|
Table 4: Empirical Results across Different CAMEL-based Classifications 1/ 2/
Variable Name |
Lags 3/ |
Benchmark (1) |
Good Banks (always 1 or 2) (2) |
Bad Banks (if ever 3, 4 or 5) (3) |
Very Bad Banks (if ever 4 or 5) (4) |
Uninsured Deposits (% of total deposits) |
15.474 (13.508) |
15.161 (13.452) |
16.185 (13.608) |
16.676 (13.974) |
UninsDep |
L1D |
0.382*** (0.064) |
0.349*** (0.074) |
0.376*** (0.093) |
0.426*** (0.111) |
L2D |
0.025 (0.036) |
0.028 (0.045) |
0.005 (0.040) |
-0.065 (0.070) |
L3D |
-0.002 (0.030) |
-0.027 (0.040) |
0.036 (0.028) |
0.041 (0.058) |
UninsIntMarg |
D |
0.880* (0.497) |
1.292* (0.746) |
0.162 (0.205) |
0.698 (0.680) |
Size |
D |
9.721*** (2.340) |
8.549*** (3.251) |
10.581*** (2.849) |
13.066*** (4.022) |
Loans & Leases |
L1D |
0.038* (0.021) |
0.067** (0.030) |
-0.008 (0.034) |
-0.043 (0.054) |
ResidLoans |
L1D |
0.061** (0.031) |
0.039 (0.039) |
0.093* (0.050) |
0.195* (0.106) |
Equity |
L1D |
0.445*** (0.133) |
0.439** (0.180) |
0.386*** (0.126) |
0.626*** (0.205) |
NonCLoans |
L1D |
-0.046 (0.040) |
-0.094 (0.061) |
-0.038 (0.052) |
-0.014 (0.069) |
GDP |
D |
0.245*** (0.047) |
0.257*** (0.059) |
0.205*** (0.061) |
0.132 (0.128) |
L1D |
0.090* (0.049) |
0.141** (0.061) |
-0.051 (0.069) |
-0.145 (0.164) |
Infl |
D |
0.599*** (0.187) |
0.908*** (0.245) |
0.140 (0.217) |
-0.352 (0.555) |
Treas_1 |
D |
-0.474*** (0.146) |
-0.716*** (0.200) |
0.031 (0.107) |
0.283 (0.313) |
Cons |
-0.048 (0.124) |
0.004 (0.174) |
-0.010 (0.128) |
-0.083 (0.209) |
Observations |
12826 |
8912 |
3914 |
1380 |
Banks |
1863 |
1307 |
556 |
207 |
Test of no AR(1) 4/ |
0.000 |
0.000 |
0.000 |
0.007 |
Test of no AR(2) 4/ |
0.597 |
0.838 |
0.456 |
0.340 |
Sargan test 5/ |
0.146 |
0.032 |
0.095 |
0.318 |
1/ Model with endogenous price mechanism.
2/ Coefficients from first-step robust estimations, unless specified otherwise.
3/ LiD = in first difference with i lag(s); and D = in first difference.
4/ AB(i) is the Arellano and Bond test of no autocorrelation of degree i.
5/ Sargan test is based on two-step regression results.
Back to document
|