- 學位論文
考量浮動門檻之最佳化Omega模型
Optimizing the Omega Portfolio Model with Floating Threshold
摘要
Omega ratio是一個有別於傳統的績效指標，不僅克服許多傳統測量方法只適用於報酬常態分佈 (Kazemi et al., 2004)的問題，並且能夠讓投資者設定衡量報酬跟損失的門檻，針對過去研究文獻大多將該指標的門檻設置為固定值，但為了更有效的反應市場波動，因此本研究依據Guastaroba et al. (2016) 提出之方法與Sharma and Mehra (2015) 的概念，將固定門檻值調整為兩種浮動門檻的機制，做出綜合性的比較並加以驗證：(1) 依照加強型指數追蹤法的方式將大盤的報酬率加上投資者期望； (2) 結合條件風險值於Omega模型的浮動門檻；從上述實驗中發現浮動門檻值確實能夠在依據市場價格波動的情況下，比固定門檻值的Omega模型帶來更好的績效，但發現結合了條件風險值模型的Omega模型，其市值表現雖優於條件風險值模型與固定門檻的Omega模型，但與整體的Omega模型相比無顯著的降低風險且績效不如使用加強型指數追蹤法的Omega模型，因此在浮動門檻機制上仍然建議使用加強型指數追蹤法較佳。本研究亦針對不同市場特性的資料集 研究Omega模型適合應用的市場，資料集分別為標準普爾500指數之成份股(S&P 500)與國家型指數基金(ETF)，發現在價格與報酬高波動之市場(S&P 500)使用浮動門檻的Omega模型時有較顯著的影響，且相較其他模型的市值表現最佳；於波動平穩之市場(ETF)，浮動門檻能夠反映市場波動但所有模型的市值表現差距不大，因此建議在報酬高波動之市場使用Omega模型，能帶來更好的績效；除此之外，本實驗亦考量了報酬不確定性下的穩健性模型，分別為最差情況Omega模型與最差情況之條件風險值模型，進而探討浮動機制之影響。
並列摘要
Omega ratio, a modern performance measure differs from classical one since it assesses the performance against a benchmark and usually apply on asymmetry distributions of return. Moreover, it enables investors to set the threshold of return and loss measurement. In most of the previous studies, the threshold of Omega ratio was set as a fixed value. To reflect the market fluctuation in a more effective way, this study changed the fixed threshold into two floating ones according to the method by Guastaroba et al. (2016) and the concept by Sharma and Mehra (2015), so as to make a comprehensive comparison for the demonstration: (1) Combined the return ratio of composite index with the investors’ expectation according to the enhanced index tracking method; (2) The floating threshold of CVaR in the Omega model was taken into consideration; it was found in the above experiment that the floating threshold indeed led to better performance than the Omega model with a fixed a threshold in case of the fluctuation in market price; it was also found that the market value of the combination of the CVaR and the Omega model was higher than that of the CVaR and the Omega model with a fixed threshold, but it didn’t have a significantly less risk than the Omega model, nor was its performance better than that of the Omega model based on the enhanced index tracking. Therefore, we still recommend the enhanced index tracking for the floating threshold mechanism. According to the data sets of different market features, this study also explored the market suitable for the Omega model. The data sets included the constituent stocks of S&P 500 as well as the ETF. It was found that the Omega model with a floating threshold had more significant effects in a market featuring great fluctuation in price and return, and its market value was better than other models; in a less fluctuated market (ETF), the floating threshold could reflect market fluctuation, but there wasn’t great difference in market value among the models. Hence, it is suggested that the market with great fluctuation in return should adopt the Omega model to achieve better performance. Besides, this experiment also probed into the prudent models with uncertain returns, including the Omega model in the worst case as well as the CVaR in the worst case, so as to analyze the effects of the floating mechanism.