Description
This paper aims to examine the relationship between Indian rupee-US dollar exchange rate
and the macroeconomic fundamentals for the post-economic reform period.
Journal of Financial Economic Policy
Macroeconomic fundamentals and dynamics of the Indian rupee-dollar exchange
rate
Chandan Sharma Rajat Setia
Article information:
To cite this document:
Chandan Sharma Rajat Setia , (2015),"Macroeconomic fundamentals and dynamics of the Indian
rupee-dollar exchange rate", J ournal of Financial Economic Policy, Vol. 7 Iss 4 pp. 301 - 326
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-11-2014-0069
Downloaded on: 24 January 2016, At: 21:53 (PT)
References: this document contains references to 65 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 106 times since 2015*
Users who downloaded this article also downloaded:
Muhammad Umar, Gang Sun, (2015),"Country risk, stock prices, and the exchange rate of the
renminbi", J ournal of Financial Economic Policy, Vol. 7 Iss 4 pp. 366-376http://dx.doi.org/10.1108/
J FEP-11-2014-0073
Chong Lee-Lee, Tan Hui-Boon, (2007),"Macroeconomic factors of exchange rate volatility: Evidence
from four neighbouring ASEAN economies", Studies in Economics and Finance, Vol. 24 Iss 4 pp.
266-285http://dx.doi.org/10.1108/10867370710831828
Sanfeng Zhang, Maoliang Bu, Huafan Yang, (2014),"Environmental Regulation and Firm Productivity
in China", Frontiers of Economics and Globalization, Vol. 14 pp. 129-152http://dx.doi.org/10.1108/
S1574-871520140000014006
Access to this document was granted through an Emerald subscription provided by emerald-
srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
*Related content and download information correct at time of
download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Macroeconomic fundamentals
and dynamics of the Indian
rupee-dollar exchange rate
Chandan Sharma
Indian Institute of Management Lucknow, Noida, India, and
Rajat Setia
National Institute of Financial Management, Faridabad, India
Abstract
Purpose – This paper aims to examine the relationship between Indian rupee-US dollar exchange rate
and the macroeconomic fundamentals for the post-economic reform period.
Design/methodology/approach – The authors have used an empirical model which includes a
range of important macroeconomic variables based on the basic monetary theories of exchange rate
determination. At the frst stage of the analysis, they have tested structural break in the data.
Subsequently, they have employed the fully modifed ordinary least square, Wald’s coeffcient
restriction and impulse response functions (IRF) to estimate the monetary model in the long- and
short-run horizons.
Findings – Results of analyses indicate that the macroeconomic fundamentals determine exchange
rate in a signifcant way, but their effect varies sizably across the periods. The IRF illustrate the
importance of interest rate in controlling exchange rate volatility.
Practical implications – The analysis of the behavior of inter-relationship among macroeconomic
variables will help policymakers in a deep-rooted understanding of this complex and time-varying
relationship.
Originality/value – Most of the existing studies have tested the impact of a single or a few
macroeconomic fundamentals on exchange rate. But in the present study, we have tested the impact of
a range of important variables, i.e. money supply, real income or output, price level and trade balance.
Further, considering the importance of structural breaks in data, they authors have employed standard
tests of structural break and incorporated the issue in the cointegration analysis.
Keywords Time series models, Macroeconomic aspects of international trade and fnance,
Macroeconomics and monetary economics
Paper type Research paper
1. Introduction
Analyzing the movement of exchange rates has always been a challenging and risky
task. The complexity of the interaction between exchange rate and macroeconomic
fundamentals has gained a great deal of attention of researchers and policymakers in the
past decades. Despite the fact that many exchange rate models and their modifcations
are developed over time, still, there is no single theory which describes the behavior of
exchange rate in its entirety. The exchange rate models empirically tested over the past
The authors thank two anonymous referees for their useful comments and helpful suggestions on
the previous versions of this paper. Any errors or omission are solely of the authors.
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1757-6385.htm
Macroeconomic
fundamentals
and dynamics
301
Received13 November 2014
Revised29 April 2015
Accepted10 June 2015
Journal of Financial Economic
Policy
Vol. 7 No. 4, 2015
pp. 301-326
©Emerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-11-2014-0069
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
three decades have shown inconsistency when presented with different data sets,
estimation methodologies and sample periods (see Meese and Rogoff, 1983, Macdonald
and Taylor, 1994).These models have been tested mainly for developed countries, while
developing and transition economies like India have unfairly received lesser attention.
Numerous theoretical and empirical models explain the impact of the macroeconomic
fundamentals on the currency exchange rate utilizing a number of factors specifc to an
economy. The fexible price model developed by Frenkel (1976) and the sticky price
model of Dornbusch (1976) had initially become dominant models of exchange rate
determination. The most signifcant blow to the monetary approach of exchange rate
determination was the fndings of the seminal work of Meese and Rogoff (1983), which
concluded that the analyzing and forecasting ability of these models are quite limited.
This spurred a large number of studies empirically testing these models and their
extensions in the past three decades, namely, MacDonald and Taylor (1994), McNown
and Wallace (1989), Mark (1995), Mark and Sul (2001), Rapach and Wohar (2002),
Cheung et al. (2005), Morley (2007), Zhang et al. (2007), Sarno and Valente (2009) and
Baharumshah et al. (2009). These studies invariably have shown that monetary models
can explain movements in exchange rate to a varied extent. Furthermore, after the
popularity of cointegration relationships since early 1990, empirical evidence which
have employed the cointegration techniques seems to indicate that the monetary models
can explain large portions of the long-run movements in the exchange rate (e.g. McNown
and Wallace, 1989; MacDonald and Taylor, 1994; Kim and Mo 1995).
Findings of MacDonald and Taylor (1994) clearly indicated that unrestricted
monetary models are valid for a long-term analysis, if analyzed in a cointegrating
framework with short-term data dynamics. In a recent study, based on the empirical
evidence, Bekiros (2014) argued that fundamentals may be important determinants of
exchange rates; however, there may be some other unobservable variables driving the
currency rates that current asset-pricing models have not yet captured. These studies
have by and large ignored the issue of structural shifts in the exchange rate and
macroeconomic variables, as the long-run cointegrated series is subjected to serious
variations in the sample period. Some studies have also highlighted that parameter
estimates are often unstable and the relationship is a subject of periodic shifts in the long
run (e.g. Florentis et al., 1994, Goldberg and Frydman, 1996). Accounting for the issue of
structural break, Beckmann et al. (2010) found that fundamentals are important for the
dollar-euro exchange rate, but their impact differs in various regimes, i.e. sign and size
of the regression coeffcients change over time. Similarly, Chang and Su (2014) explored
the linkages between exchange rates and macroeconomic fundamentals to determine
the long-run relationship for Pacifc Rim countries. In their analysis, the use of the
conventional cointegration tests failed to fnd the long-run equilibrium for any
country-pairs except Taiwan, but cointegration tests with structural breaks
demonstrated the long-run connections between exchange rates and fundamentals for
some country-pairs.
In the early 1990s, India suffered with a severe balance of payment crisis due to the
signifcant rise in oil prices, the suspension of remittances from the Gulf region and
several other exogenous developments. The country had been following an adjustable
nominal peg to a basket of currencies of major trading partners with a band. The several
measures were initiated to arrest the crisis which includes a devaluation of the Indian
rupee in July 1991. This measure was the move toward greater exchange rate fexibility.
JFEP
7,4
302
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Thereafter, India has adopted a managed fexible exchange rate regime in 1993, making
the start of an era of a market-determined exchange rate regime of the rupee with
provision for timely intervention by the central bank. The current exchange rate policy
relies on the underlying demand and supply factors to determine the exchange rate with
continuous monitoring and management by the central bank. India has been operating
with a managed fexible regime, where the management’s objective is not to achieve any
explicit or implicit target for the exchange rate but to contain volatility by ensuring
orderly market conditions. The Indian regime could be interpreted as “more fexible”
during normal market conditions with the accent shifting to “management” when the
market turns disorderly (Pattanaik and Sahoo, 2003). However, recently, the Indian
currency has been quite volatile[1]. More importantly, it seems that the government is in
mood to a frequent intervention in the exchange rate market[2].
In the Indian context, a few attempts have been made to test the monetary model of
exchange rate determination. For example, to analyze the issue, Bhanumurthy (2006)
provided the evidence based on a survey of the dealers and found that majority of the
dealers feel that short-term changes in the Indian rupee/US dollar market are basically
infuenced by the micro-variables such as information fow, market movement,
speculation, central bank intervention, etc. Dua and Ranjan (2011) have attempted to
analyze the issue and concluded that the monetary model works well in the Indian case.
More recently, Mallick (2010) have analyzed the issue and concluded that dominance of
foreign institutional investments affects the rupee-dollar exchange rate, and to a certain
extent, it was seen that the infuence of the growth rate differential also affects the
exchange rate behavior. The analysis of Tiwari et al. (2013) has indicated that oil price
and Indian rupee have a causal relationship, but the causality exists only at higher time
scales. These studies have, however, widely ignored the issues of the presence of a
structural shift in the system. Moreover, in a recent study, Hegwood and Nath (2014)
have shown that the movement of the Indian rupee suffers fromstructural breaks, which
may have seriously altered the fndings on India. Findings of Narayan (2006) using
Lagrange multiplier (LM) unit root test with structural breaks also provided evidence
that India’s exchange rate vis-a`-vis 15 out of the 16 countries is stationary, implying
support for purchasing power parity (PPP).
Against this background, we make an attempt to reassess the issue and to bridge the
existing gaps in the related literature. We mainly focus on two questions. First, what are
the determinants of the exchange rate, especially from macroeconomic perspective?
Second, is there any theory that explains the pattern of exchange rate movements
near-accurately? In this study, we attempt to provide several novelties to the related
literature. First, initially, it was understood that the development of a series of
cointegration tests will help in examining the monetary approach of exchange rate
determination. It was expected that the debate in the empirical literature will settle when
these tests will be able to explain long-term fuctuations in the exchange rate market.
However, the signs of the estimated parameters from the variety of monetary models
using the cointegration framework again conficted with widely accepted theoretical
assumptions. Recent development in time series econometrics indicates that a possible
reason for these conficting results is the presence of structural instability in the data. In
fact, the structural breaks can induce stochastic behavior similar to an integrated
process, which makes it diffcult to differentiate between the lack of cointegration and a
structural shift (Dropsy, 1996). The presence of structural changes may lead to a serious
303
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
bias in the estimated coeffcients. Moreover, it is now well-established that accurate
forecasting and empirical analysis of time series data can depend critically on
understanding the appropriate nature of structural change (e.g. Lee et al., 2006). Thus, in
the present study, we attempt to address this issue by incorporating the issue of
structural break in the analysis. This is important, as in the wake of changes in
exchange rate policy regimes, shocks to oil prices and shifting patterns of trade, there is
a need to account for structural breaks in the empirical testing of exchange rate
equations. Second, the Indian case is both interesting and relevant in the light of the fact
that polices on exchange rate management has shifted to a foating exchange rate
regime in 1993. The timing of this policy shift also coincides with embankment of
economic liberalization. The policy reforms have led to a spectacular surge in
international capital infows in India during 1990s. However, in the recent years, the
Indian rupee has witnessed excessive volatility, which is seen as a serious threat to
stability of the Indian fnancial and external sector, which, in turn, affects the country’s
overall growth trajectories. Third, this study along with the long-run linkage also
focuses on the short-run dynamics. Specifcally, the response of exchange rate to an
impulse or shock on other macroeconomic variables is generated in a vector error
correction (VEC) framework. Finally, most of the existing studies have focused on a
single theoretical model for the empirical testing; we examine several related theories
and utilize a range of important variables, i.e. money supply, real income or output, price
level and trade balance for our analysis. These models are empirically tested using
several alternative techniques to test the sensitivity and robustness of results.
Rest of this paper is divided into fve sections. Section 2 makes up the theoretical
background and set hypotheses to test. Data and methodological issues form Section 3.
Section 4 covers results and discussion. Section 5 contains the conclusion and policy
suggestions.
2. Theoretical models: monetary models of the exchange rate
The traditional model of exchange rate determination, known as the PPP theory is an
application of the “Lawof One Price”. It states that an exchange rate equalizes the price
of goods between countries. PPP focuses on general price levels and their relationship to
the nominal exchange rate as:
P ? EP* (1)
Where P and P* are the domestic and foreign price levels and E denotes the exchange
rate. There are noted deviations from PPP due to a number of theoretically implied
factors, namely, heterogeneity in the baskets of goods considered for construction of
price indices in various countries, presence of transportation cost, imperfect competition
in goods market and increase in the volume of global capital fows during the past few
decades. Although there is a mixed evidence provided by a battery of empirical tests, the
strength of the relationship between price ratios and exchange rates is not as strong as
depicted by the PPP theory (Nag and Mitra, 1998; Mishra and Sharma, 2010).
In all the monetary models, the money supply in relation to money demand
determines the exchange rate. The prominent monetary models include the fexible and
the sticky price monetary models of exchange rate as well as the real interest rate
differential (RID) model. The fexible price monetary model (Frenkel, 1976) is based on
the assumption that the prices are fully fexible, i.e. they can either increase or decrease
JFEP
7,4
304
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
from the prevalent level (also see Bilson, 1978; Hodrick, 1978 and Macdonald and
Taylor, 1994). Sticky price model of Dornbusch (1976) is based on the premise that prices
are determined in sticky price markets, i.e. they are fxed over a short period and they
tend to change only slowly over time in response to shocks such as changes in money
supply. However, exchange rates are determined in fexible price markets and respond
immediately to new developments and shocks. So exchange rates are not matched by
price movements, resulting in departures from PPP. Hooper and Morton (1982)
introduced an extension of the sticky price model by accommodating the effect of trade
balance in the exchange rate determination equation. Here, exchange rate is assumed to
be correlated with unanticipated shocks to the trade balance.
Real interest rate differential model (Frankel, 1979) combines the infationary
expectations of the fexible price monetary model with the sticky price model. It states
that the gap between the current real exchange rate and its long-run equilibrium value
is proportional to the real interest rate differential.
The basic theoretical models of exchange rate determination, such as the fexible
price model, sticky price model, real interest rate differential model and the Hooper and
Morton model, are collectively represented by the following equation:
e ? ? ? ?(m ? m*) ? ?(y ? y*) ? ?(i ? i*) ? ?(p ? p*) ? ?(tb ? tb*) (2)
Where e, m, y, i, p and tb denote nominal exchange rate, money supply, real output,
nominal interest rate, infation or price level and trade balance, respectively, and *
variables indicate the foreign counterparts.
To test the individual models, we construct hypotheses, which are as follows:
2.1 Flexible price model (? ?0, ? ?0, ? ?0, ? ?? ?0)
An increase in the domestic money supply relative to foreign money supply (? ?0) leads
to an increase in the infation rate, which will decrease the currency demand, and
currency will depreciate or e will increase. An increase in the domestic interest rate
relative to foreign interest rate (? ?0) is a refection of an expected depreciation of the
domestic currency due to an expected decrease in demand of the currency resulting from
an expected rise in infation. Thus, a rise in domestic interest rate leads to a fall in the
demand for money and hence depreciation of the domestic currency or an increment in
e. A relative increase in the output or real income means more number of money
transactions creating an excess demand for the currency, and if money supply and
interest rates are constant, then domestic prices will fall, resulting in an appreciation of
currency value to maintain PPP. So, real income differential and exchange rate move in
tandem i.e. ? ?0 will lead to a decrease in e (see Frankel, 1984 and Williamson, 2009).
2.2 Sticky price model (? ?0, ? ?0, ? ?0, ? ?? ?0)
The model argues that prices are rigid and would only adjust gradually. In response to
an increase in the money supply (? ?0 ), considering the prices are fxed in the short run,
interest rate will fall (? ? 0 ), which has to be compensated by depreciation of the
domestic currency or an increase in the e value. For this reason, the exchange rate
overshoots its long-run equilibrium value. A relative decrease in the output or real
income (? ? 0 ) will result in a decrease in the value of e (see Dornbusch 1976; Rogoff,
2009).
305
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
2.3 Real interest differential model (? ? 0, ? ? 0, ? ? 0, ? ? 0, ? ? 0)
Frankel (1979) developed an integrative approach that incorporates the sticky price
model with the notion of secular rates of infation. If the real rate of interest on domestic
bonds is greater than the real rate of interest on foreign bonds, there will be a
depreciation of the domestic currency until the real interest rates are equalized in the
long run steady state (also see Hoffmann and McDonald, 2009).
2.4 Hooper and Morton model (? ? 0, ? ? 0, ? ? 0, ? ? 0, ? ? 0)
Changing the traditional monetary model, Hooper and Morton (1982) incorporated the
real side of the economy by taking into consideration innovations in the current account.
The model utilizes trade balances as to gauge the effects of premium due to fscal
balance, international reserves holding and foreign indebtedness. Therefore, the model
equation simply adds on the effect of relative trade balance differential. If this
differential is negative (? ? 0 ), domestic currency will appreciate, and hence, e value
decreases (also see Beckmann et al., 2010).
Incorporating the issues of these models, for empirical testing, we employ a hybrid
empirical model, containing all important macroeconomic variables and their
hypothesized direction of impact as per the basic theoretical models (both monetary as
well portfolio models) of exchange rate determination mentioned above. Our benchmark
empirical model is as follows:
Y
t
? ?
t
? ?
t
X
t
? ?
t
(3)
Where variable Y
t
contains the Indian rupee-US dollar exchange rate (e) and
X
t
is a [K ? 1] vector of domestic and foreign explanatory fundamentals, i.e.
X
t
?(y, y*, p, p*, m, m*, i, i*, tb, tb*) and ?
t
is a [1 ?K] vector of respective coeffcients.
It is worth mentioning that we choose nominal exchange rate for our analysis over
real exchange rate. Theoretical and empirical studies on this issue mainly considered
nominal exchange rate, whereas studies of Kollman (1997), Kia (2013) and others have
instead chosen real exchange rate for analysis. The nominal exchange rate is preferred
because the real exchange rate converges to PPP in the long run, leading to fundamental
variables do not have any effects on the real exchange rate. Moreover, fully fexible and
the real exchange rates adjust to real disturbances and fscal policies. On the other side,
by design monetary policy indicators cannot drive the real exchange rate, as in this
equilibrium condition, price and nominal exchange rate re-adjust in the same ratio to
any monetary shock (for detail discussion, see Devereux, 1997).
3. Data description and empirical methodology
3.1 Data description
For empirical testing, this paper employs monthly data on rupee-dollar nominal
exchange rate (e) and the independent macroeconomic variables like money supply (M3
for India and M2 for USA), real income or output proxied by IIP (index of industrial
production), general price level proxied by CPI (consumer price index), short-term
interest rate (91 day treasury bill rates) and the overall trade balance (exports-imports)
for India and the USA. Data from April 1994 to March 2010 are used due to the
availability of the data for this period on all the variables required. The data on all the
variables for India are collected from IMF’s International Financial Statistics database
and Handbook of Statistics on the Indian Economy provided by Reserve Bank of India.
JFEP
7,4
306
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Data on trade balance have been converted into an index to avoid complexity arising out
of the negative magnitude of observations. The data series are taken at their level form
and seasonally adjusted (except exchange rate). The series are transformed in their
natural logarithms (except interest rates, which are in their linear form) before any
econometric treatment. Figure 1 shows the variation of exchange rate for the period of
the study. Detailed descriptions of data series and their descriptive statistics are
presented in Tables A1 and A2 respectively.
3.2 Empirical methodology
3.2.1 Unit root test with endogenous two-break. One major drawback of conventional
unit root tests is that it implicitly assumes that the model correctly specifes the
deterministic trend. Following the work of Perron (1989), one can consider that the
presence of structural change substantially reduces the power of unit root tests. Zivot
and Andrews (1992) proposed a unit root test that allows for an endogenous structural
break. Recently, Lumsdaine and Papell (1997) proposed a unit root test that allows for
two shifts in the deterministic trend at two known dates. A problem with the
Lumsdaine–Papell unit root test is that its critical values assume no breaks under the
null hypothesis. This assumption is problematic, as it may lead to conclude incorrectly
that rejection of the null is evidence of trend stationarity, when, in fact, the series is
difference stationary with breaks (Lee and Strazicich 2003, 2004). With improvement,
Lee and Strazicich (2003) have proposed the endogenous two-break LM unit root tests.
The test incorporates structural breaks under the null hypothesis, and rejection of the
minimumLMtest provides evidence for stationarity of the series. Moreover, the Lee and
Strazicich (2003) test also has higher power than the test of Lumsdaine and Papell (1997).
The model has two variants: frst, the crash model (Model A) and the break model
(Model C). Both models are based on alternative assumptions about structural breaks.
Model A allows for two shifts in the intercept, and Model C includes two shifts in the
intercept and trend. We have used the crash model for the analysis, as in our case, break
the crash model seems to be appropriate.
The crash model of Lee and Strazicich (2003) is specifed as follows:
Z
t
? [1, t, D
1t
, D
2t
] (4)
Where D
jt
?1 for t ? T
Bj
?1, j ?1, 2, and 0 otherwise. The break date is denoted byT
Bj
The null and alternative hypotheses of Model A are as follows:
0
10
20
30
40
50
60
A
p
r
-
9
4
J
a
n
-
9
5
O
c
t
-
9
5
J
u
l
-
9
6
A
p
r
-
9
7
J
a
n
-
9
8
O
c
t
-
9
8
J
u
l
-
9
9
A
p
r
-
0
0
J
a
n
-
0
1
O
c
t
-
0
1
J
u
l
-
0
2
A
p
r
-
0
3
J
a
n
-
0
4
O
c
t
-
0
4
J
u
l
-
0
5
A
p
r
-
0
6
J
a
n
-
0
7
O
c
t
-
0
7
J
u
l
-
0
8
A
p
r
-
0
9
J
a
n
-
1
0
Figure 1.
Exchange rate Indian
rupee/US dollar
307
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
H
0
:Y
t
? ?
0
? d
1
B
1t
? d
2
B
2t
? Y
t?1
? v
1t
H
1
:Y
t
? ?
0
? ?t ? d
1
B
1t
? d
2
B
2t
? v
2t
Where DT
jt
? t ? T
Bj
for t ? T
Bj
? 1, j ? 1, 2, and 0 otherwise; B
jt
? 1 for t ? T
Bj
?
1, j ?1, 2, and 0 otherwise; ?
1t
and ?
2t
denote the stationary error terms. The LMunit root
test statistic can be obtained by estimating the following:
?Y
t
? ?
=
?Z
t
? ?S
?
t?1
? ?
t
(5)
Where S
?
t
? y
t
- ?
x
- Z
t
? , t ? 2,. . . .,T; ?y
t
is regressed on ?Z
t
to obtain estimates of ?;
?
x
?y
1
?Z
1
? and the frst observations of y
t
and Z
t
are y
1
and Z
1
, respectively. The LM
test statistics are provided by ? which is the test statistic for the unit root null hypothesis
that ? ?0.
We consider a maximum lag length of fve periods and obtained the optimal lag
length on the basis of the signifcance of the last lag. The break dates are determined
where the LM test statistic is at its minimum. The critical values of this test are
tabulated in Lee and Strazicich (2003, 2004). Thus, this method is more demanding than
previously developed unit root tests with structural break because it offers more than
one break in the series.
3.2.2 Detecting structural changes: Bai-Perron test. An alternative approach to test
the structural change is applying F-test or Chowtest on the model. Bai and Perron (1998,
2003) have extended the Chow test. In dealing with a cointegrated model, the Gregory
and Hansen test (1996) is used for one unknown structural break and the Bai and Perron
(2003) (BP hereafter) test for dating multiple unknown structural breaks. Therefore, we
prefer BP test over Gregory and Hansen test, as it is highly likely that the system has
more than one break in the analysis period. The BP methodology employs a multiple
structural break model, with m breaks, i.e. (m ?1) regimes. Formally:
y
t
? x
=
t
? ? z
=
t
?
j
? u
t
, (t ? T
j?1
? 1, .....T
j
) (6)
for j ?1 ….,m?1, with T
0
?0 and T
m?1
?T, where y
t
is the dependent variable at time
t, x’
t
and z’
t
are the regressors, ? and ? are the coeffcients to be estimated and u
t
is the
error term. In this equation, only ? varies with time and ? is constant, so this is a partial
structural change model. Considering ? ? 0 will give us what is known as the pure
structural change model where all coeffcients are subject to change with time. The
linear combination of those segments of the total sample period, for which the sum of
squared residuals comes out to be minimumas per Bayesian information criteria, is the
estimated breakpoints. The test performs an algorithm which compares all possible
combinations of these segments.
Initially, the BP test was developed for stationary I(0) variables, but it equally holds
for non-stationary case as well (Kejriwal and Perron, 2008). Some recent studies (e.g.
Beckmann et al., 2010) have employed BP test for structural break analysis to
investigate the temporal stability of relationship between exchange rate and
macroeconomic fundamentals, which have yielded satisfactory results.
JFEP
7,4
308
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3.3 Cointegration and estimation of model parameters
As BP’s methodology is designed for single equation systems and the fact that the
macroeconomic variables which are used in the model exhibit the problem of
endogeneity, fully modifed ordinary least squares (FMOLS) method developed by
Phillips and Hansen (1990) is used for the estimation. FMOLS corrects the ordinary least
squares (OLS) method with regard to endogeneity and serial correlation (Phillips, 1995).
Here, the primary interest is to fnd out the long-run relationship between exchange rate
and fundamentals, without paying attention to the relationships among regressors.
To examine whether the relationship obtained from the FMOLS estimation can be
interpreted as a cointegrated relationship, we have applied the unit root tests to the
residuals. Another approach employed to test for cointegration is the Johansen and
Juselius (JJ) (1990) technique. The JJ maximum likelihood approach sets up the
non-stationary time series as the vector autoregressive process of order k in a
re-parameterized form:
?Y
t
? ? ? ?
1
?Y
t?1
? ?
2
?Y
t?2
? . . . .?
k?1
?Y
t?k?1
? ?
k
Y
t?k
? U
t
(7)
Where Y
t
is a vector of frst-order cointegrated variables, ?is a vector of constant terms
to capture the time series trend characteristics, ?is a coeffcient matrix and U
t
is a vector
of normally and independently distributed error terms.
Johansen et al. (2000) developed cointegration test, which uses the JJ framework to
allow for trend and level breaks at several known points. To use traditional
cointegration analysis, the method disregards observations after structural breaks by
including impulse dummies. The number of impulse dummies after recognizing the
breaks in the system and the inclusion of these dummies imply a reduction in the
effective sample. The technique uses two variants of the trace test for testing of
cointegration relationship among p time series. These are the Hl(r) and Hc(r) tests for
when there are (q – 1) breaks in a linear trend or in a constant level of the data,
respectively, where r denotes the cointegrating rank. The asymptotic distributions of
the test statistics depend on the values of (p – r) and the locations of the breakpoints in
the sample. These breakpoints are denoted by v
j
?(t
j
/t) , where T is the full sample size
and t
j
is the last observation of the jth sub-sample; j ?1, 2 ….., q (for details, see Giles and
Godwin, 2012).
Finally, the short-run effects of macroeconomic variables are captured using impulse
response functions (IRF) generated in a VEC framework. VEC framework adds a
restriction of cointegration error into the general vector autoregeressive framework.
Using Cholesky one standard deviation shock structure, IRF are generated.
4. Results and discussion
4.1 Results of conventional unit root tests
In this section, we explain results of our empirical analysis. The frst step in the analysis
of time series data is to examine the variables for stationarity. A stationary time series
is the one which exhibits mean reversion, has a fnite, time invariant variance and the
covariance between two values depends only on their distance apart in time and not on
the exact timing of the observation. At the frst stage, we employ three alternative tests
for testing stationarity in the variables. The augmented Dickey-Fuller (ADF) test,
Dickey-Fuller generalized least square (DF-GLS) test and Phillips-Perron (PP) test share
309
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the same null hypothesis of a unit root. If two of these three tests indicate
non-stationarity for any series, we conclude that the series has a unit root. Table I
reports the results of three tests with constant as well as with constant and trend. Our
variables have been tested for stationarity in the levels as well as at frst differences.
Table I shows that almost all the variables are found to be integrated of order one, i.e. I
(1). However, the standard unit root tests suffer from biasedness problem toward the
non-rejection of the null hypothesis, and macroeconomic series are not characterized by
a unit root but rather that persistence arises only from large and infrequent shocks
(Perron, 1989). Therefore, we need to test structural breaks in our data.
4.2 Results of unit root tests with structural breaks
To test endogenous structural break and order of integration of variables, we employ
two-break minimumLMunit root tests. Table II reports the results for the test based on
Model Aor crash model which represents two breaks in the intercept. The test statistics
of the LM unit root tests indicates that except tb, all other variables do not exceed the
critical values in absolute terms, and therefore, the unit root null cannot be rejected at the
5 per cent level. However, the t-statistics corresponding to the break dates are
statistically signifcant only for few variables. Furthermore, break dates are not
consistent across models, e.g. break model and crash model[3].
4.3 Results of the BP test
Considering the in consistency in results of endogenous structural break LM unit root
tests, we can obtain breakpoints in the system by applying BP method. If the empirical
application of the BP test proves the existence of structural breaks, we can conclude that
a stable long-run relationship among the variables does not exist. The results of the BP
test is presented in Table III. The frst breakpoint of 1996 appears to be the effect of
initiation of what we call as the Asian Financial Crisis that gripped much of Asia in July
1997. The crisis started in Thailand with the fnancial collapse of the Thai Baht caused
by the decision of the Thai Government to foat the Baht, cutting its peg to the US dollar.
As the crisis spread, most of the Southeast Asian countries and Japan witnessed
slumping currencies. The impact of the crisis has also spread further afeld, as
beginning of the Asian crisis had hurt Indian economy signifcantly. On the one hand,
the Indian exports to the rest of Asia had decelerated sharply (UNCTAD, 1998). On the
other hand, since 1993, there were fuctuations in exchange rate, which become severe
during the crisis period. Indian rupee depreciated against US dollar by 6.31 per cent
between July 1997 and March 1998 and by approximately 11 per cent from July 1997 to
December 1998 (Dua and Sinha, 2007), clearly justifying the robustness of the result.
The period from2001 to 2003 is also considered as the period of slowglobal recovery
from recession coupled with terrorist attacks on the USA, subsequent war on
Afghanistan and apprehensions about rise in oil prices due to a war-like situation
between USA and Iraq. There was a massive stock market crash during 2002 in stock
exchanges across the USA, Canada, Asia and Europe. After recovering from lows
following September 11 attacks in 2001, indices again started dipping down in March
2002. These situations could have possibly produced the break obtained in March 2001.
As the results of BP method make sense and are consistent, we consider these results for
our further analysis.
JFEP
7,4
310
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table I.
Results of unit root
tests
V
a
r
i
a
b
l
e
A
t
l
e
v
e
l
A
t
f
r
s
t
D
i
f
f
e
r
e
n
c
e
A
D
F
D
F
-
G
L
S
P
P
A
D
F
D
F
-
G
L
S
P
P
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
i
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
i
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
e
?
2
.
1
6
3
?
1
.
6
5
1
1
.
5
2
5
?
1
.
0
9
4
?
2
.
1
0
3
?
1
.
6
5
9
?
1
0
.
7
0
5
*
*
?
1
0
.
8
2
9
*
*
?
1
0
.
5
8
4
*
*
?
1
0
.
6
0
1
*
*
?
1
0
.
8
0
*
*
?
1
0
.
8
8
3
*
*
m
0
.
6
1
6
?
1
.
6
4
1
.
4
3
1
?
1
.
9
5
5
0
.
9
8
5
?
1
.
9
9
?
1
6
.
9
8
4
*
*
?
1
6
.
9
7
*
*
?
1
5
.
0
2
2
*
*
?
1
5
.
1
9
4
*
*
?
1
7
.
4
3
1
*
*
?
1
7
.
7
5
4
*
m
*
0
.
3
1
4
?
2
.
7
5
4
1
.
2
2
5
?
1
.
3
1
2
0
.
5
0
9
?
3
.
0
6
9
?
8
.
4
1
3
*
*
?
8
.
4
0
6
*
*
?
3
.
6
7
0
6
*
*
?
5
.
3
9
6
7
*
*
?
1
7
.
1
2
5
*
*
?
1
7
.
1
3
1
*
*
p
?
0
.
1
8
1
?
1
.
2
8
1
4
.
4
1
2
*
*
?
1
.
0
2
9
?
0
.
1
9
9
?
1
.
4
1
9
?
1
2
.
4
5
5
*
*
?
1
2
.
4
2
*
*
?
9
.
1
9
9
4
*
*
?
9
.
5
7
4
*
*
?
1
2
.
4
6
2
*
*
?
1
2
.
4
2
8
*
*
p
*
?
0
.
4
5
6
?
3
.
4
7
6
*
*
3
.
4
7
0
*
*
?
2
.
4
2
8
?
0
.
4
4
7
?
3
.
3
7
2
*
*
?
1
6
.
4
2
2
*
*
?
1
6
.
3
8
3
*
*
?
9
.
5
1
7
*
*
?
9
.
4
8
2
*
*
?
1
6
.
7
5
3
*
*
?
1
6
.
7
1
1
*
*
y
?
0
.
2
7
7
?
1
.
8
6
7
6
.
2
2
3
*
*
?
1
.
3
2
6
?
0
.
3
5
8
?
2
.
6
5
1
?
2
3
.
4
7
8
*
*
?
2
3
.
4
1
6
*
*
?
1
.
7
3
2
*
*
?
2
0
.
7
9
8
*
*
?
2
3
.
9
0
9
*
*
?
2
3
.
8
4
3
*
*
y
*
?
2
.
3
5
7
?
2
.
5
5
1
?
0
.
2
3
1
?
1
.
6
9
1
?
2
.
6
8
2
*
*
?
1
.
1
0
8
?
3
.
4
7
4
*
*
?
3
.
7
4
7
*
*
?
2
.
7
4
6
4
*
*
?
3
.
3
4
6
*
*
?
1
4
.
3
2
4
*
*
?
1
4
.
6
1
9
*
*
i
?
1
.
4
3
3
?
2
.
6
1
8
?
1
.
4
5
7
?
1
.
9
0
1
?
1
.
6
2
2
?
2
.
8
8
6
?
1
3
.
7
8
0
*
*
?
1
3
.
7
7
3
*
*
?
1
2
.
7
5
3
*
*
?
1
3
.
6
3
4
*
*
?
1
3
.
8
1
3
*
*
?
1
3
.
8
0
3
*
*
i
*
?
1
.
4
2
5
?
2
.
2
8
7
?
1
.
4
2
2
?
1
.
9
7
6
?
0
.
9
5
5
?
2
.
0
3
7
?
4
.
7
2
6
*
*
?
4
.
7
2
3
*
*
?
0
.
9
9
3
?
3
.
0
9
1
*
*
?
9
.
1
3
8
*
*
?
9
.
0
4
3
3
*
*
t
b
?
1
.
2
1
8
?
7
.
4
8
4
*
*
?
0
.
4
4
1
?
5
.
3
9
2
*
?
2
.
6
4
1
?
7
.
4
8
6
?
1
0
.
8
8
7
*
*
?
1
0
.
8
5
2
*
*
?
1
5
.
2
3
8
*
*
?
1
6
.
3
3
2
*
*
?
4
1
.
5
6
6
*
*
?
6
2
.
0
6
4
*
*
t
b
*
?
1
.
5
2
3
?
1
.
2
3
1
0
.
2
5
1
?
1
.
4
6
4
?
1
.
5
2
3
?
1
.
3
3
9
?
6
.
7
7
3
*
*
?
6
.
8
3
8
*
*
?
6
.
2
9
4
*
*
?
6
.
5
0
6
*
*
?
1
5
.
2
9
3
*
*
?
1
5
.
3
6
8
*
*
N
o
t
e
:
*
*
D
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
311
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4.4 Results of the model estimation: a long-run analysis
We now have two options for estimating the long-run relationship. First, we can use
dummy variables for the quantifed breaks in the model. Second, divide the sample in
sub-samples according to break in data. We have used both options to test the
robustness of our results. To examine a long-run equilibrium relationship, we estimate
equation (3) for two sub-samples[4] (April 1994-February 2001 and April 2001-March
2010) as well as for full sample using FMOLS estimator and results are reported in
Table III. Column 1 of the table reports results of full sample with and without break
dummies. Results exhibit that for the full sample period, except money supply and
interest rate, other domestic variables signifcantly determine the movement of
exchange rate. Inclusion of dummies does not signifcantly improve the results (see
Column 2). In fact, neither the size of the coeffcients nor their statistically signifcance
differ in both models. Nevertheless, the dummies for both breaks are estimated to be
positive and statistically signifcant, which further validates the results of the BP test.
Interestingly, for both of the sub-sample periods, domestic interest rate and output
are estimated to be signifcant, supporting both these variables as important
determinants of exchange rate. Furthermore, both domestic and US money supply are
found to have a crucial role in determining the exchange rate movement for the period
April 1994 to March 2010. Importantly, results indicate that domestic money supply has
Table II.
Results of Lee and
Strazicich (2003) LM
test
Variable S
t?1
Const. D1 [Break data] D2 [Break data] Lag
e ?0.027 (?2.14) 0.005 (3.11) ?0.0245 (?1.74) [2004:03] ?0.0289 (?2.03) [2005:12] 5
m ?0.055 (?2.29) 0.0125 (20.64) ?0.0204 (?2.57) [2004:08] 0.0693 (8.79) [2006:12] 0
m* ?0.0145 (?1.47) 0.005 (13.37) 0.0033 (0.93) [2001:10] 0.0035 (1.047) [2004:01] 5
p ?0.008 (?1.208) 0.006 (12.48) 0.0075 (1.19) [1999:04] ?0.014 (?2.409) [2005:04] 4
p* ?0.0801 (?3.36) 0.0025 (11.128) ?0.0012 (?0.48) [1997:04] 0.0076 (3.108) [1997:04] 1
y ?0.0301 (?1.58) 0.0058 (5.86) 0.0078 (0.56) [1998:11] ?0.0004 (?0.028) [2005:07] 5
y* ?0.0251 (?3.05) 0.0045 (3.92) ?0.006 (?1.03) [1996:06] 0.0173 (2.88) [1998:07] 5
i ?0.0731 (?2.907) 0.1804 (2.17) ?1.1702 (?2.18) [2000:03] 2.514 (4.31) [2007:06] 3
i* ?0.022 (?2.408) 0.006 (0.336) 0.789 (4.36) [1998:10] 0.112 (0.601) [2007:08] 3
tb ?0.5807 (?8.22) ?0.246 (?4.98) ?1.1757 (?2.42) [1999:02] 0.438 (0.43) [2004:04] 0
tb* ?0.0405 (?2.33) 0.0154 (2.629) 0.145 (2.49) [1996:11] 0.0757 (1.303) [1998:07] 4
Notes: This are results of Lee-Strazicich (2003) unit root test crash model with two breaks; the
numbers in parenthesis are the t-statistics for the estimated coeffcients; the coeffcient on S
t?1
tests for
the unit root; D1 and D2 equal the breaks of the slope
Table III.
Breakpoints obtained
using BP test
Year Month
1996 3
2001 3
Number of breaks: 2
Notes: These breakpoints are obtained by using Bai and Perron (1998, 2003) methodology to the
regression Y
t
??(t) ??(t)X
t
??
t
, where Y
t
denotes the Indian Rupee-US dollar exchange rate and X
t
is a 5 ? 1 vector of fve macroeconomic fundamentals; the sample period covers monthly data from
April 1994 to March 2010
JFEP
7,4
312
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
a positive effect on the exchange rate, whereas the US money supply has a negative
effect for this period. It is also noteworthy that coeffcient of US money supply is quite
sizable and statistically signifcant for all the periods. These fndings exhibit the impact
of money supply and its importance as a component of monetary management in
exchange rate determination. Results also indicate that domestic price level has a
positive coeffcient for the whole sample period as well as for the frst sub-sample period
estimation, whereas it is not found to be signifcant in case of the last period. On the other
hand, foreign price level is negatively impacting the exchange rate in the later
sub-period. Results regarding domestic output have a negative effect on exchange rate
in case of full sample and the last sub-period, whereas no signifcant effects are observed
in case of the frst sub-period. Our fndings regarding domestic interest rates indicate
that it has a positive, however, not very sizable effect on the exchange rate movement in
both sub-periods. For a robustness check, we also apply dynamic OLS (DOLS)
estimator[5], and results are presented in Table A3. These results corroborate the
fndings of FMOLS with some size variations of coeffcients (Table IV).
4.5 Results of the cointegration tests
The relationship obtained from the FMOLS estimation can be further validated for its
long-run stability by applying the idea of residual-based cointegration tests. For this
purpose, we employ unit root tests to the resulting error series, which can be interpreted
as a test of cointegration relationship. Table Vshows the ADF, DF-GLS and PPunit root
tests for the regression error terms obtained in all the study periods. Evidently, all these
tests indicate the presence of a long-run relationship in all the sample study periods. The
result is in line with the fndings of Kletzer and Kohli (2000) and Dua and Ranjan (2011).
For a robustness check, the JJ cointegration test is also conducted, and the results are
presented in Table VI. Results of trace test and the maximumeigenvalue test indicate a
cointegrating relationship, implying a long-run relationship among the variables for the
full study period as well as the two sub-periods. It is noteworthy that the number of
cointegrating equations is higher in both the sub-periods as compared to the entire study
period. Gregory and Hansen (1996) extended the Engle–Granger test and allowed for a
one-time endogenously break in the cointegrating vector. For a robustness test, we also
conduct Gregory and Hansen (1996) test of cointegration test which indicates
cointegration among our variables and indicates break in the system in July 1997 (see
Table A3). Although this test is comparatively powerful than the Engle–Granger test in
the case of presence of a single break, in our case, it is not very suitable, as it signifcantly
lacks power in the presence of multiple breaks and in the presence of multiple
cointegrating vectors. Therefore, fnally, we use Johansen et al. (2000) cointegration test,
which uses the JJ framework. Specifcally, we use its trend breaks (Hl(r)) variant for
cointegration test. The model accommodates structural breaks by including impulse
dummies for the quantifed breaks. The results are reported in Table VII, which
validates our earlier results that variables in our models have a long-run inter-linkage;
thus, they are cointegrated.
4.6 Results of the Wald restriction tests
After establishing a co-movement between exchange rate and macro-fundamentals
in the long run, we attempt to further validate the estimated linkage by applying the
Wald restriction tests to the empirical models which are estimated by FMOLS. We
313
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table IV.
FMOLS estimations
V
a
r
i
a
b
l
e
s
F
u
l
l
s
a
m
p
l
e
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
1
9
9
4
t
o
F
e
b
r
u
a
r
y
2
0
0
1
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
2
0
0
1
t
o
M
a
r
c
h
2
0
1
0
W
i
t
h
o
u
t
d
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
W
i
t
h
b
r
e
a
k
d
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
c
o
n
s
t
a
n
t
?
2
.
5
6
2
(
?
0
.
8
1
1
)
?
0
.
6
8
5
(
?
0
.
2
0
1
)
1
.
5
3
1
(
0
.
5
5
3
)
9
.
0
7
2
1
5
*
*
(
2
.
8
4
1
)
m
?
0
.
0
4
8
(
?
0
.
1
7
6
)
0
.
0
1
5
(
0
.
0
5
6
2
)
0
.
7
3
2
*
*
(
3
.
0
5
9
)
0
.
0
7
4
0
5
(
0
.
3
1
4
)
m
*
1
.
1
6
3
*
*
(
2
.
6
4
4
)
1
.
2
0
1
*
*
(
3
.
7
0
9
)
?
1
.
5
6
8
*
(
2
.
9
2
2
)
0
.
8
9
1
8
5
*
*
(
2
.
1
1
4
)
p
1
.
2
5
2
*
*
(
5
.
8
3
7
)
1
.
1
4
0
2
*
*
(
5
.
3
6
7
)
0
.
7
9
7
*
*
(
5
.
1
7
9
)
0
.
4
9
3
1
(
1
.
6
0
7
)
p
*
?
0
.
4
2
9
(
?
0
.
6
5
9
)
?
1
.
0
6
1
(
?
1
.
5
3
6
)
0
.
2
3
8
(
0
.
3
8
2
)
?
1
.
4
2
0
1
*
(
?
1
.
9
1
3
)
y
?
1
.
4
4
1
*
*
(
?
7
.
0
7
4
)
?
1
.
3
3
3
*
*
(
?
7
.
1
8
5
)
0
.
5
7
7
9
*
*
(
3
.
2
1
1
)
?
0
.
6
7
9
3
*
*
(
?
2
.
6
3
4
)
y
*
?
0
.
2
4
5
(
?
1
.
3
4
9
)
?
0
.
1
8
4
(
?
1
.
0
6
3
)
?
0
.
7
8
9
*
*
(
?
2
.
9
3
2
)
?
1
.
5
8
3
9
*
*
(
?
5
.
6
6
8
)
i
?
0
.
0
0
3
(
?
0
.
9
7
1
)
0
.
0
0
2
(
0
.
4
9
1
)
0
.
0
0
9
*
*
(
3
.
6
1
6
)
0
.
0
1
6
0
6
*
*
(
2
.
8
8
7
)
i
*
0
.
0
2
5
4
*
*
(
4
.
2
3
6
)
0
.
0
2
5
*
*
(
4
.
6
3
6
)
?
0
.
0
0
1
(
?
0
.
1
1
3
)
0
.
0
2
3
3
2
*
*
(
4
.
8
6
1
)
t
b
?
0
.
0
3
6
*
(
?
4
.
1
8
6
)
?
0
.
0
2
8
*
*
(
?
3
.
7
1
)
0
.
0
0
7
(
1
.
4
0
1
)
?
0
.
0
2
1
*
*
(
2
.
4
1
1
)
t
b
*
0
.
0
4
7
3
(
0
.
9
7
8
)
0
.
0
2
3
(
0
.
5
3
1
)
?
0
.
0
4
3
(
?
1
.
1
6
9
)
0
.
1
6
8
0
8
*
*
(
3
.
3
7
1
)
D
u
m
m
y
M
a
r
c
h
1
9
9
6
0
.
0
5
8
*
*
(
2
.
0
6
)
D
u
m
m
y
M
a
r
c
h
2
0
0
1
0
.
0
3
5
(
1
.
3
6
)
A
d
j
u
s
t
e
d
R
-
s
q
u
a
r
e
d
0
.
8
8
8
1
0
.
8
9
2
3
0
.
9
6
4
9
0
.
8
3
5
2
N
o
t
e
s
:
*
*
A
n
d
*
d
e
n
o
t
e
s
i
g
n
i
f
c
a
n
c
e
a
t
5
a
n
d
1
0
%
l
e
v
e
l
,
r
e
s
p
e
c
t
i
v
e
l
y
JFEP
7,4
314
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
test the null hypothesis for all coeffcients by individually restricting them to zero.
Table VIII presents results of Wald test for six different null hypotheses. Column 1
of the table reports results of the test in which the null hypothesis states that all
coeffcients, except the intercept termare zero. The signifcant t-statistic value at the
conventional level recommends the rejection of the null hypothesis in all three
sample periods, implying that at least one fundamental variable is signifcant in the
overall period. Null hypothesis in Column 2 states that coeffcients of national
output are restricted to zero, which is also rejected completely in all periods.
Similarly, all other null hypotheses are rejected in all cases with only one exception
where trade balance for the period April 1994 to February 2001 is restricted to zero
(see Column 6 of Table VIII). Nowcombining these results with FMOLS estimations,
we can understand the signifcance of different theoretical models in each period, as
shown in Table IX. Specifcally, our results indicate for a validity of fexible price
model, Sticky price model and real interest rate model in the Indian case. The only
exception is Hooper and Morton model as trade balance, a crucial variable in this
model, is signifcant only in the full sample period, after accounting for the
structural break in the full model; trade balance loses out its signifcance at least in
one sub-period.
4.6 Results of the IRF: a short-run analysis
To see the short-term interaction, the Cholesky one standard deviation response
functions of exchange rate over a period of 10 months are plotted for money supply,
price level, short-term interest rate and trade balance. The graphs in Figure 2a-2e
suggest that there is an inverse relationship between exchange rate and price index.
Over a period of 10 months, for the frst three periods, the response is immediate and
sharp, but beyond that, it is fxed. This inverse relationship is in accordance with the
sign conventions of RID and Hooper and Morton model. As the domestic price level
increases, the exchange rate increases or the currency depreciates. The response of
impulses to output does not show much deviation in a period of 10 months. So, none
of the models explain the effect of output on exchange rate in such a short span. It
may take a longer span of time to refect the effect of these changes. The response to
money supply and the net trade balance show a similar upward trend for an initial
period of two months, and thereafter, it attains a consistent level. All theoretical
models support the fact that as money supply increases, currency depreciates, i.e.
exchange rate, increases. The response of trade balance shock is quite ambiguous,
as with the increase in exports minus imports, the currency should appreciate and
lower down the exchange value. Finally, the interest rate shocks provide an
important fnding of direct relationship with the exchange rate as shown by the
Table V.
Results of unit root
test for error terms
Sample period ADF DF-GLS PP
April 1994 to March 2010 (full sample) ?5.908** ?4.287** ?5.829**
April 1994 to February 2001 ?3.959** ?3.953** ?3.899**
April 2001 to March 2010 ?6.117** ?5.008** ?6.117**
Note: **Denotes signifcance at 5% level
315
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table VI.
Johansen and
Juselius (1990)
cointegration test
results
H
0
r
?
0
r
?
1
r
?
2
r
?
3
r
?
4
r
?
5
r
?
6
r
?
7
r
?
8
r
?
9
r
?
1
0
P
a
n
e
l
A
:
f
u
l
l
s
a
m
p
l
e
p
e
r
i
o
d
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
E
i
g
e
n
v
a
l
u
e
0
.
4
3
6
0
.
3
9
0
0
.
3
4
9
0
.
2
9
0
0
.
2
2
8
0
.
2
0
9
0
.
1
5
0
0
.
1
3
1
0
.
0
6
7
0
.
0
3
6
0
.
0
0
0
1
?
T
r
a
c
e
4
2
8
.
3
5
*
*
3
3
8
.
9
6
*
*
2
6
1
.
8
2
*
*
1
9
4
.
4
1
*
*
1
4
1
.
2
8
*
*
1
0
0
.
9
0
*
*
6
4
.
1
7
3
8
.
7
9
1
6
.
7
1
5
.
7
5
0
.
0
2
1
?
M
a
x
8
9
.
3
8
*
*
7
7
.
1
6
*
*
6
7
.
0
0
4
*
*
5
3
.
5
0
*
*
4
0
.
3
7
3
6
.
7
3
2
5
.
3
8
2
2
.
0
7
1
0
.
9
5
5
.
7
3
0
.
0
2
P
a
n
e
l
B
:
f
r
s
t
s
u
b
-
p
e
r
i
o
d
(
A
p
r
i
l
1
9
9
4
t
o
F
e
b
r
u
a
r
y
2
0
0
1
)
E
i
g
e
n
v
a
l
u
e
0
.
7
9
6
0
.
6
4
9
0
.
6
4
5
0
.
5
8
5
0
.
4
7
6
0
.
3
9
8
0
.
3
7
5
0
.
2
6
8
0
.
2
0
1
0
.
1
7
0
7
0
.
0
4
2
?
T
r
a
c
e
4
3
1
.
0
0
5
*
*
3
3
2
.
1
4
*
*
2
6
7
.
1
5
*
*
2
0
2
.
9
4
*
*
1
4
8
.
3
5
*
*
1
0
8
.
2
3
*
*
7
6
.
7
1
*
*
4
7
.
5
0
2
8
.
1
1
1
4
.
2
8
2
.
6
7
?
M
a
x
9
8
.
8
6
*
*
6
4
.
9
9
*
*
6
4
.
2
1
*
*
5
4
.
6
1
*
*
4
0
.
0
8
3
1
.
5
1
2
9
.
2
1
1
9
.
3
8
1
3
.
8
3
1
1
.
6
0
2
.
6
7
P
a
n
e
l
C
:
s
e
c
o
n
d
s
u
b
-
p
e
r
i
o
d
(
A
p
r
i
l
2
0
0
1
t
o
M
a
r
c
h
2
0
1
0
)
E
i
g
e
n
v
a
l
u
e
0
.
6
9
7
0
.
6
3
2
0
.
5
5
0
0
.
4
3
2
0
.
3
8
5
0
.
3
0
8
0
.
2
6
9
0
.
2
3
5
0
.
1
7
5
0
.
0
7
1
0
.
0
0
0
3
?
T
r
a
c
e
5
1
0
.
3
5
*
*
3
9
4
.
4
0
*
*
2
9
6
.
5
1
*
*
2
1
8
.
9
6
*
*
1
6
3
.
8
2
*
*
1
1
6
.
6
7
*
*
8
0
.
9
3
3
*
*
5
0
.
4
6
8
*
*
2
4
.
4
2
7
.
0
7
7
0
.
0
3
7
?
M
a
x
1
1
5
.
9
4
*
*
9
7
.
8
9
*
*
7
7
.
5
4
9
*
*
5
5
.
1
3
*
*
4
7
.
1
5
8
*
*
3
5
.
7
3
3
0
.
4
6
2
6
.
0
4
3
1
7
.
3
4
7
.
0
3
9
0
.
0
3
7
N
o
t
e
:
*
*
D
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
JFEP
7,4
316
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table VII.
Results of Johansen
et al. (2000)
cointegrating test
H
0
r
?
0
r
?
1
r
?
2
r
?
3
r
?
4
r
?
5
r
?
6
r
?
7
r
?
8
r
?
9
r
?
1
0
E
i
g
e
n
v
a
l
u
e
0
.
4
7
6
2
0
.
3
9
2
5
0
.
3
2
9
8
0
.
2
7
2
7
0
.
2
5
3
8
0
.
1
9
5
1
0
.
1
8
8
9
6
0
.
1
4
3
7
0
.
1
2
7
6
0
.
0
9
2
6
0
.
0
3
3
9
?
T
r
a
c
e
5
2
3
.
1
4
*
*
4
1
0
.
6
2
*
*
3
2
3
.
8
8
*
*
2
5
4
.
2
4
*
*
1
9
8
.
8
3
*
*
1
4
7
.
8
9
*
*
1
1
0
.
1
3
*
*
7
3
.
6
8
*
*
4
6
.
6
8
2
6
5
*
*
2
2
.
9
2
5
*
*
6
.
0
1
4
7
N
o
t
e
s
:
C
r
i
t
i
c
a
l
v
a
l
u
e
s
a
r
e
c
a
l
c
u
l
a
t
e
d
b
a
s
e
d
o
n
G
i
l
e
s
a
n
d
G
o
d
w
i
n
,
2
0
1
2
;
*
*
d
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
317
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table VIII.
Wald restriction tests
on the coeffcients of
the empirical models
S
a
m
p
l
e
p
e
r
i
o
d
H
0
:
A
l
l
?
?
0
(
1
)
H
0
:
?
1
?
?
2
?
0
(
2
)
H
0
:
?
3
?
?
4
?
0
(
3
)
H
0
:
?
5
?
?
6
?
0
(
4
)
H
0
:
?
7
?
?
8
?
0
(
5
)
H
0
:
?
9
?
?
1
0
?
0
(
6
)
1
9
9
4
:
0
4
-
2
0
0
1
:
0
2
2
3
8
.
1
1
7
9
*
*
(
0
.
0
0
0
0
)
7
.
0
4
5
9
1
3
*
*
(
0
.
0
0
1
9
)
1
1
.
8
4
1
5
8
*
*
(
0
.
0
0
0
1
)
4
.
8
7
3
6
4
8
*
*
(
0
.
0
1
1
3
)
4
.
9
8
5
2
4
2
*
*
(
0
.
0
1
0
2
)
1
.
0
8
0
3
1
(
0
.
3
4
6
6
)
A
p
r
i
l
2
0
0
1
-
M
a
r
c
h
2
0
1
0
2
7
.
7
3
8
6
3
*
*
*
(
0
.
0
0
0
0
)
4
2
.
8
1
8
3
4
*
*
(
0
.
0
0
0
0
)
3
.
9
7
1
4
0
7
*
*
(
0
.
0
2
2
1
)
1
0
.
0
8
9
4
7
*
*
(
0
.
0
0
0
1
)
2
0
.
3
4
3
1
3
*
*
(
0
.
0
0
0
0
)
1
0
.
4
8
3
8
*
*
(
0
.
0
0
0
1
)
A
p
r
i
l
1
9
9
4
-
M
a
r
c
h
2
0
1
0
(
f
u
l
l
s
a
m
p
l
e
)
8
.
2
3
E
?
2
3
*
*
(
0
.
0
0
0
0
)
1
7
.
8
7
9
7
1
*
*
(
0
.
0
0
0
0
)
4
4
3
.
2
7
1
8
*
*
(
0
.
0
0
0
0
)
6
.
4
8
E
?
2
1
*
*
(
0
.
0
0
0
0
)
1
5
0
.
8
9
8
7
*
*
(
0
.
0
0
0
0
)
1
1
2
.
6
3
6
7
*
*
(
0
.
0
0
0
0
)
N
o
t
e
s
:
?
1
,
?
2
,
?
3
,
?
4
,
?
5
,
?
6
,
?
7
,
?
8
,
?
9
a
n
d
?
1
0
a
r
e
t
h
e
c
o
e
f
f
c
i
e
n
t
s
o
f
y
,
y
*
,
p
,
p
*
,
m
,
m
*
,
i
,
i
*
,
t
b
a
n
d
t
b
*
,
r
e
s
p
e
c
t
i
v
e
l
y
;
p
-
V
a
l
u
e
i
n
p
a
r
e
n
t
h
e
s
i
s
;
*
*
d
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
JFEP
7,4
318
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
steep increase in the response graph. This is in accordance to the practicability of
interest rate as an effective policy instrument in controlling exchange rate volatility.
5. Conclusion
This study has considered the possibility of structural break in the data while
examining the long-run relationship between Indian rupee-dollar exchange rate and
macroeconomic fundamentals. It is shown that fundamentals have a signifcant
impact on each sub-period; however, their size of impact differs signifcantly,
depending on various regimes and period. Our results show that macroeconomic
environment can produce breaks, which ultimately give rise to time-varying
coeffcients. This implies that for a developing country like India, the fundamentals
matter in formulating an exchange rate policy, but their behavior is not constant
over the period. Therefore, the monetary authority needs to be more fexible in terms
of controlling the exchange rate via monetary variables. The FMOLS results prove
that there is no single particular model which can explain exchange rate behavior in
the long run, and the coeffcients of the fundamentals are dynamic in the long run.
This result is in disagreement of Meese and Rogoff (1983) that monetary models fail
to perform on the empirical ground and, thus, offers support to the fndings of
MacDonald and Taylor (1994) and Beckmann et al. (2010). Importantly, our results
show that incorporating structure breaks in the model, the cointegration framework
perform well in explaining the effects of fundamentals on exchange rate movement.
In a way, this provided the evidence for the temporal stability of parameters of
exchange rate models based on macroeconomic fundamentals in the Indian currency
context. Results of the short-run analysis, i.e. impulse responses on the exchange
rate-interest rate linkage, highlight the interest rate importance as an effective
policy tool for smoothening out the volatility of exchange rate on a short-term basis.
The study of the behavior of a macroeconomic variable in different regimes have
highlighted the importance of each macroeconomic factor in movement of exchange
rate in this research, which will likely help policymakers in designing the
appropriate policy by understanding the deep-rooted, complex and time-varying
relationship.
Overall, our fndings are signifcant and relevant for a policy perspective. It is,
however, contrasting with much, if not all, of the existing empirical evidence on this
issue. Most importantly, our analysis submits that the monetary models of exchange
rate models work, but results are sensitive toward period of selection and
methodology. We also believe that the analysis may improve further if it is tested on
the next-generation tests of structural breaks in the cointegrating system[6].
Table IX.
Signifcance of
exchange rate models
in different periods
Sample period
Flexible price
model
Sticky price
model
Real interest
rate model
Hooper and Morton
model
January1994-February 2001 YES YES YES NO
April 2001-March 2010 YES YES YES NO
April 1994-March 2010 (full sample) YES YES YES YES
319
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Notes
1. The annual average exchange rate of the rupee went up from INR 45.56 per US dollar in
2010-2011 to INR 47.92 per US dollar in 2011-2012 and further to INR 54.41 per US dollar in
2012-2013. It rose to reach an average of INR 60.50 per US dollar in 2013-2014. The intra-year
levels of depreciation have been sharper in some months; but exhibit two-way movements
within the broad rising trend (MoF, 2013, p. 140).
2. Recently, the Chief Economic Advisor of Government of India, Arvind Subramanian, has
been quoted saying that “Sustaining a weaker currency by an aggressive forex reserves’
Figure 2.
IRF of exchange rate
with other
macroeconomic
variables
JFEP
7,4
320
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
build-up – $750 billion to $1 trillion in the medium-term – with letting the rupee dip at every
opportunity” (ET, 2015).
3. The break model results signifcantly vary from that of crash model. In both models,
breakpoints lack consistency as well as many of breakpoints could not clear the statistical
test. The result of the break model is not reported here to conserve the space; however, it will
be provided on a request from the corresponding author.
4. The frst breakpoint can be ignored, as it has lesser utility in predicting a relationship, as the
numbers of observations till frst breakpoint are very few, which left us with no choice but to
consider only one major breakpoint.
5. DOLS is developed by Saikkonen (1992) and Stock and Watson (1993); the method involves
augmenting the cointegrating regression with lags and leads of explanatory variables so that
the resulting cointegrating equation’s error term is orthogonal to the entire history of the
stochastic regressor innovations.
6. One such unit root test recently developed by Narayan and Poop (2010).
References
Baharumshah, A.Z., Mohd, S.H. and Ahn, S.K. (2009), “On the predictive power of monetary
exchange rate model: the case of the Malaysian ringgit/US dollar rate”, Applied Economics,
Vol. 41 No. 14, pp. 1761-1770.
Bai, J. and Perron, P. (1998), “Estimating and testing linear models with multiple structural
changes”, Econometrica, Vol. 66, pp. 47-78.
Bai, J. and Perron, P. (2003), “Computation and analysis of multiple structural change models”,
Journal of Applied Econometrics, Vol. 6, pp. 72-78.
Beckmann, J., Belke, A. and Kuhl, M. (2010), “The dollar-euro exchange rate and macroeconomic
fundamentals: a time-varying coeffcient approach”, Review of World Economics, Vol. 147,
pp. 11-40.
Bekiros, S.D. (2014), “Exchange rates and fundamentals: co-movement, long-run relationships and
short-run dynamics”, Journal of Banking & Finance, Vol. 39, pp. 117-134.
Bhanumurthy, N.R. (2006), “Macroeconomic fundamentals and exchange rate dynamics in India”,
Economic and Political Weekly, Vol. 10 No. 11, pp. 1101-1107.
Bilson, J.F.O. (1978), “The monetary approach to the exchange rate: some empirical evidence”,
IMF Staff Papers, Vol. 25, pp. 48-75.
Chang, M.-J. and Su, C.Y. (2014), “The dynamic relationship between exchange rates and
macroeconomic fundamentals: evidence from Pacifc Rim countries”, Journal of
International Financial Markets, Institutions and Money, Vol. 30, pp. 220-246.
Cheung, Y., Chin, M.D. and Pascual, A.G. (2005), “Empirical exchange rate models of the nineties:
are any ft to survive?”, Journal of International Money and Finance, Vol. 24, pp. 1150-1175.
Devereux, M.B. (1997), “Real exchange rates and macroeconomics: evidence and theory”,
Canadian Journal of Economics, Vol. 30 No. 4a, pp. 773-808.
Dornbusch, R. (1976), “Expectations and exchange rate dynamics”, Journal of Political Economy,
Vol. 84, pp. 1161-1176.
Dropsy, V. (1996), “Real exchange rates and structural breaks”, Applied Economics, Vol. 28,
pp. 209-219.
321
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Dua, P. and Ranjan, R. (2011), “Modelling and forecasting the Indian Re/US$ exchange rate”,
Working Paper no. 197, Centre for Development Economics, Delhi School of Economics,
New Delhi.
Dua, P. and Sinha, A. (2007), “India’s insulation fromthe East Asian crisis: an analysis”, Singapore
Economic Review, Vol. 52, pp. 419-443.
ET (2015), “Let rupee dip when possible for India’s long-term economic wellbeing: Arvind
Subramanian”, CEA, 2 March, available at:http://articles.economictimes.indiatimes.com/
2015-03-02/news/59683835_1_exchange-rate-capital-infows-arvind-subramanian
Florentis, G., Gordon, D.V. and Huber, P. (1994), “Monetary models of Canadian US exchange rate:
a re-examination of empirical evidence 1971-1986”, Quarterly Journal of Business and
Economics, Vol. 33, pp. 27-43.
Frankel, J. (1984), “Tests of monetary and portfolio balance models of exchange rate
determination”, in Bilson, J.F.O. and Marston, R.C. (Eds), Exchange Rate Theory and
Practice, Chicago University Press, Chicago, pp. 239-259.
Frankel, J.A. (1979), “On the mark: a theory of foating exchange rates based on real interest rate
differentials” American Economic Review, Vol. 69, pp. 610-622.
Frenkel, J.A. (1976), “A monetary approach to the exchange rate: doctrinal aspects and empirical
evidence”, Scandinavian Journal of Economics, Vol. 78, pp. 200-224.
Giles, D.E. and Godwin, R.T. (2012), “Testing for multivariate cointegration in the presence of
structural breaks: p-values and critical values”, Applied Economics Letters, Vol. 19 No. 16,
pp. 1561-1565.
Goldberg, M.D. and Frydman, R. (1996), “Empirical exchange rate models and shifts in the
co-integrating vector”, Structural Change in Economic Dynamics, Vol. 7, pp. 55-78.
Gregory, A.W. and Hansen, B.E. (1996), “Residual-based tests for cointegration in models with
regime shifts”, Journal of Econometrics, Vol. 70 No. 1, pp. 99-126.
Hegwood, N.D. and Nath, H.K. (2014), “Real exchange rate dynamics: evidence from India”,
Economic Analysis and Policy, Vol. 44 No. 4, pp. 396-404.
Hodrick, R.J. (1978), “An empirical analysis of the monetary approach to the determination of the
exchange rate”, in Frenkel, J.A. and Johnson, H.G. (Eds), The Economics of Exchange Rates,
Reading: Addison-Wesley, Boston.
Hoffmann, M. and MacDonald, R. (2009), “Real exchange rates and real interest rate differentials:
a present value interpretation”, European Economic Review, Vol. 53, pp. 952-970.
Hooper, P. and Morton, J. (1982), “Fluctuations in the dollar: a model of nominal and real exchange
rate determination”, Journal of International Money and Finance, Vol. 1, pp. 39-56.
Johansen, S. and Juselius, K. (1990), “Maximum likelihood estimation and inference on
cointegration with applications to the demand for money”, Oxford Bulletin of Economics
and Statistics, Vol. 52, pp. 169-210.
Johansen, S., Mosconi, R. and Nielsen, B. (2000), “Cointegration analysis in the presence of
structural breaks in deterministic trends”, Econometrics Journal, Vol. 3, pp. 216-249.
Kejriwal, M. and Perron, P. (2010), “Testing for multiple structural changes in cointegrated
regression models”, Journal of Business and Economic Statistics, Vol. 28, pp. 503-522.
Kia, A. (2013), “Determinants of the real exchange rate in a small open economy: evidence from
Canada”, Journal of International Financial Markets, Institutions & Money, Vol. 23
No. 2013, pp. 163-178.
Kim, B.J.C. and Mo, S. (1995), “Cointegration and the long-run forecast of exchange rates”,
Economic Letters, Vol. 48, pp. 353-359.
JFEP
7,4
322
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Kletzer, K. and Kohli, R. (2000), “Exchange rate dynamics with fnancial repression: a test of
exchange rate models for India”, Working Paper no. 52, ICRIER, New Delhi.
Kollman, R. (1997), “The exchange rate in a dynamic optimizing current account model with
nominal rigidities”, International Monetary Fund Working Paper, 97/7, IMF.
Lee, J., List, J. and Strazicich, M. (2006), “Nonrenewable resource prices: deterministic or stochastic
trends?” Journal of Environmental Economics and Management, Vol. 51, pp. 354-370.
Lee, J. and Strazicich, M. (2003), “Minimum LM unit root test with two structural breaks”, Review
of Economics and Statistics, Vol. 85, pp. 1082-1089.
Lee, J. and Strazicich, M. (2004), “Minimum LM unit root tests”, Working Paper No. 04-07,
University of Alabama.
Lumsdaine, R. and Papell, D. (1997), “Multiple trend breaks and the unit root hypothesis”, Review
of Economics and Statistics, Vol. 79, pp. 212-218.
McNown, R. and Wallace, M.S. (1989), “Cointegration tests for long-run equilibrium in the
monetary exchange rate model”, Economic Letters, Vol. 31, pp. 263-267.
Macdonald, R. and Taylor, M.P. (1994), “The monetary model of the exchange rate: long-run
relationships, short-run dynamics and howto beat a randomwalk” Journal of International
Money and Finance, Vol. 13, pp. 276-290.
Mallick, H. (2010), “Examining the behavior of exchange rate under a partial capital mobility
regime”, Forthcoming with International Economic Journal, Vol. 24 No. 2, pp. 141-157.
Mark, N.C. (1995), “Exchange rates and fundamentals: evidence on long horizon predictability”,
American Economic Review, Vol. 85, pp. 201-218.
Mark, N.C. and Sul, D. (2001), “Nominal exchange rates and monetary fundamentals: evidence
from a small post-Bretton woods panel”, Journal of International Economics, Vol. 53,
pp. 29-52.
Meese, R.A. and Rogoff, K. (1983), “Empirical exchange rate models of the seventies. Do they ft
out of sample?”, Journal of International Economics, Vol. 14, pp. 3-24.
Mishra, R.K. and Sharma, C. (2010), “Real exchange rate behavior and optimum currency area in
East Asia: evidence from generalized purchasing power parity”, International Review of
Financial Analysis, Vol. 19, pp. 205-213.
MoF (2014), “Economic survey 2013-2014”, Ministry of Finance, Government of India, Oxford
University Press, New Delhi.
Morley, B. (2007), “The monetary model of the exchange rate and equities: an ARDL bounds
testing approach”, Applied Financial Economics, Vol. 17, pp. 391-397.
Nag, A.K. and Mitra, A. (1998), “Exchange rate of rupee and purchasing power parity”, Economic
and Political Weekly, Vol. 33, pp. 1525-1532.
Narayan, P.K. (2006), “Are bilateral real exchange rates stationary? Evidence from Lagrange
Multiplier unit root tests for India”, Applied Economics, Vol. 38, pp. 63-70.
Narayan, P.K. and Popp, S. (2010), “A new unit root test with two structural breaks in level and
slope at unknown time”, Journal of Applied Statistics, Vol. 37, pp. 1425-1438.
Pattanaik, S. and Sahoo, S. (2003), “The effectiveness of intervention in India: an empirical
assessment”, RBI Occasional Papers, 22, RBI, Mumbai.
Perron, P. (1989), “The great crash, the oil price shock, and the unit root hypothesis”,
Econometrica, Vol. 57, pp. 1361-1401.
Phillips, P.C.B. (1995), “Fully modifed least squares and vector autoregression”, Econometrica,
Vol. 63, pp. 1023-1078.
323
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Phillips, P.C.B. and Hansen, B.E. (1990), “Statistical inference in instrumental variables regression
with I(1) processes”, Review of Economic Studies, Vol. 57, pp. 99-125.
Rapach, D.E. and Wohar, M.E. (2002), “Testing the monetary model of exchange rate
determination: new evidence from a century of data”, Journal of International Economics,
Vol. 58, pp. 359-385.
Rogoff, K. (2009), “Exchange rates in the modern foating era: what do we really know?”, Reviewof
World Economics, Vol. 145 No. 1, pp. 1-12.
Saikkonen, P. (1992), “Estimation and testing of cointegrated systems by an autoregressive
approximation”, Econometric Theory, Vol. 8, pp. 1-27.
Sarno, L. and Valente, J. (2009), “Exchange rates and fundamentals: footloose or evolving
relationship?”, Journal of the European Economic Association, Vol. 7, pp. 786-830.
Stock, J.H. and Watson, M. (1993), “A simple estimator of cointegrating vectors in higher order
integrated systems”, Econometrica, Vol. 61, pp. 783-820.
Tiwari, A.K., Dar, A.B. and Bhanja, N. (2013), “Oil price and exchange rates: a wavelet based
analysis for India”, Economic Modelling, Vol. 31, pp. 414-422.
UNCTAD (1998), “Trade and Development Report 1998”, United Nations Conference on Trade
and Development (UNCTAD), Geneva.
Williamson, J. (2009 ), “Exchange rate economics”, Open Economic Review, Vol. 20, pp. 123-146.
Zhang, S., Lowinger, T.C. and Tang, J. (2007), “The monetary exchange rate model: long-run,
short-run, and forecasting performance”, Journal of Economic Integration, Vol. 22,
pp. 397-406.
Zivot, E. and Andrews, D.W.K. (1992), “Further evidence on the great crash, the oil-price shock,
and the unit-root hypothesis”, Journal of Business & Economic Statistics, Vol. 10,
pp. 251-270.
Further reading
Perron, P. (1997), “Further evidence on breaking trend functions in macroeconomic variables”,
Journal of Econometrics, Vol. 80, pp. 355-385.
WEO (2015), Uneven Growth: Short- and Long-Term Factors, April, IMF, Washington, DC.
Corresponding author
Chandan Sharma can be contacted at: [email protected]
JFEP
7,4
324
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Appendix
Table AI.
Data description
Series Data description Data source
e Indian rupee/US dollar, end-of-month rates, in log form IFS
m M3 of India, in log form, seasonally adjusted IFS
m* M1 of India, in log form, seasonally adjusted IFS
p CPI of India, in log form, seasonally adjusted RBI
p* CPI of India, in log form, seasonally adjusted IFS
y IIP of India, in log form, seasonally adjusted IFS
y* IIP of USA, in log form, seasonally adjusted IFS
i 91-day treasury bill rates in India IFS
i* 91-day treasury bill rates in USA RBI
tb Trade balance of India converted in to an index, seasonally adjusted,
in log form
RBI
tb* Trade balance of USA converted in to an index, seasonally adjusted,
in log form
IFS
Table AII.
Descriptive statistics
y y* p p* e m m* tb tb* i i*
Mean 5.23 4.48 6.18 5.21 3.76 14.32 8.62 5.47 5.62 7.19 3.35
Median 5.17 4.50 6.18 5.20 3.78 14.33 8.65 5.15 5.68 7.00 4.20
Maximum 5.77 4.61 6.68 5.39 3.94 15.52 9.05 7.90 6.36 12.97 6.17
Minimum 4.76 4.27 5.71 5.02 3.45 13.17 8.17 3.10 4.56 3.22 0.03
Std. Dev. 0.28 0.09 0.24 0.11 0.11 0.67 0.27 1.23 0.53 2.36 1.90
Skewness 0.27 ?0.82 0.02 0.09 ?0.99 0.06 ?0.07 0.24 ?0.47 0.68 ?0.39
Kurtosis 1.87 3.13 2.39 1.76 3.13 1.91 1.78 1.92 1.99 3.14 1.62
Jarque-Bera 11.79 19.96 2.82 11.78 29.51 9.02 11.18 10.55 14.16 13.92 18.64
Observations 179 179 179 179 179 179 179 179 179 179 179
Table AIII.
Gregory and Hansen
(1996) cointegration
tests with structural
breaks
Break date Gregory and Hansen test statistic Existence of cointegration
Jul 1997 ?8.505** Yes
Note: **denotes statistically signifcant at 5% level
325
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table AIV.
DOLS estimations
V
a
r
i
a
b
l
e
s
F
u
l
l
s
a
m
p
l
e
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
1
9
9
4
t
o
F
e
b
r
u
a
r
y
2
0
0
1
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
2
0
0
1
t
o
M
a
r
c
h
2
0
1
0
W
i
t
h
o
u
t
D
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
W
i
t
h
B
r
e
a
k
D
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
c
o
n
s
t
a
n
t
?
1
.
0
4
5
(
?
0
2
.
3
6
)
2
.
9
8
2
(
0
.
5
7
8
)
?
3
.
4
9
7
(
0
.
6
8
3
)
1
3
.
9
1
2
*
*
(
2
.
1
3
5
)
m
0
.
1
1
9
(
0
.
3
2
5
)
0
.
3
7
3
7
2
2
(
0
.
9
9
1
)
0
.
2
2
1
*
*
(
2
.
4
4
6
)
0
.
4
5
7
(
1
.
1
8
9
)
m
*
0
.
9
6
5
8
*
(
1
.
7
8
8
)
0
.
5
7
5
*
*
(
2
.
9
3
2
)
?
0
.
5
2
3
(
?
0
.
5
1
2
)
0
.
3
6
5
*
*
(
1
.
7
9
6
)
p
1
.
0
4
4
*
*
(
3
.
9
7
5
)
0
.
9
5
4
9
*
*
(
3
.
0
6
1
)
0
.
9
0
4
(
0
.
4
8
2
)
0
.
0
8
6
(
1
.
5
5
5
)
p
*
?
0
.
2
6
1
(
?
0
.
7
9
5
)
?
1
.
4
4
8
(
1
.
4
3
2
)
0
.
4
4
2
(
0
.
4
2
5
)
?
2
.
0
3
1
(
1
.
4
7
2
)
y
?
1
.
3
3
4
*
*
(
?
4
.
5
3
7
)
?
1
.
3
0
4
*
*
(
?
4
.
2
9
9
)
0
.
5
7
5
*
*
(
2
.
1
2
7
)
?
0
.
5
6
5
*
*
(
?
2
.
1
4
5
)
y
*
?
0
.
2
6
1
(
?
1
.
3
1
6
)
?
0
.
3
0
5
(
?
1
.
4
9
2
)
?
0
.
3
4
1
*
(
?
1
.
7
4
4
)
?
1
.
7
9
7
3
*
*
(
3
.
5
3
2
)
i
?
0
.
0
0
1
(
?
0
.
4
1
2
)
?
0
.
0
0
0
8
(
?
0
.
2
1
8
)
0
.
0
0
9
*
*
(
2
.
1
0
1
)
0
.
0
2
1
*
*
(
2
.
6
2
3
)
i
*
0
.
0
2
4
*
*
(
4
.
1
8
7
)
?
0
.
0
0
0
8
*
(
1
.
7
8
1
)
?
0
.
0
0
2
(
?
0
.
2
1
5
)
0
.
3
6
5
*
*
(
2
.
4
9
6
)
t
b
?
0
.
0
5
4
*
*
(
?
4
.
4
3
3
)
?
0
.
0
4
7
*
*
(
3
.
8
5
1
)
0
.
0
1
7
(
0
.
1
5
5
)
?
0
.
0
2
7
*
(
?
1
.
7
3
7
)
t
b
*
0
.
0
6
2
1
(
1
.
0
4
6
)
0
.
0
8
2
(
1
.
3
7
1
)
?
0
.
1
1
7
*
*
(
?
2
.
1
8
2
)
0
.
2
1
8
*
*
(
2
.
3
0
3
)
D
u
m
m
y
M
a
r
c
h
1
9
9
6
0
.
0
2
7
8
*
(
1
.
8
1
7
)
D
u
m
m
y
M
a
r
c
h
2
0
0
1
0
.
0
3
2
1
(
1
.
0
4
9
)
A
d
j
u
s
t
e
d
R
-
s
q
u
a
r
e
d
0
.
9
2
9
7
0
.
9
3
5
1
0
.
9
8
2
0
.
8
7
7
E
s
t
i
m
a
t
o
r
D
O
L
S
D
O
L
S
D
O
L
S
D
O
L
S
N
o
t
e
:
*
*
D
e
n
o
t
e
s
s
t
a
t
i
s
t
i
c
a
l
l
y
s
i
g
n
i
f
c
a
n
t
a
t
5
%
l
e
v
e
l
JFEP
7,4
326
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_805623593.pdf
This paper aims to examine the relationship between Indian rupee-US dollar exchange rate
and the macroeconomic fundamentals for the post-economic reform period.
Journal of Financial Economic Policy
Macroeconomic fundamentals and dynamics of the Indian rupee-dollar exchange
rate
Chandan Sharma Rajat Setia
Article information:
To cite this document:
Chandan Sharma Rajat Setia , (2015),"Macroeconomic fundamentals and dynamics of the Indian
rupee-dollar exchange rate", J ournal of Financial Economic Policy, Vol. 7 Iss 4 pp. 301 - 326
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-11-2014-0069
Downloaded on: 24 January 2016, At: 21:53 (PT)
References: this document contains references to 65 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 106 times since 2015*
Users who downloaded this article also downloaded:
Muhammad Umar, Gang Sun, (2015),"Country risk, stock prices, and the exchange rate of the
renminbi", J ournal of Financial Economic Policy, Vol. 7 Iss 4 pp. 366-376http://dx.doi.org/10.1108/
J FEP-11-2014-0073
Chong Lee-Lee, Tan Hui-Boon, (2007),"Macroeconomic factors of exchange rate volatility: Evidence
from four neighbouring ASEAN economies", Studies in Economics and Finance, Vol. 24 Iss 4 pp.
266-285http://dx.doi.org/10.1108/10867370710831828
Sanfeng Zhang, Maoliang Bu, Huafan Yang, (2014),"Environmental Regulation and Firm Productivity
in China", Frontiers of Economics and Globalization, Vol. 14 pp. 129-152http://dx.doi.org/10.1108/
S1574-871520140000014006
Access to this document was granted through an Emerald subscription provided by emerald-
srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
*Related content and download information correct at time of
download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Macroeconomic fundamentals
and dynamics of the Indian
rupee-dollar exchange rate
Chandan Sharma
Indian Institute of Management Lucknow, Noida, India, and
Rajat Setia
National Institute of Financial Management, Faridabad, India
Abstract
Purpose – This paper aims to examine the relationship between Indian rupee-US dollar exchange rate
and the macroeconomic fundamentals for the post-economic reform period.
Design/methodology/approach – The authors have used an empirical model which includes a
range of important macroeconomic variables based on the basic monetary theories of exchange rate
determination. At the frst stage of the analysis, they have tested structural break in the data.
Subsequently, they have employed the fully modifed ordinary least square, Wald’s coeffcient
restriction and impulse response functions (IRF) to estimate the monetary model in the long- and
short-run horizons.
Findings – Results of analyses indicate that the macroeconomic fundamentals determine exchange
rate in a signifcant way, but their effect varies sizably across the periods. The IRF illustrate the
importance of interest rate in controlling exchange rate volatility.
Practical implications – The analysis of the behavior of inter-relationship among macroeconomic
variables will help policymakers in a deep-rooted understanding of this complex and time-varying
relationship.
Originality/value – Most of the existing studies have tested the impact of a single or a few
macroeconomic fundamentals on exchange rate. But in the present study, we have tested the impact of
a range of important variables, i.e. money supply, real income or output, price level and trade balance.
Further, considering the importance of structural breaks in data, they authors have employed standard
tests of structural break and incorporated the issue in the cointegration analysis.
Keywords Time series models, Macroeconomic aspects of international trade and fnance,
Macroeconomics and monetary economics
Paper type Research paper
1. Introduction
Analyzing the movement of exchange rates has always been a challenging and risky
task. The complexity of the interaction between exchange rate and macroeconomic
fundamentals has gained a great deal of attention of researchers and policymakers in the
past decades. Despite the fact that many exchange rate models and their modifcations
are developed over time, still, there is no single theory which describes the behavior of
exchange rate in its entirety. The exchange rate models empirically tested over the past
The authors thank two anonymous referees for their useful comments and helpful suggestions on
the previous versions of this paper. Any errors or omission are solely of the authors.
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1757-6385.htm
Macroeconomic
fundamentals
and dynamics
301
Received13 November 2014
Revised29 April 2015
Accepted10 June 2015
Journal of Financial Economic
Policy
Vol. 7 No. 4, 2015
pp. 301-326
©Emerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-11-2014-0069
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
three decades have shown inconsistency when presented with different data sets,
estimation methodologies and sample periods (see Meese and Rogoff, 1983, Macdonald
and Taylor, 1994).These models have been tested mainly for developed countries, while
developing and transition economies like India have unfairly received lesser attention.
Numerous theoretical and empirical models explain the impact of the macroeconomic
fundamentals on the currency exchange rate utilizing a number of factors specifc to an
economy. The fexible price model developed by Frenkel (1976) and the sticky price
model of Dornbusch (1976) had initially become dominant models of exchange rate
determination. The most signifcant blow to the monetary approach of exchange rate
determination was the fndings of the seminal work of Meese and Rogoff (1983), which
concluded that the analyzing and forecasting ability of these models are quite limited.
This spurred a large number of studies empirically testing these models and their
extensions in the past three decades, namely, MacDonald and Taylor (1994), McNown
and Wallace (1989), Mark (1995), Mark and Sul (2001), Rapach and Wohar (2002),
Cheung et al. (2005), Morley (2007), Zhang et al. (2007), Sarno and Valente (2009) and
Baharumshah et al. (2009). These studies invariably have shown that monetary models
can explain movements in exchange rate to a varied extent. Furthermore, after the
popularity of cointegration relationships since early 1990, empirical evidence which
have employed the cointegration techniques seems to indicate that the monetary models
can explain large portions of the long-run movements in the exchange rate (e.g. McNown
and Wallace, 1989; MacDonald and Taylor, 1994; Kim and Mo 1995).
Findings of MacDonald and Taylor (1994) clearly indicated that unrestricted
monetary models are valid for a long-term analysis, if analyzed in a cointegrating
framework with short-term data dynamics. In a recent study, based on the empirical
evidence, Bekiros (2014) argued that fundamentals may be important determinants of
exchange rates; however, there may be some other unobservable variables driving the
currency rates that current asset-pricing models have not yet captured. These studies
have by and large ignored the issue of structural shifts in the exchange rate and
macroeconomic variables, as the long-run cointegrated series is subjected to serious
variations in the sample period. Some studies have also highlighted that parameter
estimates are often unstable and the relationship is a subject of periodic shifts in the long
run (e.g. Florentis et al., 1994, Goldberg and Frydman, 1996). Accounting for the issue of
structural break, Beckmann et al. (2010) found that fundamentals are important for the
dollar-euro exchange rate, but their impact differs in various regimes, i.e. sign and size
of the regression coeffcients change over time. Similarly, Chang and Su (2014) explored
the linkages between exchange rates and macroeconomic fundamentals to determine
the long-run relationship for Pacifc Rim countries. In their analysis, the use of the
conventional cointegration tests failed to fnd the long-run equilibrium for any
country-pairs except Taiwan, but cointegration tests with structural breaks
demonstrated the long-run connections between exchange rates and fundamentals for
some country-pairs.
In the early 1990s, India suffered with a severe balance of payment crisis due to the
signifcant rise in oil prices, the suspension of remittances from the Gulf region and
several other exogenous developments. The country had been following an adjustable
nominal peg to a basket of currencies of major trading partners with a band. The several
measures were initiated to arrest the crisis which includes a devaluation of the Indian
rupee in July 1991. This measure was the move toward greater exchange rate fexibility.
JFEP
7,4
302
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Thereafter, India has adopted a managed fexible exchange rate regime in 1993, making
the start of an era of a market-determined exchange rate regime of the rupee with
provision for timely intervention by the central bank. The current exchange rate policy
relies on the underlying demand and supply factors to determine the exchange rate with
continuous monitoring and management by the central bank. India has been operating
with a managed fexible regime, where the management’s objective is not to achieve any
explicit or implicit target for the exchange rate but to contain volatility by ensuring
orderly market conditions. The Indian regime could be interpreted as “more fexible”
during normal market conditions with the accent shifting to “management” when the
market turns disorderly (Pattanaik and Sahoo, 2003). However, recently, the Indian
currency has been quite volatile[1]. More importantly, it seems that the government is in
mood to a frequent intervention in the exchange rate market[2].
In the Indian context, a few attempts have been made to test the monetary model of
exchange rate determination. For example, to analyze the issue, Bhanumurthy (2006)
provided the evidence based on a survey of the dealers and found that majority of the
dealers feel that short-term changes in the Indian rupee/US dollar market are basically
infuenced by the micro-variables such as information fow, market movement,
speculation, central bank intervention, etc. Dua and Ranjan (2011) have attempted to
analyze the issue and concluded that the monetary model works well in the Indian case.
More recently, Mallick (2010) have analyzed the issue and concluded that dominance of
foreign institutional investments affects the rupee-dollar exchange rate, and to a certain
extent, it was seen that the infuence of the growth rate differential also affects the
exchange rate behavior. The analysis of Tiwari et al. (2013) has indicated that oil price
and Indian rupee have a causal relationship, but the causality exists only at higher time
scales. These studies have, however, widely ignored the issues of the presence of a
structural shift in the system. Moreover, in a recent study, Hegwood and Nath (2014)
have shown that the movement of the Indian rupee suffers fromstructural breaks, which
may have seriously altered the fndings on India. Findings of Narayan (2006) using
Lagrange multiplier (LM) unit root test with structural breaks also provided evidence
that India’s exchange rate vis-a`-vis 15 out of the 16 countries is stationary, implying
support for purchasing power parity (PPP).
Against this background, we make an attempt to reassess the issue and to bridge the
existing gaps in the related literature. We mainly focus on two questions. First, what are
the determinants of the exchange rate, especially from macroeconomic perspective?
Second, is there any theory that explains the pattern of exchange rate movements
near-accurately? In this study, we attempt to provide several novelties to the related
literature. First, initially, it was understood that the development of a series of
cointegration tests will help in examining the monetary approach of exchange rate
determination. It was expected that the debate in the empirical literature will settle when
these tests will be able to explain long-term fuctuations in the exchange rate market.
However, the signs of the estimated parameters from the variety of monetary models
using the cointegration framework again conficted with widely accepted theoretical
assumptions. Recent development in time series econometrics indicates that a possible
reason for these conficting results is the presence of structural instability in the data. In
fact, the structural breaks can induce stochastic behavior similar to an integrated
process, which makes it diffcult to differentiate between the lack of cointegration and a
structural shift (Dropsy, 1996). The presence of structural changes may lead to a serious
303
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
bias in the estimated coeffcients. Moreover, it is now well-established that accurate
forecasting and empirical analysis of time series data can depend critically on
understanding the appropriate nature of structural change (e.g. Lee et al., 2006). Thus, in
the present study, we attempt to address this issue by incorporating the issue of
structural break in the analysis. This is important, as in the wake of changes in
exchange rate policy regimes, shocks to oil prices and shifting patterns of trade, there is
a need to account for structural breaks in the empirical testing of exchange rate
equations. Second, the Indian case is both interesting and relevant in the light of the fact
that polices on exchange rate management has shifted to a foating exchange rate
regime in 1993. The timing of this policy shift also coincides with embankment of
economic liberalization. The policy reforms have led to a spectacular surge in
international capital infows in India during 1990s. However, in the recent years, the
Indian rupee has witnessed excessive volatility, which is seen as a serious threat to
stability of the Indian fnancial and external sector, which, in turn, affects the country’s
overall growth trajectories. Third, this study along with the long-run linkage also
focuses on the short-run dynamics. Specifcally, the response of exchange rate to an
impulse or shock on other macroeconomic variables is generated in a vector error
correction (VEC) framework. Finally, most of the existing studies have focused on a
single theoretical model for the empirical testing; we examine several related theories
and utilize a range of important variables, i.e. money supply, real income or output, price
level and trade balance for our analysis. These models are empirically tested using
several alternative techniques to test the sensitivity and robustness of results.
Rest of this paper is divided into fve sections. Section 2 makes up the theoretical
background and set hypotheses to test. Data and methodological issues form Section 3.
Section 4 covers results and discussion. Section 5 contains the conclusion and policy
suggestions.
2. Theoretical models: monetary models of the exchange rate
The traditional model of exchange rate determination, known as the PPP theory is an
application of the “Lawof One Price”. It states that an exchange rate equalizes the price
of goods between countries. PPP focuses on general price levels and their relationship to
the nominal exchange rate as:
P ? EP* (1)
Where P and P* are the domestic and foreign price levels and E denotes the exchange
rate. There are noted deviations from PPP due to a number of theoretically implied
factors, namely, heterogeneity in the baskets of goods considered for construction of
price indices in various countries, presence of transportation cost, imperfect competition
in goods market and increase in the volume of global capital fows during the past few
decades. Although there is a mixed evidence provided by a battery of empirical tests, the
strength of the relationship between price ratios and exchange rates is not as strong as
depicted by the PPP theory (Nag and Mitra, 1998; Mishra and Sharma, 2010).
In all the monetary models, the money supply in relation to money demand
determines the exchange rate. The prominent monetary models include the fexible and
the sticky price monetary models of exchange rate as well as the real interest rate
differential (RID) model. The fexible price monetary model (Frenkel, 1976) is based on
the assumption that the prices are fully fexible, i.e. they can either increase or decrease
JFEP
7,4
304
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
from the prevalent level (also see Bilson, 1978; Hodrick, 1978 and Macdonald and
Taylor, 1994). Sticky price model of Dornbusch (1976) is based on the premise that prices
are determined in sticky price markets, i.e. they are fxed over a short period and they
tend to change only slowly over time in response to shocks such as changes in money
supply. However, exchange rates are determined in fexible price markets and respond
immediately to new developments and shocks. So exchange rates are not matched by
price movements, resulting in departures from PPP. Hooper and Morton (1982)
introduced an extension of the sticky price model by accommodating the effect of trade
balance in the exchange rate determination equation. Here, exchange rate is assumed to
be correlated with unanticipated shocks to the trade balance.
Real interest rate differential model (Frankel, 1979) combines the infationary
expectations of the fexible price monetary model with the sticky price model. It states
that the gap between the current real exchange rate and its long-run equilibrium value
is proportional to the real interest rate differential.
The basic theoretical models of exchange rate determination, such as the fexible
price model, sticky price model, real interest rate differential model and the Hooper and
Morton model, are collectively represented by the following equation:
e ? ? ? ?(m ? m*) ? ?(y ? y*) ? ?(i ? i*) ? ?(p ? p*) ? ?(tb ? tb*) (2)
Where e, m, y, i, p and tb denote nominal exchange rate, money supply, real output,
nominal interest rate, infation or price level and trade balance, respectively, and *
variables indicate the foreign counterparts.
To test the individual models, we construct hypotheses, which are as follows:
2.1 Flexible price model (? ?0, ? ?0, ? ?0, ? ?? ?0)
An increase in the domestic money supply relative to foreign money supply (? ?0) leads
to an increase in the infation rate, which will decrease the currency demand, and
currency will depreciate or e will increase. An increase in the domestic interest rate
relative to foreign interest rate (? ?0) is a refection of an expected depreciation of the
domestic currency due to an expected decrease in demand of the currency resulting from
an expected rise in infation. Thus, a rise in domestic interest rate leads to a fall in the
demand for money and hence depreciation of the domestic currency or an increment in
e. A relative increase in the output or real income means more number of money
transactions creating an excess demand for the currency, and if money supply and
interest rates are constant, then domestic prices will fall, resulting in an appreciation of
currency value to maintain PPP. So, real income differential and exchange rate move in
tandem i.e. ? ?0 will lead to a decrease in e (see Frankel, 1984 and Williamson, 2009).
2.2 Sticky price model (? ?0, ? ?0, ? ?0, ? ?? ?0)
The model argues that prices are rigid and would only adjust gradually. In response to
an increase in the money supply (? ?0 ), considering the prices are fxed in the short run,
interest rate will fall (? ? 0 ), which has to be compensated by depreciation of the
domestic currency or an increase in the e value. For this reason, the exchange rate
overshoots its long-run equilibrium value. A relative decrease in the output or real
income (? ? 0 ) will result in a decrease in the value of e (see Dornbusch 1976; Rogoff,
2009).
305
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
2.3 Real interest differential model (? ? 0, ? ? 0, ? ? 0, ? ? 0, ? ? 0)
Frankel (1979) developed an integrative approach that incorporates the sticky price
model with the notion of secular rates of infation. If the real rate of interest on domestic
bonds is greater than the real rate of interest on foreign bonds, there will be a
depreciation of the domestic currency until the real interest rates are equalized in the
long run steady state (also see Hoffmann and McDonald, 2009).
2.4 Hooper and Morton model (? ? 0, ? ? 0, ? ? 0, ? ? 0, ? ? 0)
Changing the traditional monetary model, Hooper and Morton (1982) incorporated the
real side of the economy by taking into consideration innovations in the current account.
The model utilizes trade balances as to gauge the effects of premium due to fscal
balance, international reserves holding and foreign indebtedness. Therefore, the model
equation simply adds on the effect of relative trade balance differential. If this
differential is negative (? ? 0 ), domestic currency will appreciate, and hence, e value
decreases (also see Beckmann et al., 2010).
Incorporating the issues of these models, for empirical testing, we employ a hybrid
empirical model, containing all important macroeconomic variables and their
hypothesized direction of impact as per the basic theoretical models (both monetary as
well portfolio models) of exchange rate determination mentioned above. Our benchmark
empirical model is as follows:
Y
t
? ?
t
? ?
t
X
t
? ?
t
(3)
Where variable Y
t
contains the Indian rupee-US dollar exchange rate (e) and
X
t
is a [K ? 1] vector of domestic and foreign explanatory fundamentals, i.e.
X
t
?(y, y*, p, p*, m, m*, i, i*, tb, tb*) and ?
t
is a [1 ?K] vector of respective coeffcients.
It is worth mentioning that we choose nominal exchange rate for our analysis over
real exchange rate. Theoretical and empirical studies on this issue mainly considered
nominal exchange rate, whereas studies of Kollman (1997), Kia (2013) and others have
instead chosen real exchange rate for analysis. The nominal exchange rate is preferred
because the real exchange rate converges to PPP in the long run, leading to fundamental
variables do not have any effects on the real exchange rate. Moreover, fully fexible and
the real exchange rates adjust to real disturbances and fscal policies. On the other side,
by design monetary policy indicators cannot drive the real exchange rate, as in this
equilibrium condition, price and nominal exchange rate re-adjust in the same ratio to
any monetary shock (for detail discussion, see Devereux, 1997).
3. Data description and empirical methodology
3.1 Data description
For empirical testing, this paper employs monthly data on rupee-dollar nominal
exchange rate (e) and the independent macroeconomic variables like money supply (M3
for India and M2 for USA), real income or output proxied by IIP (index of industrial
production), general price level proxied by CPI (consumer price index), short-term
interest rate (91 day treasury bill rates) and the overall trade balance (exports-imports)
for India and the USA. Data from April 1994 to March 2010 are used due to the
availability of the data for this period on all the variables required. The data on all the
variables for India are collected from IMF’s International Financial Statistics database
and Handbook of Statistics on the Indian Economy provided by Reserve Bank of India.
JFEP
7,4
306
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Data on trade balance have been converted into an index to avoid complexity arising out
of the negative magnitude of observations. The data series are taken at their level form
and seasonally adjusted (except exchange rate). The series are transformed in their
natural logarithms (except interest rates, which are in their linear form) before any
econometric treatment. Figure 1 shows the variation of exchange rate for the period of
the study. Detailed descriptions of data series and their descriptive statistics are
presented in Tables A1 and A2 respectively.
3.2 Empirical methodology
3.2.1 Unit root test with endogenous two-break. One major drawback of conventional
unit root tests is that it implicitly assumes that the model correctly specifes the
deterministic trend. Following the work of Perron (1989), one can consider that the
presence of structural change substantially reduces the power of unit root tests. Zivot
and Andrews (1992) proposed a unit root test that allows for an endogenous structural
break. Recently, Lumsdaine and Papell (1997) proposed a unit root test that allows for
two shifts in the deterministic trend at two known dates. A problem with the
Lumsdaine–Papell unit root test is that its critical values assume no breaks under the
null hypothesis. This assumption is problematic, as it may lead to conclude incorrectly
that rejection of the null is evidence of trend stationarity, when, in fact, the series is
difference stationary with breaks (Lee and Strazicich 2003, 2004). With improvement,
Lee and Strazicich (2003) have proposed the endogenous two-break LM unit root tests.
The test incorporates structural breaks under the null hypothesis, and rejection of the
minimumLMtest provides evidence for stationarity of the series. Moreover, the Lee and
Strazicich (2003) test also has higher power than the test of Lumsdaine and Papell (1997).
The model has two variants: frst, the crash model (Model A) and the break model
(Model C). Both models are based on alternative assumptions about structural breaks.
Model A allows for two shifts in the intercept, and Model C includes two shifts in the
intercept and trend. We have used the crash model for the analysis, as in our case, break
the crash model seems to be appropriate.
The crash model of Lee and Strazicich (2003) is specifed as follows:
Z
t
? [1, t, D
1t
, D
2t
] (4)
Where D
jt
?1 for t ? T
Bj
?1, j ?1, 2, and 0 otherwise. The break date is denoted byT
Bj
The null and alternative hypotheses of Model A are as follows:
0
10
20
30
40
50
60
A
p
r
-
9
4
J
a
n
-
9
5
O
c
t
-
9
5
J
u
l
-
9
6
A
p
r
-
9
7
J
a
n
-
9
8
O
c
t
-
9
8
J
u
l
-
9
9
A
p
r
-
0
0
J
a
n
-
0
1
O
c
t
-
0
1
J
u
l
-
0
2
A
p
r
-
0
3
J
a
n
-
0
4
O
c
t
-
0
4
J
u
l
-
0
5
A
p
r
-
0
6
J
a
n
-
0
7
O
c
t
-
0
7
J
u
l
-
0
8
A
p
r
-
0
9
J
a
n
-
1
0
Figure 1.
Exchange rate Indian
rupee/US dollar
307
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
H
0
:Y
t
? ?
0
? d
1
B
1t
? d
2
B
2t
? Y
t?1
? v
1t
H
1
:Y
t
? ?
0
? ?t ? d
1
B
1t
? d
2
B
2t
? v
2t
Where DT
jt
? t ? T
Bj
for t ? T
Bj
? 1, j ? 1, 2, and 0 otherwise; B
jt
? 1 for t ? T
Bj
?
1, j ?1, 2, and 0 otherwise; ?
1t
and ?
2t
denote the stationary error terms. The LMunit root
test statistic can be obtained by estimating the following:
?Y
t
? ?
=
?Z
t
? ?S
?
t?1
? ?
t
(5)
Where S
?
t
? y
t
- ?
x
- Z
t
? , t ? 2,. . . .,T; ?y
t
is regressed on ?Z
t
to obtain estimates of ?;
?
x
?y
1
?Z
1
? and the frst observations of y
t
and Z
t
are y
1
and Z
1
, respectively. The LM
test statistics are provided by ? which is the test statistic for the unit root null hypothesis
that ? ?0.
We consider a maximum lag length of fve periods and obtained the optimal lag
length on the basis of the signifcance of the last lag. The break dates are determined
where the LM test statistic is at its minimum. The critical values of this test are
tabulated in Lee and Strazicich (2003, 2004). Thus, this method is more demanding than
previously developed unit root tests with structural break because it offers more than
one break in the series.
3.2.2 Detecting structural changes: Bai-Perron test. An alternative approach to test
the structural change is applying F-test or Chowtest on the model. Bai and Perron (1998,
2003) have extended the Chow test. In dealing with a cointegrated model, the Gregory
and Hansen test (1996) is used for one unknown structural break and the Bai and Perron
(2003) (BP hereafter) test for dating multiple unknown structural breaks. Therefore, we
prefer BP test over Gregory and Hansen test, as it is highly likely that the system has
more than one break in the analysis period. The BP methodology employs a multiple
structural break model, with m breaks, i.e. (m ?1) regimes. Formally:
y
t
? x
=
t
? ? z
=
t
?
j
? u
t
, (t ? T
j?1
? 1, .....T
j
) (6)
for j ?1 ….,m?1, with T
0
?0 and T
m?1
?T, where y
t
is the dependent variable at time
t, x’
t
and z’
t
are the regressors, ? and ? are the coeffcients to be estimated and u
t
is the
error term. In this equation, only ? varies with time and ? is constant, so this is a partial
structural change model. Considering ? ? 0 will give us what is known as the pure
structural change model where all coeffcients are subject to change with time. The
linear combination of those segments of the total sample period, for which the sum of
squared residuals comes out to be minimumas per Bayesian information criteria, is the
estimated breakpoints. The test performs an algorithm which compares all possible
combinations of these segments.
Initially, the BP test was developed for stationary I(0) variables, but it equally holds
for non-stationary case as well (Kejriwal and Perron, 2008). Some recent studies (e.g.
Beckmann et al., 2010) have employed BP test for structural break analysis to
investigate the temporal stability of relationship between exchange rate and
macroeconomic fundamentals, which have yielded satisfactory results.
JFEP
7,4
308
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3.3 Cointegration and estimation of model parameters
As BP’s methodology is designed for single equation systems and the fact that the
macroeconomic variables which are used in the model exhibit the problem of
endogeneity, fully modifed ordinary least squares (FMOLS) method developed by
Phillips and Hansen (1990) is used for the estimation. FMOLS corrects the ordinary least
squares (OLS) method with regard to endogeneity and serial correlation (Phillips, 1995).
Here, the primary interest is to fnd out the long-run relationship between exchange rate
and fundamentals, without paying attention to the relationships among regressors.
To examine whether the relationship obtained from the FMOLS estimation can be
interpreted as a cointegrated relationship, we have applied the unit root tests to the
residuals. Another approach employed to test for cointegration is the Johansen and
Juselius (JJ) (1990) technique. The JJ maximum likelihood approach sets up the
non-stationary time series as the vector autoregressive process of order k in a
re-parameterized form:
?Y
t
? ? ? ?
1
?Y
t?1
? ?
2
?Y
t?2
? . . . .?
k?1
?Y
t?k?1
? ?
k
Y
t?k
? U
t
(7)
Where Y
t
is a vector of frst-order cointegrated variables, ?is a vector of constant terms
to capture the time series trend characteristics, ?is a coeffcient matrix and U
t
is a vector
of normally and independently distributed error terms.
Johansen et al. (2000) developed cointegration test, which uses the JJ framework to
allow for trend and level breaks at several known points. To use traditional
cointegration analysis, the method disregards observations after structural breaks by
including impulse dummies. The number of impulse dummies after recognizing the
breaks in the system and the inclusion of these dummies imply a reduction in the
effective sample. The technique uses two variants of the trace test for testing of
cointegration relationship among p time series. These are the Hl(r) and Hc(r) tests for
when there are (q – 1) breaks in a linear trend or in a constant level of the data,
respectively, where r denotes the cointegrating rank. The asymptotic distributions of
the test statistics depend on the values of (p – r) and the locations of the breakpoints in
the sample. These breakpoints are denoted by v
j
?(t
j
/t) , where T is the full sample size
and t
j
is the last observation of the jth sub-sample; j ?1, 2 ….., q (for details, see Giles and
Godwin, 2012).
Finally, the short-run effects of macroeconomic variables are captured using impulse
response functions (IRF) generated in a VEC framework. VEC framework adds a
restriction of cointegration error into the general vector autoregeressive framework.
Using Cholesky one standard deviation shock structure, IRF are generated.
4. Results and discussion
4.1 Results of conventional unit root tests
In this section, we explain results of our empirical analysis. The frst step in the analysis
of time series data is to examine the variables for stationarity. A stationary time series
is the one which exhibits mean reversion, has a fnite, time invariant variance and the
covariance between two values depends only on their distance apart in time and not on
the exact timing of the observation. At the frst stage, we employ three alternative tests
for testing stationarity in the variables. The augmented Dickey-Fuller (ADF) test,
Dickey-Fuller generalized least square (DF-GLS) test and Phillips-Perron (PP) test share
309
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the same null hypothesis of a unit root. If two of these three tests indicate
non-stationarity for any series, we conclude that the series has a unit root. Table I
reports the results of three tests with constant as well as with constant and trend. Our
variables have been tested for stationarity in the levels as well as at frst differences.
Table I shows that almost all the variables are found to be integrated of order one, i.e. I
(1). However, the standard unit root tests suffer from biasedness problem toward the
non-rejection of the null hypothesis, and macroeconomic series are not characterized by
a unit root but rather that persistence arises only from large and infrequent shocks
(Perron, 1989). Therefore, we need to test structural breaks in our data.
4.2 Results of unit root tests with structural breaks
To test endogenous structural break and order of integration of variables, we employ
two-break minimumLMunit root tests. Table II reports the results for the test based on
Model Aor crash model which represents two breaks in the intercept. The test statistics
of the LM unit root tests indicates that except tb, all other variables do not exceed the
critical values in absolute terms, and therefore, the unit root null cannot be rejected at the
5 per cent level. However, the t-statistics corresponding to the break dates are
statistically signifcant only for few variables. Furthermore, break dates are not
consistent across models, e.g. break model and crash model[3].
4.3 Results of the BP test
Considering the in consistency in results of endogenous structural break LM unit root
tests, we can obtain breakpoints in the system by applying BP method. If the empirical
application of the BP test proves the existence of structural breaks, we can conclude that
a stable long-run relationship among the variables does not exist. The results of the BP
test is presented in Table III. The frst breakpoint of 1996 appears to be the effect of
initiation of what we call as the Asian Financial Crisis that gripped much of Asia in July
1997. The crisis started in Thailand with the fnancial collapse of the Thai Baht caused
by the decision of the Thai Government to foat the Baht, cutting its peg to the US dollar.
As the crisis spread, most of the Southeast Asian countries and Japan witnessed
slumping currencies. The impact of the crisis has also spread further afeld, as
beginning of the Asian crisis had hurt Indian economy signifcantly. On the one hand,
the Indian exports to the rest of Asia had decelerated sharply (UNCTAD, 1998). On the
other hand, since 1993, there were fuctuations in exchange rate, which become severe
during the crisis period. Indian rupee depreciated against US dollar by 6.31 per cent
between July 1997 and March 1998 and by approximately 11 per cent from July 1997 to
December 1998 (Dua and Sinha, 2007), clearly justifying the robustness of the result.
The period from2001 to 2003 is also considered as the period of slowglobal recovery
from recession coupled with terrorist attacks on the USA, subsequent war on
Afghanistan and apprehensions about rise in oil prices due to a war-like situation
between USA and Iraq. There was a massive stock market crash during 2002 in stock
exchanges across the USA, Canada, Asia and Europe. After recovering from lows
following September 11 attacks in 2001, indices again started dipping down in March
2002. These situations could have possibly produced the break obtained in March 2001.
As the results of BP method make sense and are consistent, we consider these results for
our further analysis.
JFEP
7,4
310
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table I.
Results of unit root
tests
V
a
r
i
a
b
l
e
A
t
l
e
v
e
l
A
t
f
r
s
t
D
i
f
f
e
r
e
n
c
e
A
D
F
D
F
-
G
L
S
P
P
A
D
F
D
F
-
G
L
S
P
P
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
I
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
i
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
i
n
t
e
r
c
e
p
t
T
r
e
n
d
a
n
d
i
n
t
e
r
c
e
p
t
e
?
2
.
1
6
3
?
1
.
6
5
1
1
.
5
2
5
?
1
.
0
9
4
?
2
.
1
0
3
?
1
.
6
5
9
?
1
0
.
7
0
5
*
*
?
1
0
.
8
2
9
*
*
?
1
0
.
5
8
4
*
*
?
1
0
.
6
0
1
*
*
?
1
0
.
8
0
*
*
?
1
0
.
8
8
3
*
*
m
0
.
6
1
6
?
1
.
6
4
1
.
4
3
1
?
1
.
9
5
5
0
.
9
8
5
?
1
.
9
9
?
1
6
.
9
8
4
*
*
?
1
6
.
9
7
*
*
?
1
5
.
0
2
2
*
*
?
1
5
.
1
9
4
*
*
?
1
7
.
4
3
1
*
*
?
1
7
.
7
5
4
*
m
*
0
.
3
1
4
?
2
.
7
5
4
1
.
2
2
5
?
1
.
3
1
2
0
.
5
0
9
?
3
.
0
6
9
?
8
.
4
1
3
*
*
?
8
.
4
0
6
*
*
?
3
.
6
7
0
6
*
*
?
5
.
3
9
6
7
*
*
?
1
7
.
1
2
5
*
*
?
1
7
.
1
3
1
*
*
p
?
0
.
1
8
1
?
1
.
2
8
1
4
.
4
1
2
*
*
?
1
.
0
2
9
?
0
.
1
9
9
?
1
.
4
1
9
?
1
2
.
4
5
5
*
*
?
1
2
.
4
2
*
*
?
9
.
1
9
9
4
*
*
?
9
.
5
7
4
*
*
?
1
2
.
4
6
2
*
*
?
1
2
.
4
2
8
*
*
p
*
?
0
.
4
5
6
?
3
.
4
7
6
*
*
3
.
4
7
0
*
*
?
2
.
4
2
8
?
0
.
4
4
7
?
3
.
3
7
2
*
*
?
1
6
.
4
2
2
*
*
?
1
6
.
3
8
3
*
*
?
9
.
5
1
7
*
*
?
9
.
4
8
2
*
*
?
1
6
.
7
5
3
*
*
?
1
6
.
7
1
1
*
*
y
?
0
.
2
7
7
?
1
.
8
6
7
6
.
2
2
3
*
*
?
1
.
3
2
6
?
0
.
3
5
8
?
2
.
6
5
1
?
2
3
.
4
7
8
*
*
?
2
3
.
4
1
6
*
*
?
1
.
7
3
2
*
*
?
2
0
.
7
9
8
*
*
?
2
3
.
9
0
9
*
*
?
2
3
.
8
4
3
*
*
y
*
?
2
.
3
5
7
?
2
.
5
5
1
?
0
.
2
3
1
?
1
.
6
9
1
?
2
.
6
8
2
*
*
?
1
.
1
0
8
?
3
.
4
7
4
*
*
?
3
.
7
4
7
*
*
?
2
.
7
4
6
4
*
*
?
3
.
3
4
6
*
*
?
1
4
.
3
2
4
*
*
?
1
4
.
6
1
9
*
*
i
?
1
.
4
3
3
?
2
.
6
1
8
?
1
.
4
5
7
?
1
.
9
0
1
?
1
.
6
2
2
?
2
.
8
8
6
?
1
3
.
7
8
0
*
*
?
1
3
.
7
7
3
*
*
?
1
2
.
7
5
3
*
*
?
1
3
.
6
3
4
*
*
?
1
3
.
8
1
3
*
*
?
1
3
.
8
0
3
*
*
i
*
?
1
.
4
2
5
?
2
.
2
8
7
?
1
.
4
2
2
?
1
.
9
7
6
?
0
.
9
5
5
?
2
.
0
3
7
?
4
.
7
2
6
*
*
?
4
.
7
2
3
*
*
?
0
.
9
9
3
?
3
.
0
9
1
*
*
?
9
.
1
3
8
*
*
?
9
.
0
4
3
3
*
*
t
b
?
1
.
2
1
8
?
7
.
4
8
4
*
*
?
0
.
4
4
1
?
5
.
3
9
2
*
?
2
.
6
4
1
?
7
.
4
8
6
?
1
0
.
8
8
7
*
*
?
1
0
.
8
5
2
*
*
?
1
5
.
2
3
8
*
*
?
1
6
.
3
3
2
*
*
?
4
1
.
5
6
6
*
*
?
6
2
.
0
6
4
*
*
t
b
*
?
1
.
5
2
3
?
1
.
2
3
1
0
.
2
5
1
?
1
.
4
6
4
?
1
.
5
2
3
?
1
.
3
3
9
?
6
.
7
7
3
*
*
?
6
.
8
3
8
*
*
?
6
.
2
9
4
*
*
?
6
.
5
0
6
*
*
?
1
5
.
2
9
3
*
*
?
1
5
.
3
6
8
*
*
N
o
t
e
:
*
*
D
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
311
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4.4 Results of the model estimation: a long-run analysis
We now have two options for estimating the long-run relationship. First, we can use
dummy variables for the quantifed breaks in the model. Second, divide the sample in
sub-samples according to break in data. We have used both options to test the
robustness of our results. To examine a long-run equilibrium relationship, we estimate
equation (3) for two sub-samples[4] (April 1994-February 2001 and April 2001-March
2010) as well as for full sample using FMOLS estimator and results are reported in
Table III. Column 1 of the table reports results of full sample with and without break
dummies. Results exhibit that for the full sample period, except money supply and
interest rate, other domestic variables signifcantly determine the movement of
exchange rate. Inclusion of dummies does not signifcantly improve the results (see
Column 2). In fact, neither the size of the coeffcients nor their statistically signifcance
differ in both models. Nevertheless, the dummies for both breaks are estimated to be
positive and statistically signifcant, which further validates the results of the BP test.
Interestingly, for both of the sub-sample periods, domestic interest rate and output
are estimated to be signifcant, supporting both these variables as important
determinants of exchange rate. Furthermore, both domestic and US money supply are
found to have a crucial role in determining the exchange rate movement for the period
April 1994 to March 2010. Importantly, results indicate that domestic money supply has
Table II.
Results of Lee and
Strazicich (2003) LM
test
Variable S
t?1
Const. D1 [Break data] D2 [Break data] Lag
e ?0.027 (?2.14) 0.005 (3.11) ?0.0245 (?1.74) [2004:03] ?0.0289 (?2.03) [2005:12] 5
m ?0.055 (?2.29) 0.0125 (20.64) ?0.0204 (?2.57) [2004:08] 0.0693 (8.79) [2006:12] 0
m* ?0.0145 (?1.47) 0.005 (13.37) 0.0033 (0.93) [2001:10] 0.0035 (1.047) [2004:01] 5
p ?0.008 (?1.208) 0.006 (12.48) 0.0075 (1.19) [1999:04] ?0.014 (?2.409) [2005:04] 4
p* ?0.0801 (?3.36) 0.0025 (11.128) ?0.0012 (?0.48) [1997:04] 0.0076 (3.108) [1997:04] 1
y ?0.0301 (?1.58) 0.0058 (5.86) 0.0078 (0.56) [1998:11] ?0.0004 (?0.028) [2005:07] 5
y* ?0.0251 (?3.05) 0.0045 (3.92) ?0.006 (?1.03) [1996:06] 0.0173 (2.88) [1998:07] 5
i ?0.0731 (?2.907) 0.1804 (2.17) ?1.1702 (?2.18) [2000:03] 2.514 (4.31) [2007:06] 3
i* ?0.022 (?2.408) 0.006 (0.336) 0.789 (4.36) [1998:10] 0.112 (0.601) [2007:08] 3
tb ?0.5807 (?8.22) ?0.246 (?4.98) ?1.1757 (?2.42) [1999:02] 0.438 (0.43) [2004:04] 0
tb* ?0.0405 (?2.33) 0.0154 (2.629) 0.145 (2.49) [1996:11] 0.0757 (1.303) [1998:07] 4
Notes: This are results of Lee-Strazicich (2003) unit root test crash model with two breaks; the
numbers in parenthesis are the t-statistics for the estimated coeffcients; the coeffcient on S
t?1
tests for
the unit root; D1 and D2 equal the breaks of the slope
Table III.
Breakpoints obtained
using BP test
Year Month
1996 3
2001 3
Number of breaks: 2
Notes: These breakpoints are obtained by using Bai and Perron (1998, 2003) methodology to the
regression Y
t
??(t) ??(t)X
t
??
t
, where Y
t
denotes the Indian Rupee-US dollar exchange rate and X
t
is a 5 ? 1 vector of fve macroeconomic fundamentals; the sample period covers monthly data from
April 1994 to March 2010
JFEP
7,4
312
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
a positive effect on the exchange rate, whereas the US money supply has a negative
effect for this period. It is also noteworthy that coeffcient of US money supply is quite
sizable and statistically signifcant for all the periods. These fndings exhibit the impact
of money supply and its importance as a component of monetary management in
exchange rate determination. Results also indicate that domestic price level has a
positive coeffcient for the whole sample period as well as for the frst sub-sample period
estimation, whereas it is not found to be signifcant in case of the last period. On the other
hand, foreign price level is negatively impacting the exchange rate in the later
sub-period. Results regarding domestic output have a negative effect on exchange rate
in case of full sample and the last sub-period, whereas no signifcant effects are observed
in case of the frst sub-period. Our fndings regarding domestic interest rates indicate
that it has a positive, however, not very sizable effect on the exchange rate movement in
both sub-periods. For a robustness check, we also apply dynamic OLS (DOLS)
estimator[5], and results are presented in Table A3. These results corroborate the
fndings of FMOLS with some size variations of coeffcients (Table IV).
4.5 Results of the cointegration tests
The relationship obtained from the FMOLS estimation can be further validated for its
long-run stability by applying the idea of residual-based cointegration tests. For this
purpose, we employ unit root tests to the resulting error series, which can be interpreted
as a test of cointegration relationship. Table Vshows the ADF, DF-GLS and PPunit root
tests for the regression error terms obtained in all the study periods. Evidently, all these
tests indicate the presence of a long-run relationship in all the sample study periods. The
result is in line with the fndings of Kletzer and Kohli (2000) and Dua and Ranjan (2011).
For a robustness check, the JJ cointegration test is also conducted, and the results are
presented in Table VI. Results of trace test and the maximumeigenvalue test indicate a
cointegrating relationship, implying a long-run relationship among the variables for the
full study period as well as the two sub-periods. It is noteworthy that the number of
cointegrating equations is higher in both the sub-periods as compared to the entire study
period. Gregory and Hansen (1996) extended the Engle–Granger test and allowed for a
one-time endogenously break in the cointegrating vector. For a robustness test, we also
conduct Gregory and Hansen (1996) test of cointegration test which indicates
cointegration among our variables and indicates break in the system in July 1997 (see
Table A3). Although this test is comparatively powerful than the Engle–Granger test in
the case of presence of a single break, in our case, it is not very suitable, as it signifcantly
lacks power in the presence of multiple breaks and in the presence of multiple
cointegrating vectors. Therefore, fnally, we use Johansen et al. (2000) cointegration test,
which uses the JJ framework. Specifcally, we use its trend breaks (Hl(r)) variant for
cointegration test. The model accommodates structural breaks by including impulse
dummies for the quantifed breaks. The results are reported in Table VII, which
validates our earlier results that variables in our models have a long-run inter-linkage;
thus, they are cointegrated.
4.6 Results of the Wald restriction tests
After establishing a co-movement between exchange rate and macro-fundamentals
in the long run, we attempt to further validate the estimated linkage by applying the
Wald restriction tests to the empirical models which are estimated by FMOLS. We
313
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table IV.
FMOLS estimations
V
a
r
i
a
b
l
e
s
F
u
l
l
s
a
m
p
l
e
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
1
9
9
4
t
o
F
e
b
r
u
a
r
y
2
0
0
1
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
2
0
0
1
t
o
M
a
r
c
h
2
0
1
0
W
i
t
h
o
u
t
d
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
W
i
t
h
b
r
e
a
k
d
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
c
o
n
s
t
a
n
t
?
2
.
5
6
2
(
?
0
.
8
1
1
)
?
0
.
6
8
5
(
?
0
.
2
0
1
)
1
.
5
3
1
(
0
.
5
5
3
)
9
.
0
7
2
1
5
*
*
(
2
.
8
4
1
)
m
?
0
.
0
4
8
(
?
0
.
1
7
6
)
0
.
0
1
5
(
0
.
0
5
6
2
)
0
.
7
3
2
*
*
(
3
.
0
5
9
)
0
.
0
7
4
0
5
(
0
.
3
1
4
)
m
*
1
.
1
6
3
*
*
(
2
.
6
4
4
)
1
.
2
0
1
*
*
(
3
.
7
0
9
)
?
1
.
5
6
8
*
(
2
.
9
2
2
)
0
.
8
9
1
8
5
*
*
(
2
.
1
1
4
)
p
1
.
2
5
2
*
*
(
5
.
8
3
7
)
1
.
1
4
0
2
*
*
(
5
.
3
6
7
)
0
.
7
9
7
*
*
(
5
.
1
7
9
)
0
.
4
9
3
1
(
1
.
6
0
7
)
p
*
?
0
.
4
2
9
(
?
0
.
6
5
9
)
?
1
.
0
6
1
(
?
1
.
5
3
6
)
0
.
2
3
8
(
0
.
3
8
2
)
?
1
.
4
2
0
1
*
(
?
1
.
9
1
3
)
y
?
1
.
4
4
1
*
*
(
?
7
.
0
7
4
)
?
1
.
3
3
3
*
*
(
?
7
.
1
8
5
)
0
.
5
7
7
9
*
*
(
3
.
2
1
1
)
?
0
.
6
7
9
3
*
*
(
?
2
.
6
3
4
)
y
*
?
0
.
2
4
5
(
?
1
.
3
4
9
)
?
0
.
1
8
4
(
?
1
.
0
6
3
)
?
0
.
7
8
9
*
*
(
?
2
.
9
3
2
)
?
1
.
5
8
3
9
*
*
(
?
5
.
6
6
8
)
i
?
0
.
0
0
3
(
?
0
.
9
7
1
)
0
.
0
0
2
(
0
.
4
9
1
)
0
.
0
0
9
*
*
(
3
.
6
1
6
)
0
.
0
1
6
0
6
*
*
(
2
.
8
8
7
)
i
*
0
.
0
2
5
4
*
*
(
4
.
2
3
6
)
0
.
0
2
5
*
*
(
4
.
6
3
6
)
?
0
.
0
0
1
(
?
0
.
1
1
3
)
0
.
0
2
3
3
2
*
*
(
4
.
8
6
1
)
t
b
?
0
.
0
3
6
*
(
?
4
.
1
8
6
)
?
0
.
0
2
8
*
*
(
?
3
.
7
1
)
0
.
0
0
7
(
1
.
4
0
1
)
?
0
.
0
2
1
*
*
(
2
.
4
1
1
)
t
b
*
0
.
0
4
7
3
(
0
.
9
7
8
)
0
.
0
2
3
(
0
.
5
3
1
)
?
0
.
0
4
3
(
?
1
.
1
6
9
)
0
.
1
6
8
0
8
*
*
(
3
.
3
7
1
)
D
u
m
m
y
M
a
r
c
h
1
9
9
6
0
.
0
5
8
*
*
(
2
.
0
6
)
D
u
m
m
y
M
a
r
c
h
2
0
0
1
0
.
0
3
5
(
1
.
3
6
)
A
d
j
u
s
t
e
d
R
-
s
q
u
a
r
e
d
0
.
8
8
8
1
0
.
8
9
2
3
0
.
9
6
4
9
0
.
8
3
5
2
N
o
t
e
s
:
*
*
A
n
d
*
d
e
n
o
t
e
s
i
g
n
i
f
c
a
n
c
e
a
t
5
a
n
d
1
0
%
l
e
v
e
l
,
r
e
s
p
e
c
t
i
v
e
l
y
JFEP
7,4
314
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
test the null hypothesis for all coeffcients by individually restricting them to zero.
Table VIII presents results of Wald test for six different null hypotheses. Column 1
of the table reports results of the test in which the null hypothesis states that all
coeffcients, except the intercept termare zero. The signifcant t-statistic value at the
conventional level recommends the rejection of the null hypothesis in all three
sample periods, implying that at least one fundamental variable is signifcant in the
overall period. Null hypothesis in Column 2 states that coeffcients of national
output are restricted to zero, which is also rejected completely in all periods.
Similarly, all other null hypotheses are rejected in all cases with only one exception
where trade balance for the period April 1994 to February 2001 is restricted to zero
(see Column 6 of Table VIII). Nowcombining these results with FMOLS estimations,
we can understand the signifcance of different theoretical models in each period, as
shown in Table IX. Specifcally, our results indicate for a validity of fexible price
model, Sticky price model and real interest rate model in the Indian case. The only
exception is Hooper and Morton model as trade balance, a crucial variable in this
model, is signifcant only in the full sample period, after accounting for the
structural break in the full model; trade balance loses out its signifcance at least in
one sub-period.
4.6 Results of the IRF: a short-run analysis
To see the short-term interaction, the Cholesky one standard deviation response
functions of exchange rate over a period of 10 months are plotted for money supply,
price level, short-term interest rate and trade balance. The graphs in Figure 2a-2e
suggest that there is an inverse relationship between exchange rate and price index.
Over a period of 10 months, for the frst three periods, the response is immediate and
sharp, but beyond that, it is fxed. This inverse relationship is in accordance with the
sign conventions of RID and Hooper and Morton model. As the domestic price level
increases, the exchange rate increases or the currency depreciates. The response of
impulses to output does not show much deviation in a period of 10 months. So, none
of the models explain the effect of output on exchange rate in such a short span. It
may take a longer span of time to refect the effect of these changes. The response to
money supply and the net trade balance show a similar upward trend for an initial
period of two months, and thereafter, it attains a consistent level. All theoretical
models support the fact that as money supply increases, currency depreciates, i.e.
exchange rate, increases. The response of trade balance shock is quite ambiguous,
as with the increase in exports minus imports, the currency should appreciate and
lower down the exchange value. Finally, the interest rate shocks provide an
important fnding of direct relationship with the exchange rate as shown by the
Table V.
Results of unit root
test for error terms
Sample period ADF DF-GLS PP
April 1994 to March 2010 (full sample) ?5.908** ?4.287** ?5.829**
April 1994 to February 2001 ?3.959** ?3.953** ?3.899**
April 2001 to March 2010 ?6.117** ?5.008** ?6.117**
Note: **Denotes signifcance at 5% level
315
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table VI.
Johansen and
Juselius (1990)
cointegration test
results
H
0
r
?
0
r
?
1
r
?
2
r
?
3
r
?
4
r
?
5
r
?
6
r
?
7
r
?
8
r
?
9
r
?
1
0
P
a
n
e
l
A
:
f
u
l
l
s
a
m
p
l
e
p
e
r
i
o
d
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
E
i
g
e
n
v
a
l
u
e
0
.
4
3
6
0
.
3
9
0
0
.
3
4
9
0
.
2
9
0
0
.
2
2
8
0
.
2
0
9
0
.
1
5
0
0
.
1
3
1
0
.
0
6
7
0
.
0
3
6
0
.
0
0
0
1
?
T
r
a
c
e
4
2
8
.
3
5
*
*
3
3
8
.
9
6
*
*
2
6
1
.
8
2
*
*
1
9
4
.
4
1
*
*
1
4
1
.
2
8
*
*
1
0
0
.
9
0
*
*
6
4
.
1
7
3
8
.
7
9
1
6
.
7
1
5
.
7
5
0
.
0
2
1
?
M
a
x
8
9
.
3
8
*
*
7
7
.
1
6
*
*
6
7
.
0
0
4
*
*
5
3
.
5
0
*
*
4
0
.
3
7
3
6
.
7
3
2
5
.
3
8
2
2
.
0
7
1
0
.
9
5
5
.
7
3
0
.
0
2
P
a
n
e
l
B
:
f
r
s
t
s
u
b
-
p
e
r
i
o
d
(
A
p
r
i
l
1
9
9
4
t
o
F
e
b
r
u
a
r
y
2
0
0
1
)
E
i
g
e
n
v
a
l
u
e
0
.
7
9
6
0
.
6
4
9
0
.
6
4
5
0
.
5
8
5
0
.
4
7
6
0
.
3
9
8
0
.
3
7
5
0
.
2
6
8
0
.
2
0
1
0
.
1
7
0
7
0
.
0
4
2
?
T
r
a
c
e
4
3
1
.
0
0
5
*
*
3
3
2
.
1
4
*
*
2
6
7
.
1
5
*
*
2
0
2
.
9
4
*
*
1
4
8
.
3
5
*
*
1
0
8
.
2
3
*
*
7
6
.
7
1
*
*
4
7
.
5
0
2
8
.
1
1
1
4
.
2
8
2
.
6
7
?
M
a
x
9
8
.
8
6
*
*
6
4
.
9
9
*
*
6
4
.
2
1
*
*
5
4
.
6
1
*
*
4
0
.
0
8
3
1
.
5
1
2
9
.
2
1
1
9
.
3
8
1
3
.
8
3
1
1
.
6
0
2
.
6
7
P
a
n
e
l
C
:
s
e
c
o
n
d
s
u
b
-
p
e
r
i
o
d
(
A
p
r
i
l
2
0
0
1
t
o
M
a
r
c
h
2
0
1
0
)
E
i
g
e
n
v
a
l
u
e
0
.
6
9
7
0
.
6
3
2
0
.
5
5
0
0
.
4
3
2
0
.
3
8
5
0
.
3
0
8
0
.
2
6
9
0
.
2
3
5
0
.
1
7
5
0
.
0
7
1
0
.
0
0
0
3
?
T
r
a
c
e
5
1
0
.
3
5
*
*
3
9
4
.
4
0
*
*
2
9
6
.
5
1
*
*
2
1
8
.
9
6
*
*
1
6
3
.
8
2
*
*
1
1
6
.
6
7
*
*
8
0
.
9
3
3
*
*
5
0
.
4
6
8
*
*
2
4
.
4
2
7
.
0
7
7
0
.
0
3
7
?
M
a
x
1
1
5
.
9
4
*
*
9
7
.
8
9
*
*
7
7
.
5
4
9
*
*
5
5
.
1
3
*
*
4
7
.
1
5
8
*
*
3
5
.
7
3
3
0
.
4
6
2
6
.
0
4
3
1
7
.
3
4
7
.
0
3
9
0
.
0
3
7
N
o
t
e
:
*
*
D
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
JFEP
7,4
316
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table VII.
Results of Johansen
et al. (2000)
cointegrating test
H
0
r
?
0
r
?
1
r
?
2
r
?
3
r
?
4
r
?
5
r
?
6
r
?
7
r
?
8
r
?
9
r
?
1
0
E
i
g
e
n
v
a
l
u
e
0
.
4
7
6
2
0
.
3
9
2
5
0
.
3
2
9
8
0
.
2
7
2
7
0
.
2
5
3
8
0
.
1
9
5
1
0
.
1
8
8
9
6
0
.
1
4
3
7
0
.
1
2
7
6
0
.
0
9
2
6
0
.
0
3
3
9
?
T
r
a
c
e
5
2
3
.
1
4
*
*
4
1
0
.
6
2
*
*
3
2
3
.
8
8
*
*
2
5
4
.
2
4
*
*
1
9
8
.
8
3
*
*
1
4
7
.
8
9
*
*
1
1
0
.
1
3
*
*
7
3
.
6
8
*
*
4
6
.
6
8
2
6
5
*
*
2
2
.
9
2
5
*
*
6
.
0
1
4
7
N
o
t
e
s
:
C
r
i
t
i
c
a
l
v
a
l
u
e
s
a
r
e
c
a
l
c
u
l
a
t
e
d
b
a
s
e
d
o
n
G
i
l
e
s
a
n
d
G
o
d
w
i
n
,
2
0
1
2
;
*
*
d
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
317
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table VIII.
Wald restriction tests
on the coeffcients of
the empirical models
S
a
m
p
l
e
p
e
r
i
o
d
H
0
:
A
l
l
?
?
0
(
1
)
H
0
:
?
1
?
?
2
?
0
(
2
)
H
0
:
?
3
?
?
4
?
0
(
3
)
H
0
:
?
5
?
?
6
?
0
(
4
)
H
0
:
?
7
?
?
8
?
0
(
5
)
H
0
:
?
9
?
?
1
0
?
0
(
6
)
1
9
9
4
:
0
4
-
2
0
0
1
:
0
2
2
3
8
.
1
1
7
9
*
*
(
0
.
0
0
0
0
)
7
.
0
4
5
9
1
3
*
*
(
0
.
0
0
1
9
)
1
1
.
8
4
1
5
8
*
*
(
0
.
0
0
0
1
)
4
.
8
7
3
6
4
8
*
*
(
0
.
0
1
1
3
)
4
.
9
8
5
2
4
2
*
*
(
0
.
0
1
0
2
)
1
.
0
8
0
3
1
(
0
.
3
4
6
6
)
A
p
r
i
l
2
0
0
1
-
M
a
r
c
h
2
0
1
0
2
7
.
7
3
8
6
3
*
*
*
(
0
.
0
0
0
0
)
4
2
.
8
1
8
3
4
*
*
(
0
.
0
0
0
0
)
3
.
9
7
1
4
0
7
*
*
(
0
.
0
2
2
1
)
1
0
.
0
8
9
4
7
*
*
(
0
.
0
0
0
1
)
2
0
.
3
4
3
1
3
*
*
(
0
.
0
0
0
0
)
1
0
.
4
8
3
8
*
*
(
0
.
0
0
0
1
)
A
p
r
i
l
1
9
9
4
-
M
a
r
c
h
2
0
1
0
(
f
u
l
l
s
a
m
p
l
e
)
8
.
2
3
E
?
2
3
*
*
(
0
.
0
0
0
0
)
1
7
.
8
7
9
7
1
*
*
(
0
.
0
0
0
0
)
4
4
3
.
2
7
1
8
*
*
(
0
.
0
0
0
0
)
6
.
4
8
E
?
2
1
*
*
(
0
.
0
0
0
0
)
1
5
0
.
8
9
8
7
*
*
(
0
.
0
0
0
0
)
1
1
2
.
6
3
6
7
*
*
(
0
.
0
0
0
0
)
N
o
t
e
s
:
?
1
,
?
2
,
?
3
,
?
4
,
?
5
,
?
6
,
?
7
,
?
8
,
?
9
a
n
d
?
1
0
a
r
e
t
h
e
c
o
e
f
f
c
i
e
n
t
s
o
f
y
,
y
*
,
p
,
p
*
,
m
,
m
*
,
i
,
i
*
,
t
b
a
n
d
t
b
*
,
r
e
s
p
e
c
t
i
v
e
l
y
;
p
-
V
a
l
u
e
i
n
p
a
r
e
n
t
h
e
s
i
s
;
*
*
d
e
n
o
t
e
s
s
i
g
n
i
f
c
a
n
c
e
a
t
5
%
l
e
v
e
l
JFEP
7,4
318
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
steep increase in the response graph. This is in accordance to the practicability of
interest rate as an effective policy instrument in controlling exchange rate volatility.
5. Conclusion
This study has considered the possibility of structural break in the data while
examining the long-run relationship between Indian rupee-dollar exchange rate and
macroeconomic fundamentals. It is shown that fundamentals have a signifcant
impact on each sub-period; however, their size of impact differs signifcantly,
depending on various regimes and period. Our results show that macroeconomic
environment can produce breaks, which ultimately give rise to time-varying
coeffcients. This implies that for a developing country like India, the fundamentals
matter in formulating an exchange rate policy, but their behavior is not constant
over the period. Therefore, the monetary authority needs to be more fexible in terms
of controlling the exchange rate via monetary variables. The FMOLS results prove
that there is no single particular model which can explain exchange rate behavior in
the long run, and the coeffcients of the fundamentals are dynamic in the long run.
This result is in disagreement of Meese and Rogoff (1983) that monetary models fail
to perform on the empirical ground and, thus, offers support to the fndings of
MacDonald and Taylor (1994) and Beckmann et al. (2010). Importantly, our results
show that incorporating structure breaks in the model, the cointegration framework
perform well in explaining the effects of fundamentals on exchange rate movement.
In a way, this provided the evidence for the temporal stability of parameters of
exchange rate models based on macroeconomic fundamentals in the Indian currency
context. Results of the short-run analysis, i.e. impulse responses on the exchange
rate-interest rate linkage, highlight the interest rate importance as an effective
policy tool for smoothening out the volatility of exchange rate on a short-term basis.
The study of the behavior of a macroeconomic variable in different regimes have
highlighted the importance of each macroeconomic factor in movement of exchange
rate in this research, which will likely help policymakers in designing the
appropriate policy by understanding the deep-rooted, complex and time-varying
relationship.
Overall, our fndings are signifcant and relevant for a policy perspective. It is,
however, contrasting with much, if not all, of the existing empirical evidence on this
issue. Most importantly, our analysis submits that the monetary models of exchange
rate models work, but results are sensitive toward period of selection and
methodology. We also believe that the analysis may improve further if it is tested on
the next-generation tests of structural breaks in the cointegrating system[6].
Table IX.
Signifcance of
exchange rate models
in different periods
Sample period
Flexible price
model
Sticky price
model
Real interest
rate model
Hooper and Morton
model
January1994-February 2001 YES YES YES NO
April 2001-March 2010 YES YES YES NO
April 1994-March 2010 (full sample) YES YES YES YES
319
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Notes
1. The annual average exchange rate of the rupee went up from INR 45.56 per US dollar in
2010-2011 to INR 47.92 per US dollar in 2011-2012 and further to INR 54.41 per US dollar in
2012-2013. It rose to reach an average of INR 60.50 per US dollar in 2013-2014. The intra-year
levels of depreciation have been sharper in some months; but exhibit two-way movements
within the broad rising trend (MoF, 2013, p. 140).
2. Recently, the Chief Economic Advisor of Government of India, Arvind Subramanian, has
been quoted saying that “Sustaining a weaker currency by an aggressive forex reserves’
Figure 2.
IRF of exchange rate
with other
macroeconomic
variables
JFEP
7,4
320
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
build-up – $750 billion to $1 trillion in the medium-term – with letting the rupee dip at every
opportunity” (ET, 2015).
3. The break model results signifcantly vary from that of crash model. In both models,
breakpoints lack consistency as well as many of breakpoints could not clear the statistical
test. The result of the break model is not reported here to conserve the space; however, it will
be provided on a request from the corresponding author.
4. The frst breakpoint can be ignored, as it has lesser utility in predicting a relationship, as the
numbers of observations till frst breakpoint are very few, which left us with no choice but to
consider only one major breakpoint.
5. DOLS is developed by Saikkonen (1992) and Stock and Watson (1993); the method involves
augmenting the cointegrating regression with lags and leads of explanatory variables so that
the resulting cointegrating equation’s error term is orthogonal to the entire history of the
stochastic regressor innovations.
6. One such unit root test recently developed by Narayan and Poop (2010).
References
Baharumshah, A.Z., Mohd, S.H. and Ahn, S.K. (2009), “On the predictive power of monetary
exchange rate model: the case of the Malaysian ringgit/US dollar rate”, Applied Economics,
Vol. 41 No. 14, pp. 1761-1770.
Bai, J. and Perron, P. (1998), “Estimating and testing linear models with multiple structural
changes”, Econometrica, Vol. 66, pp. 47-78.
Bai, J. and Perron, P. (2003), “Computation and analysis of multiple structural change models”,
Journal of Applied Econometrics, Vol. 6, pp. 72-78.
Beckmann, J., Belke, A. and Kuhl, M. (2010), “The dollar-euro exchange rate and macroeconomic
fundamentals: a time-varying coeffcient approach”, Review of World Economics, Vol. 147,
pp. 11-40.
Bekiros, S.D. (2014), “Exchange rates and fundamentals: co-movement, long-run relationships and
short-run dynamics”, Journal of Banking & Finance, Vol. 39, pp. 117-134.
Bhanumurthy, N.R. (2006), “Macroeconomic fundamentals and exchange rate dynamics in India”,
Economic and Political Weekly, Vol. 10 No. 11, pp. 1101-1107.
Bilson, J.F.O. (1978), “The monetary approach to the exchange rate: some empirical evidence”,
IMF Staff Papers, Vol. 25, pp. 48-75.
Chang, M.-J. and Su, C.Y. (2014), “The dynamic relationship between exchange rates and
macroeconomic fundamentals: evidence from Pacifc Rim countries”, Journal of
International Financial Markets, Institutions and Money, Vol. 30, pp. 220-246.
Cheung, Y., Chin, M.D. and Pascual, A.G. (2005), “Empirical exchange rate models of the nineties:
are any ft to survive?”, Journal of International Money and Finance, Vol. 24, pp. 1150-1175.
Devereux, M.B. (1997), “Real exchange rates and macroeconomics: evidence and theory”,
Canadian Journal of Economics, Vol. 30 No. 4a, pp. 773-808.
Dornbusch, R. (1976), “Expectations and exchange rate dynamics”, Journal of Political Economy,
Vol. 84, pp. 1161-1176.
Dropsy, V. (1996), “Real exchange rates and structural breaks”, Applied Economics, Vol. 28,
pp. 209-219.
321
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Dua, P. and Ranjan, R. (2011), “Modelling and forecasting the Indian Re/US$ exchange rate”,
Working Paper no. 197, Centre for Development Economics, Delhi School of Economics,
New Delhi.
Dua, P. and Sinha, A. (2007), “India’s insulation fromthe East Asian crisis: an analysis”, Singapore
Economic Review, Vol. 52, pp. 419-443.
ET (2015), “Let rupee dip when possible for India’s long-term economic wellbeing: Arvind
Subramanian”, CEA, 2 March, available at:http://articles.economictimes.indiatimes.com/
2015-03-02/news/59683835_1_exchange-rate-capital-infows-arvind-subramanian
Florentis, G., Gordon, D.V. and Huber, P. (1994), “Monetary models of Canadian US exchange rate:
a re-examination of empirical evidence 1971-1986”, Quarterly Journal of Business and
Economics, Vol. 33, pp. 27-43.
Frankel, J. (1984), “Tests of monetary and portfolio balance models of exchange rate
determination”, in Bilson, J.F.O. and Marston, R.C. (Eds), Exchange Rate Theory and
Practice, Chicago University Press, Chicago, pp. 239-259.
Frankel, J.A. (1979), “On the mark: a theory of foating exchange rates based on real interest rate
differentials” American Economic Review, Vol. 69, pp. 610-622.
Frenkel, J.A. (1976), “A monetary approach to the exchange rate: doctrinal aspects and empirical
evidence”, Scandinavian Journal of Economics, Vol. 78, pp. 200-224.
Giles, D.E. and Godwin, R.T. (2012), “Testing for multivariate cointegration in the presence of
structural breaks: p-values and critical values”, Applied Economics Letters, Vol. 19 No. 16,
pp. 1561-1565.
Goldberg, M.D. and Frydman, R. (1996), “Empirical exchange rate models and shifts in the
co-integrating vector”, Structural Change in Economic Dynamics, Vol. 7, pp. 55-78.
Gregory, A.W. and Hansen, B.E. (1996), “Residual-based tests for cointegration in models with
regime shifts”, Journal of Econometrics, Vol. 70 No. 1, pp. 99-126.
Hegwood, N.D. and Nath, H.K. (2014), “Real exchange rate dynamics: evidence from India”,
Economic Analysis and Policy, Vol. 44 No. 4, pp. 396-404.
Hodrick, R.J. (1978), “An empirical analysis of the monetary approach to the determination of the
exchange rate”, in Frenkel, J.A. and Johnson, H.G. (Eds), The Economics of Exchange Rates,
Reading: Addison-Wesley, Boston.
Hoffmann, M. and MacDonald, R. (2009), “Real exchange rates and real interest rate differentials:
a present value interpretation”, European Economic Review, Vol. 53, pp. 952-970.
Hooper, P. and Morton, J. (1982), “Fluctuations in the dollar: a model of nominal and real exchange
rate determination”, Journal of International Money and Finance, Vol. 1, pp. 39-56.
Johansen, S. and Juselius, K. (1990), “Maximum likelihood estimation and inference on
cointegration with applications to the demand for money”, Oxford Bulletin of Economics
and Statistics, Vol. 52, pp. 169-210.
Johansen, S., Mosconi, R. and Nielsen, B. (2000), “Cointegration analysis in the presence of
structural breaks in deterministic trends”, Econometrics Journal, Vol. 3, pp. 216-249.
Kejriwal, M. and Perron, P. (2010), “Testing for multiple structural changes in cointegrated
regression models”, Journal of Business and Economic Statistics, Vol. 28, pp. 503-522.
Kia, A. (2013), “Determinants of the real exchange rate in a small open economy: evidence from
Canada”, Journal of International Financial Markets, Institutions & Money, Vol. 23
No. 2013, pp. 163-178.
Kim, B.J.C. and Mo, S. (1995), “Cointegration and the long-run forecast of exchange rates”,
Economic Letters, Vol. 48, pp. 353-359.
JFEP
7,4
322
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Kletzer, K. and Kohli, R. (2000), “Exchange rate dynamics with fnancial repression: a test of
exchange rate models for India”, Working Paper no. 52, ICRIER, New Delhi.
Kollman, R. (1997), “The exchange rate in a dynamic optimizing current account model with
nominal rigidities”, International Monetary Fund Working Paper, 97/7, IMF.
Lee, J., List, J. and Strazicich, M. (2006), “Nonrenewable resource prices: deterministic or stochastic
trends?” Journal of Environmental Economics and Management, Vol. 51, pp. 354-370.
Lee, J. and Strazicich, M. (2003), “Minimum LM unit root test with two structural breaks”, Review
of Economics and Statistics, Vol. 85, pp. 1082-1089.
Lee, J. and Strazicich, M. (2004), “Minimum LM unit root tests”, Working Paper No. 04-07,
University of Alabama.
Lumsdaine, R. and Papell, D. (1997), “Multiple trend breaks and the unit root hypothesis”, Review
of Economics and Statistics, Vol. 79, pp. 212-218.
McNown, R. and Wallace, M.S. (1989), “Cointegration tests for long-run equilibrium in the
monetary exchange rate model”, Economic Letters, Vol. 31, pp. 263-267.
Macdonald, R. and Taylor, M.P. (1994), “The monetary model of the exchange rate: long-run
relationships, short-run dynamics and howto beat a randomwalk” Journal of International
Money and Finance, Vol. 13, pp. 276-290.
Mallick, H. (2010), “Examining the behavior of exchange rate under a partial capital mobility
regime”, Forthcoming with International Economic Journal, Vol. 24 No. 2, pp. 141-157.
Mark, N.C. (1995), “Exchange rates and fundamentals: evidence on long horizon predictability”,
American Economic Review, Vol. 85, pp. 201-218.
Mark, N.C. and Sul, D. (2001), “Nominal exchange rates and monetary fundamentals: evidence
from a small post-Bretton woods panel”, Journal of International Economics, Vol. 53,
pp. 29-52.
Meese, R.A. and Rogoff, K. (1983), “Empirical exchange rate models of the seventies. Do they ft
out of sample?”, Journal of International Economics, Vol. 14, pp. 3-24.
Mishra, R.K. and Sharma, C. (2010), “Real exchange rate behavior and optimum currency area in
East Asia: evidence from generalized purchasing power parity”, International Review of
Financial Analysis, Vol. 19, pp. 205-213.
MoF (2014), “Economic survey 2013-2014”, Ministry of Finance, Government of India, Oxford
University Press, New Delhi.
Morley, B. (2007), “The monetary model of the exchange rate and equities: an ARDL bounds
testing approach”, Applied Financial Economics, Vol. 17, pp. 391-397.
Nag, A.K. and Mitra, A. (1998), “Exchange rate of rupee and purchasing power parity”, Economic
and Political Weekly, Vol. 33, pp. 1525-1532.
Narayan, P.K. (2006), “Are bilateral real exchange rates stationary? Evidence from Lagrange
Multiplier unit root tests for India”, Applied Economics, Vol. 38, pp. 63-70.
Narayan, P.K. and Popp, S. (2010), “A new unit root test with two structural breaks in level and
slope at unknown time”, Journal of Applied Statistics, Vol. 37, pp. 1425-1438.
Pattanaik, S. and Sahoo, S. (2003), “The effectiveness of intervention in India: an empirical
assessment”, RBI Occasional Papers, 22, RBI, Mumbai.
Perron, P. (1989), “The great crash, the oil price shock, and the unit root hypothesis”,
Econometrica, Vol. 57, pp. 1361-1401.
Phillips, P.C.B. (1995), “Fully modifed least squares and vector autoregression”, Econometrica,
Vol. 63, pp. 1023-1078.
323
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Phillips, P.C.B. and Hansen, B.E. (1990), “Statistical inference in instrumental variables regression
with I(1) processes”, Review of Economic Studies, Vol. 57, pp. 99-125.
Rapach, D.E. and Wohar, M.E. (2002), “Testing the monetary model of exchange rate
determination: new evidence from a century of data”, Journal of International Economics,
Vol. 58, pp. 359-385.
Rogoff, K. (2009), “Exchange rates in the modern foating era: what do we really know?”, Reviewof
World Economics, Vol. 145 No. 1, pp. 1-12.
Saikkonen, P. (1992), “Estimation and testing of cointegrated systems by an autoregressive
approximation”, Econometric Theory, Vol. 8, pp. 1-27.
Sarno, L. and Valente, J. (2009), “Exchange rates and fundamentals: footloose or evolving
relationship?”, Journal of the European Economic Association, Vol. 7, pp. 786-830.
Stock, J.H. and Watson, M. (1993), “A simple estimator of cointegrating vectors in higher order
integrated systems”, Econometrica, Vol. 61, pp. 783-820.
Tiwari, A.K., Dar, A.B. and Bhanja, N. (2013), “Oil price and exchange rates: a wavelet based
analysis for India”, Economic Modelling, Vol. 31, pp. 414-422.
UNCTAD (1998), “Trade and Development Report 1998”, United Nations Conference on Trade
and Development (UNCTAD), Geneva.
Williamson, J. (2009 ), “Exchange rate economics”, Open Economic Review, Vol. 20, pp. 123-146.
Zhang, S., Lowinger, T.C. and Tang, J. (2007), “The monetary exchange rate model: long-run,
short-run, and forecasting performance”, Journal of Economic Integration, Vol. 22,
pp. 397-406.
Zivot, E. and Andrews, D.W.K. (1992), “Further evidence on the great crash, the oil-price shock,
and the unit-root hypothesis”, Journal of Business & Economic Statistics, Vol. 10,
pp. 251-270.
Further reading
Perron, P. (1997), “Further evidence on breaking trend functions in macroeconomic variables”,
Journal of Econometrics, Vol. 80, pp. 355-385.
WEO (2015), Uneven Growth: Short- and Long-Term Factors, April, IMF, Washington, DC.
Corresponding author
Chandan Sharma can be contacted at: [email protected]
JFEP
7,4
324
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Appendix
Table AI.
Data description
Series Data description Data source
e Indian rupee/US dollar, end-of-month rates, in log form IFS
m M3 of India, in log form, seasonally adjusted IFS
m* M1 of India, in log form, seasonally adjusted IFS
p CPI of India, in log form, seasonally adjusted RBI
p* CPI of India, in log form, seasonally adjusted IFS
y IIP of India, in log form, seasonally adjusted IFS
y* IIP of USA, in log form, seasonally adjusted IFS
i 91-day treasury bill rates in India IFS
i* 91-day treasury bill rates in USA RBI
tb Trade balance of India converted in to an index, seasonally adjusted,
in log form
RBI
tb* Trade balance of USA converted in to an index, seasonally adjusted,
in log form
IFS
Table AII.
Descriptive statistics
y y* p p* e m m* tb tb* i i*
Mean 5.23 4.48 6.18 5.21 3.76 14.32 8.62 5.47 5.62 7.19 3.35
Median 5.17 4.50 6.18 5.20 3.78 14.33 8.65 5.15 5.68 7.00 4.20
Maximum 5.77 4.61 6.68 5.39 3.94 15.52 9.05 7.90 6.36 12.97 6.17
Minimum 4.76 4.27 5.71 5.02 3.45 13.17 8.17 3.10 4.56 3.22 0.03
Std. Dev. 0.28 0.09 0.24 0.11 0.11 0.67 0.27 1.23 0.53 2.36 1.90
Skewness 0.27 ?0.82 0.02 0.09 ?0.99 0.06 ?0.07 0.24 ?0.47 0.68 ?0.39
Kurtosis 1.87 3.13 2.39 1.76 3.13 1.91 1.78 1.92 1.99 3.14 1.62
Jarque-Bera 11.79 19.96 2.82 11.78 29.51 9.02 11.18 10.55 14.16 13.92 18.64
Observations 179 179 179 179 179 179 179 179 179 179 179
Table AIII.
Gregory and Hansen
(1996) cointegration
tests with structural
breaks
Break date Gregory and Hansen test statistic Existence of cointegration
Jul 1997 ?8.505** Yes
Note: **denotes statistically signifcant at 5% level
325
Macroeconomic
fundamentals
and dynamics
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table AIV.
DOLS estimations
V
a
r
i
a
b
l
e
s
F
u
l
l
s
a
m
p
l
e
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
1
9
9
4
t
o
F
e
b
r
u
a
r
y
2
0
0
1
S
u
b
-
s
a
m
p
l
e
A
p
r
i
l
2
0
0
1
t
o
M
a
r
c
h
2
0
1
0
W
i
t
h
o
u
t
D
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
W
i
t
h
B
r
e
a
k
D
u
m
m
y
(
A
p
r
i
l
1
9
9
4
t
o
M
a
r
c
h
2
0
1
0
)
c
o
n
s
t
a
n
t
?
1
.
0
4
5
(
?
0
2
.
3
6
)
2
.
9
8
2
(
0
.
5
7
8
)
?
3
.
4
9
7
(
0
.
6
8
3
)
1
3
.
9
1
2
*
*
(
2
.
1
3
5
)
m
0
.
1
1
9
(
0
.
3
2
5
)
0
.
3
7
3
7
2
2
(
0
.
9
9
1
)
0
.
2
2
1
*
*
(
2
.
4
4
6
)
0
.
4
5
7
(
1
.
1
8
9
)
m
*
0
.
9
6
5
8
*
(
1
.
7
8
8
)
0
.
5
7
5
*
*
(
2
.
9
3
2
)
?
0
.
5
2
3
(
?
0
.
5
1
2
)
0
.
3
6
5
*
*
(
1
.
7
9
6
)
p
1
.
0
4
4
*
*
(
3
.
9
7
5
)
0
.
9
5
4
9
*
*
(
3
.
0
6
1
)
0
.
9
0
4
(
0
.
4
8
2
)
0
.
0
8
6
(
1
.
5
5
5
)
p
*
?
0
.
2
6
1
(
?
0
.
7
9
5
)
?
1
.
4
4
8
(
1
.
4
3
2
)
0
.
4
4
2
(
0
.
4
2
5
)
?
2
.
0
3
1
(
1
.
4
7
2
)
y
?
1
.
3
3
4
*
*
(
?
4
.
5
3
7
)
?
1
.
3
0
4
*
*
(
?
4
.
2
9
9
)
0
.
5
7
5
*
*
(
2
.
1
2
7
)
?
0
.
5
6
5
*
*
(
?
2
.
1
4
5
)
y
*
?
0
.
2
6
1
(
?
1
.
3
1
6
)
?
0
.
3
0
5
(
?
1
.
4
9
2
)
?
0
.
3
4
1
*
(
?
1
.
7
4
4
)
?
1
.
7
9
7
3
*
*
(
3
.
5
3
2
)
i
?
0
.
0
0
1
(
?
0
.
4
1
2
)
?
0
.
0
0
0
8
(
?
0
.
2
1
8
)
0
.
0
0
9
*
*
(
2
.
1
0
1
)
0
.
0
2
1
*
*
(
2
.
6
2
3
)
i
*
0
.
0
2
4
*
*
(
4
.
1
8
7
)
?
0
.
0
0
0
8
*
(
1
.
7
8
1
)
?
0
.
0
0
2
(
?
0
.
2
1
5
)
0
.
3
6
5
*
*
(
2
.
4
9
6
)
t
b
?
0
.
0
5
4
*
*
(
?
4
.
4
3
3
)
?
0
.
0
4
7
*
*
(
3
.
8
5
1
)
0
.
0
1
7
(
0
.
1
5
5
)
?
0
.
0
2
7
*
(
?
1
.
7
3
7
)
t
b
*
0
.
0
6
2
1
(
1
.
0
4
6
)
0
.
0
8
2
(
1
.
3
7
1
)
?
0
.
1
1
7
*
*
(
?
2
.
1
8
2
)
0
.
2
1
8
*
*
(
2
.
3
0
3
)
D
u
m
m
y
M
a
r
c
h
1
9
9
6
0
.
0
2
7
8
*
(
1
.
8
1
7
)
D
u
m
m
y
M
a
r
c
h
2
0
0
1
0
.
0
3
2
1
(
1
.
0
4
9
)
A
d
j
u
s
t
e
d
R
-
s
q
u
a
r
e
d
0
.
9
2
9
7
0
.
9
3
5
1
0
.
9
8
2
0
.
8
7
7
E
s
t
i
m
a
t
o
r
D
O
L
S
D
O
L
S
D
O
L
S
D
O
L
S
N
o
t
e
:
*
*
D
e
n
o
t
e
s
s
t
a
t
i
s
t
i
c
a
l
l
y
s
i
g
n
i
f
c
a
n
t
a
t
5
%
l
e
v
e
l
JFEP
7,4
326
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
5
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_805623593.pdf