Table 5

Pooled OLS, fixed effects (FE), and random effects (RE) models

Variable/ItemPooled OLSFixed effects (FE)Random effects (RE)
Constant0.694** (0.031)1.183*** (0.006)0.903** (0.013)
ESG−0.006** (0.041)−0.007** (0.024)−0.007** (0.028)
LC−0.005*** (0.008)−0.004** (0.020)−0.005** (0.016)
BFS−0.003** (0.037)−0.003* (0.089)−0.003** (0.046)
SIZE−0.775*** (0.000)−0.693*** (0.004)−0.742*** (0.001)
CAP−0.028 (0.351)−0.045* (0.076)−0.049** (0.021)
GDP0.019 (0.242)−0.017 (0.328)−0.018 (0.298)
INF0.019 (0.493)0.036*** (0.010)0.038*** (0.007)
BC0.001 (0.772)−0.002 (0.715)−0.001 (0.942)
BNK−0.009 (0.528)−0.013 (0.334)−0.011 (0.421)
DCOVID−0.795** (0.042)−0.309* (0.087)−0.326* (0.069)
F stats18.12***12.67***14.01***
R-squared0.68100.53470.5786
Banks (i)626262
Panel data model statistical testsBreusch-Pagan LM: Pooled vs. RE: χ2(1) = 251.84***F-test: Pooled vs. FE: F = 12.67***Hausman test: FE vs. RE: χ2 = 1.215
Panel data model selectionReject H0: Pooled OLS inappropriate (panel effects exist)Reject H0: Pooled OLS inappropriate (unit effects exist)Fail to reject H0: No significant difference between FE and RE; thus, Random Effects (RE) is appropriate

Note(s): *, *, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. Numbers in parentheses are p-values

or Create an Account

Close Modal
Close Modal