市面上有許多軟體可以進行財務分析，多數為商業軟體，如SAS、SPSS，對使用者來說過於昂貴，因此，本文獻著重在R軟體財務套件的使用，介紹處理時間序列性質的資料套件，以及利用套件PerformanceAnalytics針對從台灣經濟新報(TEJ)下載的台灣股票型基金進行績效分析、風險分析，我們使用國內基金評比權威台大財務金融所教授邱顯比、李存修提出的四四三三法則，選出2012年較適合投資的股票型基金。最後，則是使用馬可維茲建立的投資組合理論來最佳化股票型基金投資組合；分別使用平均數-變異數與平均數-預期短缺的模型搭配穩健式估計(Robust)改善估計共變異數矩陣選出較適宜的投資組合。 結果發現，多數股票型基金的績效不如台灣加權指數，即使採用投資組合最佳化的理論調整權重，可以選出報酬率較佳的投資組合雖然最大損失以及報酬率均優於台灣加權指數，但是若是扣掉模型中未加入基金的保管費，其整體仍然不如台灣加權指數。有必要選擇其他資產或者是將資金直接存放於銀行。 It is many software of financial analysis on the market. More those of commercial are expensive for users, such as SAS and SPSS. Therefore, the thesis focuses the operation of the "R" financial suite, introducing the time series packages and performance analytics with the suit.
The Stock Funds of Taiwan, downing from Taiwan Economic Journal (TEJ), execute performance and risk analysis. We use the rule called Four-four-three-three, brought from Shean-Bii Chiu and Lee Tsun-siou--the professor of Dept. of Finance, NTC and the authority of Domestic Fund Ratings. We optimize the Portfolio of Stock Fund finally through the theory of Portfolio Found by Markowitz and improve the Estimate Covariance Matrix and choose the correct Portfolio by using the Mean, Mean-Variance and Mean-CvaR. individually.
We use the portfolio theory that establish by MarkoWitz to optimize portfolio of stock funds using the model of mean variance-CVaR and combine Robust estimate to improve the estimate of covariance matrix, then we can choice the better portfolio.
Thus, we note that many stock funds are less than TAIEX. Even if we can choice the better portfolio by using change the weight of portfolio optimization although max drawdown and return are better than TAIEX. But if we reduce the fee of funds in the model, it’s less than TAIEX. It’s necessary to select other assets or put funds into the bank.
