Conventional securities are generally offered at a fixed coupon rate that incorporates the underlying expected real rate of return in the economy, the market’s expectation at the time the security is issued of inflation over the duration of the instrument, a premium to compensate for the fact that future rates of inflation are uncertain, and an adjustment reflecting the tax treatment of interest on behalf of both the lender and the borrower. For simplicity, it is useful to abstract temporarily from the inflation risk premium and taxes, although both these factors will be discussed later.

With these simplifying assumptions, if the real rate of return in the economy is 3 per cent, and inflation is expected to remain constant at 4 per cent annually, the nominal return will be 7 per cent. If expectations should prove incorrect and inflation turns out to be lower than anticipated, say 2 per cent, investors will receive more income in present value terms than they expected and experience an increase in their real rate of return, reflecting the unanticipated decline in the inflation rate. On the other hand, if inflation turns out to be 6 per cent, then investors will receive less in real terms than expected and their real return will fall below the rate initially negotiated. If they attempt to sell the security in the higher inflationary environment, they will experience a capital loss.

Index bonds are financial instruments designed to protect investors fully against the erosion of principal and interest due to inflation. This protection is accomplished in one of two ways. Under the first option, the bond is issued at a specified real coupon rate and both coupon payment and repayment of principal are scaled up or down by the change in prices that occurs between the time that the money is borrowed and the time the payments are made. For example, if inflation is 4 per cent annually, the coupon on a five-year $1000 bond issued at a real interest rate of 3 per cent would increase from $30 in the first year to $35.10 in the fifth year. At maturity, the government then would adjust the principal for inflation over the life of the bond; thus, in the above example, the government would repay $1217 at the end of the five-year period. This approach is similar to the index bonds that have been sold in Great Britain.

Under the second approach the entire inflation adjustment is made through the coupon payment and the bondholder is repaid his original principal at maturity. For example, if the real rate is set at 3 per cent and inflation averages 4 per cent, the total annual interest cost would be 7 per cent. This approach mimics the current method of compensating the lender for inflation, except that instead of trying to predict inflation at the time of the loan and incorporating this expectation into the stated nominal interest rate, actual observations on price are used to determine annual interest payments.

Either of these two approaches will protect the investor against the risks associated with unanticipated price changes if the index bond is held to maturity; however, it is important to emphasize that neither produces a risk-free investment. As with any long-term security, bondholders selling an index bond before maturity would take a capital loss if the underlying expected real rates have increased since the date of purchase. The result is that these bonds would probably not be the ideal assets for individuals to purchase directly unless they were certain that they could hold them to maturity. For index bonds to serve as risk-free inflationprotected investments, financial intermediaries are required which will hold the bonds to maturity and offer repackaged investments free of the real-return risk.

Impact on the Government Budget

Arguments about the potential impact of index bonds on the government budget have figured prominently in debates about this form of financing and the range of opinion has been extraordinary. In Great Britain, some opponents argued that index bonds would cost the government more than fixed-interest securities to service, since they would have to be issued at positive real interest rates as opposed to the negative real returns received by investors on nominal debt during the period 1973–8 (Rutherford 1983). On the other hand, in hearings before the Joint Economic Committee in May 1985, a major proponent of index bonds projected that, because excessive inflation premiums were incorporated in current yields, the US government could save $9 billion in the first year and $135 billion over a five-year period by issuing indexed rather than conventional long-term debt (Joint Economic Committee 1985).

These conflicting statements are based on opposite assumptions about people’s ability to project future inflation. The contention that index bonds will cost money assumes that individuals will continually underestimate future inflation and always end up with lower than anticipated or negative returns; the argument that the Treasury can reduce costs with indexed debt assumes individuals will consistently overestimate inflation and demand excessive inflation premiums. It is unclear why, over the long run, individuals should systematically err on one side or another in their inflation projections.

In the last 15 years, the relationship between expected and actual inflation has varied over time; during the 1970s average expectations about near-term inflation tended to prove too low, while since 1981 inflation has generally fallen one or two percentage points below projections. Although no evidence is available on investors’ ability to forecast inflation over longer periods, of say 20 or 30 years, the same pattern is likely to emerge as swings in short-run expectations affect the longer-run outlook. Hence, the most reasonable conclusion is that in the long run forecasting errors will cancel out, and have little impact on the relative costs of indexed versus unindexed debt.

On the other hand, the uncertainty surrounding future rates of inflation means that investors demand an inflation-risk premium before they are willing to take on fixed-coupon debt. In this case, the guarantee of a real return provided by indexed securities, which eliminates the risk of reduced real returns and capital losses caused by unanticipated inflation, would lower the yield that lenders will require in order to provide their funds. In other words, the lender would be willing to accept a somewhat lower rate in return for the privilege of having the government guarantee the real return on the loan.

Little evidence exists about the size of the inflation risk premium (an exception is Bodie et al. 1986). As long as the outlook for price increases is moderate, the premium is probably relatively small; at higher and more volatile rates of inflation, the importance of risk protection would increase. Even if this premium proved to be quite small, however, its elimination could produce substantial savings in view of the enormous magnitude of government debt. The problem is that, in the short run at least, the risk premium effect is likely to be dominated by the difference between expected and actual real returns caused by errors in investors’ expectations. Hence, for any defined period of time, it would be impossible to predict whether substituting index bonds for traditional government securities would cost or save the Treasury money. In the long run, however, if errors in inflation forecasts cancel out, index bonds should save the government the inflation-risk premium on long-term securities.

While the net interest saving to the government is difficult to predict, the pattern of government borrowing would certainly be altered if the British indexing option were adopted. Even in an environment where inflationary expectations always prove correct and the inflation premium is zero, an index bond that defers the principal adjustment for inflation until maturity reduces the Treasury’s borrowing in the intervening years.

Tax Policy and Index Bonds

Uncertainty about how index bonds would be taxed has been viewed as a major impediment to their introduction. The tax questions are indeed critical, because they determine not only how index bonds would affect revenues but also who might be the likely buyers of these securities and the potential yields.

If the tax code does not distinguish between real and inflationary returns, then the most likely results would be to tax both the real component of interest and the inflation adjustment as ordinary income. This would be quite straightforward in the case of the indexing method that incorporates the inflation adjustment in the interest rate, but some complexities arise in the case of the British approach. In order to make the treatment of bonds indexed in this fashion analogous to that accorded conventional and zero-coupon bonds, the annual appreciation of principal due to inflation would have to be taxed as it accrued.

Taxing the principal adjustment as if it were received each year would make index bonds less attractive than their unindexed counterparts. Not only would owners of securities have to pay taxes on illusory gains, which they do in the case of conventional bonds, but they would also have to pay the tax before they received their inflation compensation. On the other hand, deferring the tax on the adjustment of principal until the bond is redeemed at maturity would favour the indexed over the unindexed security and result in a loss of revenue for the Treasury.

The second problem with applying current tax law to index bonds is that it would no longer be possible to guarantee a constant real after-tax rate of return. Under the current system, taxes would rise with inflation and the real after-tax return would decline. For example, if the tax rate were 30 per cent, the real return on a bond with a 3 per cent coupon would be 2.1 per cent in an environment of no inflation. If inflation should rise to 4 per cent and the nominal coupon rises to only 7 per cent, the after-tax yield is 4.9 per cent or 0.9 per cent real. The only way to avoid this problem is to exempt the nominal adjustments for inflation from taxation. This approach, however, would introduce a type of inflation indexing not found elsewhere in the tax system.

Private Issues of Index Bonds

Some sceptics charge that if index bonds were such a great idea, they would have been offered by the private sector. Indeed, theoretical work by Stanley Fischer leads to the conclusion that firms should be equally willing to issue index bonds as conventional nominal bonds (Fischer 1982). Fischer offers two possible reasons for the lack of the private sector innovation: the relatively stable rates of inflation traditionally experienced in the United States and the possibility that borrowers’ expectations about inflation have been systematically higher than those of lenders. Others contend that the issuance of index-linked debt may actually have been illegal in the United States until 1977 (McCulloch 1980). Another problem is that an aggregate price index may not correlate with prices received by an individual firm. The most persuasive reason, however, relates to the lack of indexation in the corporate income tax, which causes the effective tax rate to increase with inflation. If firms were to issue index bonds, this inverse relation between inflation and profitability would worsen, since corporations would forfeit the mitigating effect of the decline in the value of outstanding liabilities as inflation increased. Hence, the non-issuance of index bonds by the corporate sector may be one of the major casualties of an unindexed tax structure.

The only serious objection ever levelled against index bonds is that protecting bondholders from inflation might reduce public pressure to maintain price stability. If part of the pain of inflation is removed, this reasoning goes, the public’s resolve to control inflation will weaken, and inflation will ultimately get worse. On the other hand, one could argue in economic terms that index bonds might help in the fight against inflation by providing an attractive investment vehicle that would encourage saving and, as argued by Tobin, by offering the monetary authorities a tool that would strengthen their control of the economy (Tobin 1971). In political terms, it would seem that the issuance of index bonds would eliminate one of the main incentives for the government to inflate the economy. With indexed debt the government can no longer reduce the real value of its outstanding liabilities by allowing prices to rise; instead, inflation will produce an immediate increase in required expenditures. Finally, index bonds do not appear to have encouraged inflation in Great Britain; the inflation rate has declined from 15 to 5 per cent since 1981, the year the bonds were introduced.

See Also