發(fā)布時間：2018-11-14 14:56

【摘要】：金融時間序列的波動性建模經(jīng)歷了從一階矩到二階矩直到高階矩(包含三階矩和四階矩)的過程,而對于高階矩波動模型是否有助于對未來市場的波動率預測這一問題,國內(nèi)外學術(shù)界尚無文獻討論。以上證綜指長達7年的每5分鐘高頻數(shù)據(jù)樣本為例,通過構(gòu)建具有不同矩屬性的波動模型,計算了中國股票市場波動率的預測值,并利用具有bootstrap特性的SPA檢驗法,實證檢驗了不同矩屬性波動模型的波動率預測精度差異。實證結(jié)果顯示:就中國股市而言,四階矩波動模型能夠取得比二階矩波動模型更優(yōu)的波動率預測精度,而三階矩波動模型并未表現(xiàn)出比二階矩波動模型更強的預測能力;在高階矩波動模型中包含杠桿效應(yīng)項并不能提高模型的預測精度。最后提出了在金融風險管理、衍生產(chǎn)品定價等領(lǐng)域引入四階矩波動模型的研究思路。
[Abstract]:The volatility modeling of financial time series has experienced the process from first moment to second moment to high order moment (including third moment and fourth moment), but whether the high order moment volatility model is helpful to predict the volatility of the future market. There is no literature discussion at home and abroad. Taking the high frequency data samples of Shanghai Composite Index for 7 years as an example, a volatility model with different moment attributes is constructed to calculate the forecast value of volatility in Chinese stock market, and the SPA test method with bootstrap characteristics is used. The volatility prediction accuracy of different moment attribute volatility models is tested empirically. The empirical results show that the fourth-order moment volatility model can achieve better volatility prediction accuracy than the second-order moment volatility model, but the third-order moment volatility model does not show better prediction ability than the second-order moment volatility model. The prediction accuracy of the model can not be improved by including the lever effect in the higher-order moment wave model. Finally, the fourth moment volatility model is introduced in the fields of financial risk management and derivative pricing.
【作者單位】： 西南交通大學經(jīng)濟管理學院;
【基金】：國家自然科學基金(70501025;70771097)
【分類號】：F224.0;F832.51
，

本文編號：2331495