Nonlinear dynamic path on the performance persistence of banks in Taiwan: A panel smooth transition regression approach

This research adopts a panel smooth transition regression (PSTR) model and uses the average interest rate differential (IRD) of the top five banks in Taiwan as a transition variable to investigate the threshold effect on constant and variable return to scales. Empirical results show that the operative efficiency value of the financial industry changes nonlinearly with time in different regimes. The results also prove the persistent high operative performance of the financial industry. The findings reveal further that these industries are operatively stable compared with the non-financial ones, and are less influenced by monetary policies or exogenous factors. Finally, the threshold value of IRD is determined as the key factor that affects the structural change in the operation outcomes of banks. In brief, these results provide authorities with reference to modify policies or make efficient decisions.