Optimal consumption and arbitrage in incomplete,finite state security markets

We study a consistent treatment for both the multi-period portfolio selection problem and the option attainability problem by a dual approach. We assume that time is discrete, the horizon is finite, the sample space is finite and the number of securities is less than that of the possible securities price transitions, i.e. an incomplete security market. The investor is prohibited from investing stocks more than given linear investment amount constraints at any time and he maximizes an expected additive utility function for the consumption process. First we give a set of budget feasibility conditions so that a consumption process is attainable by an admissible portfolio process. To establish this relation, we used an algorithmic approach which has a close connection with the linear programming duality. Then we prove the unique existence of a primal optimal solution from the budget feasibility conditions. Finally, we formulate a dual control problem and establish the duality between primal and dual control problems.We are grateful to the editor, Hiroshi Konno, and two anonymous referees for their valuable comments and constructive suggestions on this research. We are responsible for the remaining errors. The first author is supported in part by the fund endowed to the Research Association for Financial Engineering by Toyo Trust Bank Co. and Mito Shoken Co.