Hi,

We are writing our master thesis on the effect of capital requirements on bank lending, using two different dependent variables. The first one is growth in loans to household, which runs smoothly with no problems arising. However, when we run a Pooled OLS regression on the ratio of growth in loans to households relative to growth in loans to corporations, we get an insignificant model. We have tried to include dummies for each time period, but this does not change the significance of the model. The intuition behind this ratio is to examine how banks substitute their lending towards households and corporations.

Do anyone know reasons why the model might be insignificant? And how to potentially solve it? I have tried a fixed effect and random effect model, but these do not work either.

Attached is a photo of our Stata output.



Thank you in advance.

Kind Regards,

Jørgen Beltestad