|Researchers:||Ilya Dergunov, Christoph Meinerding, Christian Schlag|
Most of the theoretical literature on equilibrium asset pricing deals with real assets like equity. Thus, by definition, inflation does not play a role in most models. Recent empirical evidence, however, suggests that inflation risk is a priced risk factor in the cross-section and in the time series of nominal as well as real stock and bond returns. As of now, the role of inflation in real asset pricing models is still largely unexplored. First of all, there is no consensus about whether (and how) inflation should be built into the real pricing kernel. Second, inflation is hard to measure precisely and the influence of its dynamics on asset prices is thus largely unexplored. Third, empirical evidence suggests that inflation can move both pro- and countercyclically, i.e. the correlation between inflation and variables related to the (real) output in the economy is time-varying and may even change signs over time. Fourth, during the recent crisis, the topic of deflation risk has been put on the agenda again. It is therefore worthwhile to analyze both the fear of inflation and the fear of deflation in a joint real asset pricing model.
This research project is supposed to explore the role of inflation in equilibrium asset pricing models in more detail. Firstly, if inflation is supposed to affect prices and price dynamics of stocks and bonds, it has to enter the pricing kernel. A rational alternative to the behaviorally motivated concept of “money illusion” could be the use of recursive utility. Inflation will then influence both nominal and real returns in the economy. Secondly, building on a recent paper by David and Veronesi (2013), we will assume that inflation has a time-varying signaling role in the economy. We will include estimates of current inflation as a state variable in an equilibrium model. Since inflation can be both procyclical and countercyclical, learning can introduce time-varying correlation between inflation and other macroeconomic fundamentals like consumption. Through this channel, learning may thus contribute to time variation (and even variation in the sign) of inflation risk premia which can be extracted from nominal and real term structures of interest rates. The model may even be able to generate a deflation risk premium at times. Finally, the goal is to verify the theoretical predictions of the model empirically.