Multi-stage stochastic wealth-semideviation-liquidity portfolio optimization under transaction costs

This paper deals with multi-stage stochastic portfolio selection problem. A three objective wealth-semi deviation-liquidity portfolio selection model is formulated that based on scenario tree in order to dealing with the markets uncertainty. The scenario tree is also an appropriate method for modelling multi-period portfolio problems since the duration, continuity of their horizon and under uncertainty. The significant criteria utilized for this problem and rebalancing are wealth, risk, threshold of assets investment, transaction cost and liquidity, and semi-deviation is considered as a measure of portfolio risk. The Node Based Modelling (NBM) is utilized to acquire efficient investment strategies for the through investment horizon. Also, the multi-objective model is transformed to single-objective by goal programming technique. A real-word empirical application with data-set from a Tehran stock exchange (TSE) is presented to illustrate functionality of proposed model for multi-period portfolio selection problem in stochastic environment.

Keywords:

Node based modelling , portfolio selection , multi-period investment , scenario tree.