演講者：黃士峰教授

單　位：國立中正大學數學系

日　期：2009年2月25日　PM 14:30

地　點：國立高雄大學理學院408室

講　題：Option Pricing and Hedging for Conditional Leptokurtic Returns

摘　要：

Conditional leptokurtosis is an important feature in financial data which can be depicted by conditional heteroscedastic models. Since conditional heteroscedastic models are incomplete market models, there exists no self-financing trading strategy to replicate a contingent claim. A hedging strategy, called η-hedging, based on minimizing the variance of the additional hedging capital at each rebalanced time is introduced. The η-hedging is shown to be more adapted to the hedging period than the commonly used delta hedging. The dynamic programming of theη-hedging is proposed as well. The initial hedging capital of the η-hedging is shown to be identical to the no-arbitrage price derived by an extended Girsanov change of measure. The risk-neutral models of conditional leptokurtic GARCH models are derived by this change of measure. And the dynamic semiparametric approach (Huang and Guo, 2008) is extended to option pricing and hedging for the conditional leptokurtic risk-neutral models. Furthermore, models risks of the GARCH-normal models in option pricing and hedging for plan vanilla and exotic options of the GARCH-t models are investigated via simulation study.