本文使用向量自我迴歸(VAR)模型探討被視為避險商品的VIX期貨之交易活動與其報酬和波動的相關影響關係。本文的研究期間將從2009年6月2日到2019年10月29日，資料頻率為以星期二為基礎的周資料。其結果發現在VIX期貨中與股價指數與商品期貨市場不同，首先，在VIX期貨也觀測到投機者反而占大多數具有50%以上，推測因其性質與商品期貨不同所導致。其次，避險者為反向動能交易者(報酬率減少則其部位增加)，並具有預測能力(部位增加則報酬率上升)且會因為波動上升增加其部位。投機者則不具有預測能力，並為正向動能交易者，其波動上升也會導致部位增加。最後，本文也觀察到小額交易者的總部位占比增加會導致VIX期貨報酬率上升。此外，本文使用VVIX與SKEW作為外生變數，發現VVIX與報酬率具有同期正相關並且會些微減少避險者跟小額投資者的部位以及些微增加投機者的部位，這表示在實務上除了參考避險者的部位變化外VVIX也需要多加留意，而SKEW以實證來看則與報酬率沒有關聯性。
This paper uses the vector self-regression (VAR) model to explore the relationship between the trading activities of VIX futures regarded as safe-haven commodities and their return and volatility. The research period of this article will be from June 2, 2009 to October 29, 2019, and the data frequency is weekly data based on Tuesday. The results show that the VIX futures are different from the stock price index and the commodity futures market. First, speculators are also observed in VIX futures to account for more than 50% of the majority, presumably due to their different nature from commodity futures. Secondly, the hedgers are reverse kinetic energy traders (the rate of return decreases when the position increases), and has the predictive ability (the position increases when the rate of return increases) and will increase its position due to rising volatility. Speculators do not have the predictive ability and are positive kinetic energy traders, whose rising volatility will also lead to an increase in positions. Finally, this article also observes that the increase in the proportion of small traders' headquarters will cause the return rate of VIX futures to rise. In addition, this article uses VVIX and SKEW as exogenous variables, and finds that VVIX and the return rate have a positive correlation over the same period and will slightly reduce the positions of hedgers and small investors and slightly increase the positions of speculators, which means that in addition to the reference avoidance in practice VVIX also needs to pay more attention to the changes in the position of the insurer, and SKEW has no correlation with the rate of return based on empirical evidence.
