A Data Science Central Community
Geert Bekaert (Columbia Business School) &
Marie Hoerova and Martin Scheicher (European Central Bank)
EUROPEAN CENTRAL BANK WORKING PAPER SERIES
EXECUTIVE SUMMARY (ABSTRACT)
In this paper, we develop a measure of time-varying risk aversion that is relatively easy to estimate or compute, so that it can be compared to the practitioners’ indices. However, the model we use is inspired by the dynamic asset pricing literature. We view risk aversion and economic uncertainty as two main drivers of asset pricing dynamics and model them as latent variables. We achieve identification by using many asset prices (as is often the case in the practitioners’ literature) and economically inspired restrictions on the dynamics of these variables.
In particular, we lean heavily on the idea that the implied volatility indices (like the VIX) should have information about risk parameters, once they are cleansed of the influence of normal volatility dynamics and uncertainty. This idea is prevalent in the reduced form stock return dynamics literature (see, e.g., Duan and Yeh, 2007). Moreover, dynamic asset pricing theory suggests that risk premiums and asset prices likely depend on both economic uncertainty and risk aversion. To measure risk aversion, macroeconomic uncertainty has to be controlled for.