| Summary: | 碩士 === 國立臺北大學 === 統計學系 === 99 === Due to the uncertainty of investments in financial markets, risk management and proper asset allocation are two important themes to consider with regards to investing. Given that financial time series data such as the return of financial products not only exist conditional heteroskedasticity but also the time-varying correlation between different products, this paper takes the dynamic conditional correlation within different investment objectives into consideration, and construct an optimal portfolio allocation model with controlling downside risk.
The model in this paper is constructed by two modules. Module I renders the Dynamic Conditional Correlation(DCC) model to estimate the conditional covariance matrix between different investment objectives. Module II is based on Markowitz’s portfolio theory, constructing an efficient frontier by the covariance matrix derived in Module I. In addition, Module II seeks to find the optimal portfolio which includes the maximum Sharpe Ratio while satisfying different downside risk constraints. Finally, we provide an empirical analysis with the stock market in Taiwan.
The results display that all of the optimal portfolios from our modules outperformed the benchmarks we chose during the same time frame. Furthermore, when the overall stock price declines, the optimal portfolios as selected in this paper had smaller loss than benchmarks. These outcomes can fully reveal that the performance as a result of taking the dynamic volatility between investment returns into account and applying the downside risk constraint to portfolios.
|
|---|
