Paper presented at the 52nd CFCS Annual Meeting, Guadeloupe, July 10-16, 2016
Table 2: Results of Regression - Christophene Wholesale Market Price
Formation Dependent Variable: LOG(CRIS_P) Method: Robust Least Squares Date: 05/06/16 Time: 21:04 Sample (adjusted): 2006M02 2015M12 Included observations: 119 after adjustments Method: MM-estimation S settings: tuning=1.547645, breakdown=0.5, trials=200, subsmpl=3, refine=2, compare=5 M settings: weight=Bisquare, tuning=4.684 Random number generator: rng=kn, seed=1258528637 Huber Type I Standard Errors & Covariance Variable
Coefficient
Std. Error
z-Statistic
Prob.
C LOG(CRIS_P(-1)) LOG(CRIS_VOL)
7.468861 0.496279 -0.617789
0.956872 0.063031 0.085741
7.805500 7.873538 -7.205280
0.0000 0.0000 0.0000
Robust Statistics R-squared Rw-squared Akaike info criterion Deviance Rn-squared statistic
0.619599 0.726550 100.0528 10.13096 242.6742
Adjusted R-squared Adjust Rw-squared Schwarz criterion Scale Prob(Rn-squared stat.)
0.613040 0.726550 109.8222 0.325730 0.000000
Non-robust Statistics Mean dependent var S.E. of regression
2.402751 0.311472
S.D. dependent var Sum squared resid
* Significant at 1% level of significance ** Significant at 5% level of significance *** Significant at 10% level of significance
34
0.559302 11.25372