本文關(guān)鍵詞： 主成分正交補(bǔ)門限方法 主成分正交補(bǔ)門限CCC-GARCH模型 高維協(xié)方差陣　出處：《統(tǒng)計(jì)與信息論壇》2016年09期 　論文類型：期刊論文

【摘要】：高維數(shù)據(jù)給傳統(tǒng)的協(xié)方差陣估計(jì)方法帶來了巨大的挑戰(zhàn),數(shù)據(jù)維度和噪聲的影響使傳統(tǒng)的CCCGARCH模型估計(jì)起來較為困難。將主成分和門限方法有效結(jié)合,應(yīng)用到CCC-GARCH模型的估計(jì)中,提出基于主成分正交補(bǔ)門限方法的CCC-GARCH模型(PTCCC-GARCH)。PTCCC模型主要通過前K個(gè)最優(yōu)主成分來刻畫大維協(xié)方差陣的信息,并通過門限函數(shù)以剔除噪聲的影響。通過模擬和實(shí)證研究發(fā)現(xiàn):較CCCGARCH模型而言,PTCCC-GARCH模型明顯提高了高維協(xié)方差陣的估計(jì)和預(yù)測效率;并且將其應(yīng)用在投資組合時(shí),投資者獲得了更高的投資收益和經(jīng)濟(jì)福利。
[Abstract]:High-dimensional data bring great challenge to the traditional covariance matrix estimation method. The influence of data dimension and noise makes the traditional CCCGARCH model difficult to estimate. The principal component and threshold method are effectively combined into the estimation of CCC-GARCH model. A CCC-GARCH model based on the principal component orthogonal complement threshold method is proposed to describe the information of the large dimensional covariance matrix by using the first K optimal principal components. And the threshold function is used to eliminate the influence of noise. Through simulation and empirical research, it is found that the PTCCC-GARCH model improves the estimation and prediction efficiency of the high-dimensional covariance matrix obviously compared with the CCCGARCH model, and applies it to the investment portfolio. Investors received higher investment returns and economic benefits.
【作者單位】： 貴州財(cái)經(jīng)大學(xué)數(shù)學(xué)與統(tǒng)計(jì)學(xué)院;西南財(cái)經(jīng)大學(xué)統(tǒng)計(jì)學(xué)院;對外經(jīng)濟(jì)貿(mào)易大學(xué)統(tǒng)計(jì)學(xué)院;
【基金】：貴州省教育廳2015年度普通本科高校自然科學(xué)研究項(xiàng)目《大維數(shù)據(jù)背景下金融協(xié)方差陣的估計(jì)及應(yīng)用》(黔教合KY字[2015]423) 2015年全國統(tǒng)計(jì)科學(xué)研究項(xiàng)目《金融動(dòng)態(tài)條件協(xié)方差陣的估計(jì)及其應(yīng)用》(2015LY19) 2015年度北京市社會(huì)科學(xué)基金青年項(xiàng)目《大數(shù)據(jù)背景下北京市網(wǎng)絡(luò)風(fēng)險(xiǎn)動(dòng)態(tài)監(jiān)測與控制機(jī)制研究》(15SHC030) 2015年度全國統(tǒng)計(jì)科學(xué)研究重大項(xiàng)目《大數(shù)據(jù)視角下我國主要宏觀經(jīng)濟(jì)指標(biāo)預(yù)判預(yù)測方法體系研究》(2015LD050)
【分類號(hào)】：F224;F830.59

【相似文獻(xiàn)】

相關(guān)期刊論文 前1條

1 吳武清;汪成杰;蔣勇;陳敏;;高維數(shù)據(jù)選元:方法比較及其在納稅評估中的應(yīng)用[J];管理評論;2013年08期

，

本文編號(hào)：1499382