The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
For release on delivery 9 30 A M , E D T June 16, 19 92 Statement by Alan Greenspan Chairman, Board of Governors of the Federal Reserve System Before the Commerce, Consumer, and Monetary Affairs Subcommittee of the Committee on Government Operations U S House of Representatives June 16, 1992 Mr. Chairman and members of the Subcommittee, I welcome this opportunity to discuss the potential issuance of indexed bonds by the Treasury This assignment touches upon a wide array of challenging analytical and policy issues, such as the appropriate tax treatment of these obligations, the technicalities of bond contracts, an assessment of investors' likely interest in these novel instruments, and the consequences for the conduct of monetary policy While I may not do justice to the range of demanding questions confronting the Treasury in its deliberations on whether to issue indexed debt, I do intend to convey the Board of Governors' current assessment of these considerations A Proposal on Indexed Debt Enthusiasm for indexation--whether of wages, entitlements, the tax schedule, or government debt--quite often may be expected to echo a government's failure to control increasing rates of inflation Indeed, some have voiced concern that by making it easier for investors to live with inflation rather than treating it as a fundamental problem, issuing indexed debt, on occasion, could appear to mark official acceptance of continuing high inflation is not the situation today The U S This economy has made considerable progress toward price stability over the past decade, trimming the core rate of inflation to below 4 percent, and it appears poised to make further advances -2Instead, we are here today to evaluate a proposal contending that continued progress in economic stabilization could be made somewhat easier That proposal is to use Treasury debt management to extract market readings on inflation expectations and real interest rates and then to use those readings to aid the conduct of monetary policy Essentially, the Treasury is being advised to split a segment of its debt issuance into two parts One part would be indexed to consumer prices and one would not. The yields on bonds that protect purchasing power could be considered measures of "real" interest rates Importantly, the gap between the yields on two issues of comparable maturity but differing protection against inflation could be viewed as a market-based assessment of inflation expectations and the risk premium associated with inflation instability I commend the Subcommittee's efforts to broaden the range of indicators examined in analyzing economic events and setting policy For my own part, I am attracted by the prospect of opening a window on the market's view of the path for inflation that potentially could provide readings of price pressures being built into wages and of real interest rates influencing spending decisions. The market provxdes many signals about the future in its current pricing of assets, and an increased menu of indicators, in principle, may offer a wider panorama on what is to come -3In a similar vein, it is helpful at times, for analytic purposes, to disentangle the movements of the Treasury yield curve into the path expected by market participants for future one-year interest rates However, those forward-rate measures are imperfect, as risk premiums built into financial returns confound attempts to take literal readings on the expected future Some of the same problems may confront analyses using indexed debt to gauge inflation expectations Moreover, changing the composition of federal debt issuance is not a matter to be taken lightly With the vast scale of Treasury indebtedness, interest expense now absorbs almost as large a share of our limited tax resources as does discretionary domestic spending Any proposal that has an impact on Treasury financing costs must clearly demonstrate that benefits exceed costs by a comfortable margin The Signals from Indexed Debt A series of hurdles must be overcome before issuance of indexed debt moves from a promising alternative to a useful policy instrument First and foremost, rigorous study is required to understand exactly what to read into the simple difference in yields between nominal and real debt. The yield on a nominal Treasury debt instrument comprises three elements a real interest rate, an inflation premium that attempts to adjust for expected changes in purchasing power over time, and a risk premium -4This last component incorporates premiums for a variety of risks assumed by the investor For a U S Treasury security denominated in dollars, default risk is negligible However, because inflation is unpredictable, there is a chance that indexed and unindexed debt will provide different payments over time to investors As a result, the market will value them differently, even in real terms The uncertainty regarding the real return provided by the unindexed debt drives a wedge between the yields on indexed and unindexed bonds in the form of different risk premiums, which may vary unpredictably over time Thus, the differential in yields likely will not serve as a pure measure of inflation expectations Still, since risk premiums with rare exceptions are positive, the differential is almost always at least as large as inflation expectations That is, the market would tend to delineate an upper bound on its prospects for inflation Second, in implementing some measure of protection for inflation to investors, the Treasury must select a single price index as the basis for that compensation and be confident that there will be no significant revisions to the referenced price index Most likely, measurement issues are not much more difficult in this regard than in the construction of cost-of- living adjustments for wages and benefits, and the not-seasonally-adjusted consumer price index will probably fit the bill However, all price -5indexes are imperfect owing to distortions and limits to their coverage To the extent that the index used by the Treasury did not adequately capture potential investors' cost of living, the estimate of the real interest rate would be comparably affected Third, the experts have to give careful consideration to tax treatment Before-tax nominal returns on coupon-bearing indexed and unindexed instruments well may have to differ to pay the same after-tax compensation to investors Since indexed debt provides protection of principal, the Internal Revenue Service likely would require investors to impute any increase in the nominal value of the principal as part of current income, as is the precedent with zero-coupon securities Thus, an investor in indexed debt may be called upon to report income not yet paid in cash In this regard, some have suggested that the Treasury issue zero-coupon securities, both nominal and real, to prevent indexed debt from being disadvantaged and make comparisons of yield differentials transparent Even if cash-flow considerations favor unindexed debt, rough estimates of the tax effects on the difference between real and nominal yields are calculable for the average investor, and hence approximate adjustments can be made However, of greater importance, those adverse cash- flow implications of zero-coupon securities now or indexed debt in the future likely renders these instruments less -6attractive to some classes of investors If holders of indexed debt are drawn from a narrow segment of the investing populace, then the real rates and implied inflation expectations derived from those instruments may not reveal economy-wide sentiments Under those circumstances, the Treasury may have to offer an elevated real return to place its indexed debt issue relative to that expected from its nominal debt, which is purchased by more investors A sufficiently elevated real rate may offset any gain to the Treasury by not having to pay investors some compensation in the form of a likely positive risk premium for inflation expectations on nominal debt Thus, at a basic level, expected financing costs to the Treasury and the value of the signal on real interest rates to the Federal Reserve importantly depend on ' investors' attraction to an untested instrument Before the fact, it is reasonable to assume that a family establishing a child's college fund or a couple planning for retirement well may pay handsomely for inflation protection After all, their anticipated future payments will certainly be influenced by movements in the general price index, and indexed debt represents an asset that at least keeps pace with the price index However, by the historical record, many of these long-planned expenses, such as tuition, do not move in lockstep with general price indexes These anticipated relative price shifts make bonds -7tied to a general price index less useful for hedging purposes Also, the imputation of taxes to the nominal increment to the value of the principal may make some investors wary of indexed debt Still, tax-favored investors probably would shift some of their investments toward indexed debt Many pension fund managers, for instance, accumulate assets to meet long-term payments tied to wages or prices Indexed Treasury securities could permit them to match their deferred liabilities of predictable real but uncertain nominal value more accurately. Of course, not all investors need, or would be willing to pay, for purchasing power protection For some hedging purposes, nominal liabilities must be matched with nominal assets Moreover, the Treasury now offers investment possibilities that provide a rough measure of compensation for inflation The simple and expedient technique of rolling over six-month Treasury bills every six months provides a stream of returns that has moved fairly closely with inflation Putting aside policy considerations, the private sector may receive direct benefits from the public example of indexed issuance The yields on Treasury securities serve as benchmarks for private rates around the world With direct quotes on indexed debt available in the broad and liquid market for government securities, private issuers may join in, issuing their own index-linked debt tailored to -8their specific needs and broadening the choice of assets available to investors However, the private sector seldom has waited for the government to lead the way in financial innovation The lack of private-sector precedent for indexed debt, as well as the short-lived experiment in trading consumer price index futures on an organized exchange, suggests that the prospects for the success of an indexed issue must be weighed carefully The thinness in that segment of the private market may simply indicate the need for the public sector to lead by example, but it instead may raise questions about investor demand and potential cost savings If, after weighing these costs and benefits, the Treasury adopts an indexed-link debt program, it will have to steer a difficult course in determining the scale of operations Splitting federal issuance in equal parts, in my view, trusts too much to the uncertain demand for these instruments. The large stakes involved given the government's need for funds surely dictates that an experiment with indexed debt must be modest in size At the same time, issuance must be large enough to attract the trading interest that would ensure an active secondary market for indexed debt. Any novel instrument initially would be less liquid and ultimately may lead to some fragmentation of trading in government securities, perhaps raising overall funding costs The prices of indexed debt -9trading in a thin market would not necessarily convey a significant amount of useful information about the economy as a who1e Foreign Experience The foreign experience divides between those developing countries that were driven by necessity to issue index debt as a means of attracting investors made wary by high inflation and a small number of developed countries that sought to save on the financing costs of the government It is difficult to find obvious lessons from that latter and more relevant group In the post-war period, the governments of several developed countries have issued debt securities offering claims that were in some way linked to a price index Two industrial countries, the United Kingdom and Canada, issue bonds for which the principal and coupon amounts are tied to a consumer price index, although the Canadian program is under one year old with only one issue on the books In 1988, Australia suspended an ambitious index-debt program begun in 1985 British index-linked gilts (the equivalent of our Treasury debt obligations) were first issued in March 1981 with a maturity of fifteen years While the ownership of index-linked gilts was initially limited to pension funds, now all investors can hold those securities Indexed-linked debt has grown more rapidly than total issuance, pushing its relative share to about one-fifth of government debt, and -10now trades in a relatively deep and liquid market Such trading provides timely quotes on real interest rates, although tax treatment and an eight-month lag in inflation compensation complicate their interpretation The Australian government indexed some of its bonds to the consumer price index between 1985 and 1988 with the stated aim that diversification might reduce interest costs Those efforts, however, were set back by weak demand that resulted in elevated real yields Trading volume in the secondary market was thin and, after several successive budget surpluses reduced the need to tap the market, the government suspended its issue of indexed-linked debt Implications for Monetary Policy Without doubt, the substantial uncertainty facing monetary policymakmg would be reduced somewhat if the market were to provide a reliable measure of current inflation expectations Indeed, the paired issuance of unindexed and indexed debt at various maturities might make it possible to offer some information on the market's expectations for the path of inflation well into the future A timely and accurate reading on inflation expectations could considerably aid in economic forecasting by casting some light on incipient wage and cost pressures and by helping to divide changes in nominal asset values into their expected real and price components -11Also, by routinely monitoring the markets for the two debt instruments, the Federal Reserve could extract the market's evaluation of the consequences of policy operations On occasion, the market's response to a policy action is difficult to interpret. A reading on real rates may make it easier to parse out the reaction of long-term nominal yields, for example, into the effects on real rates, inflation expectations, and risk But our concerns are not narrowly focused on price developments and short-term operations Indexed debt would offer other, potentially useful, information about the economy Nearly all descriptions of the economy assign important roles to real interest rates in influencing spending and investing decisions made by households and businesses As a result, economists, including policymakers at the Federal Reserve, must assess the level of real interest rates when attempting to explain or to project economy-wide developments A market for indexed debt would facilitate this process by continually updating our knowledge of investors' assessment of real interest rates and by perhaps signalling future changes in income and economic activity While those readings on real interest rates would help, monetary policy would remain a difficult job, as they would not reveal the appropriate level of real rates consistent with sustainable economic growth -12I share the view of most economists that there is no better mechanism for refining opinion and focusing attention on economic fundamentals than a competitive market. Thus, I am sympathetic to the notion that policymakers should heed the messages from markets But we must remember that there are problems associated with the issuance of indexed debt The simple difference in unmdexed and indexed returns well may convey more than a reading on inflation sentiment. At the least, economic theory suggests that a time-varying risk premium enters the picture Technical considerations may bulk large as well However, even an imperfect reading on expectations could help us to understand some aspects of the behavior of the private sector, though it in no way could supplant our other efforts to forecast inflation At times, market participants are wrong, perhaps by stubbornly holding to outmoded lessons of the past or by swinging too wildly with the latest scrap of news. Even if indexed bonds were issued, the Federal Reserve by necessity would continue to rely upon a broad array of indicators and a considerable element of judgment in determining the stance of policy Nonetheless, I am confident that we would make use of new market-based indicators of inflation and real interest rates that would be made available by the issue of indexed bonds Such measures may not mark the way as -13unambiguously as promised by their most vocal adherents, but they would help Conclusion For our part, the Federal Reserve retains responsibility for long-run price stability and fully intends to guard against reigniting inflation That commitment might be easier for us to effect or the public to monitor should the Treasury issue indexed debt Still, the benefits to monetary policy are not so obviously large as to outweigh any additional costs to the taxpayers in financing Treasury debt Thus, the decision to issue debt that provides a measure of inflation protection should remain in the domain of fiscal policy and be based primarily on the consequences for total borrowing cost