我通过Eviews软件对一组时间序列数据进行ARCH测试,得到结果如下
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | 0.980086 | 0.066359 | 14.76951 | 0 |
STD_RESID^2(-1) | -0.038851 | 0.0233 | -1.667451 | 0.0956 |
STD_RESID^2(-2) | 0.031484 | 0.023313 | 1.350467 | 0.177 |
STD_RESID^2(-3) | 0.033164 | 0.023309 | 1.422801 | 0.155 |
STD_RESID^2(-4) | -0.017047 | 0.023311 | -0.731298 | 0.4647 |
STD_RESID^2(-5) | 0.01313 | 0.023296 | 0.563608 | 0.5731 |
R-squared | 0.004142 | Mean dependent var | 1.002047 | |
Adjusted R-squared | 0.001439 | S.D. dependent var | 1.751613 | |
S.E. of regression | 1.750353 | Akaike info criterion | 3.960753 | |
Sum squared resid | 5643.4 | Schwarz criterion | 3.978681 | |
Log likelihood | -3653.736 | F-statistic | 1.532159 | |
Durbin-Watson stat | 1.999738 | Prob(F-statistic) | 0.17654 |
不知是否存在ARCH效应?