Parametric Models of the Market Pricing Kernel

Abstract

This chapter is concerned with the derivation of parametric models of the market pricing kernel, i.e. the specification of intertemporal asset pricing models. For this purpose the consumption-based equilibrium asset pricing approach is applied, which relates asset prices to aggregate consumption. Within this approach the market pricing kernel is derived from the optimal intertemporal consumption and investment choice of a representative agent. Alternative approaches attempt to generalize one-period models — in particular the Capital Asset Pricing Model (CAPM) — that usually relate asset prices to aggregate wealth, to the intertemporal context. These attempts suffer from problems that arise because in a multiperiod world aggregate wealth may not deterministically determine aggregate consumption, except for the last period. Although there exist theoretically founded generalizations they are of no practical use because they do not specify the relevant pricing factors. Therefore, the factors in the recently suggested Conditional CAPMs, which can be interpreted as linear multifactor models in the spirit of the theoretically founded generalizations, are more or less arbitrarily chosen. Moreover, in contrast to the consumption-based equilibrium asset pricing approach the conditional asset pricing approach takes the risk-free rate and the risk premia on the market portfolio and on a set of “factors” as exogenously given instead of determining them. This, however, is an important goal of the present analysis. This study therefore applies the consumption-based equilibrium asset pricing approach.