VoxEU Column Macroeconomic policy Monetary Policy

Monetary policy in Latin America: Where are we going?

Latin American central banks are facing new challenges in the form of unprecedented levels of uncertainty and exchange rate appreciation pressures. This column, focusing on Brazil, Chile, Colombia, Peru and Mexico, argues that there is an overestimation of the potential output in several Latin American economies, a lack of an explicit policy direction from central banks, and lacklustre frameworks for macroprudential policy. Although inflation targeting has served countries in Latin America well, significant risks remain.

Inflation targeting has served countries in Latin America well . They have achieved macroeconomic stability by reducing inflation and the pass-through of external shocks such as oil price and exchange rate fluctuations (cf. Mishkin and Schmidt-Hebbel 2007).

Nevertheless, central banks in the region have recently faced new challenges, including unprecedented levels of uncertainty in the global economy and recurring exchange rate appreciation pressures. Some evidence indicates that not all of them have graduated completely from procyclicality (Vegh and Vuletin 2012). In this column we discuss the current policy stance in five Latin American inflation-targeting countries – Brazil, Chile, Colombia, Peru and Mexico – and analyse some of the challenges central banks currently face in the region.

Recent trends in inflation and monetary policy

While countercyclical fiscal policy has received most attention during and after the 2009 collapse in global demand (cf. Daude et al. 2010 and their references), monetary policy has recently gained some attention from market and policy analysts. In 2009, central banks were able to use monetary policy in a countercyclical way, lowering policy interest rates at a similar pace as their advanced OECD economies’ peers. By mid-2010, Brazil, Chile and Peru had already started a tightening cycle, followed by Colombia in early 2011. The exception was Mexico, where the severity of the output collapse justified a longer period of accommodative policies.

This swift move to withdraw the monetary stimulus was in part driven by the need of central banks to re-establish credibility after inflationary spikes in 2008 pushed both headline inflation and inflation expectations beyond their target and comfort zones. While headline inflation and core inflation came down very rapidly, especially in Chile and Peru due largely to the collapse in commodity prices, inflation expectations – in Mexico and Colombia in particular – converged far more slowly towards target zones. And by mid-2009, inflation expectations started to rise again in most countries (see Figure 1).

A delicate balancing act

More recently, central banks in the region have been trying to find a delicate balance. Supply shocks and relatively high commodity prices have pushed headline inflation, while closing output gaps have been calling for a tighter policy stance, with fiscal policy remaining accommodative in most countries (Daude and Melguizo 2012). Consequently, it has often been only monetary policy that has had to do the job. At the same time, expansive monetary policies in advanced OECD economies combined with high interest rate differentials have pushed many Latin American currencies to appreciate. Thus, central banks have been facing the dilemma to raise interest rates or stop foreign exchange interventions and let the currency appreciate, which would reduce inflation but also harm competitiveness, or risk higher headline inflation and probably some loss of credibility (Lora and Powell 2011).

Figure 1. Headline inflation, core inflation and inflation expectations

The monetary policy stance: Some evidence based on Taylor rules

The most common approach to evaluate whether monetary policy is accommodative or tight is to estimate a monetary policy reaction function based on the Taylor rule that relates the policy interest rate to the inflation gap with respect to the central bank’s inflation target and the output gap (Taylor 1993). We have done so by considering a series of alternative models using inflation expectations (over the next 12 months), core inflation measures as well as headline inflation, various measures of the output gap based on monthly business cycle indicators (such as the IMACEC in Chile), including additional regressors as the real exchange rate, and several econometric specifications (including or not the lagged policy rate to account for policy smoothing motives and considering alternative estimation methods such as DOLS or GMM). The sample period goes from January 2003 to October 2012.

Figure 2. Range of predicted policy rates by Taylor rules and current rates

Our results

The results show the following:

  • For Chile and Peru all point estimates of the predicted interest rate are below the current policy rate, although they fall into the confidence intervals at standard levels of significance.

Thus, the current policy stance in these countries is slightly tight to neutral.

  • In the cases of Colombia and Mexico, monetary policy is currently within the interval of predicted levels by our estimates of Taylor rules.

Please note that this does not account for the most recent decisions in November.

  • The estimates for Brazil show that there is a high variability in the estimation of the policy rate, with point estimates varying between 7.8% and 15.1% for October 2012.

All estimates are above the current policy rate of 7.25%. Thus, monetary policy according to these estimates is between accommodative or highly accommodative.

The above estimates are in line with those presented in the latest IMF’s Regional Economic Outlook (see Magud and Tsounta 2012) and consistent also with the findings by Hofmann and Bogdanova (2012) regarding the recent divergence from Taylor rules in emerging markets, including Brazil.

Figure 3. Observed and fitted policy rates in Chile and Brazil

Taken with a pinch of salt

Of course, all these estimates should be taken with a pinch of salt, given the short samples and measurement issues (e.g. output gaps are not computed in real time). The fit for Brazil is particularly weak. Figure 3 compares the predicted and observed policy rates for Chile and Brazil using a dynamic OLS estimation, considering inflation expectations, output gaps and also the real exchange rate. Whereas Chile’s policy seems to fit the estimated Taylor rule pretty well, this is not the case for Brazil. In particular, it is interesting to observe that since end-2011 there seems to be a particular divergence between the estimated rule and the SELIC rate. One explanation could of course be that the central bank just has more or better information regarding the trajectories of output and inflation. Alternatively, there seems to be some indication that inflation is not the only objective as concerns about the exchange rate have been evident, with market participants trying to infer the BRA-USD bands that trigger foreign exchange intervention. In this context other instruments, macroprudential as well as some tightening of credit standards and fiscal policy, have been used to control inflationary pressures.

Some challenges ahead

In the short term, the baseline scenario for the rest of 2012 and early 2013 points towards more moderation of inflationary pressures in Latin America, but risks remain. In particular, there is a risk of overestimating potential output in several Latin American economies. The last decade has been extraordinarily favourable in terms of external conditions for the region but this has not necessarily translated into structural improvements. Structural policies and reforms that increase potential output via productivity gains – infrastructure, human capital and innovation – are still needed (OECD-ECLAC 2012).

For central banks, a particular problem is having various targets without an explicit policy framework regarding what these targets are and which instruments should be used to achieve them. Is financial stability, leverage or credit growth a concern? Does foreign exchange intervention aim at reducing excessive foreign exchange volatility or is there a level target? The lack of definition regarding these issues creates uncertainty that could prove harmful. In terms of communication, Turkey might be an interesting example of a country that, after struggling with it for quite some time, has recently achieved better outcomes.

In the absence of international policy coordination regarding capital flows, national macroprudential policies have a rationale for existence, but more work is needed in terms of understanding the interaction (complementarities and trade-offs) with standard monetary policy tools and defining an appropriate framework for their use.


Daude, C and A Melguizo (2012), “Fiscal policy in Latin America: How much room for manoeuvre?”, VoxEU.org, 11 September.

Daude, C, A Melguizo and A Neut (2010) “Fiscal policy in Latin America: Better after all?” VoxEU.org, 8 October.

Hofmann, B And B Bogdanova (2012), “Taylor rules and monetary policy: a global ‘Great Deviation’?”, BIS Quarterly Review, September, 37-49.

Lora, E and A Powell (2011) “More inflation or more revaluation”, VoxLACEA.org, 11 May.

Magud, N and E Tsounta (2012), “Policy Interest Rates in Latin America: Moving to Neutral?”.

Mishkin, F S and K Schmidt-Hebbel (2007), “Does Inflation Targeting Make a Difference?”, NBER Working Papers, 12876, National Bureau of Economic Research.

OECD-ECLAC (2012), “Latin American Economic Outlook 2013: SME policies for structural change”, OECD Development Centre, Paris.

Vegh, C A and G Vuletin (2012), “Graduation from monetary policy procyclicality”, VoxEU.org, 22 August.

5,039 Reads