Chapter

Chapter 8. Interest Rate and Exchange Rate Channels in Dollarized and Non-Dollarized Economies

Author(s):
R. Gelos
Published Date:
March 2014
Share
  • ShareShare
Show Summary Details
Author(s)
Santiago Acosta-Ormaechea and David Coble A version of this chapter was published in April 2013 in Revista Economía Chilena (Volume 16, Number 1). Available at: http://www.bcentral.cl/eng/studies/economia-chilena/articles.htm.

The study of the transmission of monetary policy in both advanced and developing economies has been the objective of numerous papers in recent years. This monetary transmission is broadly conceived as the analysis of how monetary policy decisions ultimately affect inflation, which is generally the primary objective of central banks.

Under an inflation targeting regime, in particular, the economic authorities conduct monetary policy by setting a reference interest rate to achieve a pre-announced inflation target, while leaving other monetary aggregates and the exchange rate to be largely determined by market forces. However, these monetary policy decisions affect inflation with a delay, as they first tend to impact on different variables and only after that, through different channels, the inflation rate.1

The objective of this chapter is to study the transmission of monetary policy empirically, comparing the cases of four small and open economies that operate under an inflation targeting regime and are exposed to different levels of dollarization. In particular, we aim to disentangle to what extent the degree of dollarization may help explain some key differences in the transmission of monetary policy in such countries.2 For this we examine two countries that run well-established inflation targeting regimes and have a negligible exposure to dollarization problems: Chile and New Zealand. The cases of Peru and Uruguay are also studied, as these countries also operate under IT regimes yet are exposed to a large degree of dollarization in their respective banking systems.3,4

To clarify the sequence through which policy decisions are transmitted to the rest of the economy, we divide the analysis into two stages. First, we examine how the policy rate affects the key interest rates of each country, notably the money market rate, the deposit rate, and the lending rate. Broadly speaking, we intend to evaluate the strength of what is in practice the first part of the transmission, which involves the pass-through from the policy rate to a number of key interest rates of each economy. In the second stage, we use a vector autoregression (VAR) model to evaluate how changes in the money market rate affect output and inflation in these economies. Our aim is to assess empirically to what extent the strength and channels of the transmission differ across countries. Finally, we evaluate through different rolling VAR models whether the degree of credit in the economy and the level of dollarization may help explain the strength of the response of inflation to policy rate innovations in these countries.

A large strand of the empirical work on monetary policy considers different VAR models applied to both advanced and emerging market economies.5 This chapter follows a similar methodology but applies it to a different set of countries, giving rise to new challenges. We found, for example, a significant price puzzle when considering the VAR specification of Kim and Roubini (2000). To correct for this problem, we modified the specification by including a corrected measure of inflation: we eliminated the foreign component of domestic inflation in our set of variables in the VAR model. This correction is important because the strength of the monetary policy transmission should be evaluated considering only those variables that are under the effective control of the monetary authority. Since this is clearly not the case for foreign inflation, this source of variability of domestic prices should be eliminated from the model. In addition, this chapter provides a comprehensive assessment of how interest rate shocks are initially transmitted to the rest of the economy through domestic credit markets, which is unusual in this literature. Likewise, it also includes an empirical evaluation of the role of financial deepening in the transmission of monetary policy. In this regard, we show that the response of inflation to the interest rate shock does not seem to depend on the degree of development of credit markets. This finding supports the view that monetary policy can in principle be effective even in the context of relatively low levels of credit in the economy.

We also found that in Chile and New Zealand there is a significant pass-through from the policy rate to the main interest rates of the economy. Accordingly, a contractionary monetary policy shock reduces inflation and output, suggesting a strong transmission of monetary policy through the traditional interest rate channel. Conversely, in Peru and Uruguay, the interest rate pass-through is rather weak, and so is the overall transmission to output and inflation. For the latter two economies, however, evidence indicates that the exchange rate channel—rather than the interest rate channel—may play a more substantial role in curbing inflationary pressures, as indicated by the relatively larger exchange rate pass-through observed in these two economies. Finally, as indicated previously, the chapter does not find conclusive evidence regarding the role of financial deepening in the transmission of monetary policy. However, it is found that as Peru and Uruguay reduced their levels of dollarization, the effectiveness of the monetary policy transmission increased somewhat.

Key Stylized Facts

Chile, New Zealand, Peru, and Uruguay are all small open economies that currently operate under inflation targeting regimes. All have relatively similar degrees of trade openness and are important commodity exporters (Figure 8.1). One key difference is, however, the degree of banking sector dollarization. In Chile and New Zealand, almost all credit and deposits are denominated in local currency, whereas in Peru and Uruguay the share of foreign-currency-denominated credit and deposits is rather large, though on a downward path.6

Figure 8.1Selected Macroeconomic Indicators

Sources: Country authorities; and authors’ calculations.

There are also additional differences regarding the performance of key fundamentals in these countries, as indicated in Table 8.1.7 For instance, Uruguay and Peru had a relatively low variability in both the nominal exchange rate (NER) against the U.S. dollar and the real effective exchange rate (REER). This may reflect a “fear of floating” type of behavior, which could result from their significant degree of dollarization. In addition, Uruguay experienced a higher degree of output volatility relative to all other countries.8 In the case of Peru, output volatility has been lower than in Uruguay, yet still above that observed in Chile and New Zealand. Somewhat surprisingly, however, the volatility of CPI inflation has recently decreased in Uruguay, contrary to the other three economies, where it has trended upward.9 Expected inflation has followed a similar pattern: in all countries except Uruguay, its volatility increased over time. Interestingly, the evidence indicates a relatively more active use of foreign exchange interventions in these three Latin American countries in recent years. This contrasts with the case of New Zealand, where the volatility of reserves has decreased over time.

Table 8.1Volatility of Selected Variable(Standard deviations of annual changes, percent)
REERNEROutputCPI

Inflation
Expected

Inflation

(12 months

ahead)
Foreign

Reserves to

GDP
(1)(2)(1)(2)(1)(2)(1)(2)(1)(2)(1)(2)
Chile6.47.65.017.21.03.81.44.10.31.17.134.2
New Zealand8.012.210.122.81.11.80.61.00.20.720.212.8
Peru2.32.93.47.31.84.70.92.30.30.615.531.4
Uruguay6.16.96.814.88.28.41.50.80.50.515.825.0
Sources: Country authorities; and authors’ calculations.Note: (1) January 2005 to December 2007; (2) January 2008 to November 2010. CPI = consumer price index; NER = nominal exchange rate; REER = real effective exchange rate.
Sources: Country authorities; and authors’ calculations.Note: (1) January 2005 to December 2007; (2) January 2008 to November 2010. CPI = consumer price index; NER = nominal exchange rate; REER = real effective exchange rate.

Although Uruguay’s consumer price index (CPI) inflation has been relatively more stable, it has also remained at a higher level relative to the other three countries, particularly during 2010 (Figure 8.2). A similar pattern arises when considering different measures of core inflation (not shown). Furthermore, actual and expected inflation have remained systematically in the upper band of (or above) the target range in Uruguay, whereas in all the other cases, both variables have hovered around their respective midpoint target ranges. As expected, inflation expectations have moved in tandem with actual inflation, yet showing a lower variability. In addition, in all four cases, the policy rate has closely tracked the inflation rate, suggesting that monetary policy has been active in trying to curb inflation pressures.

Figure 8.2Actual Inflation, Expected Inflation, Target Bands, and Policy Rates

(Year-over year percent change)

Sources: Country authorities, and authors’ calculations.

There are, however, important differences in the evolution of the policy rate.10 The sample considered here is divided into two periods: the first period extends from September 2003—the introduction of the policy rate as the main monetary policy instrument in Peru—to August 2007; the second period extends from its introduction in Uruguay—in September 2007—to the latest available observation—November 2010. Interestingly, Chile shows the most active use of monetary policy, measured by the number of policy rate changes observed during these periods. This contrasts with the case of Uruguay, which had the fewest number of modifications—although the median change was the largest whenever there was a change in the policy rate. In addition, as Uruguay has had the highest average inflation rate, the policy rate has accordingly taken the highest value after its implementation. These results differ substantially from those of Peru, where evidence suggests a more active use of monetary policy and a lower change in the policy rate whenever it was modified (Table 8.2). Overall, evidence points to the absence of a clear pattern in terms of the use of the policy rate across countries regardless of the level of dollarization and the date on which the inflation targeting regime was implemented.

Table 8.2Inflation and Policy Rate(Percent, unless otherwise specified)
CPI

Inflation
Policy RatePolicy Rate Changes1
Standard

Deviation
No. of

Times
MinimumMaximumMedian
AverageAverage(Basis Points)
September 2003–August 2007
Chile2.53.71.42965815
New Zealand2.76.70.913252525
Peru2.13.50.810252525
Uruguay7.3n.a.n.a.n.a.n.a.n.a.n.a.
September 2007–November 2010
Chile3.83.72.923826133
New Zealand2.74.72.592515050
Peru3.53.82.0192510025
Uruguay7.37.41.0825225100
Sources: Country authorities; and authors’ calculations.Note: CPI = consumer price index; n.a. = not available.

Minimum, maximum, and median are based on the absolute value of the policy rate change.

Sources: Country authorities; and authors’ calculations.Note: CPI = consumer price index; n.a. = not available.

Minimum, maximum, and median are based on the absolute value of the policy rate change.

The Interest Rate Pass-Through

Under an inflation targeting regime, the transmission of monetary policy the decisions initially takes place through the modification of the policy rate, which in turn affects interbank rates on impact. Changes in the latter are then transmitted to deposit and lending rates, thus affecting the consumption and saving decisions of individuals and firms, and hence aggregate demand and inflation. Likewise, the change in the policy rate may affect the overall availability of credit as well as different asset prices that react to short-term interest rates, thereby enhancing the transmission through the so-called credit and asset price channels. Additionally, as domestic and foreign interest rates may differ for comparable assets, arbitrage between them gives rise to nominal exchange rate fluctuations, which in turn affect inflation and economic activity through the so-called exchange rate channel.

As expected, Figure 8.3 shows that in all countries, the policy rate and the money market (interbank) rate almost completely overlap. The only exception took place in Uruguay between October 2008 and November 2008, a period when the central bank explicitly avoided large interventions in a disrupted interbank market as part of a strategy to deal with the global financial turmoil triggered by the collapse of Lehman Brothers. Besides this exception, Figure 8.3 shows that the interest rate pass-through from the policy rate to the interbank rate tends to be immediate and complete across countries.

Figure 8.3Local Currency Deposit, Lending, and Policy Rates

(Year-over-year percent change)

Source: Country authorities.

Figure 8.3 also shows that the co-movement between the policy rate and the deposit and lending rates is strong in Chile and New Zealand. This relationship is somewhat weaker in the cases of Peru and Uruguay, suggesting that the effectiveness of the policy rate in signaling the stance of monetary policy might not be that strong in these countries.

To evaluate formally the relationship between these key interest rates, the following simple regression is conducted:

where for any country i and period t, yit is the change in either the deposit or the lending rate, ci is a constant, xit is the change in the money market rate, and εit is an error term. The short-term effect of a money market rate change on the other interest rates is thus given by αi. Tables 8.3 and 8.4 summarize the main results of the estimations.

Table 8.3Pass-Through from Money Market Rate to Deposit Rate
Constant (c)Short-Term Pass-Through (α)Cross-CorrelationR-SquaredSample
Chile0.000.78*0.870.76Sept. 1999–Nov. 2010
t-Statistic−0.0820.72
New Zealand0.010.67*0.710.50Apr. 1999–Oct. 2010
t-Statistic0.7011.68
Peru−0.010.32*0.750.57Sept. 2003–Oct. 2010
t-Statistic−1.4910.54
Uruguay0.030.20*0.300.91Sept. 2007–Nov. 2010
t-Statistic0.7119.91
Source: Authors’ estimations.Note: * denotes statistical significance at the 5 percent level.
Source: Authors’ estimations.Note: * denotes statistical significance at the 5 percent level.
Table 8.4Pass-Through from Money Market Rate to Lending Rate
Constant (c)Short-Term Pass-Through (α)Cross-CorrelationR-SquaredSample
Chile−0.010.67*0.720.51Sept. 1999-Nov. 2010
t-Statistic−0.2111.88
New Zealand0.03*0.62*0.840.70Apr. 1999-Oct. 2010
t-Statistic3.1817.94
Peru−0.030.190.130.02Sept. 2003-Oct. 2010
t-Statistic0.050.16
Uruguay0.030.10*0.320.00Sept. 2007-Oct. 2010
t-Statistic0.142.05
Source: Authors’ estimations.Note: * denotes statistical significance at the 5 percent level.
Source: Authors’ estimations.Note: * denotes statistical significance at the 5 percent level.

As visually suggested by Figure 8.3, the overall interest rate pass-through in the cases of Chile and New Zealand is large. In these two countries, the change in the policy rate has a sizable and significant effect on both the deposit and lending rates through changes in the interbank rate. In contrast, for Peru and Uruguay, the policy rate pass-through tends to be much weaker, suggesting a somewhat larger impact on the deposit rather than the lending rate.11 The weaker relationship between these interest rates in the cases of Peru and Uruguay is likely to reflect differences in the structure, depth, and degree of dollarization of their financial systems.

Empirical Analysis of the Transmission of Monetary Policy

After the initial transmission of the policy rate to the different interest rates of the economy, the monetary policy change is transmitted through the various channels discussed previously to the rest of the economy. This section specifically analyzes how this monetary policy innovation affects inflation and output in each country. In a nutshell, the effectiveness of monetary policy is assessed, with consideration given to the way inflation—the primary concern of the central bank—reacts to a policy-driven increase in the money market rate (a contractionary monetary policy shock). Since the stance of aggregate demand also determines the evolution of inflation, the response of output to the same innovation is studied as well. With this objective in mind, the following VAR model is estimated:

where A(L) and B(L) are n × n and n × k polynomial matrices in the lag operator L, respectively, Yt is an n × 1 vector of endogenous variables, Xt is a k × 1 vector of exogenous variables, Ut and is an n × 1 vector of estimated residuals. Xt is included to control for those disturbances that are not directly managed by the monetary authority and may somewhat affect the dynamics of the model. The benchmark specification takes the following form:

where Rt is the money market rate, IPt is the year-over-year change of an index of economic activity, and πtπtw is a measure of the year-over-year headline inflation that is effectively under the control of the monetary authority.12 Intuitively, as these economies are small and open, world headline inflation will have a significant effect on their respective inflation rates. To effectively strip domestic inflation of external factors, the gap between domestic and external inflation is considered. This approach helps eliminate the so-called price puzzle problem often encountered in the empirical literature (see Sims, 1992). The vector of exogenous variables is in turn given by:

where FFt is the U.S. federal funds rate, WCPIt is the year-over-year change of the world commodity price index, and IPtus is the U.S. industrial production index gap in logs. In terms of identification, the model uses a standard Cholesky decomposition, with the variables ordered as in vector Yt.13 This implies, for instance, that Rt is contemporaneously affected only by its own shock, whereas πtπtw, being the most endogenous variable, is affected on impact by all structural innovations of the model.

The estimations consider monthly data, with the sample period tailored to each particular country. In the case of Chile, the sample starts with the introduction of the inflation targeting regime in September 1999 and ends in November 2010. For New Zealand, it starts in April 1999, when the so-called cash rate was set as the main policy instrument, and ends in September 2010.14 For Peru, the sample starts in September 2003, the month in which the full-fledged inflation targeting regime was implemented, and ends in November 2010. Finally, for Uruguay, even though the inflation targeting regime began to be implemented with the use of the policy rate as the main instrument only in September 2010, the sample starts in January 2006 and ends in November 2010. This was done to have a larger sample to estimate the VAR model to obtain relatively more robust results. The model is estimated with two lags for all countries, as suggested by standard tests.

Figure 8.4 illustrates that the strength of the transmission of interest rate changes varies significantly across countries. The figure presents impulse-response (IR) functions for a 100 basis point increase in the money market rate—a contractionary monetary policy shock. In New Zealand, there is a significant and persistent contraction in the growth rate of output and inflation, in line with intuition. In Chile, the negative impact on output takes about three months to materialize, yet it tends to be persistent. In addition, the effect on inflation in Chile is more immediate than on output, and it also tends to be persistent.

Figure 8.4Impulse-Response Functions for a 100 Basis Point Increase in the Policy Rate

(Percent, error bounds at 68 percent confidence)

Source: Authors’ estimations.

Note: IP = an index of economic activity (see text). All confidence intervals are computed using standard bootstrapping procedures with 1,000 replications.

The two dollarized economies analyzed here appear to have distinct results. In Peru and Uruguay, the interest rate hike leads to either a short-term increase in economic activity for about five months (Peru) or a slightly positive but not significant effect (Uruguay). In addition, the effect of the shock on inflation is either counterintuitive (Peru) or rather insignificant (Uruguay). In fact, the absence of a contraction in economic activity after the rise in the interest rate may be related to the existence of balance sheet effects in these economies. That is, the associated exchange rate appreciation that follows the interest rate hike may lead to an improvement in the balance sheets of those agents indebted in the foreign currency, thus generating an indirect positive effect on aggregate demand. This latter effect may outweigh the contractionary effect initially produced by the traditional interest rate channel (Figure 8.4).16

To analyze whether the exchange rate channel—rather than the interest rate channel—is relatively more relevant in Peru and Uruguay, a slightly different version of the VAR model is estimated. In this case, the vector of endogenous variables becomes:

where neert indicates the year-over-year change in the nominal effective exchange rate (NEER), IPt is the year-over-year change in an index of economic activity, πt represents the annual CPI inflation rate, and Rt is the money market rate, as before. The rest of the model remains as in the baseline specification.

To explore formally the role of the exchange rate in the monetary transmission, the IR functions in Figure 8.5 evaluate how a NEER depreciation affects CPI inflation and output. As expected, Peru and Uruguay have the largest exchange rate pass-through, likely reflecting their high degree of dollarization. For instance, in Peru a 10 percent depreciation of the NEER raises CPI inflation up to 1.8 percent five months after the shock. This effect tends to also be persistent. In the case of Uruguay, CPI inflation tends to increase more quickly, to about 1.2 percent after three months, and the effect remains positive for a long period of time. These results contrast with those of Chile and New Zealand, where the NEER depreciation appears to have a small and short-lived effect on inflation.17 There is also a relatively larger negative effect of the exchange rate depreciation on output in Peru and Uruguay, which again appears to be consistent with the presence of adverse balance sheet effects in these countries. In fact, in the cases of Chile and New Zealand the output response is rather limited, suggesting from a different angle the larger effectiveness of monetary policy in these countries in controlling inflation at a lower output cost.

Figure 8.5Impulse-Response Functions for a 10 Percent Depreciation in the Nominal Effective Exchange Rate

(NEER; percent, error bounds at 68 percent confidence)

Source: Authors’ estimations.

Note: CPI = consumer price index; IP = an index of economic activity (see text).

Role of Financial Factors in the Transmission of Monetary Policy

The overall strength of the monetary transmission depends on various factors, with some being specific to each country. This section explores to what extent the characteristics of the domestic financial system, notably its depth and degree of dollarization, may help explain the cross-country differences in the inflation response to the interest rate shock discussed previously.

Notice first that with a low degree of financial deepening, the traditional interest rate channel may not be operative, as a low degree of development of domestic credit markets may impinge on the ability of the monetary authority to control the flow of credit and thereby aggregate demand. In addition, if capital markets have only a modest relevance in the country, the strength of the transmission from asset prices to aggregate demand might be rather scant, thus limiting also the overall impact of the policy rate change on inflation. If the economy is also dollarized, in the sense that a large share of credit to the private sector is denominated in foreign currency, the monetary transmission may turn out to be even weaker due to the relatively lower influence of the monetary authority on modifying the key interest rates that affect foreign currency lending and thereby consumption and investment.

Table 8.5 presents a number of parameters for each country to evaluate the differences in the development of their respective financial system. The table illustrates that Uruguay ranks low in terms of credit to the private sector over GDP, a ratio that is closely followed by that of Peru. Particularly striking is the small size of the stock market capitalization in Uruguay. Not surprisingly, Uruguay also has the largest share of international debt issuance over GDP, followed by Peru, reflecting the still limited scope for funding in local markets. This trend has recently started to reverse, in line with the significant dedollarization process experienced by these two economies. Overall, the evidence suggests that credit markets are somewhat more developed in Chile and New Zealand, which may help explain the larger significance of the interest rate channel in these countries.

Table 8.5Financial System Indicators
Total Deposits1Credit to Private Sector1Stock Market

Capitalization2
International Debt

Issuance3
Percent of GDP4
Chile49.574.3154.26.3
New Zealand82.2142.829.37.4
Peru32.324.567.410.2
Uruguay44.820.90.421.9
Sources: Bank for International Settlements (BIS); country authorities; Federacion Iberoamericana de Bolsas; New Zealand Exchange; and authors’ calculations.

As of November 2010.

As of December 2010.

International debt securities of all issuers (amortizations outstanding) from BIS Securities Statistics (Table 12A) as of December 2010.

2010 GDP taken from IMF World Economic Outlook forecasts.

Sources: Bank for International Settlements (BIS); country authorities; Federacion Iberoamericana de Bolsas; New Zealand Exchange; and authors’ calculations.

As of November 2010.

As of December 2010.

International debt securities of all issuers (amortizations outstanding) from BIS Securities Statistics (Table 12A) as of December 2010.

2010 GDP taken from IMF World Economic Outlook forecasts.

To formally evaluate whether the development of domestic credit markets may affect the transmission of monetary policy, we perform a different set of exercises, considering again our benchmark VAR model. Specifically, we explore whether the three-month response of inflation to a 100 basis point increase in the money market rate is affected when the credit-to-output ratio varies across countries. Regarding the econometric approach, the VAR model is estimated recursively. That is, for each country, the first window of the sample period is set to three years (36 observations). The impulse-response function to an interest rate shock is then calculated to pick the three-month response of inflation to the shock.18 The process continues, adding one additional observation to the sample each time. The last (and largest) subsample for each country then coincides with that of the benchmark VAR model presented in the previous section. Once the three-month responses of inflation considering all the different subsamples are collected, a scatter plot is constructed including also the ratio of credit to the private sector over GDP for each country.19

Figure 8.6 illustrates that the credit-to-output ratio does not provide conclusive evidence regarding the role of credit in affecting the extent of the monetary policy transmission. The reason is that a higher credit-to-output ratio should be associated with a more negative response of inflation to the interest rate shock, because more developed financial markets should allow for better functioning of the credit channel and thus of the overall effectiveness of the interest rate shock in affecting inflation. Whereas results for the case of Chile and New Zealand are counterintuitive, those of Peru and Uruguay suggest that the effectiveness of monetary policy in curbing inflation pressures may have increased when more credit was available. In any case, the absence of conclusive results is in line with those of Saizar and Chalk (2008). They do not find clear-cut evidence showing a positive relationship between the credit-to-GDP ratio and the strength of the transmission of interest rate shocks in a set of developing economies.

Figure 8.6Credit to Private Sector and Inflation Response to a 100 Basis Point Policy Shock

(Third-period response, percent)

Source: Authors’ estimations.

A similar recursive VAR exercise is also conducted to assess whether the recent dedollarization trend in Peru and Uruguay has somewhat strengthened their transmission of monetary policy. With this objective, the three-month response of inflation to the interest rate shock is again analyzed following the steps outlined before. Additionally, two different measures of dollarization are used to construct the scatter plots used for each country: local currency credit to the private sector (as percent of total credit) and local currency deposits (as percent of total deposits).20

Results suggest that the recent dedollarization process of Peru and Uruguay is likely to have improved the effectiveness of monetary policy (Figure 8.7). In fact, the figure indicates that the positive response of inflation to the interest rate hike (that is, the so-called price puzzle) has decreased substantially when the share of either credit or deposits in local currency increased, thus suggesting a better transmission of interest rate shocks in these economies as the process of dedollarization strengthened. Notwithstanding these results, for Uruguay, the findings are still relatively weak, a fact that is likely to be associated with the relatively shorter sample used in this latter case. A relevant exercise left for further research is therefore to evaluate whether the robustness of these findings rises as more data become available.

Figure 8.7Credit and Deposits in Domestic Currency and Inflation Response to a 100 Basis Point Policy Shock

(Third-period response, percent)

Source: Authors’ estimations.

Conclusion

This chapter conducts a comparative study of the interest rate and exchange rate channels in two economies that run a well-established inflation targeting regime—Chile and New Zealand—vis-à-vis two economies that operate under relatively newer inflation targeting regimes and are exposed to a significant degree of dollarization—Peru and Uruguay. We found significant differences among these countries in terms of their respective transmission of monetary policy decisions. Whereas the traditional interest rate channel appears to operate to a large extent in Chile and New Zealand, evidence indicates instead that it is the exchange rate channel that still plays a more substantial role in controlling inflationary pressures in Peru and Uruguay. This result follows from the still limited impact of the policy rate in controlling inflation in these countries, in combination with their relatively large and persistent exchange rate pass-through. The latter is in turn likely to be associated with the substantial dollarization that still prevails in these economies. Importantly, however, since they have embarked on a significant dedollarization process, the relevance of the exchange rate channel is likely to decrease over time. In fact, the preliminary results presented in this chapter suggest that the interest rate channel is having a more significant role in their transmission of monetary policy as they tend to dedollarize their financial systems. Consequently, other channels of transmission may need to be further strengthened to curb inflation pressures more effectively over the medium term in these countries.

Appendix 8.1. Sources and Data Description

Table A8.1 provides sources, frequency, and samples for the data used in this chapter.

Table A8.1Data Sources
CountryVariableDescriptionSampleFrequencySource
ChileCPI inflationConsumer price indexJan-2005 to Nov-2010MonthlyNational Bureau of Statistics
ChileCredit to the private sectorCredit to the private sector in local currencySep-1999 to Nov-2010MonthlyCentral Bank of Chile
ChileDeposit rateDeposit rateSep-1999 to Nov-2010MonthlyInternational Financial Statistics database, IMF
ChileExpected inflationInflation expectations, 12 months aheadJan-2005 to Nov-2010MonthlyCentral Bank of Chile
ChileForeign reservesReserve assetsJan-2005 to Nov-2010MonthlyCentral Bank of Chile
ChileLending rateLending rateSep-1999 to Nov-2010MonthlyInternational Financial Statistics database, IMF
ChileMoney market rateAverage overnight interbank rateSep-1999 to Nov-2010MonthlyCentral Bank of Chile
ChileNERNominal exchange rate, local currency per U.S. dollarJan-2005 to Nov-2010MonthlyCentral Bank of Chile
ChileNominal GDPGross domestic product at current prices1995–2010AnnuallyCentral Bank of Chile
ChileOpennessExport plus imports over GDP1995–2009AnnuallyCentral Bank of Chile
ChileOutputMonthly indicator of economic activity, IMACECJan-2005 to Nov-2010MonthlyCentral Bank of Chile
ChilePolicy rateMonetary policy rateSep-1999 to Nov-2010MonthlyCentral Bank of Chile
ChileREERReal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2010MonthlyInternational Financial Statistics database, IMF
ChileNEERNominal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2011MonthlyInternational Financial Statistics database, IMF
ChileTarget bandsInflation target bandsSep-1999 to Nov-2010MonthlyCentral Bank of Chile
New ZealandCPI inflationConsumer price indexJan-2005 to Nov-2010MonthlyReserve Bank of New Zealand
New ZealandCredit to the private sectorClaims on private sectorApril-1999 to Nov-2010MonthlyInternational Financial Statistics database, IMF
New ZealandDeposit rateDeposit rateApril-1999 to Nov-2010MonthlyInternational Financial Statistics database, IMF
New ZealandExpected inflationInflation expectations, 12 months aheadJan-2005 to Nov-2010MonthlyReserve Bank of New Zealand
New ZealandForeign reservesTotal reserves minus goldJan-2005 to Nov-2010MonthlyInternational Financial Statistics database, IMF
New ZealandLending rateBase lending rateApril-1999 to Nov-2010MonthlyInternational Financial Statistics database, IMF
New ZealandMoney market rateMoney market rateApril-1999 to Nov-2010MonthlyInternational Financial Statistics database, IMF
New ZealandNERNominal exchange rate, local currency per U.S. dollarJan-2005 to Nov-2010MonthlyInternational Financial Statistics, IMF
New ZealandNominal GDPGross domestic product at current prices1995–2010AnnuallyStatistics New Zealand
New ZealandOpennessExport plus imports over GDP1995–2009AnnuallyStatistics New Zealand
New ZealandOutputGross domestic product at constant pricesJan-2005 to Sep-2010Quarterly converted to monthlyStatistics New Zealand
New ZealandPolicy rateMonetary policy rateApril-1999 to Nov-2010MonthlyReserve Bank of New Zealand
New ZealandREERReal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2010MonthlyInternational Financial Statistics database, IMF
New ZealandNEERNominal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2010MonthlyInternational Financial Statistics database, IMF
New ZealandTarget bandsInflation target bandsApril-1999 to Nov-2010MonthlyReserve Bank of New Zealand
PeruCPI inflationConsumer price indexJan-2005 to Nov-2010MonthlyNational Bureau of Statistics
PeruCredit in domestic currencyCredit to the private sector in local currencySep-2003 to Nov-2010MonthlyCentral Bank of Peru
PeruCredit in foreign currencyTotal credit in foreign currency1995–2009AnnuallyCentral Bank of Peru
PeruCredit to the private sectorTotal credit to the private sectorSep-2003 to Nov-2010MonthlyCentral Bank of Peru
PeruDeposit rateAverage deposit rate in local currency (TIPMN)Sep-2003 to Nov-2010MonthlyCentral Bank of Peru
PeruDeposits in domestic currencyLiquidity of the banking system in local currency minus vault cashSep-2003 to Nov-2010MonthlyCentral Bank of Peru
PeruDeposits in foreign currencyTotal deposits in foreign currency1995–2009AnnuallyCentral Bank of Peru
PeruExpected inflationInflation expectations, 12 months aheadJan-2005 to Nov-2010MonthlyCentral Bank of Peru
PeruForeign reservesNet international reservesJan-2005 to Nov-2010MonthlyCentral Bank of Peru
PeruLending rateAverage lending rate in local currency (TAMN)Sep-2003 to Nov-2010MonthlyCentral Bank of Peru and IMF
PeruMoney market rateInterbank rateSep-2003 to Nov-2010MonthlyCentral Bank of Peru
PeruNERNominal exchange rate, local currency per U.S. dollarJan-2005 to Nov-2010MonthlyCentral Bank of Peru
PeruNominal GDPGross domestic product at current prices1995–2010AnnuallyCentral Bank of Peru
PeruOpennessExport plus imports over GDP1995–2009AnnuallyCentral Bank of Peru
PeruOutputGross domestic product at constant pricesJan-2005 to Nov-2010MonthlyCentral Bank of Peru
PeruPolicy rateMonetary policy rateSep-2003 to Nov-2010MonthlyCentral Bank of Peru
PeruREERReal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2010MonthlyInternational Financial Statistics database, IMF
PeruNEERNominal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2010MonthlyInternational Financial Statistics database, IMF
PeruTarget bandsInflation target bandsSep-2003 to Nov-2010MonthlyCentral Bank of Peru
PeruTotal creditTotal credit1995–2009AnnuallyCentral Bank of Peru
PeruTotal depositsTotal deposits1995–2009AnnuallyCentral Bank of Peru
UruguayCPI inflationConsumer price indexJan-2005 to Nov-2010MonthlyNational Bureau of Statistics
UruguayCredit in domestic currencyCredit to the private sector in local currencyJan-2006 to Nov-2010MonthlyCentral Bank of Uruguay
UruguayCredit in foreign currencyTotal credit in foreign currency1995–2009AnnuallyCentral Bank of Uruguay
UruguayCredit to the private sectorTotal credit to the private sectorJan-2006 to Nov-2010MonthlyCentral Bank of Uruguay
UruguayDeposit rateDeposit rate in local currencySep-2007 to Nov-2010MonthlyCentral Bank of Uruguay
UruguayDeposits in domestic currencyTotal deposits in local currencyJan-2006 to Nov-2010MonthlyCentral Bank of Uruguay
UruguayDeposits in foreign currencyTotal deposits in foreign currency1995–2009AnnuallyCentral Bank of Uruguay
UruguayExpected inflationInflation expectations, 12 months aheadJan-2005 to Nov-2010MonthlyCentral Bank of Uruguay
UruguayForeign reservesReserve assetsJan-2005 to Nov-2010MonthlyCentral Bank of Uruguay
UruguayLending rateLending rate (ordinary)Sep-2007to Nov-2010MonthlyInternational Financial Statistics database, IMF
UruguayMoney market rateInterbank overnight rate (call)Sep-2007to Nov-2010MonthlyCentral Bank of Uruguay
UruguayNERNominal exchange rate, local currency per U.S. dollarJan-2005 to Nov-2010MonthlyNational Bureau of Statistics
UruguayNominal GDPGross domestic product at current prices1995–2010AnnuallyCentral Bank of Uruguay
UruguayOpennessExport plus imports over GDP1995–2009AnnuallyCentral Bank of Uruguay
UruguayOutputIndustrial production index volume (excludes refinery)Jan-2006 to Nov-2010MonthlyNational Bureau of Statistics
UruguayPolicy rateMonetary policy rateSep-2007to Nov-2010MonthlyCentral Bank of Uruguay
UruguayREERReal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2010MonthlyInternational Financial Statistics database, IMF
UruguayNEERNominal effective exchange rate, index 2005 = 100Jan-2005 to Sep-2010MonthlyInternational Financial Statistics database, IMF
UruguayTarget bandsInflation target bandsSep-2007to Nov-2010MonthlyCentral Bank of Uruguay
UruguayTotal creditTotal credit1995–2009AnnuallyCentral Bank of Uruguay
UruguayTotal depositsTotal deposits1995–2009AnnuallyCentral Bank of Uruguay
Source: Authors’ compilation.
Source: Authors’ compilation.
References

    AghionPhilippeP.Bacchetta and A.Banerjee2000A Simple Model of Monetary Policy and Currency Crises,European Economic Review44 pp. 72838.

    CéspedesLuisR.Chang and A.Velasco2004 “Balance Sheets and Exchange Rate Policy” American Economic Review Vol. 94 No. 4 pp. 1183–93.

    García-EscribanoMercedes and S.Sosa2011What Is Driving Financial De-dollarization in Latin America?IMF Working Paper No. 11/10 (Washington: International Monetary Fund).

    KimSoyoung and N.Roubini2000Exchange Rate Anomalies in the Industrial Countries: A Solution with a Structural VAR Approach,Journal of Monetary Economics Vol. 45 pp. 56186.

    LeidermanLeonardoR.Maino and E.Parrado2006Inflation Targeting in Dollarized Economies,IMF Working Paper No. 06/157 (Washington: International Monetary Fund).

    MishkinFrederic1996The Channels of Monetary Transmission: Lessons for Monetary Policy,NBER Working Paper No. 5464 (Cambridge, Massachusetts: National Bureau of Economic Research).

    PeersmanGert and F.Smets2001The Monetary Transmission Mechanism in the Euro Area: More Evidence from VAR Analysis,Working Paper No. 91 (Frankfurt: European Central Bank).

    RossiniRenzo and M.Vega2007El mecanismo de transmisión de la política monetaria en un entorno de dolarización financiera: El caso del Perú entre 1996 y 2006,Working Paper No. 17 (Lima: Banco Central de Reserva del Perú).

    SaizarAna Carolina and N.Chalk2008Is Monetary Policy Effective when Credit Is Low?IMF Working Paper No. 08/288 (Washington: International Monetary Fund).

    SimsChristopher1992Interpreting the Macroeconomic Time Series Facts,European Economic Review36 pp. 9751011.

See Mishkin (1996) for an in-depth discussion of the main channels of monetary policy transmission.

Although dollarization is an endogenous phenomenon and it will ultimately depend on the credibility of monetary policy, we take it as a given in this analysis. This is also assumed, for instance, in the works of Aghion, Bacchetta, and Banerjee (2000) and Céspedes, Chang, and Velasco (2004). This assumption is justified since dollarization (or dedollarization) is a reflection of the behavior of monetary policy over the long term. Although it will ultimately change endogenously, depending on the credibility of monetary policy, the various lags observed in such changes somewhat justify taking it as a given when studying empirically the short-term effects of monetary policy shocks.

Uruguay does not yet have a full-fledged inflation targeting regime, as the country only recently began transitioning toward this monetary arrangement. This transition took place in early 2005, when it began to conduct monetary policy with the objective of meeting a pre-announced inflation target. In addition, after September 2007, Uruguay adopted the policy rate as the main instrument for monetary policy. Previously, monetary policy was essentially run by setting objectives on monetary aggregates to meet their inflation target.

Our selection of countries is to some extent arbitrary. However, our main concern in this chapter is rather heuristic, and for this reason we decided to consider a number of cases that allow us to understand more thoroughly the monetary transmission mechanism in dollarized and non-dollarized economies. In this regard, we selected a number of countries in the Latin American region that operate under an inflation targeting regime with a similar degree of openness, which are dependent to a large extent on commodity prices and are also exposed to different levels of dollarization. With this objective in mind, we took three of the most relevant inflation targeting regimes in the region: Chile, Peru, and Uruguay. The decision to incorporate New Zealand was based on the fact that the country is a prominent benchmark in terms of running a well-established inflation targeting regime, has a strong dependence on commodity prices, and is also a small and open economy. This type of benchmarking helps assess more thoroughly the performance of the selected Latin American countries vis-à-vis key advanced economies that operate under an inflation targeting regime.

SeeGarcía-Escribano and Sosa (2011) for a discussion of the recent dedollarization trends in Peru and Uruguay.

See Appendix 8.1 for details on the variables used in Table 8.1.

The IMF’s 2010 Article IV Staff Report for Uruguay also states that Uruguay’s output volatility tends to be high relative to a number of selected peer countries.

This trend is partially explained by the evolution of certain prices in Uruguay that are somewhat controlled by economic authorities, notably transport prices and a number of selected utilities, limiting to some extent the variability of inflation in the country.

Policy rates are taken from the website of each central bank, and reflect a target publicly announced by the central bank on the overnight interbank interest rate of each country.

These results should be treated with some caution owing to the relatively short sample period for the estimation in the case of Uruguay. In any case, the results obtained here are broadly consistent with other studies that show a low pass-through from the interbank rate to active and passive interest rates in this country.

See Appendix 8.1 for details on the different variables used in the VAR model.

For robustness, estimations are compared with those obtained from a structural VAR model using an identification structure similar to that proposed in Kim and Roubini (2000), without showing major differences in results. In addition, the analysis considered all other possible orders in the Cholesky decomposition. The main conclusions of this section remain unaffected in all these cases.

In the case of New Zealand, information is mostly available on a quarterly basis. Quarterly data have been converted to monthly data by taking a linear trend between each pair of consecutive quarters. At the time of running the estimations, quarterly data were available through the third quarter of 2010.

Rossini and Vega (2007) point out that the presence of balance sheet effects may explain why economic activity seems to expand after the increase in the policy rate in the case of Peru.

Although inflation patterns in Uruguay and Chile look relatively similar in Figure 8.5, there is a key important difference worth mentioning. The initial positive response of inflation in Chile vanishes seven months after the shock and turns negative immediately afterward. This behavior contrasts with that of Uruguay, where the inflation response to the currency depreciation remains positive about 20 months after the shock. Therefore, the overall exchange rate pass-through, conceived as the response of inflation to a currency depreciation over the medium term, becomes substantially larger in Uruguay relative to the case of Chile.

The three-month response of inflation was chosen to account for the delay between the period in which the policy rate is changed and its final effect on inflation.

The credit-to-output ratio of each particular point in the scatter plot coincides with the stock of credit available at the end of the sample period of the associated rolling VAR.

Owing to data availability, we were not able to run the exercise considering only local currency dev

    Other Resources Citing This Publication