Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

Gadjah Mada International Journal of Business
Vol. 16, No. 1 (January-April 2014): 1-22

Interest Rates Targeting of Monetary Policy:
An Open Economy SVAR Study of Malaysia
Zulkefly Abdul Karima* and Bakri Abdul Karimb
a

Faculty of Economics and Management Universiti Kebangsaan Malaysia (UKM)
b
Faculty of Economics and Business, Universiti Malaysia Sarawak (UNIMAS)

Abstract: This paper examines the implementation of monetary policy during the interest rates targeting
in a small-open economy (i.e. Malaysia) by using an open-economy structural VAR (SVAR) study. It tests
the effect of foreign shocks upon domestic macroeconomic fluctuations and monetary policy, and examines how effective monetary policy is in influencing macroeconomic variables. The results show that
during interest rates targeting, monetary policy plays a significant role in affecting macroeconomics variables. This finding suggests that monetary policy has an important role as a stabilization policy in a smallopen economy.
Abstrak: Artikel ini membahas pelaksanaan kebijakan moneter selama suku bunga menargetkan dalam
perekonomian kecil terbuka (seperti Malaysia) dengan menggunakan studi VAR struktural perekonomianterbuka (SVAR). Hal ini menguji pengaruh guncangan asing pada fluktuasi makroekonomi domestik dan
kebijakan moneter, dan meneliti seberapa efektif kebijakan moneter dalam mempengaruhi variabel
makroekonomi. Hasil penelitian menunjukkan bahwa selama penargetan suku bunga, kebijakan moneter
memainkan peran penting dalam mempengaruhi variabel makroekonomi. Temuan ini menunjukkan bahwa
kebijakan moneter memiliki peran penting sebagai kebijakan stabilisasi dalam perekonomian kecil terbuka.

Keywords: interest rates targeting; monetary policy; open-economy; SVAR

* Corresponding author’s e-mail: mz@ukm.my or mzkarim72@yahoo.com
ISSN: 1141-1128
http://www.gamaijb.mmugm.ac.id/

1

Karim and Karim

Introduction
The main question relating to the implementation of monetary policy is either the
central bank has to target monetary aggregate or interest rates as an operating target.
In Malaysia’s experience, the Bank Negara
Malaysia (BNM), as the nation’s Central Bank
has changed its monetary policy strategy from
monetary targeting towards interest rate targeting since November 1995. During interest rate targeting, monetary policy operates
through short-term interest rates to attain the
ultimate target that is a sustainable long-run
economic growth, accompanied with price and
financial stability. 1
The aim of this study is to examine
empirically on the implementation of monetary policy during interest rates targeting in
a small and highly open economy, with special reference to Malaysia. This study uses a
non-recursive structural vector autoregressive
(SVAR) modelling framework in nine variables, which include three foreign variables,
two domestic monetary policy variables, and
four macroeconomic variables. By using
SVAR study, we can identify the structural
economic shocks by imposing a set of restrictions to the contemporaneous interactions of
the model variables. For example, after the

model has been specified, the impacts of foreign shocks on domestic macroeconomic and
monetary policy can be analyzed by using the
structural impulse-response function.
The main contributions of this paper
have two aspects. First, this study considers
the role of the foreign factor in modelling an
open-economy SVAR. The previous study of
monetary policy effects in Malaysia, for example Azali and Matthews (1999) and
Ibrahim (2005) have used a small-scale VAR
in a closed-economy, which does not consider
the role of foreign variables in their analysis.
Therefore, it is essential to examine the foreign shocks effects on macroeconomic fluctuations and monetary policy because the
Malaysian economy is relatively small and
highly trade-dependent.2 Second, this study
employs an open economy structural VAR
model, which permits an identification of
strategies based on economic theory rather
than the sometimes questionable assumptions
which underlie a traditional recursive VAR.
In the Malaysian context, although Tang
(2006) has considered recursive VAR in an
open-economy, and Ibrahim (2005) has used
recursive VAR in a closed-economy, however
their identification of structural shocks is
inappropriate. Thus, this study provides a
novel contribution to the monetary policy

1

During the interest rate targeting, monetary policy in Malaysia can be categorized into three main evolutions.
Firstly, from November 1995 up to September 1998, the BNM has introduced a new Base Lending Rate (BLR)
framework which takes into account the 3-month interbank rate in the BLR formula. Secondly, since September 1998,
the BNM has employed interest rate targeting with a fixed exchange rate, and modified the BLR framework taking into
account the Intervention Rate in the determination of BLR formula. At the same time, due to the currency crisis that
occurred in the East Asian region, the BNM implemented capital controls to stabilize the economy. Thirdly, since April
2004, the BNM has introduced a new interest rates framework, the Overnight Policy Rate (OPR) to signal the monetary
policy stance. During this period, the BNM has gradually liberalized capital control, and has eliminated the pegging with
the US dollar since July 2005.
2

In Malaysia, the degree of economic interdependence or economic openness as measured by the share of
exports and imports as a percentage of GDP has increased significantly from 86.88 percent in 1970 to 112.59 percent in
1980 and 146.89 percent in 1990. In fact, since 1999 it has been greater than 200 percent. These statistics indicate that the
Malaysian economy is highly dependent, and thus vulnerable to external shocks, for example foreign income shocks
from large countries.

2

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

analysis in a small-open economy (i.e. Malaysia) by identifying the structural shocks
according to the economic theory. The results
of the study indicate that monetary policy has
a significant role as a stabilization policy during interest rates targeting in stimulating macroeconomic variables, and mitigating the
negative effects of the external shock to the
domestic economy. Foreign shocks appeared
to play a prominent role in influencing domestic macroeconomic and monetary policy
variables. There is also no evidence for output puzzle, and exchange rate puzzle. However, there is a price and liquidity puzzle during the interest rates targeting.
The rest of the paper is organized as
follows. Section 2 presents a literature review
relating to the monetary policy identification
scheme in the SVAR literature. Section 3
briefly discusses the research methodology,
and Section 4 presents the empirical results
by focusing on structural impulse-response
function (SIRF). Section 5 summarizes and
concludes.

Literature Review
There are a limited number of studies
in small-open economies which examined the
effect of the monetary policy shocks by using an open SVAR approach. For example,
the most recent SVAR studies of a small-open
economy were conducted by Cushman and
Zha (1997), Brischetto and Voss (1999),
Dungey and Pagan (2000), Parrado (2001),
and Buckle et al. (2007). Most of the studies
have used block exogeneity restrictions in
modelling the international economic linkages
to the small-open economy.

Cushman and Zha (1997) and Dungey
and Pagan (2000) have imposed two blocks
in their structural equation model, which is a
block representing the international economy,
and a block representing the domestic
economy. In modelling SVAR for the Canadian economy, Cushman and Zha (1997) included four international variables, namely
the US industrial production, the US consumer prices, the US federal fund rate, and
world total commodity export prices. The
main identification scheme in their study has
three aspects. First, domestic interest rates are
assumed to react contemporaneously to foreign interest rates and commodity markets,
but not on contemporaneous output. Second,
the exchange rate is assumed to respond contemporaneously to all shocks. Third, foreign
variables are treated as a separate block,
which is block (exogenous) for the domestic
(small open) economy. This means that, domestic variables are not allowed to affect foreign variables either contemporaneously or
with a lag. The empirical findings stated that
Canadian monetary policy responded significantly and contemporaneously with home
interest rates, exchange rate, foreign interest
rates and foreign price levels. In addition,
there is no evidence of interest rates puzzles,3
exchange rates puzzles and prices puzzles,
and the external shocks become the dominant source of domestic output fluctuations.
Monetary policy shocks (an increase in interest rates) has a small but negative effect on
output.
In contrast, Dungey and Pagan (2000)
included five international variables, namely
real US GDP, real US interest rates, the Australian term of trade,4 the Dow Jones Index

3
The interest rates puzzle or output puzzle is a monetary contraction where-by an increase in interest rates is
associated with an increase in output level rather than a decrease.
4

Terms of trade refers to the relative price of exports in terms of imports, and it defined as the ratio of export
prices to import prices.

3

Karim and Karim

deflated with the US consumer price index,
and real exports. The Australian monetary
policy (cash rate) is assumed to follow a standard Taylor-rule, which is response contemporaneously with Australian gross national
expenditure and inflation. The domestic variables are assumed not to affect foreign variables either contemporaneously or with a lag.
The main findings indicated that overseas
factors are generally a substantial contributor to domestic activity, and domestic monetary policy contributes to stabilizing economic activity, but the effect is not large.
In the Malaysian context, there have
been few studies relating to the effect of
monetary policy shocks on economic activities in the existing literature. For example,
Azali and Matthews (1999) have employed
Bernanke’s (1986) contemporaneous structural VAR approach (six variables in a closedeconomy) in investigating the relationship
between money-income and credit-income
during the pre- and the post-liberalization eras.
They found that, during the pre-liberalization
period, the bank credit shock had more impact compared with the money shock in explaining output variability. In contrast, after
the post-liberalization period, money, as well
as credit innovations, were significant in explaining output shocks. In short, aggregate
demand was significantly influenced by credit
innovation during pre-liberalization, while
money innovation played an important role
during post-liberalization in explaining output variability.
Another study by Ibrahim (2005) used
recursive VAR identification in a closed
economy in examining the sectoral effect of
a monetary policy shock. His results supported the real effect of monetary policy
shocks. For example, it was seen that real
output declined during monetary tightening
4

(positive shocks of interest rates). In fact,
some sectors such as manufacturing, construction, finance, insurance, real estate, and
business services seem to decline more than
aggregate production in responding to the
interest rates shocks. In short, those sectors
that are very heavily dependent on bank loans
are more sensitive to monetary policy shocks.
In comparison, a recent study by Tang (2006)
has examined the relative importance of the
monetary policy transmission mechanism
channel (interest rates, credit, asset price, and
exchange rate channel) by using 12 variables
in an open-economy VAR model. The variables are four foreign variables (foreign
block), and eight domestic variables (domestic block), which is estimated by using a recursive VAR identification scheme. The foreign block is assumed not completely exogenous to the domestic block, in which the
domestic variables are allowed to affect the
foreign variables in a lag but not contemporaneously. His findings conclude that the interest rates channel plays a pivotal role in influencing output and inflation. In addition,
the asset price channel is also relevant for
explaining output variability, but for inflation,
the exchange rate channel is more relevant
than the asset price channel.

Methods
Data and Variables Description
All data are at a monthly frequency
spanning from November 1995 until May
2009 (during interest rates targeting regime),
and collected from the IMF’s International
Financial Statistic (IFS), except for asset price,
where the data are collected from Thompson
Data stream. All variables are transformed
into logs except for FFR, INF and IBOR,
which are stated in percentage points. Spe-

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

cifically, the endogenous variables included
in the VAR are;

etary policy stance and macroeconomic fluctuations in a small-open country economy.

Oil Price (LOIL). It refers to the log of world

Domestic Income (IPIM). Domestic in-

oil prices (US $ per barrel). The world oil price
is an indicator of adverse supply shocks (inflationary and contractionary) to the Malaysian economy. Therefore, it is expected that
positive shocks in the world oil prices will
increase the cost of production, and subsequently will trigger cost-push inflation and
reduced investment spending, subsequently
leading to a contraction in economic activity.

come is a target variable, which is proxied by
the Malaysian Industrial Production Index. In
Malaysia, the Industrial Production Index
(IPI) is a good proxy for output because it
comprises the three main indices in the
economy; the manufacturing index (weighted
70.7 %), mining index (weighted 23.4 %), and
electricity Index (weighted 6.0%).

Foreign Income (IPIUS). This study uses
the US Industrial Production Index as a proxy
for foreign income. The selection of US income as a proxy for foreign income is reasonable since the US is the major trading partner
of Malaysia. For instance, on average, from
1997 to 2008, exports to the US have constituted 20 percent of the total Malaysian exports, most of which have been concentrated
on manufacturing products such as electrical
and electronic goods and textiles, clothing and
footwear.5

Foreign Monetary Policy (FFR). The Federal Fund Rate of the US is used as a proxy
for foreign monetary policy. FFR is also an
appropriate indicator in analyzing the effect
of foreign monetary policy on a small-open
economy given that the US is a large economy
and has a powerful impact in the international
arena. Any changes in the FFR will signal the
condition of the US economy, and subsequently is expected to transmit to the mon-

Inflation (INF). Another target variable is
the inflation rate (INF) which is calculated
from the changes in the Consumer Price Indices (CPI) 6. The inclusion of INF in the
model is reasonable given that the BNM has
been targeting the inflation rate as an objective since 1990.

Monetary Aggregate (LM). The narrow
monetary aggregate M1 is also considered as
a monetary policy variable because M1 is
more liquid, and thus demanded for transactions purposes. The inclusion of M1 is also
consistent with previous studies, for example
Tang (2006) who stated that M1 is the best
candidate for a monetary aggregate because
it is relatively stable and more liquid compared to the broad monetary aggregate.

Interest Rates (IBOR). The interbank overnight rate (IBOR) is also a good candidate
for a monetary policy variable because the
BNM has directly influenced the interbank
rate through its intervention in the money
market (Domac, 1999). In fact, since April
2004, the BNM has introduced a new mon-

5
This figure is based on the author’s calculation from the Bank Negara Malaysia, Monthly Statistical Bulletin.
Besides the US, the other Malaysian major trading partners are ASEAN countries (particularly Singapore), Japan and the
European Union. For instance, in 2008 the US, Singapore, Japan and the European Union have absorbed 52.23 percent
of the total Malaysian exports.

6

The standard formula to compute the inflation rate is

CPI t  CPI t 1
 100
CPI t 1

5

Karim and Karim

etary policy framework, namely the Overnight
Policy Rate (OPR) to signal the monetary
policy stance.

Kuala Lumpur Composite Index (KLCI).
Kuala Lumpur Composite Indices (KLCI) is
also important for analyzing the effect of the
monetary policy on the asset prices, and can
examine how the economic activity responds
to the stock market development.7 For instance, an easing of monetary policy through
increasing the amount of the money supply
(or reduction in interest rates) leads to an increase in the stock market (asset price), and
subsequently will stimulate the economic
activity via two channels - a wealth effect and
a Tobin’s-q effect.

Nominal Effective Exchange Rate
(NEER). Besides asset prices, the role of
the exchange rate is also important given that
Malaysia has a highly trade-dependent
economy, and any changes in exchange rate
will significantly affect the external sectors,
particularly export and import.
In addition to the endogenous variables,
the model also includes two dummy variables
for breaks in the intercept, which coincide
with major economic events. Specifically, the
events are: the period in which the Ringgit
was pegged to the US dollar, and the Asian
financial crisis. It is assumed these two events
only influence variables in the domestic block
rather than the foreign block. In addition, the
seasonal dummy has also considered for all
endogenous variables.

7
8

Time Series Properties
Much of the SVAR literature in monetary policy has ignored possible nonstationarity of the variables in the VAR
model.8 However, ignoring nonstationarity in
the VAR system is inappropriate because only
stationary variables return to the long-run
mean after a shock. In this study, the Augmented Dickey-Fuller (ADF) test was employed to investigate the presence of unit
roots in the data. After testing for unit roots,
the optimum lag in the VAR model is chosen
by using the Akaike information criteria
(AIC).

Malaysian Structural VAR
Modelling
The endogenous variables are divided
into two blocks; namely foreign and domestic. The foreign block includes the world oil
price, foreign income, and a foreign monetary
policy measure. The domestic block includes
target variables (domestic output and inflation), policy variables (monetary aggregate and
interest rates), and two other variables namely
asset prices and the exchange rate. It is assumed that a small-open economy like Malaysia is described by a structural form representation. The dynamic relationship of the
system of equation in the structural model
can be written as follows;
A0Yt= 0D0 + A (L)Yt + t ..........(1)

An excellent literature survey about the effect of monetary policy on stock prices can be found in Sellin (2001).

Ramaswamy and Sloek (1997) have discussed extensively whether the VAR model should be estimated in level,
difference or in vector error correction model (VECM).

6

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

Where,
A0

is an invertible square matrix of coefficients relating to the structural contemporaneous interaction between the variables in the system,
is a 91 matrix that is the vector of
system variables or [LOIL LIPIUS
FFR LIPIM INF VLM IBOR
LNEER LKLCI],

Yt

D0

is a vector of deterministic variables
(constant, trend and dummy variables),

A(L) is a k th order matrix polynomial in the
lag operator L[A(L)= A1L - A2L2 - ... AkLk],
t =

[tLOIL tLIPIUS tFFR tLIPIM tINF tLM
tIBOR tLNEER tLKLCI]tLOIL is the vector of structural shocks which satisfies
the conditions that E(t  E(ts)=
 I (identity matrix) for all t= s

Equation (1) cannot be directly estimated to derive the true value of A0, A(L)
and t. However, equation (1) can be estimated by transforming to the reduced form
representation as follows;
Yt= A0-10D0 + A0-1A(L)Yt + A0-1t.....(2)
or
Yt= A0-10D0 + 1(L)Yt + t.........(3)
Where, 0= A0-10, 1= A0-1A(L), t= A0-1 and
E (tt')= A0-1A0-1'= 

The solution to the SVAR system can
be generated by recovering the relationship
between the reduced-form disturbances (  t ),
and the underlying structural shocks (  t ).
This relationship can be estimated by using

the Equation (3), which is t= A0-1t or A0ut=
t using the maximum likelihood estimates.
In matrix form, this relationship can be represented as follows;
 1
 0
 a21
a 0
 31
0
 a41
 0
a
 51
a 0
 61
0
 a71
 0
 a81
a 0
 91

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0
a32

1

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0
a54

1

0

0

0

0
0
0
a62
a63
a64

0
a65

1

0

0

0
0
0
a72
a73
a74

0
0
a75
a76

1

0

0
a82

0
a83

0
a84

0
a85

0
a87

1

0
a92

0
a93

0
a94

0
0
a95
a96

0
a86

0
0
a97
a98

0  LOIL   LOIL 
t
 t
0  LIPIUS   LIPIUS 
 t
  t

0   FFR    FFR 
t
 t
 

0   LIPIM    LIPIM 
t
 t
0   INF     INF 
 t
  t

0   tIBOR    tIBOR 

 

0  utLM    tLM 
  LNEER   LNEER 

0  t
 LKLCI   tLKLCI 
   t

1   t

...............................................................(4)
In identifying the structural VAR model,
this study employed the SVAR-A model proposed by Amisano and Giannini (1996).
Enough restrictions have to be imposed on
matrix A0. According to the order condition,
for the system to be just identified or exactly
identified, it requires K(K-1)/2= 9(8)/2= 36
zero restrictions on the contemporaneous
matrix A0. However, since the contemporaneous matrix A0 in Equation (4) has 40 zero
restrictions, the SVAR model is over-identified.
In addition to the endogenous variables,
the model also includes two dummy variables
for breaks in the intercept, which coincide
with major economic events. Specifically, the
events are: the period in which the Ringgit
was pegged to the US dollar, and the Asian
financial crisis. It is assumed these two events
only influence variables in the domestic block
rather than the foreign block. In addition, the
seasonal dummy has also considered for all
endogenous variables. By considering the two
dummy variables and seasonal dummy, it is
believed that the structural of the impulse
response function will not change and be
stable during the sample of the study. Therefore, it is not necessary to split the sample

7

Karim and Karim

according to pre-crisis and post crisis period.
In fact, by splitting the sample, it will decrease
the number of observations across the subsample, and also will lose the degree of freedom in VAR analysis after considering the
optimum lag. Thus, for that reason this study
will not split the sample size.9

Identification Scheme
The variables in foreign blocks have been
assumed to be completely exogenous to the
domestic blocks. It is common to identify that
the foreign variables do not respond contemporaneously or with a lag to the movement
in the domestic variables in a small-open
country. Specifically, the world oil price is a
structural disturbance or exogenous variable
that is uncorrelated with other contemporaneous shocks. Meanwhile, the US income and
Federal Fund Rate (FFR) can influence world
oil price in a lag. The US income is assumed
to respond contemporaneously with the world
oil price and all foreign variables lag. The US
monetary policy (Jensen and Mercer 2002) is
assumed to respond contemporaneously to
the innovations in world oil price and the US
income and all foreign variables lag. The FFR
reacts positively to the world oil price in minimizing the inflationary pressure due to the
positive shocks of world oil prices. This assumption is also consistent with prior studies, for example, Hamilton (1996) and
Bernanke et al. (1997) who found that oil
price movements have a significant forecast
power for the stance of monetary policy in
the US economy. In contrast, FFR responds
negatively with the US income in minimizing
the output gap.
9

Domestic output growth has been assumed to respond contemporaneously only
to the world oil price shocks, and respond in
the lags to all endogenous variables. Domestic output growth reacts positively to the current growth of world oil prices. This assumption is reasonable given that Malaysia is a net
exporter of oil. However, in the long run, the
relationship between oil price and output is
expected to be negative because an increase
in the world oil price will increase the cost of
production, whereas the firms will respond
by cutting the level of production or investment. In addition, it is assumed that Malaysian output does not respond contemporaneously with other variables in the system. For
instance, the monetary policy variable, that
is, monetary aggregate and interest rates, cannot influence output contemporaneously. The
main plausible justification for this assumption is that firms do not change their output
and prices instantaneously in responding to
the monetary policy signal within a month,
due to the inertia, adjustment costs and planning delay, but they will respond immediately
to the current oil prices following their markup rule (Kim and Roubini 2000). This type
of restriction is also imposed by Bernanke
and Blinder (1992), and Bernanke and Mihov
(1998).
Domestic inflation has been assumed to
respond contemporaneously to the innovations in oil prices and domestic output. The
positive innovation in oil prices and domestic output will spontaneously accelerate the
domestic price level. However, other variables in the system cannot influence domestic inflation spontaneously because inflation

In their study about foreign shocks and monetary policy in Malaysia, Zaidi and Fisher (2010) also did not split
the sample of observation into subsamples. They argues that incorporating dummy variables to represent major events
in Malaysian Economy, the shocks (Impulse Response Function) in VAR model will be more stable than splitting the
sample observation.

8

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

is a slow-moving variable. However, all endogenous variables are assumed to affect the
inflation rate in the lags.

ous studies, for example, Kim and Roubini
(2000), Cushman and Zha (1997), and Sim
and Zha (2005).

In market economies, the use of the
interest rates as a major instrument of monetary policy does not imply that it can ignore
the role of the money supply. This is because
the interest rates are determined in financial
markets. For example, if the monetary authority wants to lower the interest rates but do
not support this aim with a required increase
in the money supply, it would find that the
market interest rates would deviate from its
desired level. As a result, the intended effects
on expenditure will not be achieved. Therefore, an interest rate policy must be accompanied by an appropriate money supply
(Handa 2009). For that reason, this study will
consider two types of monetary policy
shocks, namely, money supply and interest
rates.

It is assumed that the monetary authority (the BNM) sets the money supply after
observing the current level of all foreign variables, domestic output growth, inflation, and
interest rates. This is reasonable given that
the monetary authority can observe the data
on a monthly basis and choose the amount
of money supply to offset any negative effect resulting from foreign shocks, domestic
aggregate demand and inflation shocks to the
domestic economy.

During interest rates targeting, it is assumed that the monetary authority sets the
interest rates after observing the current value
of domestic variables ie. domestic output and
inflation, and all foreign variables, but not the
current value of other variables. The inclusion of output and inflation in the monetary
policy reaction function is reasonable given
that the central bank can observe data from
these variables on a monthly basis. The inclusion of foreign monetary policy in the domestic monetary policy reaction function is
reasonable given that the Malaysian economy
is highly dependent on the US economy.
Therefore, it is reasonable to assume that the
BNM will respond positively to the US monetary policy in minimizing the capital outflow
as well as stabilizing the domestic currency.
The justification of the monetary policy reaction function is also consistent with previ-

The exchange rate is assumed to respond contemporaneously with all foreign and
all domestic variables (except stock prices).
In contrast, the asset price is assumed to respond contemporaneously to all foreign and
all domestic variables. This assumption is reasonable given that the exchange rate and
stock market are the fast-moving variables
in the system. The asset price is assumed to
respond contemporaneously to the exchange
rate because any changes to the exchange rate
will influence international capital mobility,
which in turn affects the stock market. For
instance, an appreciation in domestic currency makes domestic assets more expensive
to the international investors, and therefore
decreases the demand for domestic asset,
subsequently leading to a decline in the asset
price.

Empirical Results
Preliminary Analysis
Table 1 reports the result of the unit
root test by using the Augmented DickeyFuller test. As can be seen, only three vari-

9

Karim and Karim

ables, which are the FFR, inflation (INF), and
interbank overnight rate (IBOR) are stationary at level form at least at 10 percent significance level; meanwhile other variables are
stationary in first difference form.10 The optimum lag in the VAR system is four according to the Akaike information criterion (AIC).
Since the baseline SVAR model is over-identified, we need to test the validity of the overidentification restrictions. The value of 2
(with four degrees of freedom) is 7.97 and
the probability is 0.11, which indicates that
the over-identification restrictions are valid.11

Structural Impulse Response
Function (SIRF)
Appendix 1-Appendix 3 illustrates the
structural impulse response functions of the
endogenous variables in this study. The main
focus is to analyze the foreign shock effects
on domestic variables (monetary policy and
macroeconomic variables), the effect of domestic monetary policy shocks (monetary
aggregate and interest rates) on domestic
macroeconomics variables, and how effective
is monetary policy in mitigating the negative

Table 1. Unit Root Test: Augmented Dickey Fuller (ADF)
Variables

Level Form

First Difference

Constant and
No Trend

Constant and
Trend

Constant and
No Trend

Constant and
Trend

LOIL

-2.009 (2)

-2.477 (2)

-11.578***(1)

-11.606*** (1)

LIPIUS

-1.842 (12)

-1.870 (12)

-6.531*** (10)

-6.593*** (10)

FFR

-3.336 ***(12)

-3.896 ***(12)

-4.791** (12)

-4.822*** (12)

LIPIM

-1.571 (12)

-1.694 (12)

-4.153*** (12)

-4.323*** (12)

INF

-3.446 ***(10)

-3.365 ***(10)

-6.816*** (12)

-6.935*** (12)

LM1

0.433 (10)

-1.802 (9)

-3.054*** (12)

-3.077*** (12)

IBOR

-2.846* (6)

-3.214* (7)

-7.972*** (7)

-7.961*** (7)

LNEER

-1.373 (3)

-2.817 (3)

-8.144*** (7)

-8.592*** (2)

LKLCI

-2.008 (3)

-2.913 (3)

-10.719***(2)

-10.711*** (2)

Note: *** Denotes significant at the 1 percent level, ** significant at 5 percent level and * significant at 10 percent level
which reject of the null hypothesis on non-stationary. Critical value obtain from Fuller (1976) for constant but no
time trend is -3.45, -2.87 and -2.57 for 1 percent, 5 percent and 10 percent significant level respectively, and the
critical value for constant and time trend is -3.98, -3.42 and -3.13 for 1 percent, 5 percent and 10 percent significant
level respectively.
Number in bracket is the optimum lagged based on Akaike Information Criterion (AIC).
10

According to Blanchard and Quah (1989), Gali (1992), and Bjørnland and Leitemo (2009) the level and
difference form variables can be combined in the VAR system as long as the system is stationary. An excellent survey
about the issue of stationary systems, non-stationary I(1) systems, and a mixture of I(1) and I(0) variables in the SVAR
model can be found in Levtchenkova et al. (1998).
11

10

The full result of SVAR contemporaneous coefficient estimation is available upon request.

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

effects of external shocks on domestic macroeconomics. The solid line represents the
estimated responses; meanwhile the two
dashed lines represent the confident bands
or error bands. The error bands of the SIRF
are derived from Hall’s bootstrapping methodology, which has a 68 percent confidence
interval with 300 being the number of bootstrap replications.12 Accumulated impulse-response functions will be discussed for the first
difference variables responses, while the
usual IRF will be used for the level form variable responses. By accumulating the responses
of the first-difference variable on its structural shocks, we can interpret the impact of
structural shocks on the level form of endogenous (Y) variables.

Interest rate response is significantly
positive to foreign income shocks with the
highest response being 0.3 percentage points
within 25 months. However, after 25 months
the positive response of the interest rate begins to reduce and decays after 60 months.
The money supply response is negative to
foreign income shocks. For instance, an accumulated response of the money supply, in
response to a one percent increase in foreign
income growth is appropriately -0.6 percent
after a 25 month period. One possible explanation is that the monetary authority has been
concerned with stabilizing the domestic price
level, since the expansion of foreign income
has triggered an increase in domestic inflation due to an increase in external demand.

Foreign Shocks Effects on Domestic
Variables
Response of domestic monetary policy.

The domestic interest rate has responded positively to the FFR shocks, with
the highest response being 0.15 percentage
points change in 15 months. The money supply has negatively responded to the positive
innovation in FFR. This finding indicates that
the monetary authority is concerned about the
capital outflow from the domestic country to
the US due to the monetary tightening in the
US. Therefore, to ensure that portfolio investment in Malaysia is competitive, the BNM
has to respond by increasing the domestic
interest rates.

Panel A in Appendix 1 shows the response
of interest rates to the innovation in foreign
shocks, and Panel B describes the accumulated response of the money supply to the
innovation in foreign shocks. As can be seen
in Appendix 1 (Panel A), the response of the
interest rates to the innovation in world oil
price growth is negative, decaying after 30
months. The accumulated response of the
money supply has a positive effect to the innovation in world oil price growth. These
findings indicate that the monetary authority’s
concern is the negative effects of world oil
price shock on domestic output, and therefore implementing an easing monetary policy
through a decrease in interest rates, and an
increase in money supply. This strategy is crucial to offset the adverse supply shocks on
economic activity through stimulating the
aggregate demand.
12

Response of domestic macroeconomic
variables. Appendix 2 in Panel A-Panel C
describes the response of domestic macroeconomic variables to the innovation in foreign shocks. As depicted in Appendix 2 in
Panel A, a positive innovation in world oil
price growth leads to an increase in domestic
output in within a two month period. However, after four months the positive innovation of the world oil price growth leads to a

The SVAR model has been estimated by using J-Multi statistical software developed by Lutkepohl and Kratzig

(2004).

11

Karim and Karim

decline in domestic output. There is no significant effect of world oil price on domestic
output after 12 months. The positive response
of output within a two month period is reasonable given that Malaysia is a net oil exporting country. An increase in the world oil
price has generated higher income for the petroleum industry, and subsequently leads to
an increase in domestic output. However,
after four months, output responds negatively
to the world oil price because of the adverse
supply shocks. For example, an increase in
the world oil price will lead to an increase in
firms’ production costs, and subsequently the
firms will respond by reducing their investment spending. Domestic inflation also positively responds to the positive shocks in the
growth of world oil prices. For instance, the
highest effect is recorded in a 5 month period, at which point every 1 percent increase
in world oil prices growth leads to an increase
in the inflation rate of 0.25 percent. However, after 5 months the response of inflation has gradually declined, and vanishes
within 40 months.
The exchange rate responds positively
and significantly to the positive shock in the
world oil price growth, and the accumulated
effect is approximately 0.6 percent in 60
months. In other words, an increase in the
world oil price triggers an appreciation in the
domestic currency relative to other currencies because Malaysian is a net exporter of
oil. The stock market responds positively to
the positive shock in the world oil price
growth, but decays after 40 months.
Panel B in Appendix 2 plots the response of domestic macroeconomic variables
to innovation in foreign income (US output).
As can be seen, domestic output growth positively responds to the foreign demand shock.
The positive response of domestic income is
reasonable since most of the Malaysian ex12

ports are concentrated in the US market. The
effect of the US income shock to domestic
inflation is negative within 20 months,
whereas after 20 months domestic inflation
has responded positively. This indicated that
after 20 months, foreign income shocks can
influence domestic inflation through an increased external demand, which is subsequently stimulating the aggregate demand.
A positive shock in the US income has
caused the exchange rate to respond negatively within 25 months; however, after 25
months the response is positive. The positive response of exchange rate (appreciation
in domestic currency) occurs because an increase in foreign income has stimulated domestic exports, and afterwards has increased
the demand for domestic currency. Asset price
has responded positively to the positive innovation in the foreign income, which is increased by 0.08 percent within a 60 months
period.
Panel C in Appendix 2 shows the effect
of foreign monetary shocks on domestic macroeconomic variables. As can be seen, the
accumulated response of domestic output is
positive within a 5 month period; however,
the effect is relatively small. After 5 months,
the accumulated response of domestic output is negative. For example, within a 20
month period; the accumulated response of
domestic output has decreased by 0.4 percent in responding to a one percentage point
increase in FFR. A possible explanation is that
an increase in the foreign interest rate has
contracted foreign economic activity, which
afterwards decreased foreign demand, and
subsequently contracted the domestic
economy. Domestic inflation has responded
negatively to positive innovation in FFR after 2 months, and decayed after 25 months.
For example, in 10 months, domestic inflation has decreased by 0.1 percent in response

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

to a one percentage point increase in foreign
monetary policy.
The accumulated response of exchange
rates to the innovation in FFR is positive. A
possible reason is that the BNM has increased
the domestic interest rates to offset the capital outflow to the foreign country. The accumulated response of the stock market to the
FFR shocks is positive and relatively small
in 5 months, which indicates that foreign
monetary policy is not important in influencing the domestic stock market. However, after 7 months, the stock price has responded
negatively with the positive innovation in
FFR.

Monetary Policy Effects on Domestic
Variables
Interest Rate Shocks. Appendix 3 in Panel
A indicates the effect of the interest rate
shock on domestic macroeconomic variables.
As can be seen, domestic output growth has
responded negatively to a positive shock in
interest rates. For instance, the accumulated
response of domestic output is -0.003 percent after 20 months, which is indicated that
a one percentage point increase in interest
rates (IBOR) leads to a contraction in domestic output of 0.3 percent. The negative response of domestic output to monetary policy
tightening can be explained by standard economic theory, by which an increase in interest rates increases the firms’ capital costs,
while in turn the firms will respond by reducing their capacity of fixed investment. The
response of inflation to monetary policy tightening is positive, which indicates the existence of a price puzzle. Barth and Ramey
(2000) have provided an alternative explanation of the price puzzle, where they argued
that monetary policy tightening operates on
aggregate supply as well as aggregate demand. Specifically, an increase in interest rates

raises the cost of holding inventories (cost
channel of monetary policy). This negative
supply effect raises domestic inflation and
lowers output. Money supply has responded
negatively to a positive shock in interest rates,
where the accumulated effect is -0.2 percent
after 20 months.
The exchange rate responds positively
to the positive innovation in interest rates.
For instance, after 10 months, the accumulated response of the nominal exchange rate
is 0.012, which indicates that a one percentage point increase in interest rates leads to
an appreciation in the nominal exchange rate
by 1.2 percent. A possible reason is that an
increase in interest rates has stimulated the
capital inflows to the domestic country, and
subsequently increased the demand for the
domestic currency, and fuels the exchange rate
appreciation. The stock market also responds
positively within 5 months, however after 5
months, stock price response is negative to
the innovation in interest rates. One possible
reason is that an increase in interest rates has
contracted economic activity and dividend
payments, which in turn decreased the asset
price.

Money Supply Shocks. Appendix 3 in Panel
B plots the money supply (narrow money M1)
impulses on domestic macroeconomic variables. As can be seen, in one month, the accumulated response of domestic output is
negative in response to positive innovation
in money supply growth. However, after one
month, there is a positive cumulative effect
of the shocks in money supply growth on
domestic output. For instance, the highest
cumulative effect is shown in 3 months,
where one percent increase in money supply
growth leads to an increase in domestic output by approximately 0.3 percent. However,
after five months, the positive accumulated
response of domestic output has decreased.
13

Karim and Karim

Specifically, after 15 months, the accumulated
response of domestic output is 0.1 percent.
This finding indicates that the money supply
plays an important role in stimulating domestic
output growth. The response of inflation on
money supply growth shows a hump-shape
pattern, and returns to the equilibrium path
after 15 months.
Interest rates respond positively to the
innovation in money supply growth, which
indicates the existence of a liquidity puzzle.
Some empirical studies, for example
Reichenstein (1987), and Leeper and Gordon
(1991) have supported the existence of the
liquidity puzzle. In addition, the positive effect of interest rates gradually decreases and
vanishes after 30 months. The effect of
money on the exchange rate is not statistically significant. In comparison, the stock
market has a positive response to money supply shocks, which indicates that an expansionary monetary policy has stimulated the
asset price to increase.

Robustness Checking 13

(i.e. Malaysia) by using an open-economy
SVAR study. Specifically, this study investigates the effect of external shock, namely the
world oil price and the US international transmission (US income and US monetary policy)
to domestic monetary policy and macroeconomic fluctuations. In addition, this study
examines the effectiveness of domestic monetary policy in stimulating the macroeconomic variables (output and inflation).
Several major conclusions can be drawn
from the empirical findings. First, the world
oil price shock has generated a significant
effect in influencing domestic monetary
policy implementation and macroeconomic
variables in the Malaysian economy. Second,
the US transmission in terms of the US income and the US monetary policy shocks have
also played a prominent role in influencing
domestic monetary policy and macroeconomic fluctuations. Third, domestic monetary
policy, either money supply or interest rates
have played a significant role in stabilizing
economic activity (output and inflation).

Summary and Conclusions

These findings have several implications for domestic monetary policy implementation. First, since the foreign shocks can influence domestic macroeconomic fluctuations, the monetary authority (in particular,
the BNM) has to make an accurate assessment about the external events by monitoring international conditions in implementing
monetary policy. This is very important for
minimizing the negative effect of the foreign
shocks on the macroeconomic fluctuations
through proactive monetary policy measures.

The focal point of this study is to examine the role of monetary policy during interest rates targeting in a small-open economy

Second, there is a relevant role for monetary policy instruments (money supply and
interest rates) in stimulating macroeconomic

For robustness checking, several alternative procedures have been considered in
this study; (i) estimating the recursive SVAR
model; (ii) alternative contemporaneous
structural identification schemes; (iii) model
with money demand; and (iv) model without
money. Overall, the structural IRF are robust
with the baseline restriction.

13

request.

14

I do not report the full result of robustness checking to save space. However, the full results are available upon

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

variables, which suggests that monetary
policy can be used as a stabilization policy.
Third, during the interest rate targeting
regime, domestic monetary policy has positively responded to the foreign monetary
policy shocks. It seems that the BNM has to
follow the US monetary policy. This strategy
is crucial under perfect capital mobility to
maintain the competitiveness of the domestic financial assets, and it can stabilize the
domestic currency. As noted by Umizaki
(2007), the BNM not only has to consider
the domestic factors such as inflation rates
and output but also foreign variables, particularly foreign monetary policy and exchange
rates, in designing an appropriate monetary
policy rule.

Fourth, since domestic output has responded negatively to interest rates, the BNM
has a greater opportunity to stimulate economic activity by controlling the interbank
overnight interest rates. Finally, the monetary
policy also plays an important role in stabilizing economic activity via the exchange rate
and stock market effects. For example, the
central bank can stimulate the export sector
by implementing an easing of monetary policy
(either by an increase in money supply or decrease in interest rates). This strategy will
depreciate domestic currency, and subsequently stimulate external demand and economic activity.

References
Amisano, G., and C. Giannini. 1996. Topics in Structural VAR Econometrics. Berlin: Springer.
Azali, M., and K. Matthews. 1999. Money-Income and credit-income relationships during the pre-and the
post-liberalization periods: Evidence from Malaysia. Applied Economics 31: 1161-1170.
Barth, M. J., and V. A. Ramey. 2000. The cost channel of monetary transmission. NBER Working Paper
Series. Cambridge: National Bureau of Economic Research.
Bernanke, B. S. 1986. Alternative explanations of the money-income correlation. Carnegie-Rochester Conference Series on Public Policy 25: 49-100.
Bernanke, B. S., and A. S. Blinder. 1992. The federal funds rate and the channel of monetary transmission.
The American Economic Review 82: 901-921.
Bernanke, B. S., M. Gertler, and M. Watson. 1997. Systematic monetary policy and the effects of oil
shocks. Brookings Papers on Economic Activity 1: 91-157.
Bernanke, B. S., and I. Mihov. 1998. Measuring monetary policy. The Quarterly Journal of Economics 113:
869-902.
Bjørnland, H. C., and K. Leitemo. 2009. Identifying the interdependence between US monetary policy
and the stock market. Journal of Monetary Economics 56: 275-282.
Blanchard, O., and D. Quah. 1989. The dynamic effects of aggregate demand and supply disturbances.
American Economic Review 79: 655-673.
Brischetto, A., and G. Voss. 1999. A structural vector auto regression model of monetary policy in
Australia. Research Discussion Paper. Reserve Bank of Australia.

15

Karim and Karim

Bucle, R. A., K. Kim, H. Kirkham, N. McLellan, and J. Sharma. 2007. A Structural VAR business cycle
model for a volatile small open economy. Economic Modeling 24: 990-1017.
Cushman, D. O., T. Zha. 1997. Identifying monetary policy in a small open economy under flexible
exchange rates. Journal of Monetary Economics 39: 433-448.
Domac, I. 1999. The Distributional Consequences of Monetary Policy: Evidence from Malaysia. Policy
Research Working Paper the World Bank. East Asia and Pacific Region.
Dungey, M., and A. Pagan. 2000. A Structural VAR model of the Australian economy. The Economic Record
76: 321-342.
Fuller, W. A. 1976. Introduction to Statistical Time Series. New York: John Wiley.
Gali, J. 1992. How well does the IS-LM model fit postwar U.S data? . The Quarterly Journal of Economics
107: 709-738.
Hamilton, J. 1996. This is what happened to the oil price macro economy relationship. Journal of Monetary
Economics 38: 215-220.
Handa, J. 2009. Monetary Economics. London and New York: Routledge, Taylor & Francis Group.
Ibrahim, M. 2005. Sectoral effect of monetary policy: Evidence from Malaysia. Asian Economic Journal 19:
83-102.
Jensen, G. R., and J. M. Mercer. 2002. Monetary policy and the cross-section of expected stock return. The
Journal of Financial Research XXV: 125-139.
Kim, S., and N. Roubini. 2000. Exchange rate anomalies in the industrial countries: A solution with a
structural VAR approach. Journal of Monetary Economics 45: 561-586.
Leeper, E. M., and D. B. Gordon. 1991. In search of the liquidity effect. Journal of Monetary Economics 29:
341-369.
Levtchenkova, S., A. R. Pagan, and J. C. Robertson. 1998. Shocking stories. Journal of Economic Survey 12:
507-532.
Levy J., and I. Halikias. 1997. Aspects of monetary transmission mechanism under exchange rate targeting: the case of France. IMF Working Paper Series.
Ludvigson, S., C. Steindel, and M. Lettau. 2002. Monetary policy transmission through the consumptionwealth channel. FRBNY Economic Policy Review: 117-133.
Lutkepohl, H., And M. Kratzig (eds.) 2004. Applied Time Series Econometrics. New York: Cambridge University Press.
Parrado, E. 2001. Effects of foreign and domestic monetary policy in a small open economy: The case
of Chile. Working papers (108). Central Bank of Chile.
Ramaswamy, R., and T. Sloek. 1997. The real effects of monetary policy in the European Union: What
are the differences? IMF Working Paper (97/160).
Ramey, V. A. 1993. How important is the credit channel in the transmission of monetary policy? NBER
Working Paper Series. Massachusetts Avenue, Cambridge.
Reichenstein, W. 1987. The impact of money on short-term interest rates. Economic Inquiry 25: 67-82.
Sellin, P. 2001. Monetary policy and the stock market: theory and empirical evidence. Journal of Economic
Surveys 15: 491-541.

16

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

Sim, C. A, and T. Zha. 2005. Does monetary policy generate recession? Macroeconomic Dynamics 14: 1-42.
Tang, H. C. 2006. The relative important of monetary policy transmission channels in Malaysia. CAMA
Working Paper Series. The Australian National University.
Umizaki, S. 2007. Monetary policy in a small open economy: The case of Malaysia. The Developing Economics XLV (4): 437-464.
Zaidi, M. A. S., and L. A. Fisher. 2010. Monetary policy and foreign shocks: A SVAR analysis for Malaysia.
Korea and the World Economy 1 (3): 527-550.

17

Karim and Karim

Appendix 1. Structural Impulse-Response Function: The Effect of Foreign Shocks on
Monetary Policy Variables
accumulated OIL_log_d1_M1_log_d1

OIL_log_d1IBOR

0.032

0.4

0.028
0.024

0.2

0.020
0.0

0.016
0.012

-0.2

0.008
-0.4
-0.6
0

0.004

10

20

30

40

50

60

IPIUS_log_d1IBOR

0.000
0

10

20

30

40

50

60

accumulated IPIUS_log_d1M1_log_d1
0.12

0.8

0.08
0.6
0.04
0.4

0.00
-0.04

0.2

-0.08

0.0

-0.12
-0.2
-0.4
0

-0.16
10

20

30

40

50

60

0

10

20

30

40

50

60

accumulated FFRM1_log_d1

FFRIBOR

0.4

-0.20

0.12
0.08

0.3

0.04

0.2

0.00

0.1

-0.04
0.0
-0.08
-0.1

-0.12

-0.2
-0.3
0

18

-0.16
10

20

30

40

50

60

-0.20

0

10

20

30

40

50

60

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

Appendix 2. Structural Impulse-Response Function: The Effect of Foreign Shocks on
Macroeconomic
accumulated OIL_log_d1IPIM_log_d1

accumulated IPIUS_log_d1IPIM_log_d1

0.03

0.05

0.012

accumulated FFRIPIM_log_d1

0.02

0.04

0.008

0.01

0.03

0.000

0.00

0.02

-0.004

-0.01

0.01

-0.02

0.00

0.004

-0.03
0

10

20

30 40

50

60

-0.01
0

-0.008
-0.012
-0.016
10

OIL_log_d1INF

20 30 40 50 60

-0.020

IPIUS_log_d1INF

FFRINF

0.05

0.03

0.15

0.04

0.02

0.10

0.03

0.01

0.02

0.00

0.05
0.00
-0.05

-0.01

0.01

-0.10

-0.02

-0.15

0.0

-0.03

-0.20

-0.1

-0.04

-0.2

-0.05
0

-0.25
-0.30

0

10

20

30 40

50

60

0 10 20 30 40 50 60

10

20 30 40 50 60

0 10 20 30 40 50 60

accumulated OIL_log_d1NEER_log_d1

0.05
0.04

0.05

0.06

0.04

0.04

0.03

0.03

0.02
0.02

0.02

0.01

0.00

0.0

-0.04

0.01
0.00

-0.1
-0.2

-0.04
0

-0.01
10

0 10 20 30 40 50 60

20 30 40 50 60

accumulated OIL_log_d1KLCI_log_d1

0.06

0.22

0.04

0.04

0.18

0.02

0.02

0.14

0.00

0.00

0.10

-0.02

-0.02

0.06

-0.04

-0.04

0.02

-0.06

-0.06

-0.02
0

-0.08
10

20 30 40 50 60

0 10 20 30 40 50 60

19

Appendix 3. Structural Impulse-Response Function: The Effect of Monetary Policy Shocks on Domestic Variables

Karim and Karim

20

Appendix 3 (Continued)

Gadjah Mada International Journal of Business - January-April, Vol. 16, No. 1, 2014

21

Appendix 3 (Continued)

Karim and Karim

22