預測風險中立機率之高階動差及其選擇權投資組合:臺灣加權指數選擇權之實證分析

Abstract

本研究採用臺灣加權指數期貨與臺灣加權指數選擇權每日收盤資料，推算風險中立動差，包括波動度偏態及峰態。再利用多變量ARIMA模型預測出隔日的各階動差變化量，並檢定預測結果是否具預測性。最後使用其預測出的動差做為指標進行動差交易法，測試在考慮成本的情況下是否能獲得顯著的正報酬。本論文發現在資料區間內，多變量ARIMA模型對於標的物的動差變化方向具有預測能力；使用預測的動差變化方向做為交易策略的決策指標，該交易策略在考慮交易成本下具顯著有獲利能力。

In this research, we first calculate risk-neutral moments including volatility, skewness and kurtosis using daily close price of TAIEX (Taiwan Stock Exchange Capitalization Weighted Stock Index) futures and options risk-neutral moments including volatility, skewness and kurtosis. Next, we employ multivariate ARIMA model to predict the moment change on the next day, and examine its predictability. Lastly, we explore the moment trading strategies using the predicted moments as indicators to investigate if these strategies can earn significantly positive returns considering transaction costs. We find that the multivariate ARIMA model is able to predict the change of risk-neutral moments. Considering transaction costs, the trading profit is significant when we use the predicted changes of risk-neutral moments as the trading rules.