Comment on GSE Debt and the Decline in the Treasury Debt Market

Abstract

Listen

Translate article

Comment on GSE Debt and the Decline in the Treasury Debt Market Toni Gravelle (bio) Let me start by providing some general comments that will prove useful later on in my discussion. As the authors note, reductions in the size of the U.S. Treasury debt stock and of other sovereign debt stocks is an important issue because government debt is central to the implementation of monetary policy and to the allocative efficiency (in the welfare sense) of fixed income markets more broadly. What do people mean when they attribute benchmark status to a fixed income security? This, of course, depends on who you ask, since benchmark status has several context-sensitive dimensions. First, U.S. Treasury securities and, more generally, government securities in most developed economies are perceived as being free of default risk and thus allow the yields on these government securities to proxy for nominal risk-free interest rates. Second, the large borrowing needs of governments and their essentially risk-free status allows them to offer a broader range of maturities than is the case for most any other class of debt issuer. Third, the fungibility of various bond issues and the typically large size of each issue makes the government securities market the most liquid segment of the fixed income market. This factor supports the existence of well-developed repo and derivative markets for government securities that in turn serve to enhance the liquidity of the government securities market. These affiliated markets enable market participants to take on short and long positions that reflect their views of future interest rate movements. Fourth, government securities also serve as reference securities for pricing other debt instruments and as hedging vehicles for other debt securities.1 A less-liquid government securities market has a negative impact on its status as a benchmark market. Moreover, there is a self-fulfilling aspect to liquidity, which could quickly cause the benchmark status of U.S. Treasury securities to degenerate. As such, to the extent that a reduction in the outstanding stock of government securities has a negative impact on this market's benchmark status, then policymakers are rightly concerned about the declining allocative efficiency of fixed income markets [End Page 840] that may result. In the end, if capital markets become less efficient, due to the lack of a benchmark debt security, economic growth and investor welfare is likely to be negatively affected. The majority of my discussion will focus directly on the analysis contained in the Ambrose and King paper. However, I will also provide brief comments on issues that I suspect lie outside the scope of this paper. Paper Summary Ambrose and King carry out an empirical study whose main goal is to examine whether or not government-sponsored enterprise securities—also known as agency securities—are increasingly being used as a substitute for U.S. Treasury securities by investors seeking benchmark debt securities. They do so by examining the relationship between the outstanding stock of U.S. Treasury (TSY) and government-sponsored enterprise (GSE) securities, and the (market) liquidity of these GSE bonds. In analyzing GSE bond yield characteristics, they use monthly observations of duration-matched yield spreads under the assumption that, after controlling for other factors, declines in the yield spread proxy for a smaller liquidity premium (and thus improved liquidity) for GSE securities. Ambrose and King's analysis can be broken into three parts centered on three specific regressions. First, Ambrose and King regress the change in these spreads on a set of variables that includes the change in the stock of TSY and GSE securities. The authors control for other factors that might explain yield spreads and include variables that control for changes in the interest rate environment and issue specific characteristics. Ambrose and King show that GSE yield spreads are significantly positively related to the stock of TSY securities, but that there is no statistical relation to the level of GSE debt. In order to check that GSE securities are, in fact, increasingly replacing TSY securities as the investors' choice of benchmark debt instrument, they carry out the same regression analysis with duration-matched yield spreads between corporate and TSY debt securities instead. The authors...