Long-term U.S. government bond yields have been very volatile in recent decades: yet, analysts can say surprisingly little about the sources of the yield fluctuations. Analysts know that changes in nominal yields from period to period must stem from some combination of changes in investor expectations of future inflation and changes in the real returns that investors demand.
These components of bond yields are unobservable, however, and analyst cannot say what really has caused the fluctuations in yields. In this study, we used surveys of economists' long-term forecasts to estimate the components of bond yields and tentatively identify the causes of changes in yields.
Data from surveys of economists' forecasts can be used to decompose long-term yields into three parts: expected inflation, expected real short-term interest rates, and the required bond risk premium. All three components have contributed to the decline trend in yields in the past one and a half decades.
Moreover, we can show that fluctuations in the series over time have a plausible economic rationale. We believe our approach of using long-term consensus forecasts provides analysts with better information about the bond market than does studying realized returns or assuming constant relationships in the structure of yields.
Graham and Dodd Award of Excellence 1998