Measures of Monetary Policy in Latin America
Trung Thanh Bui
Falcuty of Economics and Business Administration,
University of Szeged, Szeged, Hungary
Keywords: Monetary Policy, Instruments, Interest Rate, Money Supply, Latin America.
Abstract: Although the instrumentation of monetary policy is still constantly debated in the existing literature, there is
a paucity of studies investigating this problem in emerging economies, especially after the recent global
financial crisis. The objective of the paper is to investigate the role of interest rate and money supply as an
overall measure of monetary policy in four major Latin America economies that follow the inflation targeting
framework. Evidence from causality test and the analysis of impulse response function shows that both
indicators have explanation for price movement. The price puzzle is clearly visible follows positive shocks to
interest rate. These findings suggest that a composite index can be a better measure of monetary policy and
other means should be conducted to improve the performance of the interest rate policy.
1 INTRODUCTION
The choice of an appropriate measure of monetary
policy is of importance in the analysis of monetary
policy (Bernanke–Mihov 1998). There are two main
reasons (Romer–Romer 2004). Firstly, it alleviates
the effect of the endogenous interaction between
changes in monetary policy and changes in the state
of the economy, thereby alleviating the problem of
underestimating the effect of monetary policy on
output and prices. Secondly, a representative
monetary policy indicator also helps to reveal the true
interaction between monetary policy and
macroeconomic outcomes.
Most of studies on monetary policy in emerging
economies have based on the prior that monetary
policy is properly measured by a single indicator such
as interest rate. See, for instance, Cermeño et al.
(2012) for Mexico; Furlani et al. (2010), Sánchez-
Fung (2011), Jawadi et al. (2014) for Brazil; or De
MelloMoccero (2011) for 4 Latin America
countries. However, the practical conduct of
monetary policy in emerging economies raises doubts
on the effectiveness of interest rate as the sole
measure of monetary policy, especially during the
post-crisis period. Although Latin America countries
have decided on interest rates as an official
operational target since the adoption of inflation-
targeting framework in the 1990s, they also depend
on other instruments to affect reserve money such as
reserve requirements, discount windows, and
exchange rate interventions. Such a multiple
instrument framework can stem from the insufficient
knowledge about the structure of the economy or the
distortion effect of objectives other than price
stability. For instance, the Central Bank of Brazil
simultaneously pursued several targets after crisis,
including inflation-targeting, flexible exchange rate,
and macroprudential regulations (Jawadi et al. 2014).
Apart from price stability, Bank of Mexico also
implicitly aimed at objectives such as output stability
(Cermeño et al. 2012).
Furthermore, previous studies are limited to the
pre-crisis period and, to the best knowledge of the
author, there is no studies investigating the relative
performance of interest rate and monetary aggregate
as an overall measure of monetary policy in Latin
America. Therefore, the performance of the two
indicators remain ambiguous in the last decade. Since
the choice of an appropriate monetary policy
indicator is the first step to analyse further issues of
monetary policy such as effectiveness, monetary
policy rules, or transmission channels, it is of
importance to have a rigorous study on the
effectiveness of various instruments.
This paper sheds light on some crucial issues
related to indicator problem of the monetary policy
analysis in four emerging economies in Latin
America, including Brazil, Chile, Colombia, and
Mexico. What is the superior indicator in Latin
Bui, T.
Measures of Monetary Policy in Latin America.
DOI: 10.5220/0009343600270036
In Proceedings of the 2nd International Conference on Finance, Economics, Management and IT Business (FEMIB 2020), pages 27-36
ISBN: 978-989-758-422-0
Copyright
c
2020 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
27
America? Is the usefulness of monetary policy
indicators different after the global financial crisis?
What is the role of monetary aggregates in the
conduct of monetary policy? The investigation of
these questions provides evidence about the role of
various monetary policy indicators in emerging
economies that follow inflation targeting framework.
Furthermore, understanding the instrumentation over
different time horizons contributes to the effective
implementation of monetary policy.
To compares the significance of money supply
and interest rate as an overall measure of monetary
policy on two bases. First, the causality analysis is
conducted to investigate the predictive power of
changes in a monetary policy indicator on inflation.
Such an analysis is of importance to examine the
tightness between the indicator and inflation. Second,
the analysis of the impulse response of inflation to a
monetary policy indicator indicates the magnitude of
the effect of monetary policy on inflation. The
preferred indicator of monetary policy should show
the dominance in fulfilling the objective of price
stability.
Turning to the key findings, the paper found that
there is shifts in the causal effect of interest rate
instrument in Latin America after crisis. In addition,
evidence from impulse response function indicates
that price puzzle is clearly visible after positive
interest rate shocks. On the contrary, inflation
response to monetary aggregates is more consistent
with the prediction of monetary theories. The results
of causality and impulse response analysis suggest
that neither interest rates nor money supply
sufficiently measures the stance of monetary policy in
Latin America. Therefore, a composite measure can
be a better measure of monetary policy. Other
suggestions are related to the improvement of the
compliance of the basic principle of inflation
targeting such as increasing the independence of
central banks or the performance of inflation forecast.
The rest of the paper is organized as follows.
Section 2 provides theoretical background for the
optimal choice of monetary policy indicators and
empirical studies of instrument problems in emerging
economies and Latin America in particular. Section 2
discusses how to investigate the relative significance
of interest rate and money supply instrument. Section
4 presents and discusses empirical results. Section 5
is conclusions and policy implications.
2 LITERATURE REVIEW
2.1 Monetary Policy Indicator
The indicator problem of monetary policy refers to
the controversy about the effectiveness of interest rate
and money supply in signalling the stance of
monetary policy. It arises because of incomplete
knowledge about the structure of the economy as well
as the existence of lagged effect of monetary policy
on economic targets.
Poole (1970) analysed the optimality of interest
rate and money supply based on a simple IS/LM
framework. Two primary assumptions of the analysis
are that monetary authorities have no errors in
controlling interest rate or money supply and they
must choose only one instrument to minimize the
volatility of output. The conclusion is that money
supply is preferred to deal with shocks from the real
sector while interest rate is superior in dealing with
shocks from monetary sector.
Following studies (Bhattacharya–Singh 2008,
Singh–Subramanian 2008) reached a similar
consensus. Likewise, Atkeson et al. (2007) argued
that monetary policy instruments can be ranked in
terms of tightness or transparency. They define
tightness as the strength of the linkage between
monetary policy instruments and target variables such
as inflation or output growth and define transparency
as the observability of instrument adjustments by the
public. Based on these criteria, they rank interest rates
as the best instrument because it has natural
advantages over exchange rate instrument and money
supply instrument in term of tightness and
transparency. Exchange rate is less preferred
instruments and money supply is at the bottom.
However, the consensus under Poole (1970)
framework may fall if real/ monetary shocks are
serially related or monetary authorities have
imperfect knowledge about the economy (Howells–
Bain 2003). The serial correlation is likely to happen
because monetary authorities concern about
smoothing the path of interest rate/money supply. The
violation also happens when there is lag in the system
or changes in the slope of IS and LM curve.
Moreover, Poole (1970) derives the consensus from
an ad hoc macroeconomic models, whereby the
derivation of aggregate demand and aggregate supply
does not base on consistent assumptions about the
behaviour of consumers and firms.
In the context of emerging economies, Poole
(1970) analysis has three primary limitations. First,
monetary authorities in emerging economies cannot
control monetary policy instruments as well as
FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business
28
counterparts in advanced economies. The low
performance is caused by the underdevelopment of
the financial system or the low expertise of
policymakers. Therefore, errors in controlling
instruments can be large and the conclusion about the
superiority of interest rate and money supply is not
clear-cut as the simple analysis of Poole (1970).
Second, output stabilization may not be the exclusive
objective of monetary policy, especially for central
banks that follows inflation targeting framework. The
focus can be the deviation of inflation from the target
or any weighted combination of output stabilization
and price stability. Therefore, the robustness of Poole
(1970) consensus is open to question for cases that
central banks have preference to other
macroeconomic outcome beyond output stabilization.
Final, monetary authorities in emerging economies
can choose both instruments rather than one. One
reason is that they are unsure about the source of
uncertainty. It is difficult to conclude whether
changes in the economy is from the real sector or the
monetary sector. It is highly likely that both monetary
shocks and real shocks have an effect on the
economy. Furthermore, in practice, the money supply
and interest rate are not necessarily competing but
they can be complementary. For instance, reserve
instrument can support interest rate instrument when
the level of financial friction is high (Sensarma–
Bhattacharyya 2016). Recently, central banks in
emerging economies have an additional task of
securing financial stability; therefore, they opt to use
reserve requirement instrument (Glocker–Towbin
2012).
Furthermore, several studies show that the use of
interest rate instrument does not always follow Poole
(1970) criteria. Higher output volatility when
targeting interest rates allows individuals and firms
more room to optimize the utility. In some cases,
interest rate can be employed because of political
pressure coped by monetary authorities (Cover–
VanHoose 2000). Particularly, monetary authorities
can lose credibility because of political pressure if
there is high degree of error in the control of reserve
instrument. This implies that interest rate instrument
can be employed even though it is suboptimal.
Small open economies also face more challenges
when deciding whether money supply or interest rate
is optimal. With interest rate parity assumption,
Gardner (1983) argued that monetary authorities in
small open economies encounter the trade-off
between money supply instrument and exchange rate
instrument when exchange rates are not fixed. When
the parity holds, it is equivalent in controlling interest
rate and exchange rate, thereby these instruments are
equivalent. In other words, monetary authorities have
two instruments at their disposal, exchange rate/
interest rate and money supply. However, there is no
general conclusion about the optimal choice based on
an ad hoc loss function that minimize sum of squares
of deviation of money supply and exchange rate from
their targets. The optimal choice of instruments
depends on the knowledge of the money demand and
money supply as well as the relative weight put on the
control of exchange rate. If monetary authorities have
perfect knowledge about money demand, interest rate
instrument is superior to reserve instrument. If they
know money supply perfectly, reserve instrument is
superior. However, when exchange rate movement is
of great concern, interest rate instrument can be
preferred even though monetary authorities
completely know the process of money supply. Under
New Keynesian framework, Singh–Subramanian
(2008) examined the superiority of money supply
instrument and interest rate instrument under
different types of shocks. Based on welfare yardstick,
they found that targeting money supply is preferred to
deal with demand (fiscal) shock whereas interest rate
is best to respond to supply (productivity) shock or
monetary (velocity) shock.
2.2 Empirical Choice of Monetary
Policy Indicator in Emerging
Economies
Empirical studies use both quantity-based indicators
such as monetary aggregates and price-based
indicators such as short-term interest rates to measure
monetary policy. Follow the seminal Sims (1972)
work, many studies use VAR-based innovations to
the growth rate of monetary aggregates as surprised
changes in monetary policy. In the 1990s, however,
many countries shifted to inflation-targeting
monetary policy (Howells–Bain 2003, Peters 2016).
The measure of monetary policy also changed, with
the focus moving onto price-based indicators.
While there is no agreement on the optimal choice
of monetary policy instrument, the prior that interest
rates are an appropriate measure of monetary policy
is popular for studies of emerging economies or Latin
America. Among many others, Furlani et al. (2010),
Sánchez-Fung (2011), De Mello–Moccero (2011),
Cermeño et al. (2012), Jawadi et al. (2014), (Aragón–
de Medeiros 2015) are studies that employ interest
rate as measure of monetary policy for investing the
performance of Taylor rule in Latin America.
However, the paucity of studies on indicator problem
results in the ambiguity of the relative effectiveness
Measures of Monetary Policy in Latin America
29
of monetary aggregates and interest rates in
measuring monetary policy.
3 METHODOLOGY AND DATA
3.1 Methodology
This paper compares the performance of different
indicators of monetary policy on two bases. First, the
monetary policy indicator should be causally related
to the objective of price stability. Second, the
preferred indicator of monetary policy should show
the dominance in fulfilling the objective of price
stability.
3.1.1 Causality Analysis
An indicator is effective when its adjustments result
in changes in targeted variables. Therefore, we can
determine the usefulness of a monetary policy
indicator by investigating its causal effect on target
variables. According to Sun–Ma (2004), if the
causality runs from instruments to prices/output, the
instruments are effective for price/output stability. By
contrast, if the causality runs from target variables to
the instrument, the instrument is considered as
endogenous. Since four Latin America countries in
the sample follow inflation-targeting framework, the
focus of the analysis is on how a monetary policy
indicator leads to inflation.
Granger (1969) causality test is a pioneering
method for examining the causality between
variables. Its VAR representation is:
01122
...
tttptpt
YYY Y
ββ β β
ε
−−
=+ + ++ +
(1)
Where
t
Y is a vector of k endogenous variables
and
t
ε
is white noise. Since this paper investigates
the causality between monetary policy indicators and
inflation,
t
Y consists of inflation and a monetary
policy indicator.
The standard VAR copes with the stationarity and
cointegration condition. However, if variables are
integrated or cointegrated, it leads to the violation of
the standard distribution of the Wald test in VAR
model (Toda–Yamamoto 1995). To overcome this
issue, Toda–Yamamoto (1995) suggested adding the
maximum integration order
d
into the selected lag of
the standard VAR
p , then estimating the VAR
system with the surplus lag
pd+ , and finally
conducting Granger test up to standard lag
p . The
additional lag
d
ensures that the Wald test of VAR
coefficients is asymptotic. The surplus lag VAR is:
01122
...
tttpdtpdt
YYY Y
ββ β β
ε
−− +
=+ + ++ +
(2)
This paper employs TodaYamamoto (1995)
method to test the causal effect of monetary policy
indicators on inflation in Latin America. One reason
for such a choice is that variables under investigation
are unlikely to be stationary at the same level for four
Latin America. Another reason is to ensure the
comparability of the results when simultaneously
considering several countries.
3.1.2 Relationship Analysis
The literature suggests that interest rate and money
supply can provide numeric information about the
direction and size of changes in the monetary policy.
To examine the impact of these measures on inflation,
we use a VAR model of four endogenous variables as
follows:
t
Y =[DLCPI DLY DLNEER DLM1/DLM2/R] (3)
Monetary policy indicators in equation (3)
include three variables: M1, and M2 and policy rate.
Therefore, equation (3) is regressed three times, each
time with one monetary indicator. Because of
stationary condition, variables that enter the VAR
model are first difference of their logarithm,
excepting for interest rate. In particular, DLCPI,
DLY, DLNEER, DLM1, DLM2 are the first
difference of the logarithm of consumer price index,
industrial production index, nominal effective
exchange rate, M1, and M2 respectively. R is the
policy rate.
The strength of the linkage between monetary
policy indicators and inflation is analysed through
impulse response function (IRF). IRFs indicate the
direction and the magnitude of the effect of
exogenous changes in monetary policy indicators on
inflation.
It should be noted that the VAR model is
recursive with the ordering specified in Equation (3).
The given ordering implies that inflation, output, and
exchange rate have a contemporaneous effect on a
monetary policy indicator while current changes in
the monetary policy indicator causes other variables
to changes in the future. Such a recursive causal
ordering requires minimum assumptions about the
structure in the VAR model.
FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business
30
3.2 Data
The paper examines the performance of various
monetary policy indicators in four Latin America
countries: Brazil, Chile, Colombia, and Mexico. The
study period starts at January 2002 for Brazil and
January 2000 for other countries. The sample ends at
June 2018 for all countries. We split the sample into
two subsamples to examine the influence of crisis on
the performance of monetary policy instruments. The
selected break point is June 2008.
Monetary policy indictors include both monetary
aggregates and interest rates. Monetary policy
aggregates are narrow money supply (M1), broad
money supply M2 (M2). Interest rate measure of
monetary policy (R) is Selic rate for Brazil, monetary
policy rate for Chile, central bank policy rate for
Colombia, and 91 days TIIE rate for Mexico. The data
is collected from website of corresponding central
banks.
It should be noted that four Latin America
countries adopted inflation targeting framework in
the late 1990s. Currently, Brazil has the target
inflation of 4.5% and 2% tolerance while other
countries aim for the target of 3% with 1% tolerance.
The performance of inflation stabilization is not quite
good in the region. Inflation rate are volatile,
especially at the begin of inflation targeting and after
the global financial crisis. Compared to other
countries, Brazil experienced a lengthy period of
stable inflation. These facts raise doubts about the
effectiveness of interest rate instrument in stabilizing
inflation in these countries.
4 EMPIRICAL RESULTS
4.1 Causality
This section discusses the causal relationship between
monetary policy indicators and inflation in Latin
America. The analysis is of importance because it
shows whether changes in a monetary policy
indicator lead to changes in inflation. Moreover, it
fills the weakness of correlation analysis in previous
studies. We present the results of Toda–Yamamoto
(1995) test since it accounts for the nonstationarity of
variables. As shown in Table 1, variables are not
integrated at the same level across countries. The
majority is stationary at first difference (superscript
a). Some variables are stationary at level (superscript
b).
Table 1: ADF test for the stationarity of variables.
Variable Brazil Chile Colombia Mexico
LY -7.49
*(b)
-
12.11
*(b)
-11.82
*(b)
-8.56
*(b)
LCPI -5.24
*(b)
-8.15
*(b)
-8.02
*(b)
-8.75
*(b)
LM1 -3.25
**(a)
-8.63
*(b)
-13.99
*(b)
-
16.69
*(b)
LM2 -3.57
*(b)
-8.11
*(b)
-17.09
*(b)
-
16.25
*(b)
R -4.97
*(b)
-3.05
**(a)
-4.96
*(a)
-4.68
*(a)
LNEER -7.92
*(b)
-3.11
**(a)
-9.16
*(b)
-9.21
*(b)
Source: Author’s calculation
Notes:
*
,
**
,
***
denote significance at 1%, 5%, 10%.
(a)
: unit root test at level.
(b)
: unit root test at first
difference. Lag is selected by SBIC criterion.
Table 2: The causal effect of monetary policy indicators on
price.
Before crisis Brazil Chile Colombia Mexico
R → LCPI 22.06
**
3.26 20.67
**
32.94
*
LM1→ LCPI 38.31
*
26.48
**
50.43
*
83.15
*
LM2 →
LCPI
130.76
*
26.89
*
85.72
*
10.88
After crisis
R → LCPI 11.06
*
47.6
***
12.85 57.5
***
LM1→ LCPI 0.17 27.69
**
43.86
***
71.16
***
LM2 →
LCPI
6.04 1.48 33.15
***
47.49
***
Source: Author’s estimation.
Notes:
*
,
**
,
***
denote significance at 1%, 5%, 10%.
As shown in Table 2, there is a shift in the
significance of the causality between monetary policy
indicators and price after crisis. The occurrence of the
global financial crisis leads to changes in the
significance of the interest rate instrument in two
countries, Chile and Colombia. While Colombia
copes with the loss in the causal effect of policy rate
on price, the reverse happens with Chile, whereby the
causality becomes statistically significant. For other
countries, changes in policy rate cause price to
change. Turning to monetary aggregates, the causal
effect of M1 on price is significant for Latin America
economies (except for Brazil during the post-crisis
period). Its significance does not alter during the post-
crisis period in most countries. M2 has a significant
causal effect on price in all countries excepting for
Mexico before crisis. This causality is statistically
significant for Colombia and Mexico after crisis.
Measures of Monetary Policy in Latin America
31
In summary, the recent global financial crisis has
a trivial effect on the causality between monetary
policy indicators and monetary policy objectives.
Overall, changes in both interest rate and money
supply lead to changes in prices in four Latin America
countries.
4.2 Impulse Response Analysis
We proceed by separately investigating the dynamic
effect of monetary aggregates and interest rates on
inflation (see Figure 1). Since policy rate and
logarithm of other variables have different order of
integration, we estimate recursive VAR as specified
in equation (3) by using interest rate and first
difference of other variables. Such a transform does
not affect the interpretation of the empirical results.
As shown in Figure 1, there are two panels
corresponding to the pre-crisis and post-crisis period.
Each panel indicates the response of inflation to
interest rate, M1, and M2.
The results show that interest rate shocks have
positive effect on inflation, indicating the presence of
price puzzle, a phenomenon labelled by Sims (1992).
This means that a restrictive monetary policy
constructed by raising interest rate does not lead to a
fall but a rise in inflation, which is counterintuitive.
For Brazil, the price puzzle is pronounced and
observable after the crisis while being muted before
crisis. This pattern can be a result of deconstructing
credibility of the Central Bank of Brazil (CBB).
According to Aragón–de Medeiros (2015), CBB
became less and less responsive to current and
expected inflation and eventually violated the Taylor
principle since the mid-2010. Cortes–Paiva (2017)
also pointed out that CBB follows excessively loose
monetary policy during the first administration of
Rousseff president, from 2011 to 2014. Other reason
might be the reluctance of CBB in fighting inflation
(Moura–de Carvalho 2010).
Similar results emerge for Chile, Colombia and
Mexico. It should be noted that the number of
instruments in Colombia have increased over time,
which is important for the attainment of multiple
targets such as price stability, economic growth,
financial stability, exchange rate stability, and
adequacy of international reserve. As a result, interest
rate is weak in representing monetary policy in
Colombia. For Mexico, Bank of Mexico has more
indirect influence on market interest rate before crisis.
At the beginning of inflation targeting, it uses two
instruments, Corto and minimum interest rate, to
signal the stance of monetary policy. As noted by
Garcia-Iglesias et al. (2013), overnight interbank is an
official operational target in Mexico after 2004 and it
only replaced Corto instrument after January 2008.
Corto refers to the system of target balances that
commercial banks must reserve at the central bank.
The central bank can announce a negative balance
target to signal a restrictive stance, which motivates
banks to chase for funds and increase market interest
rates. The existence of two instrument also indicates
that changes in interest rates show a part of changes
in the stance of monetary policy. After crisis,
however, the price puzzle is not persistent, reflecting
improvement in the performance of the interest rate
instrument.
The existing literature also suggests some
explanation for the existence of the price puzzle.
First, a disadvantage of a VAR model is its small
scale; therefore, it is likely that the VAR model fails
to capture important information that monetary
authorities use to forecast the movement of inflation
in the future (Sims 1992). The existing literature
(Sims 1992, Bernanke–Mihov 1998) suggests that the
inclusion of additional variables such as commodity
prices or asset prices can eliminate the problem of
price puzzle. However, this remedy is not always
effective, especially for the case of Latin America
under investigation (see Section 4.3).
Second, price puzzle can emerge because of
factors other than model misspecifications. First,
price puzzle can be a result of the effect of monetary
policy on the supply side of the economy. Interest
rates can be considered as capital cost of productive
production; thereby raising it leads to a rise in the cost
of borrowing and this cost will pass on consumers.
This implies a rise in the price level after an increase
in interest rate. If this effect dominates the effect of
demand reduction on prices, prices are higher rather
lower. Such a mechanism is also termed as cost
channel (Barth–Ramey 2001). Other reason is the
influence of information asymmetry. It is likely that
monetary authorities may have more information
about price movement than the private sector and they
will increase interest rates when expecting a rise in
the price level. However, the absence of complete or
perfect information leads to the fact that their
responses are insufficient to or too late to curb
inflation. Therefore, inflation increases rather than
fall after a rise in interest rates (Walsh 2010).
Furthermore, Latin America countries do not strictly
follow inflation targeting. This means that their
increase in interest rate is not larger than the increase
in inflation. According to Moura–de Carvalho (2010),
while Brazil and Mexico are more responsive to
inflation expectation; Chile is less responsive; and
Colombia is almost irresponsive. For Chile and
FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business
32
Figure 1: Inflation response to monetary policy indicators.
Colombia, the response of interest rate to inflation is
less than one-for-one. The weak inflation response
can also be a result of the high inflation expectation
of economic agents, which prolongs the process of
disinflation (Mackiewicz-Łyziak 2016).
Turning to monetary aggregates, Figure 1 shows
that inflation positively reacts to shocks to money
supply. This implies that a rise in money supply
causes inflationary pressure on the economy. Such a
response is of expected sign and is according to the
monetary theory. However, the negative response of
inflation to money supply shocks is found in Brazil
and Chile, which is counterintuitive.
Overall, the empirical results show the
ineffectiveness of monetary policy for price stability
in Latin America when measuring monetary policy by
a single indicator, either interest rates or monetary
aggregates. The existing literature suggests two
justifications for the insignificant effect of monetary
policy on prices. Firstly, the incorrect identification
leads to a dirty measure of exogenous shocks in the
system, which is less likely to happen in the paper.
Secondly, it is the choice of an inappropriate
indicators of monetary policy. The analysis of
causality and IRF provides evidence that the latter
may be an applicable explanation.
4.3 Robustness Tests
The literature (see, for instance, Sims 1992,
Bernanke–Mihov 1998) suggests that the presence of
price puzzle can be a result of the failure to
incorporate useful information for the forecast of
inflation. Therefore, we conduct a robustness test of
Before crisis
R
Brazil Chile Colombia Mexico
M1
M2
After crisis
R
M1
M2
Measures of Monetary Policy in Latin America
33
the model (3) by augmenting shocks to commodity or
oil prices. Since Latin America countries are small-
open economies, the evolution of commodity or oil
prices have impacts on domestic prices. However,
these countries are not likely to affect the price of the
world commodity or oil; therefore, shocks to these
variables are considered as exogenous. This means
that their shocks have contemporaneous effect on
domestic economic activities and changes in a
monetary policy indicator, but not the reserve. The
results (not shown, available upon request),
demonstrate that there is little difference in the
impulse response. The price puzzle is still present,
reflecting the positive effect of interest rate on
inflation. The effect of monetary supply on inflation
is positive, which is consistent with the prediction of
the monetary theory.
Another robustness test involves changing the
measure of inflation. Following Acosta-Ormaechea–
Coble (2011), we replace inflation measure in the
baseline model (equation 3) by the differential
between domestic inflation and the US inflation. This
approach is believed to isolate domestic inflation
from the effect of external factors, thereby removing
the presence of price puzzle. In this paper, we choose
US inflation because Latin America countries use US
dollar as an anchor currency. Other reason is that US
is a large economy that can affect the price of Latin
America economies, its small neighbours. We
estimate how changes in a monetary policy indicator
affect the inflation gap with and without considering
the influence of commodity or oil prices. The results
show that these changes do not solve the problem of
price puzzle and provide no general consensus about
the superiority of either policy rate or monetary
aggregates.
In summary, the paper provides evidence in
support of the argument that the price puzzle is a
result of low representative power of interest rate
other than model misspecification. To put it
differently, neither interest rate nor monetary
aggregate can summarize enough information about
changes in monetary policy.
5 CONCLUSIONS
What should be the primary indicator of monetary
policy: interest rate or some monetary aggregates? It
is a controversial issue in the analysis of monetary
policy and is limitedly investigated in emerging
economies. This paper sheds light on the usefulness
of interest rate and monetary aggregates as an overall
measure of monetary policy in four emerging
economies in Latin America. The results of causal
analysis and impulse response function demonstrate
that both policy rate and monetary aggregates explain
the movement of inflation. Moreover, monetary
aggregates dominate interest rate. There is strong
evidence suggesting the presence of price puzzle
following a positive shock to interest rate instrument.
The presence of price puzzle in Latin America
provides some crucial implications for the
implementation of the interest rate policy. As argued
by Torres (2003), interest rates must change with
larger amount when monetary authorities react to
both inflation and output gap than when they focus
exclusively on inflation. This implies that monetary
authorities have two possible options. First, they
should be more responsive to inflationary pressure.
Interest rate should be raised with larger magnitude to
create a contractionary effect on aggregate demand
and thus reducing prices. However, this approach
comes with the cost of greater variation in interest
rates, eventually increasing the volatility of expected
inflation (De Mello–Moccero 2009). Secondly,
monetary authorities should focus more on inflation
if they want to lower the degree of interest rate
volatility. This requires an increase in the dependence
for monetary authorities and substantial changes in
institutional setting for some countries in the region.
For instance, Bank of Brazil has low level of
independence and accountability (Barbosa‐Filho
2008); therefore, institutions should specify the role
of the central bank in choosing the target and
instruments as well as the penalty when the target is
not fulfilled. High level of independence is also
crucial for the maintenance of credibility, which takes
time for successful construction.
Furthermore, monetary authorities should
increase the effectiveness of forecasting inflation to
improve the performance of the interest rate policy
for price stability. When monetary authorities are
forward-looking, changes in interest rate depends on
the expectation and the forecast of inflation. If
monetary authorities systematically underestimate
the expected level of inflation, interest rates show
smaller response to changes in inflation. This results
in a reduction in the role of the interest rate as a
nominal anchor. Several tools are available for
effective management of inflation expectation of the
public: (1) obtaining greater insight into determinants
of inflation or the structure of the Phillips curve and
(2) using forward guidance to improve the
transparency of monetary policy (Mackiewicz-
Łyziak 2016).
Finally, the evidence that both monetary
aggregates and policy contains information about
FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business
34
changes in monetary policy suggests that a composite
measure is better than single indicator in capturing the
stance of monetary policy. Although the construction
of the composite measure is outside the scope of this
paper, this topic is deserved for deeper investigation.
It should also be noted that the paper is subjected
to some drawbacks. First, the sample is small, which
concludes only four emerging economies. Further
studies should add more countries to enrich the
information of the research sample. Second, the
parameters can be time-varying, which is out of the
scope of the paper. However, this issue should put
more emphasis in future studies.
REFERENCES
Acosta-Ormaechea, S. & Coble, D. (2011): Monetary
transmission in dollarized and non-dollarized
economies: The cases of Chile, New Zealand, Peru and
Uruguay. International Monetary Fund, Washington
DC.
Aragón, E. K. D. S. B. & De Medeiros, G. B. (2015):
Monetary policy in Brazil: evidence of a reaction
function with time-varying parameters and endogenous
regressors. Empirical Economics, 48(2), pp. 557-575.
Atkeson, A.- Chari, V. V. & Kehoe, P. J. (2007): On the
optimal choice of a monetary policy instrument.
National Bureau of Economic Research, New York.
BarbosaFilho, N. H. (2008): Inflation targeting in Brazil:
1999 2006. International Review of Applied
Economics, 22(2), pp. 187-200.
Barth, M. J. & Ramey, V. A. (2001): The cost channel of
monetary transmission. NBER macroeconomics
annual, 16, pp. 199-240.
Bernanke, B. S. & Mihov, I. (1998): Measuring monetary
policy. The Quarterly Journal of Economics, 113(3),
pp. 869-902.
Bhattacharya, J. & Singh, R. (2008): Optimal choice of
monetary policy instruments in an economy with real
and liquidity shocks. Journal of Economic Dynamics
and Control, 32(4), pp. 1273-1311.
Cermeño, R.- Villagómez, F. A. & Polo, J. O. (2012):
Monetary Policy Rules In A Small Open Economy: An
Application to Mexico. Journal of Applied Economics,
15(2), pp. 259-286.
Cortes, G. S. & Paiva, C. A. (2017): Deconstructing
credibility: The breaking of monetary policy rules in
Brazil. Journal of International Money and Finance,
74(6), pp. 31-52.
Cover, J. P. & Vanhoose, D. D. (2000): Political pressures
and the choice of the optimal monetary policy
instrument. Journal of Economics and Business, 52(4),
pp. 325-341.
De Mello, L. & Moccero, D. (2009): Monetary Policy and
Inflation Expectations in Latin America: LongRun
Effects and Volatility Spillovers. Journal of money,
Credit and Banking, 41(8), pp. 1671-1690.
De Mello, L. & Moccero, D. (2011): Monetary policy and
macroeconomic stability in Latin America: The cases
of Brazil, Chile, Colombia and Mexico. Journal of
International Money and Finance, 30(1), pp. 229-245.
Furlani, L. G. C.- Portugal, M. S. & Laurini, M. P. (2010):
Exchange rate movements and monetary policy in
Brazil: Econometric and simulation evidence.
Economic Modelling, 27(1), pp. 284-295.
Garcia-Iglesias, J. M.- Muñoz Torres, R. & Saridakis, G.
(2013): Did the Bank of Mexico follow a systematic
behaviour in its transition to an inflation targeting
regime? Applied Financial Economics, 23(14), pp.
1205-1213.
Gardner, G. W. (1983): The choice of monetary policy
instruments in an open economy. Journal of
International Money and Finance, 2(3), pp. 347-354.
Glocker, C. & Towbin, P. (2012): Reserve Requirements
for Price and Financial Stability-When are they
effective? International Journal of Central Banking,
8(1), pp. 65-114.
Granger, W. J. C. (1969): Investigating causal relations by
econometric models and cross-spectral methods.
Econometrica: Journal of the Econometric Society, pp.
424-438.
Howells, P. & Bain, K. (2003): Monetary Economics:
Policy and its Theoretical Basis. Palgrave Macmillan,
New York.
Jawadi, F.- Mallick, S. K. & Sousa, R. M. (2014): Nonlinear
monetary policy reaction functions in large emerging
economies: the case of Brazil and China. Applied
Economics, 46(9), pp. 973-984.
Mackiewicz-Łyziak, J. (2016): Central bank credibility:
determinants and measurement. A cross-country study.
Acta Oeconomica, 66(1), pp. 125-151.
Moura, M. L. & De Carvalho, A. (2010): What can Taylor
rules say about monetary policy in Latin America?
Journal of Macroeconomics, 32(1), pp. 392-404.
Peters, A. C. (2016): Monetary policy, exchange rate
targeting and fear of floating in emerging market
economies. International economics and economic
policy, 13(2), pp. 255-281.
Poole, W. (1970): Optimal choice of monetary policy
instruments in a simple stochastic macro model. The
Quarterly Journal of Economics, 84(2), pp. 197-216.
Romer, C. D. & Romer, D. H. (2004): A New Measure of
Monetary Shocks: Derivation and Implications. The
American Economic Review, 94(4), pp. 1055-1084.
Sánchez-Fung, J. R. (2011): Estimating monetary policy
reaction functions for emerging market economies: The
case of Brazil. Economic Modelling, 28(4), pp. 1730-
1738.
Sensarma, R. & Bhattacharyya, I. (2016): Measuring
monetary policy and its impact on the bond market of
an emerging economy. Macroeconomics and Finance
in Emerging Market Economies, 9(2), pp. 109-130.
Sims, C. A. (1972): Money, income, and causality. The
American Economic Review, 62(4), pp. 540-552.
Sims, C. A. (1992): Interpreting the macroeconomic time
series facts: The effects of monetary policy. European
Economic Review, 36(5), pp. 975-1000.
Measures of Monetary Policy in Latin America
35
Singh, R. & Subramanian, C. (2008): The optimal choice of
monetary policy instruments in a small open economy.
Canadian Journal of Economics/Revue canadienne
d'économique, 41(1), pp. 105-137.
Sun, H. & Ma, Y. (2004): Money and price relationship in
China. Journal of Chinese Economic and Business
Studies, 2(3), pp. 225-247.
Toda, H. Y. & Yamamoto, T. (1995): Statistical inference
in vector autoregressions with possibly integrated
processes. Journal of econometrics, 66(1-2), pp. 225-
250.
Torres, A. (2003): Monetary policy and interest rates:
evidence from Mexico. The North American Journal of
Economics and Finance, 14(3), pp. 357-379.
Walsh, C. E. (2010): Monetary theory and policy. 3
rd
. MIT
press, Massachusetts.
FEMIB 2020 - 2nd International Conference on Finance, Economics, Management and IT Business
36