May 8th, 2008

1. Creeping Inflation: How much of a concern? What should the response be? XXVII Meeting of the Latin American Network of Central Banks and Finance Ministries Alberto Torres Banco de México May 8th, 2008

2. Outline • 1. Real Food and Oil Price Changes • 2. Implications for Monetary Policy • 3. Additional Considerations

3. 1. Real Food and Oil Price Changes Commodity Price Indexes (Index Jan 2003 = 100) Source: IMF.

4. 1. Real Food and Oil Price Changes Real Food Price Changes (Annual % Change of Real Price) Source: IMF. Calculations by Banco de México.

5. 1. Real Food and Oil Price Changes Distribution of Real Food Price Changes (Annual % Change of Real Price) 40 20 0 -20 Jan.81-Nov.85 Dec.85-Jan.97 Feb.97-Dec.99 Jan.00-Aug.02 Sep.02-Mar.08 Source: IMF. Calculations by Banco de México.

6. 1. Real Food and Oil Price Changes Real Food Price Changes (Proportion of Time Above Threshold) Source: IMF. Calculations by Banco de México.

7. 1. Real Food and Oil Price Changes Real Food Price Changes (Mean Above Threshold) Source: IMF. Calculations by Banco de México.

8. 1. Real Food and Oil Price Changes Real Crude Oil Price Changes (Annual % Change of Real Price) Source: IMF. Calculations by Banco de México.

9. 150 100 50 0 -50 Jan.81-Jan.87 Feb.87-Jul.91 Aug.91-Mar.99 Abr.99-Mar.08 1. Real Food and Oil Price Changes Distribution of Real Crude Oil Price Changes (Annual % Change of Real Price) Source: IMF. Calculations by Banco de México.

10. 1. Real Food and Oil Price Changes Food (Annual % Change of Real Price) Crude Oil (Annual % Change of Real Price)

11. Outline • 1. Real Food and Oil Price Changes • 2. Implications for Monetary Policy • 3. Additional Considerations

12. 2. Implications for Monetary Policy • As is well-known in the economic literature (e.g. Clarida, Gali and Gertler, 1999; Walsh, 2003; Woodford, 2003; and others) the optimal monetary policy response to any shock is the solution to a problem where: • The policy maker minimizes a quadratic loss function that reflects society’s preferences for inflation and output stabilization. • This optimization is made subject to the structure of the economy, which is typically assumed linear (or log-linear). • In addition, shocks are assumed to follow an AR(1) process with white noise disturbances. • The solution to this problem provides an optimal feedback rule for the monetary policy instrument which will be a linear, time-invariant function of all the variables and shocks that form part of the economy.

13. 2. Implications for Monetary Policy • However, we are not currently in this type of framework: • For many central banks, high inflation may be more costly than low inflation. • The economy seems to behave differently in recessions than in booms (e.g., Hamilton (1989)). • The shocks we are currently facing, as shown above, follow a very complex stochastic process, with regime shifts, time varying variances and extreme values. • In addition, there is uncertainty about the duration of the “high inflation regimes” in commodity prices. • In this context, the optimal response of monetary policy is likely to be non-linear and time-varying, and will tend to focus on risk management.

14. 2. Implications for Monetary Policy • Following the risk management approach, the response to this environment should be preemptive, significant and transparent: • Failure to act timely may lead to a contamination of the price-setting process; in particular, may affect the formation of inflation expectations. • The response has to be strong, as central banks may prefer to insure against low probability, costly outcomes. • The central bank has to clearly explain the rationale for its policy actions. • But, as we will now see, how preemptive and strong the response should be, depends on the particular circumstances of each economy.

15. Uru Bol Ven Pan Nic Par Col Chi Ecu Mex Peru Arg Sal Gua Bra CR Jam 2. Implications for Monetary Policy 2005 – 2007 Inflation Change and Inflation Persistence 4 2 Sin HK Ire Ger Jap Por 0 UK Can Tai Ita Gre Mal USA Kor Spa Fra Change in Inflation from 2005 to 2007 -2 Tha Ind -4 Phi All Countries -6 Only LA -8 0.90 1.00 0.20 0.30 0.40 0.50 0.60 0.70 0.00 0.10 0.80 Inflation Persistence Source: WEO, IMF. Calculations by Banco de México.

16. 2. Implications for Monetary Policy 2005 – 2007 Inflation Change and Low Inflation Regime 4 Uru Bol Ven 2 Pan Sin HK Par Chi Nic Col Ger Ire UK Peru Ecu Ita Jap Por Can 0 Kor Tai Spa Fra Gre Mex Mal Arg Sal USA Change in Inflation from 2005 to 2007 -2 Gua Thai Bra -4 Ind Phil Bol Jam -6 All Countries Only LA -8 0 5 10 15 20 30 25 Duration of Low Inflation Periods (Years) Source: WEO, IMF. Calculations by Banco de México.

17. 2. Implications for Monetary Policy 2005 – 2007 Inflation Change and CPI Food Weights 4 Uru Bol Ven 2 Pan Chi Sin Nic Ire Ger HK Par Can Col Jap Mex Peru Ita Por UK Ecu 0 Kor Fra Tai Gre Mal Spa USA Arg Change in Inflation from 2005 to 2007 -2 Thai Gua Bra -4 Ind CR Phil All Countries -6 Jam Only LA -8 0 50 60 20 30 40 10 CPI Food Weight Source: WEO, IMF. Calculations by Banco de México.

18. 2. Implications for Monetary Policy Cumulative Change in Policy Target Rates and Inflation Persistence 300 HK Uru Col 200 Gua Gre Chi Ire Spa Peru 100 Ger Por Mex Kor Mal Ita 0 Jap UK Fra -100 Can Bra Ind Thai -200 Points) Cumulative Change in Policy Rate (Basis USA -300 -400 -500 All Countries -600 Only LA CR -700 0.9 0.2 0.3 0.4 0.5 0.6 0.0 0.1 0.7 0.8 1.0 Inflation Persistence Source: WEO, IMF. Calculations by Banco de México.

19. 2. Implications for Monetary Policy Cumulative Change in Policy Target Rates and Low Inflation Regime 300 HK Uru Col 200 Gua Chi Ire Fra 100 Ger Kor Ita Per Jap Gre Por Mex Spa 0 Mal UK -100 Can Bra Cumulative Change in Policy Rate Ind Tha (Basis Points) -200 USA -300 -400 -500 All Countries -600 Only LA CR -700 5 0 10 15 20 30 25 Duration of Low Inflation Periods (Years) Source: WEO, IMF. Calculations by Banco de México.

20. 2. Implications for Monetary Policy Cumulative Change in Policy Target Rates and CPI Food Weights 600 400 Uru HK Fra 200 Col Gua Kor Spa Ita Gre Chi Per Ger Mex Por Ire Jap 0 Mal UK Cumulative Change in Policy Rate (Basis Points) Can Ind -200 Tha Bra US -400 All Countries -600 Only LA CR -800 0 50 60 20 30 40 10 CPI Food Weight Source: WEO, IMF. Calculations by Banco de México.

21. 2. Implications for Monetary Policy • The optimal monetary policy response to “creeping inflation” depends on each country’s particular circumstances: • The inflationary history of the country and, in particular, the extent to which a low inflation equilibrium has been sustained. • The weight that food and energy goods have on the CPI. • The degree of persistence that each economy’s inflation process exhibits to price shocks. • The phase of the cycle the country is located in. • The impact of the current environment on the country’s terms of trade.

22. 2. Implications for Monetary Policy • The evidence suggests that the impact on inflation levels and inflation volatility of the worldwide shocks we have observed in the past years tends to be heterogeneous across countries. In particular: • Inflation seems to have responded more to the price shocks in countries that still exhibit a higher degree of inflation persistence. • Inflation volatility seems to be larger in those countries where the weight of foodstuffs in their consumer prices is higher, as well as in countries where the attainment of a low inflation equilibrium has been more recent.

23. 2. Implications for Monetary Policy • Thus, the same external shocks may imply different monetary responses in different countries, given their own specific characteristics. • In particular, countries that face higher risks of suffering large inflation increases and price volatility (due to their consumption baskets and their inflation history) may have found themselves in the need to respond more aggressively to the current price shocks than other countries. • Indeed, the evidence does suggest that countries that are more vulnerable in these terms have acted accordingly, by restricting their monetary policy stance.

24. 2. Implications for Monetary Policy • Given the evidence in the first part of this talk, a relevant question we should try to address with more research is: for how long will we be facing the current high commodity inflation regime? • Up to this point, the discussion has centered on the policy responses to the current environment, given the policy frameworks that the central banks currently have. • The answer to the question posed above could nonetheless have implications for the issue of whether central banks should rethink the definition of their monetary policy objectives.

25. Outline • 1. Real Food and Oil Price Changes • 2. Implications for Monetary Policy • 3. Additional Considerations

26. 3. Additional Considerations • Creeping inflation from the current shocks to energy and food prices may imply a restrictive bias in the monetary policy stance: • This is especially true in countries where inflation is persistent, the low inflation equilibrium has not been sustained for a long time and where food and energy have a large weight in consumers’ spending. • This, in turn, reflects the high costs that a breakdown of expectations-based nominal anchors that sustain the low-inflation equilibrium may have on these economies’ performance. • Indeed, some countries may face a combination of the above mentioned features that may lead them to maintain a restrictive bias in their monetary policy, even when they may be currently facing the low phase of their business cycle.

27. 3. Additional Considerations • The issues raised above focus on the implications of this high inflation regime for monetary policy. • We have to ask to what extent we need to consider a more broad view of the policy response that we should undertake, given this environment. In particular we need to ask: • Which other policy instruments are available to face the current environment? • What is the optimal mix of policy responses to this environment, in order to minimize negative effects on welfare?

28. 3. Additional Considerations • The policy mix clearly depends on the structure of each economy and on the shocks that each country has faced: • Those who faced an improvement in terms of trade may use specific policies to redistribute the aggregate gains of the shocks to those groups that may have been hurt from price changes. • Those that have faced a terms of trade deterioration also need to determine the optimal policy mix to minimize the welfare costs of this shock.

29. 3. Additional Considerations • Fiscal policy • May be helpful to redistribute losses (or gains, if terms of trade have improved), across different economic agents. • Competition policy • A non-competitive environment may exacerbate the impact of price shocks on domestic markets. Thus, antitrust agencies should avoid collusive practices in this environment of high price increases.

30. Appendix Real Crude Oil Price Changes (Proportion of Time Above Threshold) Source: IMF. Calculations by Banco de México.

31. Appendix Real Crude Oil Price Changes (Mean Above Threshold) Source: IMF. Calculations by Banco de México.