Skip to main content

Table 4 GMM system two-step estimates to see Non-Performing Loan with Loan Growth

From: Does abnormal lending behavior increase bank riskiness? Evidence from Turkey

Variables

Model 1

Model 2

Model 3

NPL i,t-1

0.7136a (0.0047)

0.7514a (0 .0458)

0.8124a (0.0357)

LG it

0.0451a (0 .0048)

0.0561a (0.0023)

0.0412a (0.0071)

Size it

−0.3187a (0.0648)

−0.6548a (0.0460)

− 0.6814a (0.0924)

LEV it

− 0.3147b (0.1531)

−0.2145c (0.1112)

− 0.1981a (0.0063)

EFF it

0.9891a (0.0786)

0.8954a (0.0258)

0.4871a (0.0569)

GDP t

−0.8791a (0.0784)

−0.7451a (0.0956)

− 0.5644a (0.0984)

INF t

−0.6889a (0.1245)

− 0.6541a (0.01478)

 

UNEMPL t

  

0.7843a (0.0658)

ROL t

−0.9156a (0.0613)

−0.6587a (0.0048)

 

POLST t

−0.8954a (0.0459)

 

−0.9546a (0.0148)

Observation

375

375

375

Sargan Test

0.795

0.864

0.826

AR1

0.048

0.089

0.079

AR2

0.946

0.756

0.265

  1. Note: Parentheses showing standard errors, a. b. c indicate 1%, 5% and 10% level of significant respectively. L_NPLit-1 represent the lag of non-performing loans those calculated by NPL to gross loans. LGit is loan growth collected from bankscope which is our main independent variable in the model that influence on the dependent variable; we also add Lag of loan growth to find previous year effect on non-performing loans. Sizeit denotes natural logarithm of total assets. LEVit stands for a leverage ratio that is calculated as total equity to total assets. We expected an adverse relationship between NPLs and leverage ratio. EFFit: non-interest expenses to total assets represent the efficiency ratio. NPLs will increase due to an increase in the provision of loan losses occur due to bad loaning. GDPt, INFt, UNEMPLt, ROLt, and POLSTt denotes gross domestic product, inflation, unemployment, rule of law and political stability respectively used as a macro (control variables) in the model for the purpose of most robust results of coefficients. For further descriptions of these variables, please see (Table 1 Main variables)