VIX選擇權交易量與未來VIX指數變動之關係

Abstract

本文旨在探討VIX選擇權交易量(trading volume)、現金交易量(dollar volume)及交易筆數(number of transaction)對未來VIX指數變動的預測能力。首先以Lee and Ready (1991)的演算法區分VIX選擇權的買賣單，從而將上述之解釋變數分為隱含資訊及未隱含資訊二類，並比較其預測能力。此外，由於樣本期間包含2008年的金融海嘯期間，本文亦檢驗在該期間VIX選擇權之交易量、現金交易量及交易筆數對未來VIX指數變動的預測能力。又因價外(out-of-the-money)選擇權具有較高財務槓桿，交易人可能會選擇操作VIX價外選擇權，故本文亦對VIX價外選擇權的預測能力進行探討。實證結果發現，VIX選擇權的交易筆數對未來VIX指數變動的預測能力最佳。此外，不論解釋變數為交易量、現金交易量或是交易筆數，隱含資訊類的變數對VIX指數未來變動的預測能力較未隱含資訊類的變數佳。

The purpose of this research was to empirically investigate the information role of VIX options variables, namely trading volume, dollar volume, and number of transaction, with regard to the prediction of future VIX index change. Firstly, we use the Lee and Ready (1990) algorithm to classify VIX option trades into buyer- or seller-initiated in order to construct information based variables and compare the significance of them with standard-definition variables. Besides, since the sample period covers financial crisis period, we also examined the difference of predictability in financial crisis period. In addition, we investigated whether out-of-the-money VIX option provided more information in prediction of future VIX index change. We found that number of transaction of VIX option is the most informative in predicting future VIX index. Furthermore, information-based variables provide more information than standard-definition variables, especially in financial crisis period.