| Summary: | 碩士 === 銘傳大學 === 財務金融學系碩士在職專班 === 93 === Abstract
It is often found that some fear-index seems to be linearly co-moved with stock price index. For instance, the holdings of call-option relative to put-option (CPR) positively move with the stock price index. With comparable striking price, the call-option’s implied volatility is higher than the put-option’s volatility. It practically accounts for the fear of market crash. As a matter of facts, the future value of option should be reflected by the implied volatility of underlying assets. Whaley (2000) suggested that the fear index can be properly measured by volatility index (VIX). The study is to examine how the deviation from the linearity of VIX and future index affects the information content of future index.
With referring to the essence of the last version of VIX of CBOE, the paper builds up a feasible VIX in accordance with the trading traits of Nikki futures market using the five-minutes intraday data fully quoted on the Nikkei SEC from January 2, 2003 to December 31, 2004. The rolling-over contract’s prices and exercising prices will be adequately revised to act well with the behavior of realized Nikkei VIX (NVIX). With allowing for the effects of the deviation from the linearity of NVIX and Nikkei 225 future index (NFI), the threshold error correction model (TECM, Hansen and Seo, 2000) will be employed to re-examine the Granger cause-effects between NVIX and NFI.
The empirical evidences support the existence of threshold effects. It is found that the NVIX Granger does not significantly precede the NFI when the NVIX deviates a bit from the NFI (i.e. the error correction term stays at low regime); however, the NVIX surely Granger precedes the NK255 when the NVIX highly deviates from the NFI (i.e. the error correction term stays at high regime).
|
|---|
