We investigate the macroeconomic impacts of changes in capital adequacy requirements, as developed in the Basel Capital Accords, on Brazil and Mexico. Changes in the capital adequacy requirements of international and domestic banks are considered, since the former adopted the Basel Capital Accord in 1988 and the latter in the mid-90s. Unlike most papers in the budding literature on the effects of the Basel Capital Accords on developing countries, we adopt an empirical approach, grounded in a general equilibrium macroeconometric model, which allows us to examine indirect transmission mechanisms. We first estimate a reduced financial block for Brazil and Mexico, which we integrate into the National Institute's General Equilibrium Model (NiGEM). We then simulate a shock to domestic and international capital adequacy ratios.
The simulations show that an increase in capital adequacy ratios-either domestic or international-has adverse impacts on Brazilian and Mexican GDPs. A moderate credit crunch occurs in both cases and in both countries and is accompanied by a rise in lending rates. However, there are important differences in banks' reaction to tighter solvency ratios in each country. In Brazil, international and domestic banks adjust their portfolios by switching from higher-risk loans (private sector) to zero-risk loans (sovereign and public sector), instead of increasing their capital provisions. Sovereign lending, and hence government spending, thus rises sharply in Brazil. This offsets the negative impacts of the fall in private investment that follows the credit crunch. In Mexico, sovereign lending from domestic banks remains largely unaffected by changes in capital adequacy ratios, whereas foreign loans to the Mexican public sector decrease. In both cases, the Mexican private sector bears the bulk of the adjustment of domestic and foreign banks to the new regulatory rules. These findings suggest the existence of a financial "crowding-out", where government borrowing replaces private sector borrowing in domestic banks loans portfolios.
Household borrowing including housing loans represents around 5 per cent of GDP in Mexico and about 8 per cent of GDP in Brazil, on average over 1997-2004. These ratios are considerably lower than those of countries such as the UK and the US. In 2000, for instance, total consumer credit in the UK and the US amounted to 73 and 78 per cent of GDP respectively (See Byrne and Davis 2003). This may account for our finding that consumer credit in both countries is not sensitive to changes in solvency ratios. Nonetheless, our simulations show that household consumption in Brazil and Mexico drops following a rise in capital adequacy ratios. The transmission mechanism is carried out through household net wealth. Higher solvency ratios lead to higher interest rates, which, other things unchanged, increase net interest payments of households and thus their net financial wealth. In our model, lower financial wealth results into lower consumption. Overall, given an increase of _ percentage point in solvency ratios, we found that GDP falls by 3.5 per cent in Brazil, and by 2.2 per cent in Mexico.