Alternative approaches for econometric modeling of panel data using mixture distributions

Journal of Statistical Distributions and Applications

Table 3 Estimates of the parameters of the components distributions for Models \(\mathcal {M}_{1}\), \(\mathcal {M}_{3}\), and \(\mathcal {M}_{8}\)

	CF _t ^a	CF _t−1	AS _t+1 ^b	AS _t	AS _t−1	sigma	π	ρ
Model \(\mathcal {M}_{1}\)
Component 1
Estimates	0.274	0.004	0.180	0.238	0.125	0.281	0.333	___
ste	0.012	0.003	0.055	0.061	0.055	0.003	0.005	___
Component 2
Estimates	0.113	0.016	0.023	0.042	0.027	0.073	0.667	___
ste	0.005	0.002	0.013	0.015	0.015	0.001	0.005	___
Model \(\mathcal {M}_{3}\)
Component 1
Estimates	0.136	0.040	0.024	0.049	0.032	0.071		0.031
ste	0.006	0.004	0.013	0.020	0.018	0.001		0.050
Component 2
Estimates	0.231	0.006	0.143	0.156	0.145	0.281		-0.146
ste	0.010	0.003	0.055	0.052	0.051	0.003		0.050
Model \(\mathcal {M}_{8}\)
Component 1
Estimates	0.225	0.004	0.137	0.133	0.124	0.291	0.491	0.447
ste	0.012	0.003	0.257	0.577	0.571	0.006	0.387	0.126
Component 2
Estimates	0.139	0.037	0.024	0.049	0.036	0.072	0.509	0.293
ste	0.006	0.005	0.022	0.033	0.033	0.001	0.387	0.093

^aCash flows
^bAsset Sales at time t+1
Notes: For each component, π is the prior probability of belonging to the component, σ is volatility of the change in investment for firms belonging to the component, and ρ is the correlation coefficient between the change in investment and financial status.
The vector of explanatory variables does not include lags of the dependent variable, but includes time dummies and other control variables whose coefficients are not reported to save space. The dependent variable is the first difference of investment-to-capital ratio