Research And Application On Portfolio With Patulous Entropy Principle

Under the increasingly sophisticated political and economicsituations of international and domestic environment, financial market is facing greatopportunities and challenges. And meanwhile the matter that how investors could gainhigher expected profits and reduce lower risk in this special background has been paidclose attention for investors. Some intelligent investors select portfolio to spread therisk they tolerated, so researches on risk in portfolio have become one of the majorissues in academic circles.Current portfolio theories are built to satisfy the demands of investor decisionmaking based on the classical portfolio theory advanced by Harry Markowitz,moreover entropy theory has been familiar with more and more scholars in measuringthe market risk, and this theory could better measure the risk objectively so as tocompensate the lack of other models. This thesis is put forward under mentionedbackground. Then inverse entropy principle and generalized entropy principle arebuilt in this thesis, which not only expends entropy theory, in addition the riskaversion degree is also considered in generalized entropy principle, but also be closeto personal preference. So this thesis does a research on patulous entropy and containssix chapters.In Chapter1, the background and significance are introduced and each theory andmodel on portfolio is devoted, at last the innovations of this thesis are elaborated.In Chapter2, some comprehensive analysis on entropy optimal theory isdiscussed, and the feasibility and practicability are issued in portfolio field later,meanwhile a few important entropy laws are mentioned.In Chapter3, the inverse entropy principle is defined based on inverse problemsin physics and mathematics, and then the inverse entropy model in portfolio is built,which could provide necessary order measurements.In Chapter4, the generalized entropy principle with risk aversion degree is builtbased on Csiszer directed divergence theory, which could provide more effectiveselecting rules through empirical analysis.In Chapter5, some comparative analysis between inverse entropy principle andgeneralized entropy principle are explained later, which could provide some objectivepreference for investors.In Chapter6, some perspective are put forward to point at the insufficient ininverse entropy and generalized entropy models, finally this Master thesis ends.