Diminishing Treasury Supply: Implications and Benchmark Alternatives.
Over the last year, the fixed income markets have become increasingly subject to technical distortions as the development of a U.S. government surplus has resulted in a diminishing supply of U.S. Treasury securities. This technical pressure has substantially inverted the yield curve, increased levels and volatility in yield spreads between government bonds and other fixed income securities, and created a great deal of confusion among market participants regarding appropriate benchmarks. The supply-induced inversion affects corporate finance, risk management, valuation and asset allocation, and will force the adoption of alternative benchmarks.
For several decades, persistent federal government deficits forced significant issuance of U.S. Treasury securities. From 1980 to the peak in 1997, U.S. Treasury debt outstanding grew from $1.6 trillion to $3.8 trillion (an annualized growth rate of 5.2 percent). Accelerating economic growth throughout the late 1990's and relative fiscal austerity have resulted in declining levels of net federal borrowing since 1992, culminating in a budget surplus in 1999. The Office of Management and Budget now forecasts surpluses to continue, to the tune of $300 billion a year until 2010 and $400-550 billion through 2012. Treasury debt outstanding now stands at around $3.2 trillion, and is shrinking rapidly--particularly in comparison to private sector debt. (See Figures 1 and 2.) Given the extent of the budget surpluses, the U.S. Treasury has reduced the size and frequency of its borrowings. It is also repurchasing its own securities--over $30 billion of notes and bonds is likely to be repurchased this year alone. Late i n 1999, the magnitude of this shortage became apparent, particularly in the long end of the yield curve (beyond ten-year maturities). As a result, investors rushed into Treasuries, particularly into longer-dated securities. The Treasury curve inverted sharply--first from ten-year notes to thirty-year bonds, then eventually along the entire curve from two-year notes all the way out to the thirty-year bond. (See Figure 3.)
Other fixed income instruments have been significantly resistant to this inversion, even in relatively short maturities. This supply dynamic, in conjunction with fundamental concerns regarding credit quality and the late stage of the economic cycle, has resulted in a dramatic widening in the spreads between risky fixed income assets and Treasury securities (beyond even the panic-induced levels of the 1998 meltdown). (See Figure 4.)
Treasury debt will not disappear overnight, and, forecasts aside, it is far from certain that all Treasury debt will be completely retired, as that outcome depends on future economic growth and fiscal policy. However, the unusual supply-and-demand dynamic in the market, and the resultant discrepancy in yield levels and curve shape, have brought into question the Treasury's benchmark status. Among the most likely alternative benchmarks are U.S. agencies (Fannie Mae, Freddie Mac, and the Federal Home Loan Banks), large private borrowers, and the interest rate swap market.
Until recently, even sophisticated market participants have taken for granted the central role of Treasury securities in the global financial markets. These extremely liquid, risk-free assets play critical roles in measuring monetary conditions and inflation expectations, in valuing corporations, risky securities, and markets, and in facilitating trading and risk management on a global basis. Finding a suitable alternative benchmark will not be easy.
Why Treasuries Are a Good Benchmark
Treasury securities play three basic roles as a benchmark. First, backed by the full faith and credit of the U.S. government, Treasury yields can be considered the risk-free rate, thus becoming the critical input when discounting cash flows of other securities. Examples include valuing non-government debt, equity, derivatives, and currencies. The risk-free rate is also the basis for corporations and other investors evaluating projects or potential acquisition candidates using discounted cash flow analysis. This role transcends borders, as a key characteristic of a good benchmark is to facilitate such analysis across currencies.
Second, Treasuries provide the benchmark rate for asset allocation decisions. As investors evaluate relative risk and return, government bonds are typically considered the fixed income alternative to equity investments. Capital will flow from one asset class to another based on expectations of price return, yield, and volatility. Of course, the anticipated interest rate environment itself contributes to these expectations, and the shape of the Treasury yield curve is a benchmark for portfolio decisions on duration.
Finally, Treasury securities play an important role in the capital markets as tools of risk management, trade facilitation, and relative value comparison. Securities dealers use Treasury securities to hedge interest rate risk embedded in corporate bonds, mortgage-backed securities, and a vast variety of other fixed income instruments. Dealers and fixed income investors also use Treasury securities as benchmarks in the pricing of new bond issues and secondary transactions throughout the course of every trading day. Yield spreads relative to Treasuries, whether expressed in nominal terms or as derived by an option-adjusted spread (OAS) model for bonds with embedded options, are the most commonly used means of comparing the relative attractiveness of securities across the maturity and risk spectrums.
Treasury securities have reached this degree of acceptance as a benchmark because they are U.S. government obligations, and hence the timely return of principle and interest is unquestioned. Furthermore, these bonds are extremely liquid--they trade with an extremely tight bid-ask spread in significant size by market participants of all varieties, aided by the ability to short and borrow through the repurchase market. Trading and pricing of Treasury bonds is also transparent, facilitating their use as a pricing determinant for new transactions. Finally, the long historic data set created as a result of price transparency reinforces the benchmark role.
Implications of a "Benchmarkless" World
The yield distortions caused by the unusual supply conditions and the resultant evolution toward alternative benchmarks create a host of concerns for financial market participants.
In corporate finance, the discounting of cash flow (DCF) at the risk-free rate is an elemental function in the valuation of corporations and projects. The magnitude of valuation errors introduced by irregularities in the Treasury curve could be significant. Consider a cash flow stream of one hundred dollars growing at twelve percent per annum for ten years, and then at six percent thereafter (perhaps analogous to the cash flows expected to be produced by a certain corporation or project). A sixty basis-point distortion in the coupon curve (in this example a linear inversion from two years to thirty years, which would manifest itself into an eighty-three basis-point distortion in the zero coupon curve) reduces the value of that cash flow stream by 6.7 percent relative to a normal yield curve.
Public market investors (particularly "value"-oriented money managers and analysts) also use DCF analysis to look at equities. They need to be very careful that they are not overstating valuations. For other investors, market valuation models will also tend to overstate fair value given artificially low Treasury rates. Dividend discount models--such as the one Federal Reserve Chairman Alan Greenspan is said to monitor and those used by equity market strategists Byron Wien and Peter Canelo at Morgan Stanley Dean Witter--essentially value growing annuities and, therefore, will be similarly skewed to overstate the fair valuation of the market. Simpler models comparing dividend yields (whether of market averages or sub-sectors such as utilities) or earnings yields versus bond yields will also be distorted.
These concerns regarding the valuation of equity are due to the necessity of comparing risky cash flows to a stream of fixed (and guaranteed) payments, essentially a measurement of opportunity cost. That measurement also guides asset allocation. Portfolio selection among bonds, equities, cash, and other investments requires making assumptions about the relative return and risk potential of these asset classes. The potential for a scarcity value to affect bond returns complicates this analysis, although it is relevant to note that past periods of heavy Treasury issuance did not coincide with periods of rising yields (and therefore a negative correlation to price returns). (See Figure 5.)
As Table 1 shows, there were thirteen quarters in which net Treasury issuance exceeded five percent of the existing outstandings, but again no strong pattern is seen in the direction of yields (in fact, yields typically declined).
This relative supply dynamic is even more daunting for fixed income investors than for equity investors. Consider the thirty-year debt of the Ford Motor Company (the largest U.S. corporate bond issuer). During 1999 (a year noted for high spread volatility), these bonds traded at a yield spread to long Treasuries ranging from 103 basis points to 153 basis points, and closed for the year at a spread of 118 basis points off their long Treasury benchmark. As the Treasury curve inverted in early 2000, that spread gapped out to levels in excess of 200 basis points off Treasuries. This phenomenon has been experienced by the entire spectrum of fixed income assets. (See Figure 6.)
Institutional fixed income investors must contend with the likelihood that this pressure will continue as long as the fiscal picture remains so strong. Therefore, despite the fact that yield spreads now seem wide, particularly if one views spreads simply as compensation for default risk, fixed income investors benchmarked versus an index that contains Treasury securities should recognize that underweighting such a technically strong asset class comes with some performance risk.
Portfolio managers, financial institutions, and financial regulators must face the reality that the U.S. bond market is de facto becoming riskier as the supply of Treasuries dwindles and the supply of other securities continues to grow. They must adjust practices and parameters for risk management and the allocation of institutional capital, adjustments that will have to take place around the globe. (See Table 2.)
These parties will be forced to take on incremental credit risk, optionality, or derivative structures. The capital markets will likely create long-duration instruments for the hedging of long-dated liabilities by insurance companies (which generally favor non-Treasury instruments anyway). On the other hand, investors who have used long Treasunes as a deflation hedge for equity portfolios may find substitutes difficult to come by, as cataclysmic events that have resulted in Treasury bond rallies typically lead to widening in spreads of other fixed-income assets. David Swensen's new book has an excellent discussion of the role of Treasury bonds in the context of a broad asset portfolio.  In the government arena, the Federal Open Market Committee is studying how its open market operations will be affected by the dwindling Treasury supply.
Capital markets' applications will also be affected. The vast majority of U.S. agency and corporate bond transactions are still priced versus Treasuries, and Treasuries are often exchanged between dealers and investors in these transactions (though a growing number of agency transactions are priced off the interest rate swap curve). However, in the new issue market, dealers increasingly give preliminary pricing indications relative to one of the several alternative benchmarks, typically either the swap curve or agencies. Thus, financial managers and investment bankers need to be careful when evaluating financing costs. Historic spread relationships relative to the Treasury market are generally not anywhere near achievable.
Economists, policy makers, and analysts may find it difficult to determine real interest rates and inflation premiums, and it will be difficult to gauge the effectiveness of monetary policy actions. If one defines an inflation premium as the differential between the ten-year Treasury bond and the real interest rate (perhaps defined as the yield on the Treasury's own inflation-indexed securities), it is clear that yield curve distortions would understate the inflation premium. Similarly, the real rate of interest would be underestimated if it is calculated as the differential between the ten-year Treasury bond and some inflation measure. As to monetary policy actions, consider the period from January 1, 2000 to the date of the Fed's most recent rate action, May 15, 2000. During that Lime, the Fed raised rates three times, lifting the Fed Funds rate by one hundred basis points. At the same time, thirty-year Treasury bonds actually declined in yield by forty-five basis points, implying monetary conditions as measured out the yield curve were actually less restrictive. However, a look at private market rates tells a different story, as mortgage rates were up forty-eight basis points and rates on Ford Motor's ten-year debt were up seventy basis points. Private market rates showed that the Fed had made some headway in raising borrowing costs. (See Figure 7)
Regulators around the globe must also evaluate risks to the financial system if the Treasury shortage accelerates, particularly with regard to the banking system. U.S. and international banks are significant holders of Treasury securities. The banks will likely be forced to shift these assets towards agency debt, mortgage-backed securities, corporate bonds, or loans. Policy makers have already voiced concerns over the banks' large holdings of agencies, while the other choices will raise the risk profile of the banks' balance sheets.
So, we are left scrambling for new benchmarks. There are three contenders to be crowned the "new" benchmark: U.S. agency debt, the most liquid corporate debt issues, and LIBOR (as represented out the maturity spectrum by the interest rate swap curve). The advantages and disadvantages of these instruments are outlined in Table 3.
Each of these potential benchmarks has already begun to play the role in certain circumstances. Market participants are starting to use both agency and swap spreads as reference points for new deals and secondary trades. Institutional fixed-income investors are starting to use the swap curve as the benchmark for the riskier (or "spread") sectors. Greenspan has even noted the necessity of looking at private market rates to glean market insights. No one benchmark has yet gained full market acceptance, and it is likely that each of the alternatives will be used for some purpose.
Each potential benchmark comes with its own shortcomings. Further, the shape and slope of the agency, corporate, and swap curves have all been affected by the Treasury curve inversion. Therefore, particularly in corporate finance and market valuation applications, the answer may be some combination of estimating the risk-free rate off a short point on the curve and estimating curve slope, while looking at alternative curves for further guidance. (See Figure 8)
As the evolution of the fixed income market from an interest rate market to a credit market continues, expect to see LIBOR make further advances as the benchmark of choice. LIBOR is the cleanest expression of credit due to its importance as a funding benchmark for issuers, many investors, and dealers, while the LIBOR-setting process makes it relatively immune to idiosyncratic credit developments. The size, liquidity, and relative transparency of the swap market should give market participants of all stripes confidence in its benchmark role. Add the ability to use the swap curve to make relative value comparisons across currencies and maturities, and expect to see more LIBOR benchmarking.
U.S. markets have used Treasury benchmarks and hedges to tie performance to general and risk-free interest rate levels. This approaches a deterministic ideal whose appeal may have been overemphasized. Changing the benchmark represents a bold leap into a greater uncertainty; we need to turn the focus from absolute to incremental default premium. These new benchmarks will have default or credit-driven characteristics. The rest of the world already deals with that possibility; now the U.S. may as well. We just need to readjust our point of reference.
Steven A. Zamsky is Principal/U.S. Corporate Bond Strategist with Morgan Stanley Dean Witter. He was ranked #2 in Investment Grade Strategy in both Institutional Investor's 2000 All-America Fixed Income Research team as well as their Global Fixed Income team. He holds a BS from Oregon State University and MBA from the University of Chicago.
(4.) David F. Swensen. 2000. Pioneering Portfolio Management: An Unconventional Approach to Institutional Investment. New York: Free Press.
HIGHEST QUARTERS OF NET ISSUANCE OF TREASURY DEBT AND CHANGE IN YIELDS Net Issuance of Treasury Debt as a Percentage of Outstanding Change in Yield Quarter Treasury Debt (percent) 2Q75 5.01 1.56 3Q91 5.21 -6.45 4Q81 5.24 -22.09 4Q85 5.49 -5.69 4Q84 5.78 -9.96 1Q75 5.80 -8.21 1Q81 5.84 7.61 1Q76 5.85 -0.14 1Q83 6.11 -2.77 4Q82 6.59 -9.47 4Q75 6.83 -8.44 3Q75 7.04 6.94 3Q82 7.50 -15.16 OWNERS OF US TREASURY SECURITIES in $ billions Federal Reserve and Government 2,542.2 Foreign 1,268.8 Pension Funds 445.1 Mutual Funds 350.9 Individual and Others 334.5 State and Local Governments 266.8 Commercial Banks, S&L, Credit Unions 245.1 US Savings Bonds 186.5 Insurance Companies 136.2
ADVANTAGES AND DISADVANTAGES OF ALTERNATIVE BENCHMARKS
Large, liquid market with tight bid/offer spreads that allow investors to trade in size
High price and volume transparency with readily available historical data
Highest credit quality (AAA) among all the with least Treasury spread volatility
Large range of points on the yield curve (e.g. FNMA committed to the 2,3,5,7,10, and 30 yr points)
Open to a large range of investors, with growing repo market liquidity futures market
Commitment to a fixed issuance calendar
Spreads are highly correlated with riskier assets
U.S. centric benchmark
Uncertainty regarding ongoing regulatory status of agencies creates idiosyncratic risk
Uncertainty regarding future funding needs, (i.e. current situation of Treasuries)
Repo market is quite technical, raising funding risk
Component of credit risk in benchmark would reduce spread volatility of risky assets
Best approximation of funding costs for private borrowers
Difficult to trade in size with little price and volume transparency and poor historical data
Price movements subject to credit quality of one issuer or relatively few issuers
Difficult to short sell given higher yields and problems associated with covering short positions
Repo market is quite technical, raising funding risk
Ability to have consistent spread framework across currencies and countries
Index-like nature due to lack of credit-specific risk
High price and volume transparency with readily available historical data
Swap curve is continuous, as opposed to discrete cash markets
Ideal for moving towards an OAS framework (robust pricing for all points on the curve)
Isolated from future funding risk
Availability of tradable forwards (spreadlocks) and options on swap spreads
Somewhat limited investor base that can use interest-rate swaps due to investment policies on derivatives and tedious documentation process
LIBOR spreads affected by technical market dynamics such as demand for fixed, change rates
Swap markets in emerging markets are relatively immature
|Printer friendly Cite/link Email Feedback|
|Author:||Zamsky, Steven A.|
|Article Type:||Statistical Data Included|
|Date:||Oct 1, 2000|
|Previous Article:||The Demise of the 30-Year Treasury Bond as a Benchmark or Pricing Fixe -- Income Securities.|
|Next Article:||Bribery in International Business Transactions and the OECD Convention: Benefits and Limitations.|