Jess B. Yawitz

Measuring the Effect of Callability on Bond Yields

Most CORPORATE AND MANY government bond issues contain call provisions. Since both buyer and seller are concerned with the effect that the terms of call provisions ought to have on yields, considerable effort has been devoted to developing models of the relationship between yields and callability. Also of interest are the benefits from optimal designl and exercise of the call. In addition, researchers concerned with the impact on yields of credit risk, issue size, and other bond characteristics have had to make some provision in empirical models for callability. Generally, empirical models of the effect of callability have not controlled for the other factors determining yields. Neither have studies of other yield determinants attempted to construct sophisticated measures to represent callability. In this paper we develop a theoretically sound model of the effect on yield spreads of callability and test that specification within the context of a comprehensive empirical model of yield spreads first estimated by Cook and Hendershott [3] (henceforth, C-H). C-H provide for the effects of taxes, default risk, relative security supply, and, through a simple proxy, for callability. Substituting our more sophisticated measure of the effect of call, we are able to test our model against the

Risk and Return in the Government Bond Market

has been devoted to testing the so-called capital asset pricing model (CAPM) of asset returns. While the various attempts to validate this model have addressed a wide range of questions, we are not aware of a single test that employs the CAPM to measure the return performance in the U.S. government bond market. This is the purpose of our paper. Our findings indicate that the risk-return performance of the government bond market from 1953 through 1972 is not unlike that of the equity market. Specifically, we conclude that 1) while risk has a generally consistent effect on the structure of returns, there appears to be a systematic tendency for longer maturity bonds to offer lower yields than would be expected given their exposure to interest rate risk; 2) purchase yield the so-called yield to maturity at time of pu:rchase serves as a better estimate of expected return than does actual return; and 3) standard deviation in return (or mean absolute change) does at least as well as the so-called beta coefficient in explaining the structure of bond returns. Several unique characteristics of the govemment bond market make it particularly well-suited as a vehicle for testing the CAPM: 1. The C A M is specified in terms of a securitys expected, or ex ante, return. Since there is no readily available measure of ex ante return, researchers have typically taken historical return performance to be the appropriate proxy. For the government bond market, we have two candidates for expected returns actual return or purchase yield. Actual return is obviously an ex post measure, since price change may be anticipated but cannot be known until the conclusion of the holding period. On the

Evaluating the Decision to Issue Original Issue Discount Bonds: Term Structure and Tax Effects

* There is considerable evidence that during periods of high interest rates participants in the financial markets rely more heavily on innovative financing methods. In an attempt to deal with the high interest rates of recent years, corporate borrowers have devised a series of special characteristics for corporate bond issues including put options, detachable warrants, and variable coupon rates.2 Certainly one of the more important innovations employed by corporate borrowers has been the original issue discount bond (OID).3 The OID has recently been the subject of a considerable amount of academic research. Most notably, studies by Kalotay [61 and Pyle [8] have documented many

Lower Bounds on Portfolio Performance: An Extension of the Immunization Strategy

The concept of immunization is currently one of the most widely discussed innovations in bond portfolio management. The interest in immunization, which allows one to nearly guarantee a minimum return performance over a specified investment horizon, is due in large measure to the desire on the part of bond managers to reduce the risk which has accompanied the recent swings in interest rates.

The term structure and callable bond yield spreads

T he inclusion of call provisions on various types of bonds is widespread. Virtually all corporate bonds and many government bonds contain call features. Since these features are commonplace, both buyers and sellers of bonds must be concerned with the effect of callability on prices and yields. In addition, issuers of callable bonds must be concerned with the optimal design of the call feature, such as call price and length of call protection, and with whether it is optimal to exercise the call option at a particular moment or to delay the decision. These two issues, discussed elsewhere,’ ‘are not the subject of this paper. Here we consider only those factors that determine the value of the call provision and, hence, the yield spread between otherwise identical callable and non-callable securities. The conventional wisdom regarding the value of a call option is summarized well by Van Horne (1978). It is generally believed that the level of interest rates exerts a positive effect on the compensation in yield required for investors to be indifferent between otherwise identical callable and non-callable bonds. In particular, the higher the general level of interest rates, the higher (lower) is the price (yield) on a non-callable bond relative to an identical callable bond. The intuition underlying this approach is based upon the assumed relationship between the level of interest rates and investors’ expectations about future interest rates. When rates are historically high, so the argument goes, there is a greater probability that rates will fall in the future, at which time the issuer of the callable bond will find it beneficial to exercise the op-