ByMichelle Clark Neely
Taking a cue from Canada, the United Kingdom and other countries, the U.S. Treasury decided last year to begin offering inflation-indexed bonds to investors looking to protect their savings from unexpected increases in inflation. (See table below.) With the first auction scheduled for early 1997, economists, finance professionals and policymakers are cheering the change. But will indexed bonds shake the market, or merely cause a stir?
|Country of Issuance||Year of Adoption||Inflation Rate Before Introduction||Indexed Bonds as a % of Total Marketable Debt|
Unlike a conventional, or nominal bond, an inflation indexed, or real, bond promises to pay its holder a fixed real rate of return—a return that is unaffected by unexpected changes in the inflation rate. While a conventional bond repays an investor principal plus some stated interest, an indexed bond repays principal adjusted for inflation and a fixed interest rate applied to the adjusted principal. Investors value such protection because large increases in unanticipated inflation can eat away at an investment's real return.
Expected inflation, real returns and nominal returns are linked by a simple relationship called the Fisher equation, which states that the real return on a bond is roughly equivalent to the nominal interest rate minus the expected inflation rate.1 For example, if an investor purchases a Treasury security with a 6 percent nominal interest rate, and inflation is expected to be zero during the investment period, the real expected return would be 6 percent. In the real world, however, inflation is usually positive, so in most cases the real rate of return will be less than the nominal return.
Because investors understand this relationship between inflation and real returns and want, therefore, to be compensated for any decline in purchasing power, nominal interest rates tend to rise when investors expect the inflation rate to worsen, and vice versa. But what happens if actual inflation is higher than expected inflation? An investor purchasing a conventional bond at 7 percent expects a real return of 5 percent if inflation is expected to be 2 percent during the investment period. If actual inflation turns out to be 4 percent, however, the bond's real return drops to 3 percent. If the actual inflation rate is high enough, the real return can even turn negative, causing the investor to pay the borrower for the privilege of using his money, rather than the other way around.
An inflation risk premium is built into nominal bond yields to compensate investors for the risk that inflation will be higher than expected. Of course, inflation risk can work the other way: If actual inflation is less than expected inflation, the investor gains while the issuer loses. Because investors are thought to be risk averse—they dislike surprise losses more than they like surprise gains of equal magnitude—the inflation risk premium in nominal interest rates is positive.
Indexed bonds eliminate inflation uncertainty. A holder of an indexed bond is assured that the real cash flow of the bond (principal plus interest) will not be affected by inflation. On the surface, at least, indexing appears to be a win-win proposition. Investors gain because their capital is protected, while issuers gain because they do not have to pay the inflation risk premium. Moreover, this joint gain increases with the term of the bond, for two reasons. First, there is a higher risk that expected inflation will differ from actual inflation the further out into the future you go, and second, any inflation forecast error is magnified with longer-term bonds because of interest compounding.
Although the Treasury's initial auction of inflation-indexed bonds is to be for 10-year notes starting at $1,000 denominations, by early 1998, the Department plans to offer indexed securities of other maturities, as well as indexed savings bonds.2 The 10-year indexed notes will be auctioned quarterly in a uniform price auction similar to that used for other marketable Treasury securities.
The Treasury's indexed bond is structured like the Canadian real return bond. The Department will calculate semi-annual interest payments by adjusting the principal for inflation and applying the auction determined, fixed real interest rate to the adjusted principal. The inflation adjustment will be based on the Consumer Price Index for all Urban Consumers (CPI-U), a widely used, though flawed, measure of U.S. inflation.3 To ensure that investors will not come up short from any deflation occurring during the investment period, the Treasury has promised that the final principal payment will be at least equal to the original paramount of the security at issuance.4
Investors, issuers and policymakers—especially monetary policymakers—all stand to gain from indexed bonds. For investors, the major benefit is the guarantee of a real yield.5 In the past, government bond investors have been burned when inflation exceeded nominal interest rates, resulting in negative real returns. Although the inflation rate has been relatively low for the past several years—hovering around 3 percent—there is always a chance that poor economic policy or an external shock could drive it higher. Consequently, the Treasury Department is promoting the securities to conservative investors who can ill-afford capital losses, like those saving for impending retirement or college costs and those living on fixed incomes.
The potential benefits for the Treasury are many. Indexed bonds could substantially reduce inflation risk and—depending on their share of total government debt—stabilize the Treasury's real funding costs. While unexpectedly high inflation benefits the Treasury by lowering the real return it has to pay investors, unexpectedly low inflation increases the government's funding costs. For example, the Treasury is currently paying a 15.75 percent coupon on a 20-year bond it issued in 1981 when CPI-U inflation was 10.4 percent; the real cost to the Treasury for that bond at issuance was 5.35 percent. Today, however, with inflation averaging about 3 percent, the real cost of the bond is close to 13 percent. Many analysts believe the Treasury will pay real rates of 3 percent to 4 percent on long-term indexed bonds.
The more certain benefit for the Treasury is the money it will save by eliminating the inflation risk premium on some portion of its debt. Although the size of the premium is debatable, most economists estimate it to be at least 50 basis points for short-term bonds and even more for longer-term bonds.6 Because the Treasury borrows about $200 billion a year, the potential savings could be substantial, even if just a small portion of new debt were indexed.
One of the subtler, yet equally important, benefits brought about by indexing government debt is the information the process would provide policymakers about inflation expectations and real interest rates. For their part, monetary policymakers could as certain a market-based estimate of inflation expectations by observing the difference in interest rates on conventional and indexed bonds of the same maturities. They could then use these estimates to assess how well the central bank is doing its job.
Despite these wide-reaching benefits, the success of indexed bonds is by no means certain. Demand for them will be dampened by several factors, not the least of which is the tax treatment they are subject to. The tax consequences are twofold. First, because the current U.S. tax code does not distinguish between increases in real income and increases in nominal income due to inflation, the indexed bond holder's tax liabilities will increase, lowering the after-tax real yields on these securities. Second, the Treasury has determined that investors will pay taxes on the inflation-adjusted increase in principal accrued each year (as well as the interest received), even though it is not paid out to maturity. A surge in inflation, therefore, could result in a tax liability that would swamp the current cash income from the bond. Because of this unfavorable tax treatment, many analysts think the demand for these securities will be limited to tax-deferred financial assets, like IRAs and 401(k) plans.
Another factor working against the success of indexed bonds is that, even after adjusting for inflation and risk, they still will be outperformed by stocks and many corporate bonds, especially over the long term. That's why they're likely to make up only a small portion of most investors' portfolios. But even if inflation-indexed bonds fail to dazzle the securities world, they're still likely to quench the thirst of conservative investors—without leaving them on the rocks.
Berry, John M., and Eric Pianin. "Hill Panel Says Inflation Overstated," Washington Post (December 5, 1996).
Campbell, John Y., and Robert J. Shiller. "A Scorecard for Indexed Government Debt," National Bureau of Economic Research Working Paper No. 5587 (April 1996).
Federal Register, U.S. Department of the Treasury. 31 CFR Part 356. "Sale and Issue of Marketable Book-Entry Treasury Bills, Notes, and Bonds" (Department of the Treasury Circular, Public Debt Series No. 1-93); Proposed Rule (September 27, 1996).
Hetzel, Robert L. "Indexed Bonds as an Aid to Monetary Policy," Economic Review, Federal Reserve Bank of Richmond (January/February 1992), pp. 13-23.
Shen, Pu. "Benefits and Limitations of Inflation Indexed Treasury Bonds," Economic Review, Federal Reserve Bank of Kansas City (Third Quarter 1995), pp. 41-56.