1 |
利用GARCH模型預測VIX ETN並建構避險策略 / VIX ETNs hedging strategies using GARCH models吳培菱 Unknown Date (has links)
自從2008年金融危機爆發後,黑天鵝事件相繼出現,VIX成為投資人衡量股市波動度的重要指標。但是若投資人想使用VIX避險,僅能透過限專業投資人參與的VIX期貨。而在近年ETF產品盛行的背景下,投資標的更加多元的交易所交易債券(ETN)也應運而生,使一般投資人得以進入以往難以觸及或交易成本高昂的市場。本研究採用兩檔交易量較大之VIX ETN,分別追蹤VIX短期與中期期貨指數之VXX與VXZ,希望透過建構GARCH模型用以預測其隔日價格,並以此預測的價格近一步建構避險策略,目標係在預期空頭即將發生時,提供投資人除了VIX期貨和波動相對平穩的債券以外的避險替代工具。
建構GARCH模型的部分,本研究主要參考Kambouroudis和McMillan(2013)的文獻,在變異數方程式中加入輔助變數,可以增加模型的預測能力,故本研究在VIX ETN之GARCH模型的變異數方程式中加入VIX、短期VIX指數及中期VIX指數。實證結果顯示,在VIX ETN的GARCH模型中同時加入VIX相關指數,確實能提高配適程度並增進預測能力,尤其當加入的輔助變數與VIX ETN追蹤標的的到期期限相符時,此改善模型的效果最為顯著。
本研究接者參考Alexander和Korovilas(2012)的VIX ETN避險研究,文獻顯示,在S&P 500 ETF投資組合中加入VXX與VXZ避險可提高夏普比率。本研究在此基礎上,額外考量了不同的持有期間、進場條件、股債混合的投資組合,並分別比較兩種ETN的避險效果。本研究發現只在VIX大於20時才進場建構避險部位的策略,提前買入VIX ETN確實可以做為良好的避險工具。此外,在此策略下,VIX ETN亦則可達到比持有債券更佳的避險效果。而本研究所測試的兩種VIX ETN中,又以VXX 避險效果更佳,因VXX乃是追蹤VIX短期期貨指數,更能反映市場短期的變化,搭配滾動的避險比率,能更加精準的反應空頭時期劇烈的波動。 / Since the 2008 financial crisis, along with the black swan events, the volatility of global stock market has intensified, and VIX index becomes an important indicator for investors to measure the volatility of the stock market. However, if investors would like to use VIX index for hedge, they could only use VIX futures, which is only for professional investors to participate. In recent years, the prevalence and popularity of the various ETPs lead to the booming of VIX ETNs, which has become an alternative for regular investors to invest in VIX index. Therefore, this study hopes to build GARCH model for VIX ETN and predict their prices of the next day, and use the prediction to build hedging strategies.
In this paper, this study mainly refers to the paper of Kambouroudis and McMillan (2013) to construct the VXX and VXZ prediction models. Because the two VIX ETNs track the S&P 500 VIX short-term and medium-term futures index respectively, the study add the VIX index, short-term VIX index and medium-term VIX index in the GARCH models. The empirical results show that the addition of VIX and other relevant VIX indices in the VIX ETN GARCH models can improve the forecasting ability. In particular, when the maturity of the VIX index is consistent with the maturity of the VIX ETN’s tracking target, it would improve the prediction power the most.
Based on the predicted VIX ETN prices, this study then constructs the hedging strategies, considering the different holding period, the entry condition and the stock and debt mixed portfolio, and also compares the hedging effect of VXX and VXZ respectively. This study found that under the strategy that only enter the VIX ETN market when VIX was greater than 20, VIX ETN can indeed be a good hedge tool and reduce the standard deviation of the portfolio. In addition, under this strategy, if investors use VIX ETN to hedge, investors can achieve a higher return and lower standard deviation than holding a bond to hedge. Finally, among the two VIX ETNs tested in this study, VXX is a better hedge tool against VXX. It is because VXX tracks the VIX short-term futures index which reflects the short-term changes in the market and hence could reflect the short-term volatility better.
|
2 |
應用機器學習於標準普爾指數期貨 / An application of machine learning to Standard & Poor's 500 index future.林雋鈜, Lin, Jyun-Hong Unknown Date (has links)
本系統係藉由分析歷史交易資料來預測S&P500期貨市場之漲幅。 我們改進了Tsaih et al. (1998)提出的混和式AI系統。 該系統結合了Rule Base 系統以及類神經網路作為其預測之機制。我們針對該系統在以下幾點進行改善:(1) 將原本的日期資料改為使用分鐘資料作為輸入。(2) 本研究採用了“移動視窗”的技術,在移動視窗的概念下,每一個視窗我們希望能夠在60分鐘內訓練完成。(3)在擴增了額外的變數 – VIX價格做為系統的輸入。(4) 由於運算量上升,因此本研究利用TensorFlow 以及GPU運算來改進系統之運作效能。
我們發現VIX變數確實可以改善系統之預測精準度,但訓練的時間雖然平均低於60分鐘,但仍有部分視窗的時間會小幅超過60分鐘。 / The system is made to predict the Futures’ trend through analyzing the transaction data in the past, and gives advices to the investors who are hesitating to make decisions. We improved the system proposed by Tsaih et al. (1998), which was called hybrid AI system. It was combined with rule-based system and artificial neural network system, which can give suggestions depends on the past data. We improved the hybrid system with the following aspects: (1) The index data are changed from daily-based in into the minute-based in this study. (2) The “moving-window” mechanism is adopted in this study. For each window, we hope we can finish training in 60 minutes. (3) There is one extra variable VIX, which is calculated by the VIX in this study. (4) Due to the more computation demand, TensorFlow and GPU computing is applied in our system.
We discover that the VIX can obviously has positively influence of the predicting performance of our proposed system. The average training time is lower than 60 minutes, however, some of the windows still cost more than 60 minutes to train.
|
Page generated in 0.0135 seconds