Is monetary policy non-linear in Latin America? a quantile regression approach to Brazil, Chile, Mexico and Peru

There is a growing literature on nonlinear monetary policy in industrialized countries. These investigations have revealed that the reaction of central banks to economic variables depends on the level of such variables, confirming the nonlinearity of policy in the industrialized world. Researching whether monetary policy is nonlinear in Latin America, however, is hampered by the lack of data as stable monetary regimes have existed in many Latin American nations for only about fifteen years. We investigate monetary policy for four Latin American nations with quantile regression. Quantile regression does not require the additional parameters of traditional switching or threshold models; thus quantile regression can reveal nonlinearities without the long data span typically required in Markov Switching or Threshold estimation. Our results indicate that all of the nations display nonlinearities. For three, Chile, Mexico and Peru, policy becomes systematically tighter as the level of inflation rises, mirroring results found previously for the US and Japan.