The compensation principle extends the notion of endogenous money and exogenous short-term interest rate to the open economy. Thus, it demonstrates that there is no operational constraint on the monetary authority's ability to set the interest rate and that the quantity of money in the economy is neither controlled by this institution nor altered by foreign reserve flows. Yet, it is not always the case that this principle is taken into account when limits to developing economies' monetary policy are under scrutiny, even in some branches of the heterodox literature.

In order to contribute to understanding the limits of the Brazilian monetary policy, this paper aims to empirically test the validity of the compensation principle in the country. By doing so, the paper suggests a renewed discussion on macroeconomic policy autonomy, thus indicating in a more precise manner the external constraints over this economy and the range of monetary policy possibilities.

Initially, a literature review is presented, in which both the theoretical and empirical discussions are summarized. Then, the empirical model is constructed based on data from the Brazilian Central Bank (BCB) for the period from 1994 to 2017. The sample consists of monthly data from the BCB balance sheet, such as the foreign reserves, Treasury securities, and bank reserves balance. As previously done in the literature for other economies, the compensation thesis is tested in two manners: the cointegration between the foreign reserves and Treasury securities and between the foreign reserves and bank reserves balance are estimated through an ARDL model. A positive cointegration between the foreign reserves and Treasury securities and no cointegration between the foreign reserves and bank reserves balance indicate that the compensation principle applies. Despite the managed float regime by the BCB since 1999, the period of 1994 to 1999 is tested as a robustness check, as in this period the exchange rate was pegged.

The preliminary results suggest that the compensation principle is valid in the Brazilian economy. This means that, in operational terms, foreign reserve flows affect neither the ability of the BCB to set the short-term interest rate nor the quantity of bank reserves. Moreover, the BCB is able to tame the fluctuations in the exchange rate, which are particularly intense in the country, without inevitable implications to the short-term interest rate as long as it is willing to act to keep the latter on its own established target. Therefore, if one is looking to understand the external influences on the developing countries' monetary policy, our conclusions suggest that these influences do not lie on technical operational limits to setting the interest rate, but rather on political economy issues.