity, geographic location of the mortgage borrowers and issue year. A reference portfolio might be, for example, all 30-year fixed-rate Fannie Maes with a 10-percent coupon rate originated in California in 1990. Under various scenarios, the computer models, programmed with past prepayment experience by mortgage type, predict the rate of prepayment. The Public Securities Association (PSA) helped this analysis by creating in 1985 a new measure for the rate of prepayment, the "PSA." Thus, a benchmark of 100 PSA for the market speed for prepayments was born, a run-off level of about 6 percent per year.
Next came the problem of creating a high degree of correlation needed for hedge accounting. This required a better understanding of how servicing portfolios can be expected to prepay under a variety of interest-rate and economic conditions. Harley Bassman, a managing director at Merrill Lynch who has developed option-derived hedges, uses historical prepayment performance and models to track the performance volatility of mortgage servicing. Like others on Wall Street, Bassman quickly saw that the return on investment for mortgage portfolios has what students of finance call "convexity." Bassman, a former professor, likes to explain convexity to the uninitiated in the Socratic method, by asking questions, to help the listener come to understand what is, for most in the industry, a foreign concept. When all the questions are answered, a definition appears. An asset whose return on investment has convexity performs out of sync with changing market conditions. Assets with positive convexity tend to have more upside under a range of market conditions over time, and those with negative convexity have more downside.
The curse of convexity
While most may not be familiar with the term, there's hardly a mortgage banker in America who has not had sleepless nights worrying about mortgage banking's peculiar curse of negative convexity. The rate of return on the servicing portfolio also possesses this feature. If the expected average return is around 15 percent on a portfolio, then, should adversity strike--a sudden, sharp drop in interest rates--then a servicing portfolio's return can fall as low as minus 10 percent, a horrifying 25-point plunge. The upside, should rates rise an equal amount, however, may only be a gain of 5 points in the rate of return, to a maximum of 20 percent under the best of conditions. As Scott Shay, managing director of Rainieri Wilson & Company, Inc., puts it, "In mortgage banking you make money in nickels and dimes--and lose it in dollars."
Because the return on servicing has negative convexity, it needs to be hedged by something that has positive convexity, explains McGranery at J. P. Morgan Securities. This is why investment bankers turned to option technology to create and structure their hedges--options have convexity, and the right option can have a positive convexity of a servicing portfolio.
While First Interstate continues to offer primarily the single product, a cash flow hedge that functions as a prepayment cap, Wall Street, in typical fashion, keeps proliferating a wide variety of hedges. First Interstate charges an up-front, lump-sum payment, as with any call or put option, which then becomes the floor for future losses. When anticipated income does not accrue due to early prepayments, First Interstate provides the missing cash. Most of the Wall Street products, however, do not require up-front lump sums, even the option-derived hedges. However, most of the firms will offer clients the choice of making a lump-sum payment, if they wish.
Although many of the second-generation products are derived from option technology, mortgage bankers tend to think of them more as a type of insurance. That's how Sy Naqvi, president of Sears Mortgage Securities Corporation, views the products, for example.
"Our approach is to protect ourselves from catastrophic risks," says Naqvi, who heads up Sears' hedging operations for both the mortgage pipeline and the servicing portfolio. Naqvi uses hedges sparingly because they can be expensive. Some are "totally cost prohibitive at 10 to 12 basis points" or 25 percent of the annual income stream from the servicing spread, according to Naqvi. But, according to a Wall Street source, the second-generation hedges currently cost about 2 to 4 basis points or 5 to 10 percent of the servicing spread. Tilghman at J. P. Morgan Securities says that if a mortgage banker comes to them willing to spend only 1 basis point, they can devise a product at that price. Naqvi says Sears does not cover all runoffs greater than the current market speed, only risks 25 PSA or more than current market levels, for example. This is covered by a regular premium and is usually bought to cover portfolios of $1 billion or more.
Tilghman says that Sears Mortgage has typically purchased a product that gives Sears 100 percent of the book value, not the cash-flow value, of prepayments exceeding 200 PSA. By covering only runoffs above 200 PSA, Sears is, in effect, taking a deductible on its insurance coverage, Naqvi explains. The coverage, too, is usually limited to the first four to six years. The insurance hedges are better than POs, Naqvi says, because "no asset is bought and there is a lower capital outlay." He finds that "on the whole, they are pretty exciting" and hopes that they will become less "pricey" as competition increases. So far, Naqvi says, "the investment bankers have offered a good response to a genuine need." Sears Mortgage has, in fact, bought the prepayment cap.
Weeding through the hedges
If one wanted to simplify Wall Street's cornucopia of second-generation hedges, it would be best to divide them into two broad categories, according to Stuckey at Salomon Brothers--symmetrical and asymmetrical. With the symmetrical hedge, "you give up future benefits in return for protection" of prepayment risks, Stuckey says. "If prepayments speed up, you get monthly payments from the hedge. If they slow down, you make a payment," he explains. Prepayment swaps fall into this category.
The hedges in the asymmetrical category, in their purest form, require a one-time cash payment, which may be financed over the term of the contract with monthly payments. If prepayments speed up, the hedge will pay the mortgage banker. But, if prepayments slow, there are no further losses. These are known as prepayment caps and they come in two major subcategories. One is the book-value prepayment cap, which protects the full book value of the hedge, sometimes with a deductible. The cash-flow prepayment cap covers only the lost revenue during the term of the hedge. This is the type offered by First Interstate. Some of the firms give their hedges service marks. Merrill Lynch, for example, calls two of their most popular prepayment hedges the PreServ Principal Cap and the PreServ Revenue Cap.
Mortgage bankers and depository institutions should also be aware that the current crop of hedges, although priced fairly low today, when compared to the price of the first generation hedges, are likely to rise sharply in the future when the outlook for accelerating prepayments increases. As Naqvi notes, "The higher the note rate is in relation to the current market, the higher the hedge cost. The time to buy insurance, of course, is before the fire, not after the barn burns down." So, there is no rest for the whizzes on Wall Street, who even now are dreaming up new products to meet the price pressure coming from the mortgage banking industry, a pressure that is likely to intensify.
While cost is paramount to most mortgage bankers, apparently some of the biggest mortgage servicers are not balking at prices that most mortgage bankers would consider irrational--that is, near, or above, 25 percent of the servicing spread income stream, or 10 to 20 basis points. The appeal of these hedges is their increased efficiency. In fact, one of them developed by Goldman Sachs is 100-percent efficient, making it a perfect hedge. While the official name is being kept a secret for now, pending a service-mark award, it could be called a mortgage-indexed general obligation (MIGO). It monetizes prepayment risk for servicing pools, and Hinkle says a few of the big ten mortgage servicers are already using it. The MIGO can reduce the need to borrow from banks to buy servicing. That type of borrowing creates a liability that has to be paid down in tandem with the servicing portfolio. By monetizing the prepayment risk, the mortgage banker can fund the acquisition of the servicing, and, at the same time, "the equity insulates itself from prepayment risk," Hinkle says.
In spite of all the growing excitement over hedges by some mortgage bankers, the overwhelming majority remain convinced they should not buy them. In particular, the idea of hedging unmonetized servicing originated in-house has yet to set the world on fire. Yet, Bassman at Merrill Lynch is optimistic that mortgage bankers and thrifts will eventually want to hedge more of their unmonetized servicing. He thinks that if more mortgage bankers understand the negative convexity of returns on servicing they will begin to hedge it. "They have an economic loss, even if it's on the book at zero," he says. Then, too, notes Stuckey at Salomon, mortgage bankers, whether they admit it or not, are betting their business on the future course of prepayments. "It's quite silly to bet on something you can't predict or control," he says. Hedging this risk, then, becomes just the cost of doing business. But, as Naqvi at Sears points out, most mortgage bankers are still unaware of the dangers of prepayment. When prepayments speed up, they focus their attention on the crush of new originations, ignoring the loss of "their most important asset," servicing.
Some industry analysts who know full well what hedges can do urge caution to those who are considering buying them. Indeed, some even believe that the road to ruin is paved with unneeded hedges. One of these critics is Hunter Wolcott of Reserve Financial Management in Miami, who advises mortgage bankers how to pick a hedge without overinvesting in such precautions. "The value of purchased servicing is independent of its source," he warns. "Servicing is an internal hedge all by itself for mortgage banking companies. It behaves like an IO strip." At the same time that a plunge in interest rates causes a runoff in the servicing portfolio, it also drives new originations to replace them. "Thus, to hedge a hedge could demonstrate a faulty definition."
Wolcott suspects that mortgage bankers may be buying hedges they don't need because they worry too much about prepayment risk. "People's expectations of prepayments are more damaging than prepayments themselves," he says. A mortgage banker can cover his market value or accounting risk with a properly balanced position in Treasury-bond futures or Ginnie Mae futures, Wolcott says. That, of course, carries considerable basis risk, the investment bankers counter. That's just the point, Wolcott shoots back. Every hedge has basis risk--a fundamental flaw that escapes mortgage bankers," Wolcott says. The reason that returns on servicing are so high, at 17 percent, is that there are extraordinary risks involved, Wolcott explains. Eliminate all those risks and the return will be no greater than Treasuries.
Another concern about hedges is that too often the mortgage servicer "can not unwind them," Wolcott says. POs, for example, are for 30-year terms--an eternity for most businesses. Because the secondary market for POs is highly illiquid, then, if there is a 100-basis point drop in interest rates, then not everyone who wants out of the market will be able to get out. The prepayment caps are fixed-term contracts and the prepayment swaps are mandatory future contracts, making it impossible to get out of them until their term expires. Only those hedges that are paid with an up-front lump sum are totally worry-free. Once you get over the initial fee, there's nothing else to pay--ever.
Critics say servicing hedges cover too long of a term to work like traditional hedges. Hedges on farm products and commodities, or even the mortgage pipeline, were very limited in time--therefore, they were not so prohibitively expensive, Wolcott points out. But, servicing hedges, at their shortest, are four years. "Over time, the sellers will win and the buyers [will] lose," Wolcott warns.
Shay at Rainieri Wilson & Company notes that prepayment is only one of the risks of servicing. "While you can hedge a lot of things," he explains, "there are a lot of unhedgeable risks in servicing, including foreclosures, delinquencies and operating costs." He predicts that the use of hedges will not likely spread beyond the big players--because they're the only ones with sufficient equity and credit standing to execute hedges without putting up expensive collateral.
The lure of hedging
Both the hedging skeptics and the Wall Street innovators agree that the hedging of servicing, if widely adopted, will drive down the return on mortgage banking and lead to the securitization of mortgage servicing. Wolcott, however, sees this as no great gain for mortgage bankers. He says it will bring new capital into the industry that will lead to a consolidation of servicing in fewer firms.
The Wall Street innovators believe that the fact mortgage servicing yields remain unnaturally high is proof-positive that the industry is capital-starved. Hinkle says that because the returns on servicing have risen even as excess supply has been introduced by the Resolution Trust Corporation, this "shows that more capital is needed to complete the restructuring of the mortgage banking assets in defunct S&Ls." As more sophisticated capital flows into mortgage servicing, it will shift the focus from the yield of the investment to one of risk-adjusted return on equity, Hinkle says. "The more sophisticated the capital flowing into the market, the less yield on return on investment will matter and the more risk-adjusted return on equity will matter," Hinkle predicts.
A change in the tax treatment of servicing income could also begin to make securitization irresistible to mortgage bankers, Hinkle says. The Internal Revenue Service (IRS) has ruled recently that mortgage bankers will have to pay tax on "excess" servicing fee "income." The IRS defines excess servicing income as that which is in excess of minimum servicing fees set by the different agencies. Mortgage bankers, as a result, could be more tempted to securitize servicing in order to create a cash flow to pay for this tax, Hinkle says.
Ultimately, Fannie Mae and Freddie Mac may decide just how far the
monetization of servicing may proceed. The regulators have traditionally wanted active origination and vertically integrated servicing operations. Fannie Mae and Freddie Mac, thus, believe firmly that "mortgage servicing is not an investment--it's a business," Wolcott says. The agencies may find that the quality of servicing may decline if it becomes just an investment, which "takes it out of relational financial servicing into transactional financial servicing," Wolcott says. Wall Street, too, has reason to preserve a fully integrated mortgage banking business. Notes Hinkle, "At the point the capital market loses faith in the integrity of the processing behind the securities, that's when they lose confidence in the securities themselves."
Robert Stowe England is a freelance writer based in Washington, D.C.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||way to protect the economic value of mortgage servicing from prepayment|
|Author:||England, Robert Stowe|
|Article Type:||Cover Story|
|Date:||Sep 1, 1991|
|Previous Article:||Servicing under a microscope.|
|Next Article:||The world according to GAAP.|