Convex duality in optimal investment under illiquidity (Q484140)

The author studies the problem of optimal investment in a discrete-time financial market by using the general conjugate duality theory of convex analysis. This allows for various generalizations to classical models (of liquid markets) based on stochastic integration. The author obtains a dual representation of the optimal value function in the presence of portfolio constraints and nonlinear trading costs. The financial position of an agent is described by a sequence of cash-flows which should be delivered by him (her). Such a presentation is essential in markets without a numerical asset when pricing swap contracts and other financial instruments with multiple payout dates. In the special case of perfectly liquid markets with proportional transaction costs, the dual expressions are obtained in terms of martingale measures.