We demonstrate that prevalent empirical implementations of asset pricing are inconsistent with dynamic equilibria [multiperiod equilibria with no single-period (static) representations]. Specifically, empirical implementations are misspecified with respect to three essential asset pricing questions (TEQ) on: dependency on higher moments, complexity of risk premia, and mean-variance efficiency of the “market portfolio” (the pricing kernel/SDF). While we already know that “Merton models” and their derivatives differ from static models in all TEQ, we show that this is the case for all dynamic equilibria, even for the minimal dynamic equilibria (dynamic equilibria with the simplest structure).