15679

1598 دكتري

دكتري

صنايع

اصفهان : دانشگاه صنعتي اصفهان

1398

چهارده، 166ص. : مصور، جدول

مهدي بيجاري

مهدي خاشعي

علي زينل همداني، سعيد فتحي، محمدتقي جهانديده

1399/05/12

كتابنامه

مهندسي صنايع

مهندسي صنايع و سيستم ها

1399/05/13

2624704

058IUTDegree Doctor of PhilosophyTitle A novel class of financial risk measurement methods based on the predictability ofinvestment returnsAuthor Zahra Moeini NajafabadiSupervisor Professor Mehdi BijariDepartment Industrial and Systems EngineeringDate February 29 2020Language PersianAbstractInvestment decision making is one of the most important tasks of financial managers which havealways been exposed to uncertainty over financial markets So that measuring and managing risk areinvestors main concerns to make profit from financial markets The accuracy of the investment riskhas a significant impact on the efficiency of the investment decisions Therefore many studies havebeen done in defining and calculating risk such as Markowitz theory His researches have been thebase of many other studies and many researchers have proposed new approaches to risk calculationusing basic Markowitz concepts But in spite of all the benefits of the Markowitz based methods noneof them consider the predictability of the under studied systems in calculating risk However thepredictability of a system is inversely related to its risk Therefore the purpose of this study is toimprove investment decisions by changing the method of calculating risk based on the predictability ofthe expected return of financial assets For this purpose six risk measurement methods were proposedbased on time series forecasting models Then the proposed methods were used in making investmentdecisions In order to evaluate the proposed models the data of 30 stocks in Tehran Stock Exchange were used tomake investment decisions on independent assets as well as 9 stock exchange indices in portfolioselection problem Then the expected return and risk on each financial asset or portfolio werecalculated using proposed models over consecutive weeks and became the basis for investmentdecisions The empirical results have shown that the use of forecasting based models by increasing the accuracyof risk and reducing its amount can make investment decisions at lower risk levels and at the sametime increase the profitability of decisions Thus the annual profit of investment decisions on eachasset was improved from 3 68 in the AR Cai and ARMA Cai methods up to 35 88 in the ANN Konno method depending on the risk calculation approach and the characteristics of the time seriesforecasting methods In addition using the proposed risk measurement methods based on directionalanalysis was improved the annual profit than the Markowitz method from 11 44 up to 41 6 depending on the forecasting method used Finaly the results also have shown the improvement of theefficient frontier using the proposed methods about 200 up to 300 Keywords Risk Investment Portfolio selection problem Efficient frontier Time seriesforecasting models

مهدي بيجاري

مهدي خاشعي

علي زينل همداني، سعيد فتحي، محمدتقي جهانديده