噪聲交易者風(fēng)險存在性與補(bǔ)償性

1、噪聲交易者風(fēng)險存在性與補(bǔ)償性：基于中國封閉式基金市場的實(shí)證分析Existence and Pricing of Noise Trader Risk:Empirical Analysis of Closed-end Fund Market in China王建軍指導(dǎo)教師姓名： 任宇助理教授專業(yè)名稱： 金融學(xué)摘要噪聲交易能夠增加資產(chǎn)的流動性，但是噪聲交易過度會造成資產(chǎn)價格的過分波動。如果一個金融市場中的噪聲交易泛濫，那么該金融市場中的投資者在投資交易時將可能面臨一種新的風(fēng)險，即噪聲交易者風(fēng)險。本文以中國封閉式基金市場為研究對象，探討了這種新風(fēng)險在中國金融市場的存在性和補(bǔ)償性。統(tǒng)計數(shù)據(jù)顯示，在200

2、2年前中國封閉式基金持有人以個人投資者為主，在2002年后機(jī)構(gòu)投資者迅速成為新的主要持有人，而且滬市和深市封閉式基金市場的換手率、貼水率和持有人結(jié)構(gòu)都存在一定的差異。因此本文在實(shí)證分析時把封閉式基金市場分為四部分，1999年1月至2001年12月的滬市和深市封閉式基金市場，以及2003年1月至2009年8月的滬市和深市封閉式基金市場。為了分析噪聲交易者風(fēng)險的存在性，本文先引用GJR-GARCH(1,1)-M模型檢驗了封閉式基金市場是否支持時變理性假說，然后利用封閉式基金樣本檢驗了基金價格的過度波動現(xiàn)象和均值回歸效應(yīng)。兩個實(shí)證結(jié)果顯示，第一，雖然兩個時間段的封閉式基金持有人結(jié)構(gòu)有顯著的差別，但是

3、兩個時間段的封閉式基金市場都存在噪聲交易者風(fēng)險；第二，2003年1月至2009年8月的滬深封閉式基金價格的過度波動現(xiàn)象比1999年1月至2001年12月更為顯著，這意味著機(jī)構(gòu)投資者可能存在非理性交易行為；第三，同一時間段的滬市和深市封閉式基金市場的檢驗結(jié)果無明顯差異。在實(shí)證分析的最后部分，基于風(fēng)險補(bǔ)償與收益率之間的關(guān)系，本文探討了噪聲交易者風(fēng)險的定價問題。結(jié)果顯示，兩個時間段的滬深兩市封閉式基金市場的投資者沒有因為承受噪聲交易者風(fēng)險而獲得相應(yīng)的風(fēng)險補(bǔ)償。關(guān)鍵詞：噪聲交易；封閉式基金；GJR-GARCH-MAbstractNoise Trade increases the liquidity o

4、f financial assets, and also the volatility of assetsprice. Theoretically, under some conditions, Noise trade will create a new risk in financial market. The risk is called noise trader risk. In this thesis, existence and pricing of the risk are discussed with closed-end fund market in China. The da

5、ta shows thatthe individual-holding ratio, exchange ratio and discount ratio of closed-end fund are significantly different between the period from 1999.1 to 2001.12 and the period from 2003.1 to 2009.8, and the three ratios also different between closed-end fund in Shanghai stock market and in Shen

6、zhen stock market. Hence, the data of closed-end fund used in the thesis is divided into four parts. The four parts are the data of closed-end fund in Shanghai stock market and in Shenzhen stock market from 1999.1 to 2001.12,and the data of closed-end fund in Shanghai stock market and in Shenzhen st

7、ock market from 2003.1 to 2009.8. As for the discussion about the existence of noise trader risk,the model GJR-GARCH(1,1)-M is used to test the time-varying rational expectation in four closed-end fund market, the testing of excess volatility and mean reversion of price too is used during the discus

8、sion. The results are as follows: first, there is noise trader risk in four closed-end fund markets; second, there is more excess volatility in the closed-end fund market from 2003.1 to 2009.8 than the period from 1999.1 to 2001.12,which means that institute investors are possible to do irrational trades; third, no significantdifference is found between the closed-end fund in Shanghai stock market and in Shenzhen stock market in the same period. In the last part of empirical analysis, based on the relationship between risk and return, pricing of noise trader risk is discussed. The resu