Ethem Canakoglu and Süleyman Özekici. Portfolio optimization with imperfect information

Abstract. We consider a multiperiod utility based portfolio selection problem where the parameters change according to a Markovian market which is not perfectly observable. The market consists of a riskless asset and several risky assets whose returns depend on the state of the unobserved market. The stochastic market represents the prevailing economic, social, political and other relevant factors that affect the financial parameters. Investment decisions are based on what is observed. We consider two different models that describe the imperfect information flow between the unobserved stochastic market and the observed process one using HMM and other using sufficient statistics.