A Comparison of Stock Index Futures Valuation under Different GARCH Volatility Models

[[abstract]]本文研究採用五種GARCH家族波動性估計方法來估計Hemler and Longstaff模式中的股價波動性，以檢示何種GARCH方法技能捕捉五個樣本國家股價指數的波動性，進而提升Hemler and Longstaff模式定價績效。其次，本文也比較已開發市場和開發中市場之Hemler and Longstaff (1991) 模式與持有成本模式定價績效。 本文的實證結果顯示：(1)不同期間的絕對價格誤差率比較，結果顯示，在Nikkei 225指數期貨、TAIFEX指數期貨和KOSPI 200指數期貨中，越晚期間存在越大的絕對價格誤差率。S&P 500指數期貨和Heng Sang指數期貨則未出現這樣的情形。(2)無論在任何期間，Hemler and Longstaff模式中，任何波動性估計方法所估計出的定價誤差率均無顯著差異。(3)考量股價波動性Hemler and Longstaff定價模式，運用在開發中市場指數期貨優於持有成本定價模式。[[abstract]]The purposes of this study are (1) to test which GARCH model can capture stock index volatility and enhance the performance of the Hemler and Longstaff (1991) model by using five GARCH volatility models to estimate stock index volatility in the Hemler and Longstaff (1991) model; and (2) to compare the price performance of the cost of carry and Hemler and Longstaff (1991) models in developed and developing markets. The empirical results can be summarized as follows :( 1) The absolute percentage error of the later period is larger than that of the earlier period for Nikkei 225 futures, TAIFEX futures and KOSPI 200 futures. However, S&P 500 futures and Heng Sang futures markets do not show above situation. (2) The mean absolute percentage errors of the Hemler and Lonstaff model do not differ significantly for the five GARCH (1, 1) volatility models. (3) Overall, the performance of the Hemler and Lonstaff model that incorporates stochastic market volatility is better than that of the cost of carry model for the developing countries