Empirical Risk Decomposition of Break-even Inflation

Abstract: Using principal component analysis, we find that there are three principal components that can explain almost all of the variation in the 5-20 year maturities of break-even inflation, defined as the spread between nominal Treasury and TIPS yields. The three principal components correspond to the factors regularly used in the empirical literature to describe the term structure of interest rates, namely, level, slope and curvature. The first principal component, explaining over 90 percent of the variation in the data, is very similar to 10-year break-even inflation. We regress both of these series on variables aimed at embodying different sources of risk relevant to break-even inflation, and find that short-term inflation expectations and differences in liquidity have the highest explanatory power. In accordance with our theoretical discussion of risk premia, we also include measures for the inflation risk premium and volatility risk. Although these measures do not add much to the analysis of the first principal component, they contribute significantly to the explanation of the second principal component. We interpret this as an indication that the inflation risk premium is an important driver of the slope of break-even inflation. We also find that changes in liquidity conditions impacts the slope of the term structure, but propose that this merely reflects a change in short-term inflation expectations.