The question of why some Wall Street AI applications are successful while others fail is explored. The Wall Street problem domain is modeled from a knowledge-based perspective, and the environment is represented from both an object-oriented view and a process view. Both models demonstrate the complexity inherent in formally defining the Wall Street environment. Analytic and cognitive models of the Wall Street environment are discussed, focusing on the shortcomings of each approach. The use of a new type of intelligent tool, called a collaborative system, that combines analytic models, human judgment, and cognitive models is proposed.