The advantages of investing in collateralized mortgage obligations are attractive yield, high credit quality, and favorable regulatory treatment under risk-based capital rules.
Unfortunately, the disadvantages would require much more ink.
Investment banks provide extremely valuable services in creating and distributing innovative CMO products.
They also offer research, analysis, and commentary to help investors understand and evaluate these products. (The information is usually free to clients.)
Nevertheless, some individuals and firms do not adequately understand the behavior of mortgage products and the accompanying risks.
Let the buyer beware.
Investors before making purchases. They should not rely excessively on the broker-dealer, no matter how well intentioned, for advice on choosing securities.
Ultimately, it is the bank that must live with the consequences of the investment decision. Unfortunate surprises that may occur in a CMO's performance are the bank's problem.
Cash Flow Uncertainty
The primary risk in CMO investing is the uncertainty of cash flow timing.
Because homeowners have the option to prepay mortgages at any time, most mortgage products contain some element of cash flow uncertainty.
With some CMO products, such as VADMs (very accurately defined maturity tranches), this uncertainty is substantially reduced.
With other products, such as PACs (planned amortization class tranches), substantial uncertainty is removed if prepayments are reasonably close to the original assumption.
I will focus here on the traditional "payer" tranches.
Once investors have analyzed the two key determinants of cash flow risk in CMOs (the structure of the deal and the characteristics of the underlying collateral), they are ready to proceed to "dealing with the dealer."
Players in the CMO market engage in many games.
If potential investors do not understand the games, they are likely to overpay for CMO products and may not achieve the performance they expected.
The Speed Game
Analysis of most CMO tranches revolves around an assumed prepayment speed. This assumption critically affects the period over which an investor can expect to receive cash flows.
It is to the seller's advantage to assume a relatively fast prepayment speed, especially when the yield curve is positive. This results in an investment with a shorter expected average life and can therefore be priced higher than over a lower-yielding portion of the interest rate curve.
Because the prepayment assumption is only an educated guess of future results, the investor cannot prove that any particular assumption is right or wrong.
Potential investors must arrive at an estimate that provides them with at least some degree of confidence.
A study of the performance of various types of mortgage collateral during previous periods of alternative interest rate scenarios, coupled with the investor's view of future financial and demographic trends, will lead to a prepayment assumption which may differ radically from that of the seller.
It is interesting to note that various investment bankers using their own models arfive at wildely different results.
A Complex Problem
Also, a firm's traders may not necessarily use the results generated by their own firm's economists and statisticians.
The analysis is complex and must include type of mortgage, geographic dispersion, age of the loan, seasonal factors, burnout, gross and net weighted average coupon, and a host of other financial and demographic considerations.
The more homogeneous the collateral, the easier it is to analyze. Conversely, the more diverse the collateral, the more difficult (if not impossible) it is to achieve any degree of certainty about the potential prepayment behavior.
Because the yield curve is so steep at this time, a prepayment speed assumption used by a dealer that is slightly too aggressive can easily result in an average life that is significantly shorter than the investor might actually achieve.
A shortening of one year in the three- to four-year portion of the yield curve can result in the CMO being overpriced by as much as 1.0%.
The Rounding Game
Dealers usually price CMOs against the U.S. Treasury yield curve, with the CMO's average life rounded to the nearest whole year.
The result is that CMOs with average lives between years can be priced off a shorter maturity portion of the curve.
If the yield curve is as steep as it is now, the price of the security could easily be 0.5% too high in the three-year sector.
I am sure that it is no coincidence that many of the CMOs coming to market this year have an average life of 3.4 years, and even 3.48 and 3.49 years.
This market tradition is difficult to justify.
In a positive yield curve environment, would you be willing to purchase a Treasury note due in 3.5 years at the same yield as a substantially similar Treasury note due in three years?
This tactic, which might also be called the "on the run" game, results in the investor overpaying. The game is to base prices on certain Treasury securities that are no longer appropriate, because of the passage of time.
This game can be especially treacherous in the three- and four-year maturity or average life area of the yield curve.
On Aug. 6 we were attempting to purchase CMOs with an average life in the range of three to five years.
At that time the "current" or "on the run" three-year Treasury note was the 7.0% coupon due May 15, 1994, a date that was almost three months short of a full three years at that time.
Differences Add Up
Again, this technique on the part of the dealer results in a higher price when the yield curve is positive.
On Aug. 6 the "current" three-year was quoted to yield 6.88%. On the same date, the new "when issued" three-year was trading at 6.93% bid, and the old 8.75% coupon due Aug. 15, 1994 was also quoted at 6.93%.
Any CMO purchased on that date, based on a spread over the "current" three-year, would have resulted in a yield of five basis points less then an investor should have expected and thus a price about 5/32 too high.
The situation in the four-year area was even worse. There is no four-year Treasury quoted on the Bloomberg page USD.
Not to worry. On the PX1 page, the current four-year is conveniently shown. On Aug. 6 the current four-year was the 7.625% coupon due Dec. 15, 1994.
What a deal! The four-year has magically shrunk to three years and four months and is quoted at a yield of 7.02%.
The Blending Subgame
This, of course, is a serious miscarriage of justice, which leads into a subgame I call the "blended" or "interpolated curve" game.
In the four-year area, this game consists of averaging the yields of the "current" three-year and the "current" five-year Treasury, resulting in a yield of 7.21% on Aug. 6.
This is certainly much better than the previously calculated 7.02%, but again this method suffers from the fact that the "current" three- and five-year may not in fact be so current.
So, you ask, why not find a Treasury security that really has four years to maturity? Maybe a 10-year note sold six years ago or a 30-year bond sold 26 years ago would give us some guidance.
It just so happens that we had two from which to choose: the 10.5% coupon, due Aug. 15, 1095, yielding 7.32%, and the 8.5% coupon also due that day, yielding 7.31%.
These yields are certainly more representative of the real four-year area of the curve and are obviously more attractive to the buyer than either of the two previous calculations.
Long Settlement Game
It is fairly common for newly issued CMOs to settle as much as eight weeks after they come to market.
Assuming that the investor has money available and places it in fed funds at 5.5% until settlement day, the actual spread achieved on the investment could be substantially less than expeced if the investor does not reques a supplement to the spread that is quoted for similar CMO tranches with corporate settlement (five business days).
Additional compensation of 10 basis points of yield spread would not be unreasonable in today's yield environment.
Investing with Eyes Open
At this point it is fair to ask how all this game playing is affecting potential investors, and what if anything can be done about it.
Most important, we need to realize that these games are being played and that they have a substantial impact on the average life, price, yield, and spread over the Treasury curve that is actually achieved.
We may not be able to do anything about the games, but at least we should not allow ourselves to be deceived about the actual performance to expect from our CMO investments.
I frequently voice my displeasure to the broker-dealer community about these games. If enough investors show their displeasure, maybe we can get the rules changed.
Mr. Giuliani is vice president and investment officer of Security Bancorp, Southgate, Mich.