Bond market innovations and financial intermediation.
For many years the bond market was the realm of blue chip issuers and conservative investors. Its sedate pace suited those whose primary objectives were steady income and preservation of capital.
The upheaval began gradually. Increased market volatility in the late 1960s and 1970s led to growing awareness of interest rate and currency risk. This in turn sparked a demand for new types of market instruments, new strategies for managing risk, and new ways to exploit interest rate differentials across markets.
In the 1980s, the flow of innovation has become a flood. New instruments have proliferated in a bewildering variety. Financial engineering allows the uncoupling of investor demand from the instruments that issuers want to supply and has fundamentally altered the structure of financial intermediation. Bond markets have become truly global in scope, with trading occurring around the clock and around the world.
This article will describe the common characteristics of these bond market innovations and the underlying forces that have caused the recent wave of change. It first analyzes the general determinants of bond market demand and supply. It then describes some of the major innovations of the 1970s and 1980s, explaining how they represent a pattern of accelerating response to changing market conditions.
BOND MARKET DEMAND AND SUPPLY
A bond or any other financial instrument can be viewed as a package of characteristics such as return, risk, and liquidity. Investors, of course, prefer high returns, low risk and high liquidity, but issuers will not be able to produce that package at reasonable cost. Hence, tradeoffs will be determined by supply and demand.
On the demand side, these tradeoffs are influenced by the distribution of investable wealth, tax rules, regulatory restrictions, and perceptions of the relative importance of various risk categories. In the 1980s, for example, Japanese financial institutions have accumulated an increased share of the world's total investable wealth. It is not surprising, then, that the tax and regulatory rules facing these institutions have affected the form of bond market instruments.
Changing perceptions of risk have influenced investors' views of bonds in both the 1970s and 1980s. Credit risk was once thought to be the most important category of bond risk, but investors have now become more keenly aware of interest rate and currency risk. The corporate restructuring wave of the 1980s has also given rise to "event risk," that of a sudden decline in credit quality resulting from a takeover, divestiture, or recapitalization.
Increased sensitivity to these risks has in turn spurred a demand for securities with protective features. Examples include floating-rate notes, zero-coupon bonds, and put bonds. New investment strategies designed to tailor the overall risk of a portfolio more precisely, such as immunization and hedging strategies using futures and swaps, have also been developed. Finally, the market volatility generating these risks has increased the demand for liquidity, thus enhancing the appeal of public securities markets. Investors have discovered that risk-management strategies require a liquid portfolio.
On the supply side, bond issuers face a set of cost conditions for providing financial services. They will try to issue securities that minimize the cost of providing a given financial services package. These costs include both the explicit costs of executing transactions and the implicit costs of bearing financial risk, and they are affected by available technology, tax and regulatory rules, and the issuer's ability to offset various forms of risk.
For example, the relative costs of issuing debt in the public and private markets have been affected by Securities and Exchange Commission Rule 415, which makes it faster and easier for large companies to bring a public issue to market. Advances in communications and information processing technology have helped allow continuous and flexible access to debt markets.
Tax considerations have clear effects on the costs of servicing bond issues and can thus influence their design. A primary motivation for the introduction of zero-coupon and other original issue discount bonds in 1981 was the advantageous way that issuers could deduct the discount amortization from annual tax payments.
An issuer's ability to bear particular risks will influence the degree of risk protection offered to investors. Commercial banks have been frequent issuers of floating-rate notes, since their interest rate risk is hedged by the tendency of their revenues to vary directly with interest rates. Producers of commodities, such as silver or oil, have been natural issuers of commodity-backed bonds, whose ultimate payoff is indexed to a specified commodity price level.
The demand and supply sides of the bond market can be joined in two ways. They can meet directly when the issuers supply bonds with characteristics ultimate investors desire. This has been the traditional sphere of the public bond markets. Alternatively, they can meet indirectly through a financial intermediary. The intermediary purchases the issuer's debt security and in turn issues a claim on itself having characteristics that are more highly valued by the ultimate investors. This has been the traditional role of commercial banks, which purchase debt securities from businesses (make loans to them) and issue more liquid, less risky deposits to investors.
One of the most important developments is the way in which traditional methods for joining the two sides of the market have been completely transformed. The increasingly global scope of bond markets has afforded more opportunities to issue securities tailored to particular investor tastes, tax rules, or regulatory restrictions. Formerly, the issuer may not have been willing to bear the specialized risks such securities entail. Beginning in the 1970s and increasingly in the 1980s, however, markets have developed for reallocating these risks. The swaps market, for example, allows issuers to offer securities in either fixed or floating-rate form in any currency and then swap the associated profile of interest rate and currency risk for another, more desired profile.
Traditional patterns of intermediation have also changed radically. Investment bankers now play a more overt role in financial intermediation by buying market securities and repackaging them for sale to investors (stripped Treasury bonds or collateralized mortgage obligations, for example). These activities have impinged on the traditional business of commercial banks, who have been forced in turn to define new roles for themselves, such as the market-maker role many large commercial banks have assumed in the swaps market.
BOND MARKET INNOVATIONS OF THE 1970S
World financial markets faced several new sources of upheaval in the 1970s. Inflation rates grew higher and more volatile. Interest rates went through the same changes. In addition, the demise of the Bretton Woods Agreement in 1971 introduced a new element of volatility to world exchange rates.
Motivated largely by these new sources of risk, a number of innovations either were introduced or first became prominent in the 1970s. Six of these will be addressed here: floating-rate notes, Eurobonds, the early stages of the junk bond market, mortgage pass-through securities, parallel loans, and financial futures contracts.
Floating-rate notes are debt instruments on which the coupon rate is periodically reset to bring it into line with some easily observed market rate. For example, the contract formula might call for the coupon rate to be reset every quarter to equal the prevailing 91-day U.S. Treasury bill rate plus 2 percent. Such instruments had been issued elsewhere, but they did not enter U.S. bond markets until 1973, and they did not attract widespread attention until Citicorp's $850 million issue in 1974. The impetus for the innovation was the sudden volatility of U.S. inflation and interest rates, which imposed losses on holders of long-term, fixed-rate bonds. Long-term investors, aware of the need to protect themselves against increases in interest rates, wanted an instrument whose coupon rate would automatically adjust with the market. At the same time, banks and finance companies were natural issuers of such securities; their assets were primarily short-term or even floating-rate loans, so their revenues afforded a built-in hedge against floating-rate liabilities. Government regulation provided a further incentive for innovation, since U.S. commercial banks were still subject to interest rate ceilings on deposits. By issuing floating-rate notes instead of certificates of deposit, banks could pay market rates yet still offer investors a relatively safe and liquid investment. In this case, a new source of variability gave rise to natural clienteles of both investors and issuers, and floating-rate notes allowed the two sides to be brought together.
Another security that attained its first real prominence in the 1970s was the Eurobond, a bond issued outside the regulatory confines of any particular market. Its currency denomination may or may not be the same as the domestic market of the issuer. For example, a U.S. corporation can issue Euroyen or Eurodollar bonds in London, and neither issue will be subject to U.S. securities regulation. Although the first Eurobond was issued in 1957, the market remained relatively small until the mid-1970s, when several forces combined to fuel its rapid growth. For corporations in the U.S. and elsewhere volatile inflation caused sharper fluctuations in the availability of internal funds. Since they were forced to tap the external funds markets more frequently during a time of relatively high interest rates, corporate treasurers looked for new ways to economize on both interest and issue costs. One way to reduce interest costs is to seek an untapped clientele of investors who place special value on your securities. Since conditions in other countries had created demand outside the U.S. for dollar denominated securities, large and well recognized U.S. companies sought to lure foreign investors. Securities issued in the U.S. to these investors could not be easily sold, however. Since the late 1960s, the federal government had imposed a 30 percent withholding tax on interest paid to foreign holders of domestically issued securities. It was natural to turn to the Eurobond market where no such tax existed and where bonds could be issued in bearer form to preserve investor anonymity. In addition, issuers in this market did not face the SEC's costly and time-consuming registration and disclosure requirements. Aided by these advantages, annual Eurobond issues grew from $2.1 billion in 1974 to $24 billion in 1980. In this case, a more urgent need to economize on the costs of raising funds helped spur the trend toward global integration of securities markets.
Although its spectacular growth phase did not occur until the mid-1980s the junk bond market should be mentioned in this same context. Junk bonds (or "high-yield" bonds, as their promoters prefer to call them), are publicly issued bonds that are either rated below investment grade by the rating agencies or unrated altogether. Junk bonds have long existed. They accounted for 17 percent of total bonds issue proceeds to U.S. corporations during the years 1909-43. In the wake of the Great Depression, however, the new issue market dried up and became dormant. It was not until 1977 that it revived. The same emphasis on economy and flexibility that was important in the Eurobond market's growth provided an impetus for this revival. Lower-grade issuers formerly confined to negotiated loans from commercial banks, finance companies, or insurance companies discovered they could economize by going directly to the public market. And public issues typically contained fewer restrictive covenants than negotiated loans. The move to the public markets also illustrates the growing competition among different types of financial institutions during the 1970s. Drexel Burnham Lambert, the investment banking firm credited with reviving the junk bond market, was able to win business away from commercial banks and others by taking debt contracts that would formerly have been held as bank loans and selling them in a public market. Thus the growth of the junk bonds market can be seen as part of the more general "securitization" phenomenon, in which nonmarketable credit instruments that intermediaries would once have held privately become publicly traded securities.
For securitization to work, of course, public investors must demand the securities. Several factors interacted to create the demand for junk bonds. First investors discovered during the 1970s that high credit quality would not protect them from inflation and interest rate fluctuations. Junk bonds were attractive on this score because of their high yields. In addition, their higher default risk causes junk bond values to fluctuate more with their issuers' asset values. Consequently, junk bonds are relatively less susceptible to pure interest rate risk than are investment grade bonds. Finally, the ability of Drexel Burnham and other securities firms to make relatively liquid markets in junk bonds greatly enhanced their appeal to investors.
Securitization manifested itself in a different form in the development of mortgage pass-through securities. The first Government National Mortgage Association (GNMA, or "Ginnie Mae") pass-throughs appeared in 1970. These securities are backed by portfolios, or pools, of government-insured home mortgages, all of them bearing the same contract interest rate. Monthly payments of principal and interest are passed through proportionately to holders of the securities at an interest rate 50 basis points below the contract rate. The securities are issued directly by mortgage originators, such as mortgage banks, commercial banks or thrift institutions, but timely payment of the scheduled interest and principal is guaranteed by GNMA. A fee of six basis points is paid to GNMA for providing this guarantee; the remaining 44 basis points of the difference between the contract rate and the pass-through rate goes to the issuer as a servicing fee
In this case, securitization is helped by the standardized features of the pass-throughs. The U.S. home mortgage market had been quite segmented Lenders were hindered in their attempt to exploit regional interest rate differentials by the difficulty of evaluating the credit quality of home mortgage far removed from their own territory. The government guarantee and restrictions on the mortgages eligible for the pool made pass-throughs far more homogeneous than the typical direct mortgage investment. This homogeneity in turn allowed the growth of secondary trading, making pass-through more liquid. Pass-throughs helped to satisfy investors' increased demand for real estate-related securities. Thus, the mortgage market was opened to a broader spectrum of investors and national integration of the market was furthered.
In addition to new instruments and broader markets for existing instruments, the volatility of the 1970s also gave rise to new risk-management strategies. One such strategy, the parallel loan, was the precursor of the booming 1980s market in interest rate and currency swaps. Parallel loans were designed primarily to circumvent exchange controls and their attendant exchange rate risk. Suppose, for example, that Company A does business in Country B and faces a stream of future receipts denominated in B's currency. Because of exchange restrictions, Company A will not be able to freely convert these receipts into its own currency and thus will be exposed to future exchange rate risk. If Company B, operating in Country A, faces the analogous situation, however, the two companies can eliminate exchange risk through a parallel loan. Company A lends Company B an amount denominated in A's currency. Simultaneously, Company B lends Company A an amount denominated in B's currency. Since the initial loan amounts are equivalent, no money changes hands at the outset. In the future, B repays its loan with its A-denominated receipts and A repays its loan with its B-denominated receipts. The two companies have effectively swapped their future receipts, thus fixing them in their own currencies and insulating themselves against future exchange rate risk.
In this instance, the increased importance of exchange rate risk, in combination with regulatory restrictions, gave rise to an innovative strategy. But that strategy had the same limitation as barter: each party needed a counter-party with exactly offsetting needs. Because of this limitation, and because the loan agreement exposed each party to default risk, even as it protected against exchange rate risk, parallel loans remained a small and specialized feature of financial market activity throughout the 1970s.
Financial futures markets, which also began in the 1970s, addressed exactly this type of weakness by allowing parties to interact anonymously through an exchange clearinghouse mechanism. A financial futures con tract is an agreement that sets the term on which a financial instrument will b bought or sold on a specified future date. Unlike a forward contract, it is traded on an organized futures exchange, and buyers and sellers strike their bargains with the exchange clearinghouse, not with each other. The exchange sets rules about contract terms, thus ensuring the degree of standardization needed to promote anonymous auction trading. Furthermore, the contracts are "marked to market" each day by the exchange. That is, parties on both sides of the market have their accounts debited or credited by the amount of daily price movements, and they must maintain sufficient "margin" in their accounts to protect the exchange from default by either side.
The first financial futures contracts in the U.S. were for currencies traded on the International Monetary Market in Chicago. These were soon followed by contracts on GNMA pass-throughs and U.S. Treasury bills and bonds. Their popularity was attributable to the increased volatility of both exchange rates and interest rates in the 1970s, as they afforded new opportunities for both investors and bond issuers to either hedge or speculate on market fluctuations. But, although these contracts were conducive to high-volume trading and liquid markets, their marking-to-market feature was cumbersome to large corporate and institutional users. In addition, the heavy volume needed to justify the market's operation meant trading was confined to a limited number of instruments and future dates. Thus, an opportunity remained for a risk-management mechanism that was somewhere between the highly specialized parallel loan and the highly standardized futures contract.
To summarize, the major bond market innovations of the 1970s offered new ways to deal with interest rate and currency risk in an environment of uncertain inflation. Liquidity and flexibility are important in such an environment. Therefore, several of the innovations were aimed at tapping the advantages afforded by public markets. In addition, some of the innovations exhibited the trends toward globalization and securitization that would become hallmarks of bond markets in the 1980s.
INNOVATIONS OF THE 1980S
An important feature of financial markets in the 1980s has been the continuation and further evolution of trends that began in the 1970s. This includes the natural growth of existing markets and instruments in the face of increased demand, their adaptation to new uses, and their alteration to address perceived problems.
Financial futures trading, for example, has mushroomed, both in the U.S. and abroad. Available contracts now cover not only traded securities, but also market indexes. The junk bond market has continued its tremendous growth: by the end of 1988, the amount outstanding exceeded $180 billion, more than 20 percent of the total public corporate bond market. This growth was fueled, in part, by the realization that the flexibility and rapid access to capital afforded by junk bonds were ideally suited to leveraged buyouts, takeovers, and other corporate restructuring.
Market instruments have undergone continuous tinkering in an attempt to meet the needs of investors and issuers more efficiently. Although floating-rate notes protect investors against interest rate risk, for example, they do not protect against credit risk. A bond whose coupon rate is set at a fixed spread over the Treasury bill rate will still decline in value if the issuer's credit quality deteriorates. Hence, there have been some issues of "puttable" or "extendable" bonds. These are floating-rate bonds that give the investor the option to periodically put (resell) the bonds to the issuer at par. As long as the issuer can honor this option, then, the extent to which the bonds' value can fall below par will be limited. In a similar vein, "reset" bonds call for the interest rate to be periodically reset by the underwriter to a level that will allow the bonds to trade at par.
Other variations in contract design call for adjustments to interest or principal payments that are geared to some index other than market interest rates. Examples include commodity price levels, such as oil, gold, and silver; general price levels, such as the Consumer Price Index; and even business activity levels, such as New York Stock Exchange trading volume. The motivation for such wrinkles is to better match investors' desires to protect against price risk with the issuers' own ability to hedge these risks.
The market for any security type will always be limited if the desires of investors and issuers must be precisely matched. In our opinion, then, the most significant financial innovations of the 1980s have been those that allow an uncoupling of investor and issuer desires. This has been accomplished through a radical change in the role of investment bankers and in patterns of financial intermediation. We will illustrate this development by considering, in some detail, three of its primary manifestations: zero-coupon bonds, collateralized mortgage obligations, and the use of specialized securities in combination with swaps to exploit market niches around the world.
A zero-coupon bond, which promises a single payment on a specified future date, is the most basic security imaginable. indeed, any conventional bond, with its stream of coupon and principal payments, can be considered a series of zero-coupon bonds. Because of its intermediate payments, a conventional bond returns cash to the purchaser more quickly than does a zero-coupon bond of identical maturity. This implies that the conventional bond has a shorter duration, which means that its value is less sensitive to changes in market yields than is the identical-maturity zero. On the other hand, the zero-coupon bond allows the investor to lock in the stated yield to maturity: since there are no coupons, no assumption need be made about the rates at which intermediate payments will be reinvested.
Despite their simplicity, zeros were virtually unheard of until the 1980s even though there have never been any legal or regulatory restrictions against issuing or investing in them. The absence of zeros is perhaps as important to understanding innovation as is their introduction.
In theory, there should be ready buyers for zeros as long as there is some difference of opinion on future interest rates. Some short-term speculators might like the long-duration aspect of zeros, since their prices will rise faster than those of conventional bonds in the event of a decline in rates. Likewise, some long-term, buy-and-hold investors might like the absence of coupon reinvestment risk and the ability to lock in the promised yield to maturity. In particular, the passive investment strategy known as immunization is easier and less costly to implement with zero than with coupon bonds. But this latent demand for zeros was not really sparked until the cyclical peak in interest rate in 1982. Persistent inflation kept up ward pressure on market rates through out the late 1970s. In that environment investors were interested in shortening, not lengthening, the average maturity of their portfolios. Floating-rate notes, which protect against interest rate increases, were more likely to attract investor attention than zero-coupon bonds.
In addition, there was virtually no natural supply of zero-coupon securities before 1980. Since most investment projects generate a stream of revenues, few corporate or governmental funding needs call for single, lump-sum payments at specific future dates. corporate borrower issuing zero-coupon debt would also have to convince skeptical investors that the lack of coupons was not a sign of financial weakness.
The impetus for the rise of the zerocoupon bond market in the 1980s came from corporate tax strategy rather than from interest rate risk management. According to IRS rules prevailing at the time, the issuer of an original issue discount (O.I.D.) bond(1 ) had to amortize the discount on a straight-line basis over the lifetime of the bond. The amortized amount is treated as interest expense by the issuer and interest income by the investor, so the issuer has a tax write-off and the investor has a tax obligation even though no cash is exchanged. The use of straight-line rather than compound-interest amortization on the O.I.D. bond allowed the corporate issuer to have a tax-deductible expense larger than the true financial expense, lowering its effective cost of funds. That tax treatment did not suddenly become available in 1980, but the value of the net tax advantage to corporate issuers was attractive only in periods of high interest rates. Thus there was a large volume of corporate zero-coupon issues in 1981 and early 1982.
The initial demand for zeros was also influenced by taxes. In a number of countries, most notably Japan and the United Kingdom, the increase in value as the zero approaches maturity was treated as a capital gain rather than an accrual of interest. In Japan such capital gains were untaxed, and thus zeros had a substantial appeal to taxable investors.
The Tax Equity and Fiscal Responsibility Act of 1982 removed the tax advantage for U.S. corporations issuing zeros. Any O.I.D. debt issued after July 2,1982 had to use compound-interest amortization. That might have spelled the end of the zero-coupon market, but the dramatic decline in interest rates in mid-1982 spurred additional demand for zeros that had been absent in previous years. The demand was focused, however, on Treasury zeros, which are not sullied by credit risk considerations.
Treasury-backed zero-coupon bonds are a prototype of the financial engineering activities that have burgeoned in the 1980s. There was substantial demand for long-term Treasury zeros, but limited supply. A market in stripping and selling coupons from bearer Treasury bonds existed, but this market was small and illiquid because of the necessity for physically handling small-denomination coupon certificates. Then, in August of 1982, Merrill Lynch introduced TIGRs (Treasury Investment Growth Receipts), and other investment banks soon followed with their own versions-for example, Saloman Bros.' CATS (Certificates of Accrual on Treasury Securities) or Lehman Bros.' LIONs (Lehman Investment Opportunity Notes).
TIGRs and CATS are remarkably simple in design, as illustrated in Figure 1. The investment bank buys conventional Treasury bonds and places them in a single-purpose dedicated trust. The trust, which is managed by a custodian commercial bank, issues the zero-coupon securities. The coupon and principal cash flows from the conventional bonds are used to pay the principal on the derivative zeros at their maturity dates. The investment bank profits by purchasing the coupon Treasuries for less than the total sale price of the zeros. In the declining interest rate environment of late 1982 and 1983, buyers were willing to pay extra for a long-duration zero with no credit risk. The investment bank operates as a traditional financial intermediary, a role usually taken by a commercial bank or mutual fund. One security's characteristics are transformed by the intermediation process to supply the market with a security it prefers to hold.
Given the investment banks' profit from creating zeros via coupon stripping, one might wonder why the Treasury did not, itself, immediately issue zero-coupon debt (other than nonmarketable Series EE Savings Bonds). However, the Treasury has never sought the role of financial innovator. Because of its large scale, it intentionally seeks to be predictable with respect to the timing, amount, and type of its debt issues. The Treasury auctions its bills, notes, and bonds at scheduled offerings in announced quantities so the private sector can work around its dominating presence. Nevertheless, the unprecedented size and carrying cost of the national debt in the 1980s has forced the Treasury to respond to market pressures and design its securities with (increasingly overseas) investors in mind.
The Treasury finally addressed the overwhelming success of the financially engineered zero-coupon market by introducing its STRIPS (Separate Trading of Registered Interest and Principal of Securities) program in February 1985. Any financial institution having access to the Federal Reserve's book entry system can participate by buying specially designated Treasury notes or bonds and converting them to STRIPS. This effectively requires separate registered owners for each of the cash flows and allows the individual coupon and principal payments to be sold as zero-coupon instruments. The program eliminates the need for the custodian bank and dedicated trust in the CATS or TIGRs framework. STRIPS now have such an established place in the Treasury market that their prices and yields are quoted in the Wall Street Journal alongside the traditional bonds, notes, and bills.
Collateralized Mortgage Obligations
The broad decline in market interest rates after mid-1982 also induced issuers of callable debt to exercise their prepayment options and refinance at the lower prevailing rates. This especially hurt investors in mortgage securities. Whereas corporate bonds usually have a call protection period and are callable at some premium over par value, mortgages can be paid off at any time, often without a prepayment penalty.
Investors in GNMA pass-throughs found out about prepayment risk in the early 1980s. Prepayment risk, of course, had always been present, but in the rising interest rate environment of the 1970s voluntary refinancings of mortgages were rare. In fact, some creative real estate financing strategies were designed to pass on an existing, low-rate mortgage to the home buyer. GNMAs seemed to institutional investors to offer a low-risk way to pick up some extra basis points relative to Treasury securities, especially since the default risk was viewed to be equivalent. But unexpected prepayments meant that realized returns fell short of expectations, as more funds had to be reinvested sooner at lower rates.
Investment banks soon recognized the market demand for a mortgage-backed security that had less prepayment risk than the traditional GNMA pass-through. Each investor in a GNMA pass-through owns a share of the underlying mortgage pool and receives all payments, including principal prepayments, on a pro rata basis. Therefore, all shares of the GNMA mortgage pool have the same maturity and prepayment risk. The idea behind collateralized mortgage obligations (CMOs), which were introduced in 1983, is to redistribute the timing and amount of interest and principal cash flows across the investor base. This provides a range of expected maturities and redistributes the prepayment risk.
To create a CMO, an investment bank first buys a pool of GNMAs, say $100 million of 30-year pass-throughs with a 10 percent coupon rate (GNMA 10s). Then, following the TIGRs and CATS prototype, the GNMAs are placed in a single-purpose, dedicated trust that issues various classes of CMOS. Like Treasury-backed zeros, the CMOs are derivative securities: the funds to pay the interest and principal on the CMOs are derived from the cash receipts on the underlying GNMAs. This process is pictured in Figure 2.
The CMO structure has turned out to be a remarkably flexible way to transform GNMA cash flows. The first CMOs divided up the principal payments according to time, basically letting the various classes "queue up" for receipt of principal. For example, the CMO could have two $50 million classes, A and B. Class A receives all scheduled and unscheduled principal paid on the underlying GNMAs while Class B receives only interest. Class B has "prepayment protection" until Class A is fully paid off. In technical terms, the structure divided a 30-year GNMA with a duration of, say, 12 years into two smaller pieces with durations of, say, 6 and 18 years. Typically, these CMOs have three or more classes, sometimes with different coupon rates and often including a zero-coupon class.
The investment bank profits when the CMOs are sold for more than the purchase price of the GNMAS, so the key for the intermediary is to identify investor market segments that have different hedging needs or different views on future market conditions. The basic CMO structure creates a "mortgage yield curve" and allows investors to specialize in particular maturity ranges. Thrift institutions were drawn to the "fast-pay" classes to match their relatively short-term deposits. Pension funds and insurance companies were drawn to the "slow-pay" classes to match their long-term liabilities. The 1980s trend toward global integration of financial markets has created many other opportunities to serve specialized market segments; CMOs have afforded a pliable device for realizing these opportunities.
Another relatively simple CMO structure divides the underlying GNMAs into "high coupon" and "low coupon" classes. For example, the pool of GNMA 10s could be divided into two $50 million classes, one with a 12 percent coupon rate priced at a premium, and the other with an 8 percent coupon rate priced at a discount. These classes will perform differently depending on whether prepayments are faster or slower than originally expected. Investors with different views on future market rates will prefer one class over another.
An extreme version of this type of innovative structure is the "interest only/principal only" (IO/PO) CMO. The GNMA 10s again could be divided into two classes, one that receives all of the interest payments, and the other that receives all of the principal. Each class would have a principal of $100 million (a "notional" principal in the case of the 10) and would be priced at a discount. These classes turn out to be extremely sensitive to changes in market interest rates. If rates plummet and mortgage refinancings soar, the PO class would jump in value, since cash flows are received much sooner than expected. The 10 class would decline in value, since future interest receipts disappear altogether as principal is prepaid. The organizer of the CMO is, of course, unconcerned about this interest rate risk as long as the separate pieces can be fully sold at prices that exceed the purchase price of the GNMAs.
CMOs can also be used to create a floating-rate class, even though the underlying securities are fixed-rate mortgages. This requires creation of another innovative security, an inverse floating-rate note ("bull floater"), whose coupon rate goes up when market rates fall. Such a security would experience a large increase in value when market rates decline and hence would be attractive to investors who are bullish on bonds. Creation of these two classes is possible as long as the combined interest payments to the CMO classes never exceed those on the underlying GNMAs.
Floating-rate CMOs proved to be very popular after their introduction in 1986 and are credited with supplanting the unsecured floating-rate note in the Eurodollar market. The main reason is that the CMO offers much lower credit risk than the typical corporate obligation, since the underlying GNMAs are insured. This structure also deals with prepayment risk in a novel way. As long as the floating-rate class trades at or close to par value, principal prepayment is of little consequence since the funds can be reinvested in equivalent securities.
The basic CMO structure has also been adapted to other types of credit instruments, such as auto loans and credit card receivables. Because of its flexibility in parceling out cash flow streams, it can securitize virtually any type of loan agreement.
Specialized Securities with Swaps
Currency swaps evolved from parallel loans to address the problem of default risk. If one of the parties to a parallel loan defaults the other cannot easily suspend its own obligation since there are separate loan agreements. A currency swap is, in principle, merely an exchange of the principal and coupon cash flows on bonds of equal value but denominated in different currencies. The coupons can be either fixed or floating rate. In the swap agreement, performance of one party depends on the continued performance of the other; if one side defaults, the other can suspend its payments.
The interest rate swap market developed in the 1980s from currency swaps. An interest rate swap is an exchange of coupon cash flows on bonds of equal value. For example, one coupon stream might represent fixed-rate payments, while the other might represent floating-rate payments. Since the bonds are denominated in the same currency, an exchange of principal is unnecessary.
A swap can be viewed as a series of forward contracts, since it sets the terms on which cash will change hands on specified future dates. However, the agreement is made without the margin accounts and daily marking-to-market procedures that characterize exchange-traded futures contracts. Swaps, therefore, are more flexible but less liquid than futures contracts, and have greater default risk than futures.
Interest rate and currency swaps are key factors in integrating worldwide capital markets. They provide an efficient tool to transform the nature of liabilities-a firm can issue in the market and currency where it is most advantageous and then convert that debt to the desired currency or coupon pattern via swap agreements. This continuous search for arbitrage opportunities brings rates into alignment, making the markets more efficient.
Swaps can be used to lower funding costs by identifying investors with specialized needs and designing innovative securities for them. If a security's cash flow stream does not meet the issuer's needs, it can simply be swapped. For example, suppose that a firm can issue an inverse floating-rate note at 21 percent minus the London Interbank Offer Rate (LIBOR) in the Eurodollar market. It can then enter an interest rate swap with a commercial bank to receive a fixed rate of 10 percent and pay a floating rate of LIBOR. The financially engineered package, as exhibited in Figure 3, results in fixed-rate funding at approximately 11 percent. A firm would do this if the cost of funds, including the default risk it bears on the swap, is lower than its alternative fixed rate cost. This might occur in the case of the inverse floater if some pocket of investors is particularly bullish on bonds and expects LIBOR to fall.
In the 1980s, a large number of innovative securities have been aimed at Japanese institutional investors. The increasing pool of accumulated savings in Japan has been an attractive target for the security issues of governments and corporations around the world. It is only natural, then, that new debt instruments and markets have been developed to address the unique needs of Japanese investors, especially the large life insurance companies. Two particular regulations facing these institutions have affected the design of debt instruments: first, they face a limit on the percentage of foreign (non-yen-denominated) assets they can hold; second, they are required to pay dividends only from coupon interest income and not from capital gains.
Dual-currency bonds are a classic example of a security that comes from investor demand and not issuer supply. These bonds pay coupon interest in one currency and principal in another. The usual pattern has been coupons in a relatively low-interest currency, such as the Swiss franc or Japanese yen, and principal in a high-interest currency, such as the U.S. dollar. A dollar/yen dual currency bond deals with the regulatory constraints on Japanese life insurers in a clever manner. First, the Japanese Ministry of Finance has ruled that such a bond is "primarily yen-denominated" and so does not fall within the foreign asset constraint. Second, since the U.S. dollar is expected to depreciate against the yen, such a bond must bear a higher coupon rate than a straight yen bond. The increased interest income in turn helps the Japanese life insurance companies pay dividends and essentially represents a tradeoff between the (desirable) coupon and (less desirable) principal components of total return.
The corporate issuer, on the other hand, is unlikely to have a project generating that unique stream of yen and dollar cash flows. Thus, the corporate issuer would not have found it attractive to issue a dual-currency bond prior to the development of the swaps market. Now, however, a currency swap can be used to hedge the issuer's yen exposure, since it sets the terms on which dollars will be exchanged for yen on future coupon dates.
Japanese institutional investors' appetite for high coupon debt also led to the development of "weak currency" capital markets such as the Australian dollar and New Zealand dollar. In 1987, U.S. corporations issued more than $1 billion in Eurobonds denominated in these two currencies. This represented approximately 5 percent of their total Euromarket borrowing in that year; U.S. corporate borrowing in those currencies had been negligible as recently as 1984. The globalization of capital markets in the 1980s has enabled corporate borrowers to issue debt in any currency, whether or not they have matching income streams. Currency swaps have been the vehicle for this capability: the firm can always swap for the desired liability. Hence, a large percentage of Eurobond issues in both the traditional and newly integrated currencies have been accompanied by swaps. Tokyo has been the ultimate destination of much of these debt issues, in part because of the perceived need for global diversification and in part because of the need for high coupon income.
THE NEW STRUCTURE OF FINANCIAL INTERMEDIATION
The pace of financial innovation in recent years has been breathtaking. Sparked by technological advances, changing risk-management needs, taxes and regulation, and a changing distribution of the world's investable wealth, new instruments and strategies have poured forth in seemingly endless variety. The bond markets have become truly global in scope. In our view, however, the outcome of bond market innovation in the 1980s can best be summarized by examining the changes it has brought about in the structure of financial intermediation.
A financial intermediary-a commercial bank, thrift institution, credit union, mutual fund, or pension fund-sells securities to ultimate lenders, using the proceeds to buy other securities from ultimate borrowers. The key element is that securities held as assets differ from those held as liabilities along one or more dimensions, such as maturity, default risk, denomination, and marketability. The intermediation process serves to transform the securities and allows borrowers and lenders to transact with financial instruments that meet their particular needs. In its most classic form, large-denomination, individually risky, nonmarketable business loans are transformed by commercial banks into small-denomination, low-risk, highly liquid consumer deposits.
In this type of intermediation, the bank purchases debt claims for which there is little or no organized market. During the 1980s, however, investment bankers have found new ways to bring formerly nonmarketable credit instruments within the scope of organized securities markets. With junk bonds they have done so simply by acting as market makers and providing liquidity. In other cases, the investment banks have served as true financial intermediaries.
Traditionally this was not the case. Investment banks underwrote new issues or developed secondary markets in existing securities, but did not transform the securities they purchased. This has changed in the 1980s as the very nature of investment banking has changed. Investment bank intermediation, as in the creation of Treasury-backed zeros and CMOs or the combining of securities issues with swaps, differs from traditional intermediation in several ways. In commercial banking, the transactions are on the balance sheet and are primarily designed to meet the needs of the domestic retail depositor. Intermediation for an investment bank is aimed at exploiting some unmet, perhaps short-lived, market demand. The transactions are primarily off the balance sheet (via single-purpose dedicated trusts or swap agreements) and are aimed at the needs of the global institutional investor.
Investment banks have been drawn to this activity in part because their own traditional business has shifted its emphasis. Starting with the advent of shelf registration in 1982, investment banking has increasingly been based more on the ability to execute transactions than on customer relationships. The day of the long-standing, almost inviolate, connection between corporation and investment banker has passed. Instead, corporations now seek out the least-cost bidder to provide them with financial services.
One result of this has been an emphasis on innovation as a means to increase market share. The investment bank that is first to develop (and name) a new type of security or strategy can win new business that in previous days would have been tied by customer relationships to other firms. Others will naturally imitate the innovation, but the originator has a head start. In this environment, investment banks have sought to establish reputations for creativity and expertise in financial engineering. They have increasingly hired professionals with advanced technical training to staff their new-product research departments.
As a result of this emphasis on innovation, an institutional investor can now hold securities no primary borrower would consider issuing. Some innovative instruments, such as an IO/PO mortgage-backed security, would be virtually impossible, perhaps even illegal, to issue directly. Others, such as an inverse floater, would test the limits of project finance. However, the newfound ability to transform cash flows with respect to timing and amount-types of intermediation far beyond the classic role of a commercial bank-irrevocably breaks any necessary link between ultimate borrower and ultimate investor.
Commercial banks, of course, have not sat idly by while investment banks chipped away at their business. They, too, have adapted to a more transactions-based environment by increasingly separating and repackaging different facets of their traditional intermediation function. Even when they choose not to hold the loans on their balance sheets, for example, commercial banks and thrift institutions still earn fee income by originating and servicing the mortgages underlying various types of mortgage-backed securities. Commercial banks have also separated the credit analysis component of their traditional business in a variety of innovative ways. When a company goes directly to the public debt markets, a bank can support the issue with a letter of credit. Many large, money center banks have assumed a market-making role in the swap market. They serve as the direct counter-party in swap agreements and then manage their interest rate and currency risk exposure by entering into offsetting agreements with other parties. However, since they do not set up the same system of margin accounts and marking-to-market that characterizes futures exchanges, this role in turn requires that banks assess the credit-worthiness of any party with whom it has made a swap agreement. In capacities such as these, then, the banks act as facilitators of market transactions rather than as true financial intermediaries.
Whether the pace of bond market innovation will be as rapid in the 1990s remains to be seen. It will depend on changes in risk perceptions, tax laws and regulations, and technological advances. Whatever these changes may be, however, the innovations of the 1980s have fundamentally changed the way that market participants conceive of the process of financial intermediation.
1. An O.I.D. bond is one in which the issue price is less than 100 minus 0.25 times the number of years to maturity. For example, if the price of a 10-year bond is less than 97.5 percent of the face value at issuance, it is considered O.I.D. for tax purposes.
Keith C. Brown and Donald J. Smith, "Recent Innovations in Interest Rate Risk Management and the Reintermediation of Commercial Banking," Financial Management, Winter 1988, pp. 45-58.
Ian Cooper, "Innovations: New Market Instruments," Oxford Review of Economic Policy, Winter 1986, pp. 1-17.
John D. Finnerty, "Financial Engineering in Corporate Finance: An Overview," Financial Management, Winter 1988, pp. 14-33.
Richard M. Levich, "Financial Innovations in International Financial Markets," National Bureau of Economic Research Working Paper No. 2227, June 1987, Cambridge Mass.
Merton H. Miller, "Financial Innovation: The Last Twenty Years and the Next," Journal of Financial and Quantitative Analysis, December, 1986, pp. 459-471.
Kevin J. Perry and Robert A. Taggart, Jr., "The Growing Role of junk Bonds in Corporate Finance," Continental Bank Journal of Applied Corporate Finance, Spring 1988, pp. 37-45.
Donald J. Smith, "The Pricing of Bull and Bear Floating-Rate Notes: An Application of Financial Engineering," Financial Management, Winter 1988, pp. 72-81.
Marcia Stigum and Frank L. Fabozzi, The Dow Jones-Irwin Guide to Bond and Money Market Investments (Homewood, Ill.: Dow Jones-Irwin, 1987).
Dimitri Vittas, "The New Market Menagerie," The Banker, June 1986, pp. 16-27.
|Printer friendly Cite/link Email Feedback|
|Author:||Smith, Donald J.; Taggart, Robert A., Jr.|
|Date:||Nov 1, 1989|
|Previous Article:||The decline of public equity: the return to private enterprise?|
|Next Article:||The globalization of world financial markets.|