In Search of an Optimal Portfolio Strategy for Institutional Investors

Autor: MING-YUAN LIU, 劉明源
Rok vydání: 2014
Druh dokumentu: 學位論文 ; thesis
Popis: 102
In order to find an optimal strategy for long-term investor, this paper applies Mean-Variance Portfolio model introduced by Markowitz(1952), and choose Tangency portfolio on efficient frontier. The study tries to find out how long the window width (sample period) fits window step length (holding period) best within recent 5 years sample. The performances of portfolio were also estimated and observed by Shrinkage covariance matrix along with sample covariance matrix. With those findingsthe researcher begins with a measurement of portfolio returns from 1991 to 2014. And it finally turned out MV portfolio an optimal strategy against Dow Jones Industrial Average(DJIA) which is used as a benchmark. There also follows some further results: 1. Is not easy to tell the differents of those two kinds covariance matrix method. ShrinkEstimator can indeed reduce the maximum loss. But its performance is weaker than covEstimator when total return is less than the benchmark. 2. In 18 years investment,MV portfolio has made every one dollar into $3.34 compared with DJIA the benchmark turned it into $2.36. It means 98.75% premium. Three major crises lies in the way, especially the 2008 subprime catastrophe, which should make our portfolio suffered serious loss because the bankruptcy of GM, one of the 30 components of DJIA. Fortunately the MV model is keeping away from these traps, and outperforms 10.61% finally during those hard times.
Databáze: Networked Digital Library of Theses & Dissertations