市場上評估基金績效的方法很多,但多僅限於單一因子或二至三個因子的分析,未能詳細分析基金經理人在不同構面上的表現,本研究乃運用近兩年引進國內之BARRA基金績效拆解分析系統,並舉日盛投信公司所管理的六檔股票型基金為例,說明此一決策支援系統如何幫助經理人改善操作績效,及提供投信管理階層在行銷基金上的賣點(Saleskit)。
BARRA為一種多因子計量分析模型,可將基金在評估期間內的報酬率拆解成選時能力(Market Timing)、操作風格(Risk Indices)、產業選擇能力(Industries)、選股能力(Asset Selection)、月中交易換股能力(Trading)及交易成本(Transaction Cost)等六大項目,並配合t統計量的檢定,以分析基金經理人在各方面的表現,及這些表現是否處於顯著的水準。
本研究運用BARRA績效分析決策支援系統,評估日盛投信公司六檔股票型基金在93年1月1日至93年12月30日止的期間內,該基金經理人在各項構面上的表現及是否處於顯著的水準,並依據分析所得的結果對各基金經理人提出日後在操作上改進的建議。同時亦歸納出日盛投信公司在股票型基金上操作的優缺點,,以作為日後在在推廣基金上的參考。
研究結果發現BARRA系統在投信公司管理基金經理人及基金行銷上,能提供明確且有效的資訊。後續研究者可加強資料的取得以作更精確的分析;或拉長分析期間以分析某一投信旗下所有股票型基金以歸納出該投信公司的強勢及弱勢;或運用BARRA系統的其他模型來提升基金績效。
There are many ways to evaluate fund performance, but most of them are limited to sin-gle, two, or three factor models. They fail to provide thorough analysis of fund managers in different perspectives. Our research is based on Aegis, MSCI Barra’s performance attribu-tion system. We use Jih Sun Investment Trust’s (JSIT)six domestic equity funds as an example and explain how this decision support system helps fund managers improve their performance as well as fund’s selling point.
With the calculation of t statistics, fund managers thus can be identified if they have the abilities to outperform in various ways, including market timing, risk indices, industries, asset selection, trading, as well as transaction cost.
Based on MSCI Barra’s system, we evaluated JSIT’s six domestic equity funds from Jan 1st to Dec. 30th of Y2004. We would like to know if fund manager’s t-test result in each of the factor is statistically significant. This result will help fund manager improve their port-folio management. Alternatively, we can also identify JS funds’ aggregated strength and weakness, thus promote them more efficiently.