Permanent income hypothesis in emerging markets : some Brazilian evidence

This paper uses a dataset from one of the largest credit card issuers in Brazil to investigate the extent of credit constraints for one of the world's largest emerging market economies. In order to deal with the endogeneity problem from an observational dataset, an identification strategy based on the issuer actual policies was used to better identify the long run effects of a credit card limit increase on debt. The results point to a long run effect of credit limit on credit card debts of 0.1144, a similar size to the ones found in Gross and Souleles (2001). This value indicates the Permanent Income Hypothesis is rejected for this sample, and some evidence supporting a variation of buffer stock behavior is found. Furthermore, results indicate the degree of heterogeneity in effects is much higher than similar results for developed countries, lending credence to the conclusion credit constraints are much stronger for lower income groups. This conclusion is in line to previous studies in Brazil, such as de Lucinda and Vieira (2014).