Domestic saving and investment revised: Can the Feldstein-Horioka equation be used for policy analysis?

Based on the relation between investment and domestic saving proposed by Feldstein and Horioka (1980) to verify capital mobility, this study performs some exogeneity tests in order to determine the capacity of the FH equation of supporting and implementing economic policies in Brazil. We then use the result of weak exogeneity test to identify a structural vector autoregressive (SVAR) involving investment and saving in order to evaluate the effect of exogenous shocks through impulse response functions (IRFs) on both variables. The main findings of this paper are: a) the elasticity of domestic saving estimated using appropriate methods points out to high capital mobility for Brazil; b) domestic saving is weakly exogenous in the FH equation; c) domestic saving is not strongly exogenous, therefore this equation should not be used to make forecasts for the Brazilian economy; d) superexogeneity is accepted for domestic saving, meaning that Lucas' criticism does not apply; and e) the IRFs showed that investment is sensitive to contemporaneous innovation on saving and this effect lasts for a long time. Regarding to domestic saving, the response of this variable to a non-expected shock on investment has a more is more complicate description. Initially domestic saving goes down. After some lags this movement changes and domestic saving begins to react positively to the shock.