頁籤選單縮合
題名 | 股票型共同基金相關性預測模型之比較=Forecasting the Correlation among Equity Mutual Funds |
---|---|
作者 | 陳振遠; 高蘭芬; 吳香蘭; Chen, Roger C. Y.; Kao, Lanfeng; Wu, Hsianglan; |
期刊 | 輔仁管理評論 |
出版日期 | 20050500 |
卷期 | 12:2 2005.05[民94.05] |
頁次 | 頁127-156 |
分類號 | 563.538 |
語文 | chi |
關鍵詞 | 相關性預測; 共同基金; 風格指數; Correlation forecasting; Mutual fund; Style index; |
中文摘要 | 以往的研究指出,持有多種基金之投資策略將有助於投資人分散風險,然而其前提乃在於投資人必須能精確預測共同基金報酬間的相關性。本研究除了比較Ahmed (2001)所使用的相關性預測模型之預測績效外,另依Buetow, Johnson&Runkle (2000)之觀點,設計產業多重風格指數預測模型,並將其一併納入預測模型之績效比較。實證結果顯示,本研究所建立之產業多重風格指數模型,由於較符合我國股市交易的特性,故成為最佳的預測模型。此一結果呼應Buetow, Johnson&Runkle (2000)所指出,在使用歷史報酬率法時,應先檢視經理人的資產配置策略,如此才可有效的建構指數以進行風格分析。此外,一般多重風格指數模型及Fama-French三因子及動態模型,亦有不錯的預測績效及穩定性。整體而言,多數模型之預測能力皆較歷史模型為佳,亦即透過一些估計技巧對基金間未來相關性作預測,會比直接以歷史相關性作為預測值較為精確。 |
英文摘要 | Previous studies show that a multi-fund portfolio is far less risky than its single-fund counterpart and will enable investors to diversify effectively. To successfully implement diversification strategies, investors must obtain accurate estimates of the correlation among mutual fund returns. This paper forecasts mutual fund correlatoin using the models discussed in Ahmed (2001). Moreover, following Buetow, Johnson and Runkle (2000), we use an industry-based style index model to capture the characteristics of trading behavior in Taiwan’s stock market. We evaluate the performance of each model in forecasting correlation among equity mutual funds in Taiwan. Results show that the estimate of future correlation from the industry-based style index model has the lowest prediction errors. Such result is consistent with the findings of Buetow, Johnson and Runkle (2000), which shows that the return-based style analysis is a useful tool when the investment philosophy of the portfolio manager is properly captured by a set of asset classes. The multi-style index model, Fama-French 3-factor model and dynamic model also perform well in forecasting future correlation among mutual funds. In short, most of the models examined by this paper perform better than the historical model. It shows that we can forecast the correlation among mutual funds more precisely through some techniques than using the historical correlation directly. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。