The purpose of this paper is to enlarge a generally accepted and often used fundamental approach of capital market theory in the evolutionary sense. Our intension is to develop the basics of a model which is oriented on the evolutionary theory and rooted in the stringency of the neoclassical decision theory. In most cases this is the lack of the evolutionary model conceptions in our days. For this purpose, we take recourse to the State Preference Model which is regarded as the basic model of the neoclassical capital market theory. In doing this, we overtake certain basic assumptions concerning the decision and risk behavior of the agents, but we do not accept the neoclassical condition that the agents are always optimally adapted to exogenous given environmental states. However, due to their actions, in our model the agents generate the state realisations themselfes. This creates the condition for evolutionary extensions. Mainly these are: Endogenous adaption processes because of heterogeneous and interacting agents, the development of individual and aggregate state spaces, shown itself by the subjective price expectations of the market participants, the enlargement of the mechanically optimization behaviour to cognitive founded decision routines, or the integration of the model in more complex modelling techniques, like the neural network approach. Furthermore, we can show that an evolutionary oriented model on the one hand can be logically constructed on neoclassical insights, but on the other hand can generate results which are beyond the neoclassical findings. As we show in the paper, the applicability of this model for simulation runs allows a more realistic and evident description how capital markets work.