Duration: Its Development and Use in Bond Portfolio Management

This survey article reviews the historical development of duration and its uses in (1) summarizing in one variable the cash flow characteristics of bonds, (2) approximating the price sensitivity of bonds, and (3) developing bond portfolio strategies, particularly those that attempt to immunize against interest rate risk. In all these uses, duration has been shown to be superior to term to maturity. Research and experience to date suggest that single-factor duration models (which assume that changes in interest rates for all maturities are perfectly correlated) are useful in practice. In particular, the evidence suggests that the Macaulay measure of duration performs reasonably well in comparison to its more sophisticated counterparts and, because of its simplicity, appears to be cost-effective. For portfolio managers who wish to immunize portfolios of default- and optionfree bonds against interest rate risk, or who wish to use duration in formulating active strategies, single-factor models should outperform more naive maturity models while matching the performance of more complex multifactor models. Duration is particularly useful to managers of financial institutions. As a measure of interest rate exposure, it is more accurate than maturity information. Fairly recent research has indicated that duration is related to beta, hence may prove a useful tool for equity, as well as bond, management. As duration increases toward and beyond the length of the investor's planning period, beta first decreases, troughing at zero when duration is equal to the planning period, then increases again. To the extent investors have different planning horizons, this finding casts doubt on the uniqueness of beta or any other risk measure of a security.