【摘要】：以滬深300股指期貨仿真交易的5分鐘高頻數(shù)據(jù)為例,運用滾動時間窗的樣本外預測和具有Bootstrap特性的SPA檢驗法,全面對比了基于日收益數(shù)據(jù)的歷史波動率(historical volatility)模型和基于高頻數(shù)據(jù)的已實現(xiàn)波動率(realized volatility)模型對波動率的刻畫和預測能力.主要實證結(jié)果顯示,已實現(xiàn)波動率模型以及加入附加解釋變量的擴展隨機波動模型是預測精度較高的波動模型,而在學術(shù)界和實務(wù)界常用的GARCH及其擴展模型對滬深300股指期貨的波動率預測能力最弱.
[Abstract]:Taking the 5-minute high-frequency data of Shanghai-Shenzhen 300 stock index futures as an example, this paper uses the out-of-sample prediction of rolling time window and the SPA test method with Bootstrap characteristics. In this paper, historical volatility (historical volatility) model based on daily income data and realized volatility (realized volatility) model based on high frequency data are comprehensively compared to describe and predict volatility. The main empirical results show that the realized volatility model and the extended stochastic fluctuation model with additional explanatory variables are the volatility models with high prediction accuracy. However, GARCH and its extended model, which are commonly used in academic and practical circles, have the weakest ability to predict volatility of Shanghai-Shenzhen 300 stock index futures.
【作者單位】： 西南交通大學經(jīng)濟管理學院;
【基金】：國家自然科學基金資助項目(70501025;70771097;70771095) 教育部新世紀優(yōu)秀人才支持計劃資助項目(NCET-08-0826) 教育部創(chuàng)新團隊發(fā)展計劃資助項目(PCSIRT0860)
【分類號】：F224;F832.51

【相似文獻】

相關(guān)碩士學位論文 前1條

1 彭珍;我國股指期貨推出對股票市場的影響[D];對外經(jīng)濟貿(mào)易大學;2007年

，

本文編號：2471766