本文關(guān)鍵詞：分位數(shù)回歸視角下我國商品期貨市場羊群效應(yīng)實(shí)證分析　出處：《浙江工商大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 羊群效應(yīng) 商品期貨 CCK模型 分位數(shù)回歸

【摘要】：在理性的資本資產(chǎn)定價(jià)模型CAPM (Capital Asset Pricing Model)理論支持下,CCK模型(Chang、Cheng和Khorana(2000)提出)通過橫截面絕對偏離度CSAD(Cross-sectional Absolute Deviation)和資產(chǎn)組合收益率的線性關(guān)系來檢測羊群效應(yīng)是否存在。基于此模型,眾多學(xué)者對股票市場進(jìn)行了研究,發(fā)現(xiàn)了不同程度的羊群效應(yīng),并且國內(nèi)的羊群效應(yīng)尤為明顯。目前,國內(nèi)外對期貨市場羊群效應(yīng)的研究相對較少,僅有的一些文獻(xiàn),也只是對個(gè)別品種的研究。本文從檢測我國商品期貨市場整體羊群效應(yīng)角度出發(fā),先利用C(x模型對期貨市場進(jìn)行最小二乘回歸作為分位數(shù)結(jié)果的對比,得出我國商品期貨市場整體沒有羊群效應(yīng)。由于傳統(tǒng)的最小二乘回歸只考慮條件均值的回歸,無法探測到尾部信息,為了更加全面地檢測羊群效應(yīng)對我國商品期貨市場的影響,需要利用分位數(shù)對模型進(jìn)一步回歸,結(jié)果仍然沒有發(fā)現(xiàn)羊群效應(yīng)�？紤]到板塊之間的特征差異,現(xiàn)實(shí)中羊群效應(yīng)更可能發(fā)生在局部,因此本文進(jìn)一步把整個(gè)市場分成若干板塊進(jìn)行實(shí)證。最小二乘回歸得出金屬板塊和油脂板塊有微弱的羊群效應(yīng),分位數(shù)回歸得出在橫截面離散度CSAD的低分位點(diǎn)有較為劇烈的羊群效應(yīng)。另外,在最小二乘回歸結(jié)果下沒有羊群效應(yīng)的板塊,通過分位數(shù)回歸,在低分位點(diǎn)也發(fā)現(xiàn)了微弱的羊群效應(yīng)。這體現(xiàn)了分位數(shù)的優(yōu)越性,它能更加全面地檢測局部信息,合理利用這些不同的分位數(shù)區(qū)間結(jié)論的差異性,投資者可以得出更好的投資組合,增加了期貨市場的有效性。
[Abstract]:In the capital asset pricing model CAPM (Capital Asset Pricing Model rational) theory support, CCK model (Chang, Cheng and Khorana (2000) proposed by CSAD) CSAD (Cross-sectional Absolute Deviation) and the linear relationship of portfolio returns to detect whether there is herding effect. Based on this model, many scholars the research on the stock market, found the herding effect in different degree, and the domestic herd effect is particularly obvious. At present, the domestic and foreign research on the herding effect of futures market is relatively small, only some of the literature, but also research on individual species. This article from the whole herd effect detection of China's commodity futures market perspective. By using the C X model (least squares regression for comparison, quantile results of the futures market from China's commodity futures market as a whole is not due to herding. The traditional least squares regression regression only considers the conditional mean, cannot detect the tail information, in order to comprehensively detect herding of commodity futures market in China, need to use quantile regression model to further results, still did not find herding. Considering the differences between the characteristics of the plate, the herding effect is more likely to occur in reality in part, this paper further put the whole market into several empirical plates. Least squares regression that metal plate and oil plate with herding weak, low quantile quantile regression obtained in CSAD dispersion cross section are herding more intense. In addition, no herding in the least squares regression results under the plate. Quantile regression, herding weak is also found in the lower quantiles. This reflects the superiority of the quantile, it can be more fully Local information can be detected locally, and the difference between the different quantile intervals can be rationally utilized. Investors can draw a better portfolio and increase the effectiveness of futures market.

【學(xué)位授予單位】：浙江工商大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：F724.5;F224

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 趙保國,甘茂智;市場壓力與羊群行為——基于中國證券市場實(shí)證分析的一個(gè)改進(jìn)[J];財(cái)經(jīng)問題研究;2005年11期

2 趙家敏,彭虹;我國證券投資基金羊群行為及其對股價(jià)影響的實(shí)證研究[J];系統(tǒng)工程;2004年07期

3 郭海星;萬迪f ;;期貨市場中的信息精度與羊群行為實(shí)驗(yàn)研究[J];系統(tǒng)工程;2010年10期

4 程慶生;;中國B股市場羊群效應(yīng)檢驗(yàn)[J];河南金融管理干部學(xué)院學(xué)報(bào);2008年01期

5 劉榮金;;我國股指期貨市場羊群行為實(shí)證研究[J];商業(yè)經(jīng)濟(jì);2011年24期

6 宋軍,吳沖鋒;基于分散度的金融市場的羊群行為研究[J];經(jīng)濟(jì)研究;2001年11期

7 孫培源,施東暉;基于CAPM的中國股市羊群行為研究——兼與宋軍、吳沖鋒先生商榷[J];經(jīng)濟(jì)研究;2002年02期

8 邵永同;高旺盛;;中國小麥期貨市場羊群行為實(shí)證研究[J];內(nèi)蒙古農(nóng)業(yè)大學(xué)學(xué)報(bào)(社會(huì)科學(xué)版);2009年03期

9 孫建全;陳小龍;;上證A股與B股市場的羊群效應(yīng)比較研究[J];山東財(cái)政學(xué)院學(xué)報(bào);2011年05期

10 邵永同;高旺盛;;基于CSAD的期貨市場羊群效應(yīng)實(shí)證分析——以CZCE棉花期貨為例[J];金融發(fā)展研究;2009年02期

，

本文編號(hào)：1365358