O impacto econômico do fim de um monopólio estatal: o caso do mercado de resseguros no Brasil

Abstract

The present study seeks to measure the impact of an exogenous shock on the supply of a reinsurance market on the penetration of the main insurance business, using the the case of Complementary Law 126/2007, which ended the monopoly of the Instituto de Reinsurance from Brazil (IRB). Economic theory indicates that the practical consequence of breaking the monopoly is the introduction of new reinsurers into the market, so that the increase in offer of reinsurers generates an indirect impact on the total offer in the direct market of insurance (BLAZENKO, 1986). The literature lists the two main transmission channels for the proposed relationship: the supply shock on prices (BLAZENKO, 1986) and the improvement of technical efficiency over marginal cost (ANAND; LEVERTY; WUNDER, 2020; COLLECTION; MCCULLOUGH, 2006). To carry out the proposed evaluation, we used the already established synthetic control. Additionally, as a way of checking the robustness of the results, the factorial method of Hsiao, Ching and Wan (2012) and the estimator of difference in differences to construct counterfactuals for insurance penetration in the Brazilian product. Still, in order to use an adequate method of selection of the best comparison model for Brazil without the intervention of the deconcentration policy market, we adopted the criteria suggested by Ferman, Pinto and Possebom (2016) to the synthetic control and by Li and Bell (2017) for the panel estimator. The results found corroborate our hypotheses discussed in the theoretical section and show that there was significant increase in insurance activity in Brazil from 2008 to 2016. models estimated through the synthetic control method account for an effect positive impact on insurance penetration in the order of 42% to 52%. The other models estimated by the difference-in-differences and factorial methods of Hsiao, Ching and Wan (2012) also had positive results, but with a very high variance of results. higher, between 13% and 68%.