台灣高科技共同基金之績效評估─風險值的應用

Performance Evaluation of Taiwan Hi-Tech Mutual Funds－An Application of Value-at-Risk

羅健榮(Lo Chien-Jung)

指導教授 ： 胡次熙

亞洲大學/管理學院/國際企業學系碩士班/碩士(2004年)

摘要

基金經理人與投資人最關心的議題是投資的績效與評估，投資的盈與虧可以清楚的看見，投資績效的優劣卻較不容易評定，其關鍵因素在於評比者對風險認知的不同。在眾多績效指標中，夏普指標最常被投資人使用。大部分績效指標皆建構在報酬率為常態分配的假設上，所以當基金報酬率不為常態分配時，指標將會產生偏誤。本文針對國內高科技基金進行常態分配檢定，發現所選取的樣本，在樣本期間內報酬率是常態分配，因此在傳統績效指標的運用上較不容易產生偏誤。本研究另外引入風險值的觀念，進行國內高科技基金績效評估，結論如下:
1.當報酬率為正，應用風險值的績效指標確實能有效的評估基金績效。當報酬率為負時，績效指標可能會產生偏誤。
2.當以相對標竿市場的報酬率取代無風險利率，及以相對標竿市場的風險值取代原市場的風險值所做的績效指標，可以了解基金在相對市場的績效。
3.在績效指標的相關性上，相關係數皆相當接近1，表示彼此存在高度正相關。
4.風險值模型方面，就研究發現delta-normal法接受度較高，但仍存在可能發生的模型風險。

關鍵字

高科技基金 ； 風險值 ； 報酬率 ； 夏普指標

並列摘要

The measurement of investment performance has always been a main concern for managers and investors. While the profit and loss of investment could be clearly seen, the performance of investment fund is not easily determined due to a key factor that people have different acknowledgement toward risk. In most cases, Sharpe index is widely used by investors. Most performance indices assume that returns follow normal distribution. When return distribution is not normally distributed, the index will not be theoretically correct. This paper tests the normal distribution assumption for Taiwan Hi-Tech mutual funds, we find return of our sample mutual funds in sample period follow the normal distribution, so an application of traditional performance index is not inappropriate. This paper includes an application of Value-at-Risk to the performance evaluation of Taiwan Hi-Tech mutual funds and we reach the following conclusions:
1. When return is a positive, VaR-based performance index is indeed effective in evaluating the mutual funds performance. When return is negative, the index will become biased.
2. Substituting benchmark relative market return for risk-less return and benchmark relative market VaR for original market VaR can compare funds performance in different markets.
3. To test performance index correlation, our test results display high positive correlation.
4. In back testing, delta-normal approach performs better than historical simulation approach, but it still shows model risk.

並列關鍵字

Hi-Tech mutual funds ； VaR ； return ； Sharpe index

參考文獻

6. 張有若。(2002)。「全球共同基金群組風險與績效評估─以風險值修正夏普指標之應用─」中原大學企業管理研究所碩士論文。

1. Alexander, C.O., and C.T. Leigh. (1997). On the Covariance Metrices Used in Value at Risk Models. The Journal of Derivatives, 4(3): 50-62.

2. Bollerslev, T. (1986). Generalized Autoregressive Conditional Heteroskedasticity. Journal of Econometrics, 31(3): 306-327.

3. Beder, T. (1995). VaR: Seductive but dangerous. Financial Analysts Journal, 5(1): 12-24.

6. Dowd. (1999). A Value at Risk Approach to Risk-Return Analysis. Journal of Portfolio Management, 25(4): 60-67.