Comments on Chapters 5 and 6

Adrián Armas, Eduardo Levy Yeyati, and Alain Ize
Published Date:
July 2006
  • ShareShare
Show Summary Details
Klaus Schmidt-Hebbel

Dollarization presents a daunting task to central banks in many emerging market economies – all those countries where domestic money coexists with a foreign currency to which the domestic population had fled in the past because of domestic monetary mismanagement. Monetary authorities in dollarized economies face a dual challenge. First, they have to conduct monetary policy under the continuous threat of large portfolio shifts between both currencies. Second, they have to bring about monetary stability that is similar to or greater than that of the competing foreign currency, in order to convince their local populations to gradually shift back to using and holding their domestic currency again.

In the pursuit of the latter objectives, central banks in dollarized economies are aided by the fact that much of domestic income is in local currency (or in non-tradable sectors), reinforcing the demand of local currency for transactions and savings. Moreover, high exchange-rate volatility and low inflation (volatility) strengthen the demand for a stable domestic currency that is used for carrying out transactions and building up domestic assets that match income, consumption and balance sheets linked to the domestic economy. Hence a related challenge faced by central banks in dollarized economies is to let the exchange rate float (freely), make their monetary policy more independent of exchange rate shocks and support market development of financial instruments to hedge exchange rate risk.

The two chapters on this subject on which I am about to comment are closely related. Leiderman, Maino and Parrado (Chapter 5) assess differences in monetary policies and outcomes between dollarized and non-dollarized Latin American economies, with and without inflation targeting, with some stronger focus on Peru. Armas and Grippa (Chapter 6) present a full-blown case study of Peru’s unique experience, hitherto the only significantly dollarized economy that has adopted IT. After commenting on each of the two chapters, I will briefly discuss the more general issue of Tear of floating’ among IT countries in Latin America and end with some remarks on the Peruvian experience.

Chapter 5

This excellent chapter provides a unique analysis of the challenge faced by current (Peru) and prospective (Bolivia) IT countries in the face of large financial dollarization, in comparison with non-dollarized inflation targeters in South America (Chile and Colombia). In doing this, the chapter provides new empirical evidence on the interplay between monetary regimes (IT and exchange rate targeting), financial structure (dollarization, financial fragility), monetary policy and inflation. The cross-country and cross-regime evidence provides underpinnings to the notion that dollarization and financial stress are endogenous to monetary regimes and monetary policy performance (inflation) and that – the chapter’s main policy inference – ‘high dollarization per se does not preclude the use of IT as an effective policy regime’.

Although the authors justify ‘leaning against-the-wind’ interventions as consistent with – even as a strengthening device of – IT, I will hold a different point of view. Exchange rate (ER) interventions provide implicit insurance against exchange rate volatility and risk, and therefore inhibit de-dollarization and market development of ER risk hedging instruments, leading to larger currency imbalances in balance sheets and hence to higher financial fragility and stronger contractionary effects of ER depreciations. By signalling central bank concern for the ER, such interventions potentially weaken the credibility of the inflation target. Moreover, the political economy of public support of ER interventions leads to pressures for one-sided interventions (when the domestic currency is perceived as over-appreciated). Finally, the effectiveness of interventions ranges from small and temporary to dubious and nil (e.g., Tapia and Tokman [2004] on Chile’s experience).

My final comments on this chapter are on the monetary policy reaction functions. Its ad hoc specification makes it hard to infer which is the implied neutral policy rate – literally, a variable that combines the constant and the terms involving the Federal Funds rate and (if non-zero) the trend changes in the real effective exchange rate and in nominal net international reserves. It is also not clear why Bolivia’s rate of ER crawl should respond to the same variables as those specified for the monetary policy reaction function for the three IT countries, and which should be the expected signs on their coefficients. An expanded set of results for alternative specifications and data samples may be valuable to convince the reader about the robustness of the reported results.

Chapter 6

This excellent chapter reports new evidence and an insightful analysis of monetary policy in Peru. Presented by two important insiders, they should be commended for their balanced and yet insightful discussion of policies and results of the BCRP.

The chapter documents the triple transition led by the BCRP from partial to full-fledged IT, from fear of floating to less fear, and from dollarization towards what I call ‘soli(di)fication’ (i.e., de-dollarization or making the nuevo sol more solid and attractive to local money holders). The authors also document BCRP’s internal projection model, which should eventually be published. Finally, the paper discusses the remaining policy challenges.

The significant progress in Peru’s market-based financial soli(di)fication is particularly surprising because it has not been based on mandatory de-dollarization (like, for example, Argentina’s) but is a voluntary shift back into sol financial assets, in response to successful monetary policy. This success is likely to be the result of BCRP’s shift from money to the interest rate as its monetary policy instrument (or operational target), the achievement of low inflation and small target misses since 2000, a strong anchoring of inflation expectations to target, a massive decline in volatility of nominal and real interest rates, as well as of inflation, and high reserve requirements on bank dollar liabilities. Moreover, the latter success and BCRP’s institutional strength has enabled the bank to pursue an active countercyclical monetary policy, reflected in a strong expansionary stance since 2002 – in all likelihood a result of its independence and strong credibility (as documented for a panel of emerging countries by Calderón and Schmidt-Hebbel [2003]).

Some evidence on fear to float among inflation targeters in Latin America

Four main factors lead to ‘fear of floating’ in emerging market economies, including some countries that have adopted IT and others that are still considering IT adoption: (i) concerns about high pass-through from ER to inflation; (ii) concerns about financial vulnerabilities arising from highly dollarized liabilities on balance sheets; (iii) fears about loss of export competitiveness; and (iv) concerns about losing a transparent nominal anchor to guide expectations. Do Latin American inflation targeters:

  • fear floating à la Calvo and Reinhart?

  • exhibit large devaluation-inflation pass-through?

  • react strongly to the ER in their conduct of monetary policy?

I will address each of these questions.

Result 1: Macroeconomic volatilities have been consistent with a floating regime since 1999.

Under a floating regime, the ER should gain importance as an adjustment mechanism, while reserves and interest rates should become more stable. In Chile and Mexico, ER volatility has risen significantly relative to reserves volatility. In Brazil it has declined, but is still very high. And in all three Latin American countries, relative ERs to reserve volatilities are similar to or higher than those observed in Australia, Canada and New Zealand (Figure 6C.1).

Figure 6.C1Ratio of exchange rate and international reserve volatilities in six countries

Result 2: ER pass-through to inflation has declined.

Pass-through from ER depreciation to inflation is larger under low central bank credibility, a high degree of openness and a history of high inflation. Pass-through coefficients (measured as simple correlation coefficients) have fallen significantly in IT countries in Latin America, from levels between 50 and 80 per cent before or at the start of IT to levels that range from 20 per cent in Chile to 50 per cent in Brazil (Schmidt-Hebbel and Werner, 2002; García and Restrepo, 2002).

Result 3: ER is not a significant determinant in Taylor rules of Latin American central banks.

Even under low ER pass-through, monetary policy could react to the ER out of concern for devaluation-induced financial crises and recessions or excessive ER volatility. To test for this, the ER is included as an additional argument in a conventional Taylor rule. The results for three Latin American IT countries (Schmidt-Hebbel and Werner [2002], reproduced here as Table 6C.1) show that the ER is not a significant driver of monetary policy, except in special circumstances, as exemplified by rolling coefficient estimates (Figure 6C.2).

Figure 6.C2Monetary policy response to the exchange rate in Brazil, Chile and Mexico (rolling coefficient estimates)

Table 6.C1Estimations of Taylor rules for Brazil, Chile and Mexico
Real interest rate1999-20011991-20011997-20011997-20011999-2001
Real interest rate lag0.890.920.920.470.68
Expected inflation-inflation target0.300.010.010.08-0.66
Output gap0.
Nominal exchange rate-0.01-
Long-term government bond0.320.030.021.981.49
Trade deficit lag (exports-imports)-0.01-0.01-0.01-0.01-0.01
Summary statistic
Adjusted R20.750.880.870.800.81
Durbin-Watson statistic1.921.641.682.171.95

Final Remarks on Peru

Peru has gone a long way toward full-fledged IT, reaping many of the benefits that world evidence on IT shows can be obtained (Mishkin and Schmidt-Hebbel, 2005). Peru’s progress in price stabilization has been impressive: inflation is at low stationary levels, inflation targets are met and expectations are anchored to inflation targets; BCRP’s switch of operational target from money to the interest rate has increased effectiveness of monetary policy and contributed to soli(di)fication. And, in contrast to other mandatory de-dollarization experiences, soli(di)fication has been based on the markets’ voluntary response to policy changes and correct incentives.

Peru faces one important policy challenge towards completion of its full-fledged IT framework. This is accelerating its pace towards more floating, through less frequent exchange market interventions, more support of issuance of sol-denominated liabilities and support of market development of financial instruments that protect against ER risk. Peru’s policy framework and achievements are ripe for taking these steps, which will reinforce soli(di)fication and monetary policy effectiveness.


    CalderónC. and K.Schmidt-Hebbel (2003) ‘Macroeconomic Policies and Performance in Latin America’Journal of International Money and FinanceVol. 22 pp. 895923.

    GarcíaC. and J.Restrepo (2002) ‘Price Inflation and Exchange Rate Passthrough in Chile’Working Paper No. 128 (Santiago: Central Bank of Chile).

    IzeA. and E.Levy Yeyati (2003) ‘Financial Dollarization’Journal of International EconomicsVol. 59 (March) pp. 32347.

    MishkinF. and K.Schmidt-Hebbel (2005) ‘Does Inflation Targeting Make a Difference?’, paper prepared for the Monetary Policy Conduct under Inflation Targeting Conference20-1OctoberSantiagoCentral Bank of Chile.

    Schmidt-HebbelK. and A.Werner (2002) ‘Inflation Targeting in Brazil, Chile and Mexico: Performance, Credibility, and the Exchange Rate’EconomiaVol. 2No. 2 pp. 3089.

    TapiaM. and A.Tokman (2004) ‘Effects of Foreign Exchange Intervention under Public Information: The Chilean Case’EconomiaVol. 4No. 2 pp. 21545.

    Other Resources Citing This Publication