The Taylor rule is a simple formula that can be used for an evaluation of the stance of monetary policy in any country. The Taylor rule is usually presented in terms of the nominal interest rate. However, for Latin American countries with histories of high inflation it is more practical to represent the same Taylor rule but in the real interest rate:
real interest rate = real natural interest rate + 0.5 * inflation gap + 0.5 * output gap
Graph 1 shows the real policy interest rate compared with the one prescribed by the Taylor rule formula for four Latin American countries which are now inflation targeters. A band of plus and minus two percentage points has been added to allow for uncertainty.
We can see in the graph that the observed real policy rate deviates from the prescription of the real Taylor rule especially at times of crisis, for example, at the end of the century and at the current global financial crisis. Comparing these two episodes we can see that the direction of the deviation has changed – this is the “sea change.” At the turn of the century the policy rate was above the one prescribed by the real Taylor rule; in contrast, during the current global financial crisis the real policy rates are heading below the real Taylor rule. In other words, at the turn of the century monetary policy was very pro cyclical, today it intends to be countercyclical, even in real terms.
One reason for this sea change is less fear of floating. During the last big regional crisis of the turn of the century, many Latin American countries had crawling bands that initially they decided to defend. Within today’s inflation targeting regimes, exchange rates float, hence, monetary policy need not be pro cyclical. Also, as a consequence of exchange rate rigidity, then Latin American countries went into that earlier crisis with more exposure to short-term, foreign currency denominated external debt, and that limited the amount of depreciation that they could allow. (This is why we are seeing procyclical monetary policy in Eastern Europe now). Furthermore, the inflation targeting central banks in the region have probably become more forward looking. This means that they respond more to the forecasted effect of the crisis on inflation into the near future and less to actual inflation, which would be the case with the simple Taylor rule. Finally, according to flexible inflation targeting, an inflation targeting central bank may have some scope – especially at times of crisis – to lower rates to anticipate and react to future recessions.
But one thing is the stance of the policy interest rate and another is the transmission of this stance to those interest rates that are important for economic activity. The risk structure of interest rates states that interest rates on loans are the sum of a risk free rate which is the policy interest rate, plus a default risk premium, plus a term premium. A decrease in the real policy rate may be offset by an increase in credit risk spreads, in the premium for sovereign risk and other sources of credit risk. Graph 2 shows that, with the exception of Peru, the spread between the interest rate on loans and the policy rate has been increasing. The interest rates are on different types of private sector borrowing so spreads are roughly comparable between countries over time, although not in terms of levels. In the cases of Brazil, Chile and Mexico, the increase in credit risk spreads may offset the current counter cyclical stance that monetary policy makers are taking.
The conclusion is that, despite the sea change, Latin American monetary policy may still need to take the risk structure of interest rates into account. The appropriate monetary policy could be conditional on how these risks are expected to evolve, in a way not envisaged by the simple Taylor rule.