《上证综指对数收益率月度数据的特征分析》由会员分享，可在线阅读，更多相关《上证综指对数收益率月度数据的特征分析（9页珍藏版）》请在金锄头文库上搜索。
1、上证综指对数收益率月度数据的特征分析（1991.1-2014.9）目录1. 数据处理：收益率对数化2. 数据导入：3. 对数收益率特征分析（1） 简单描述性统计（2） 平稳性检验（3） 自相关性分析（4） 损益的不对称性（5） 分布的尖峰厚尾分析（6） 波动聚集效应检验4. 对数收益率可预测性分析（1） 短期（2） 中期（3） 长期一、 数据处理Lnrt=ln(1+rt*)其中rt*为普通收益率二、 数据导入FilenewworkfileDated-regular frequency; monthlyObjectnew objectgroupg1三、 对数收益率特征分析（1）简单描述性统计ln
2、rt窗口-viewdescriptive statistics & testshistogram & stats or stats tableMean0.010218Median0.006760Maximum1.019664Minimum-0.373282Std. Dev.0.130937Skewness2.438914Kurtosis21.00333Jarque-Bera4131.466Probability0.000000Sum2.912210Sum Sq. Dev.4.869011Observations285作图：viewgragh-line（2）平稳性检验ADF-testViewun
3、it root testlevelnone Null Hypothesis: LNRT has a unit rootExogenous: NoneLag Length: 0 (Automatic - based on SIC, maxlag=15)t-StatisticProb.*Augmented Dickey-Fuller test statistic-17.523360.0000Test critical values:1% level-2.5731305% level-1.94194510% level-1.615953*MacKinnon (1996) one-sided p-va
4、lues.Augmented Dickey-Fuller Test EquationDependent Variable: D(LNRT)Method: Least SquaresDate: 09/29/14 Time: 01:34Sample (adjusted): 1991M02 2014M09Included observations: 284 after adjustmentsVariableCoefficientStd. Errort-StatisticProb.LNRT(-1)-1.0410890.059411-17.523360.0000R-squared0.520394Mean
5、 dependent var0.000137Adjusted R-squared0.520394S.D. dependent var0.189813S.E. of regression0.131452Akaike info criterion-1.216828Sum squared resid4.890168Schwarz criterion-1.203980Log likelihood173.7896Hannan-Quinn criter.-1.211677Durbin-Watson stat1.992620（3）自相关性分析：ACF&PACFViewcorrelogramlevel36（4
6、）损益不对称作出分布图：viewgraghdistribution偏度：见（1）（5）尖峰厚尾见分布特征、偏度、峰度、及J-B检验结果（6）波动率聚集Genr vt=lnrt2检验vt序列的自相关：ACF相关系数检验相关系数VTVT1VT2VT3VT1.0000000.0647410.0498720.054298VT10.0647411.0000000.0646720.049843VT20.0498720.0646721.0000000.064714VT30.0542980.0498430.0647141.000000收益率可预测性分析短期：1、2、3、4ls lnrt c ar(1) ar(
7、2) ar(3) ar(4)与ls lnrt c lnrt(-1) lnrt(-2) lnrt(-3) lnrt(-4) 是一样的Dependent Variable: LNRTMethod: Least SquaresDate: 09/29/14 Time: 01:55Sample (adjusted): 1991M02 2014M09Included observations: 284 after adjustmentsConvergence achieved after 3 iterationsVariableCoefficientStd. Errort-StatisticProb.C0
8、.0101840.0074371.3693870.1720AR(1)-0.0473080.059496-0.7951590.4272R-squared0.002237Mean dependent var0.010190Adjusted R-squared-0.001301S.D. dependent var0.131167S.E. of regression0.131252Akaike info criterion-1.216375Sum squared resid4.858053Schwarz criterion-1.190678Log likelihood174.7252Hannan-Qu
9、inn criter.-1.206072F-statistic0.632278Durbin-Watson stat1.992328Prob(F-statistic)0.427190Inverted AR Roots-.05中期：6、9、12ls lnrt c ar(6) ar(9) ar(12)Dependent Variable: LNRTMethod: Least SquaresDate: 09/29/14 Time: 01:57Sample (adjusted): 1992M01 2014M09Included observations: 273 after adjustmentsCon
10、vergence achieved after 3 iterationsVariableCoefficientStd. Errort-StatisticProb.C0.0071490.0090960.7860240.4325AR(6)0.0071370.0587900.1214010.9035AR(9)0.2053360.0586043.5037720.0005AR(12)-0.0716970.058559-1.2243580.2219R-squared0.048799Mean dependent var0.007626Adjusted R-squared0.038191S.D. depend
11、ent var0.131559S.E. of regression0.129022Akaike info criterion-1.243117Sum squared resid4.477986Schwarz criterion-1.190231Log likelihood173.6855Hannan-Quinn criter.-1.221888F-statistic4.600115Durbin-Watson stat2.082147Prob(F-statistic)0.003701Inverted AR Roots.74+.08i.74-.08i.64-.58i.64+.58i.18-.85i
12、.18+.85i-.30-.68i-.30+.68i-.43+.60i-.43-.60i-.83+.27i-.83-.27iDependent Variable: LNRTMethod: Least SquaresDate: 09/29/14 Time: 02:00Sample (adjusted): 1991M10 2014M09Included observations: 276 after adjustmentsConvergence achieved after 3 iterationsVariableCoefficientStd. Errort-StatisticProb.C0.0090600.0097650.9278770.3543AR(9)0.2030650.0
