Taxation and Corporate Debt: Are Banks any Different?

IMF Working Papers, October 29, 2013

Source details

The paper investigates whether corporate tax bias toward debt finance differs between banks and nonbanks using a large panel of micro data.
On average, there is no significant difference between banks and non-banks.
The marginal tax effect for both banks and non-banks is close to 0.2.
Responsiveness to tax incentives differs considerably across the asset size distribution and the conditional leverage distribution.

Nonbanks:
- Exhibit a U-shaped relationship between asset size and tax responsiveness.
- The U-shaped pattern does not hold universally across the conditional leverage distribution.
Banks:
- Tax responsiveness declines linearly in asset size.
- Quantile regressions indicate capital-tight banks are significantly less responsive than capital-abundant banks.
- Even the largest banks with high conditional leverage ratios still display a significant, positive tax response.

Quantile regressions reveal heterogeneity in tax responsiveness across the conditional leverage distribution:
- Capital-tight banks: significantly less responsive to tax incentives.
- Capital-abundant banks: more responsive to tax incentives.
- The same pattern of lower responsiveness for capital-tight firms holds for the largest non-banks.
Despite lower responsiveness in capital-tight segments, significant positive tax responses remain for the largest banks with high conditional leverage ratios.

Subjects: Bank levy, Banking, Collateral, Corporate income tax, Debt bias, Financial institutions, Inflation, Prices, Tax policy, Taxes
Keywords: bank, bank assets, bank choice, bank dummy, Bank levy, banks, capital structure regression, Collateral, Corporate income tax, Corporate tax, debt, debt bias, Global, Inflation, leverage, leverage ratio, non-financial firms, quantile, quantile regressions, size distribution, WP