Inflation Targeting and Predictive Power of Term Spreads

Abstract

Many studies have investigated the predictive contents of the term spreads for future inflation and real activity since the 1990s, which was the time when some countries adopted inflation targeting (IT). As notable as the number of studies are the diversity and heterogeneity of the empirical findings on this issue across countries and over time. In this paper, we attempt to assess whether IT is responsible for the changes in the forecasting power of the term spreads for inflation. If future inflation is anchored to the target rate under an IT regime, then changes in the spread would contain information other than the future inflation, thereby reducing the predictive contents for future inflation. We test this hypothesis by (1) comparing the forecasting power of the term spread for inflation in a group of inflation-targeting countries before and after adopting IT and (2) comparing this power with the forecasting power in a group of countries that have not adopted IT. Findings support the hypothesis overall, but results are weakened when the interest rate is included in the forecasting equation.Keywords: Inflation targeting, Term spread, Predictability, Monetary policyJEL Classification: E52, E58, G12(ProQuest: ... denotes formulae omitted.)I. Introduction Over the past two decades, many studies have analyzed whether the yield spread contains predictive contents for future real activity and inflation in many industrialized countries. The empirical evidence for the predictive power of the yield spread for real activity is relatively strong, but the predictive power for future inflation appears to have decreased significantly or disappeared over time in most countries. In this paper, we show that the adoption of inflation targeting (IT) is responsible for the time variation in the predictability of inflation using the term spread in relation to that of output growth across countries and over time.Around the 1980s, the term structure of interest rates emerged as an active major research topic in several academic fields. The modern asset pricing literature pertaining to the factors underlying the term structure was initiated by the work of Vasicek (1977) and Cox, Ingersoll, and Ross (1985). From a theoretical point of view, the forward-looking nature is inherently an essential characteristic of the bond yields, and many empirical studies have investigated the predictive power of the term structure for the future output growth and inflation since the 1990s. Another recent body of research, the macro-finance literature that started in the 2000s, primarily focuses on the relationship among the level, slope, and curvature factors in the term structure and the macro-variables, such as inflation target of the central bank, actual inflation, and real activity. Rudebusch and Wu (2008), Ho?rdahl, Tristani and Vestin (2006), and Bekaert, Cho, and Moreno (2010) found that the level factor, which is primarily driven by the short-term interest rates, accounts for much of the inflation target or the long-run inflation rate, whereas the slope, which is often proxied by the term spread, has a marginal relationship with inflation or real activity.The keys to understanding the term structure are the dynamics of the short-term nominal interest rates and the expectation formation of the market participants on future short-term interest rates, both of which are heavily influenced by monetary policy. Most industrialized countries replaced their primary monetary policy instrument, monetary aggregates, with the short-term interest rate in the 1980s. This policy change implies that the equilibrium of short-term interest rates is directly controlled by the central bank. Accordingly, the long-term bond market participants have changed the way they form expectations of future short-term interest rates. This explains that the objective, instrument, and conduct of monetary policy have been taken into account in a vast majority of research papers in the aforementioned literature. …