Capital inflows, inflation, and the exchange rate volatility: an investigation for linear and nonlinear causal linkages.
Rashid, Abdul ; Husain, Fazal
This paper empirically examines the effect of foreign capital
inflows on domestic price levels, monetary expansion, and the exchange
rate volatility for Pakistan using linear and nonlinear causality tests.
The key message emerging from the analysis is that there is a
significant inflationary impact of capital inflows, in particular during
the period of surges in capital inflows. Specifically, we find evidence
of a significant nonlinear Granger causality running from capital
inflows to the change in domestic prices. We also show that domestic
prices are nonlinearly caused (in Granger sense) by the growth of
domestic debt and money supply-to-GDP ratio. Our results, however,
suggest that the market interest rate and the nominal exchange rate do
not have significant relationships with domestic prices. The findings
suggest that there is a need to manage the capital inflows in such a way
that they should neither create an inflationary pressure in the economy
nor fuel the exchange rate volatility.
JEL Classification: C22, C32, F21, F31, F32
Keywords: Capital Inflows, Inflationary Pressures, the Exchange
Rate Volatility, Monetary Expansion, Nonlinear Dynamics
1. INTRODUCTION
Examination of how macroeconomic indicators respond to foreign
capital inflows is important to understand the role of foreign funds in
host countries. Several studies have empirically examined the effects of
ebbs and surges in capital inflows on macroeconomic performance of host
countries. The findings of these studies are inconclusive at best,
however. On the one hand, large number of studies have documented that
surges in foreign capital inflows help promote investments, stimulate
economic development, improve resource allocation, interact human
capital, deepen domestic financial sector, and encourage positive growth
externalities. Examples of these studies include, among others, De Mello
(1996, 1997), Reisen and Soto (2001), Hermes and Lensink (2003), Alfaro,
et at. (2004), Buch, el al. (2005), Adams (2009), Wang and Wong (2009),
Choong, et al. (2010), and Azman-Saini, et al. (2010). Researchers have
also shown that access to international funds help countries in
attaining sustainable economic growth, provide benefits associated with
international financial integration, and ensure domestic macroeconomic
stability [Kose, et al. (2009) and Obstfeld (2009)].
On the other hand, several studies have argued that the abrupt
integration of emerging market countries with international capital
markets has created some problems for the host economies. In particular,
researchers have observed that foreign capital inflows create
difficulties for the recipient countries [e.g., Rodrik and Velasco
(1999), Aghion, et al. (2000), Ventura (2002), Eichengreen (2004),
Caballero and Krishnamurthy (2006), Baharumshah and Thanoon (2006),
Edwards (2007a, 2007b), Mendoza and Terrones (2008), Reinhart and
Reinhart (2008), Caballero, et al. (2008), Hegerty (2009), Cardarelli,
et al. (2010), Kim and Yang (2011), Furceri, et al. (2011), Cecen and
Xiao (2012), Sethi and Sucharita (2012), Caballero (2012), and Furceri,
et al. (2012)]. These difficulties generally include appreciation in
currencies and in turn loss of foreign competitiveness, high inflation
rates, and increased vulnerability to banking crises. Large capital
inflows also help fuel headwinds in financial markets, surges in money
supply, excessive private credit growth, spending booms, asset market
bubbles, and the undermining of a strategy to achieve monetary stability
by pegging the exchange rate. Further, some studies such as Bernanke
(2005) argue that a glut of global savings leads to large international
trade imbalances. (1)
There is a growing agreement in the literature that preserving
stability of real and financial sectors during episodes of surges in
international capital inflows requires effective absorption and
sterilisation of foreign capital inflows. (2) For instance, the central
bank should intervene in the foreign exchange market in order to absorb
the foreign exchange brought in by the capital inflows. However, such
policy measures are not costless. For example, buildup of foreign
reserves as a result of the central bank's foreign exchange
purchases not only helps increase the monetary base of the economy but
also expands bank deposits and loans. Such surges in the money supply
result in excessive private credit growth and in a sequence generate
inflationary dynamics. Further, the expansion of bank balance sheets
owing to international capital inflows may increase the fragility of the
banking system if bank supervision is weak.
In theory, the effects of capital inflows on domestic financial and
real indicators depend on the ways in which they flow into an economy. T
he effects also depend on whether the inflows are sustainable or
temporary. Theoretically, the forces driving capital inflows differ from
country to country and can be classified into three clusters: (1) an
exogenous increase in the domestic productivity of capital, (2) an
autonomous increase in the domestic money demand function, and (3)
external factors, such as a reduction in international interest rates.
The former two are known as "pull" factors and the latter one
is called "push" factor. (3)
This paper aims to examine how domestic prices respond to foreign
capital inflows. Specifically, we propose a simple empirical model of
the equilibrium price by incorporating foreign capital inflows into the
standard classical quantity theory of demand for money. We also
empirically study the inflationary effects of capital inflows for a
relatively small open economy, namely Pakistan, using monthly data
covering the period 1990-2012. In particular, the paper investigates the
causal linkages between capital inflows, domestic price levels, the
growth of domestic debt, money supply, the market interest rate, and the
nominal exchange rate using the linear and nonlinear cointegration and
Granger causality tests. The paper also examines the impact of capital
inflows on the exchange rate volatility. The full sample period is
divided into two sub-samples in order to examine the differential
effects of capital inflows across episodes of low and high capital
inflows. Three different measures of foreign capital inflows are used in
empirical investigation.
The results of the paper suggest a significant inflationary impact
of foreign capital inflows, in particular during the period of surges in
capital inflows. Specifically, we show that there is a significant
co-movement in capital inflows and the price level. Results concerning
short-run dynamics indicate that there is significant linear as well as
nonlinear Granger causality running from capital inflows to the rate of
inflation. Our regression results also reveal that domestic prices are
nonlinearly caused (in Granger sense) by the growth of domestic debt and
the money supply-to-GDP ratio. However, our results suggest that the
market interest rate and the nominal exchange rate do not have
significant relationships with domestic prices. We also observe that
capital inflows amplify the volatility of real effective exchange rate
irrespective of whether the influx of foreign capital is low or high.
The rest of the paper proceeds as follows. Section 2 reviews the
inflow of foreign funds and the rate of inflation in Pakistan. Section 3
describes the empirical model, the empirical methodology, and the data
used to assess the relationship between capital inflow surges and the
price level. Section 4 presents the empirical results. Section 5
concludes the paper.
2. FOREIGN FUNDS AND THE RATE OF INFLATION: PAKISTANI CONTEXT
We start our empirical investigation by estimating correlations
between foreign capital inflows and the other variables included in the
analysis. We divide the full-sample period into two sub-periods. The
first sub-sample period ranges from January 1990 to December 2000, while
the second sub-sample runs from January 2001 to June 2012. This division
seems rational because there was a large capital surge during 2001 to
2012. The correlation matrices for first and second sub-sample periods
are presented in Tables 1 and 2, respectively. (4,5)
The correlation estimates suggest that the relationship among the
variables has changed dramatically during the massive capital surge
episode in 2001-2012. For instance, the ratio of money supply to GDP is
significantly correlated (it is also interesting to note that the
magnitude is negative) with the capital account to GDP ratio during the
period 1990-2000 when the inflow of foreign funds was stumpy and
inconsistent. The net foreign assets to GDP ratio and the foreign
reserves to GDP ratio, however, are not significantly related to money
supply during the period 1990-2000. During the period of relatively
large capital inflows (2001 to 2012), not only the magnitude of
correlation between the money supply-to-GDP ratio, the net foreign
assets-to-GDP ratio and the foreign reserves-to-GDP ratio has
considerably increased but also the correlation appears statistically
significant. This implies that after the year 2001, the foreign capital
inflows have played a significant role in expanding the monetary base of
Pakistan's economy.
The estimates of the correlation between the rate of inflation and
the net capital inflows-to-GDP ratio, the balance of capital
account-to-GDP ratio, and the foreign reserves-to-GDP ratio provide
fascinating insight about the association of foreign funds and
inflationary pressures. The inflation rate is significantly correlated
with the three ratios with a positive sign during the period of
2001-2012, whereas, it was only significantly related to the capital
account-to-GDP ratio over the period 1990-2000. The growth in domestic
debt is approximately 50 percent correlated with the monetary base of
the economy during the latter sub-period, though both were independent
of each other in earlier period.
In sum, the coefficients of correlation provide some preliminary
evidence of the dynamic interactions between capital inflows and
inflationary pressures: a theme that is explored in this paper.
Moreover, the estimates of correlation clearly indicate that there is a
structure break in 2001. Thus, it is very likely that nonlinearities
exist in the salient economic relationships. This motivates us to apply
the nonlinear cointegration and Granger causality test to examine the
long- and short-run linkages among the variables.
The correlation coefficients presented in Tables 1 and 2 provide
insights about the ineffectiveness of the policy used by the State Bank
of Pakistan (SBP) to manage the foreign capital inflows, particularly,
during the second sub-period. Theoretically, the change in monetary base
driven by capital inflows depends on the central bank's decision to
maintain a fixed exchange rate or to allow it to float freely with no
intervention. If there is an intervention, then an accumulation of
international reserves results in an increase in the net foreign
exchange assets of the central bank and directly affects the monetary
base of the economy. The inefficient intervention by the central bank
further aggravates the problem of expansion in the monetary base.
For effective absorption and sterilisation of foreign exchange
reserves, it is necessary to know whether the relationships between
foreign capital inflows, the monetary base of the economy, and the price
level, are stable in the long run or just short-term in nature. This
paper tries to address this question. If there is a significant
causation running from capital inflows to the rate of inflation, then,
definitely, the continuity of the existing foreign exchange management
policy could spell trouble for the economy.
Our paper contributes to the existing literature in at least four
major dimensions. First, we propose a simple model for equilibrium
prices, which predicts a positive impact of capital inflows on domestic
price levels. Second, we empirically examine the influence of foreign
capital inflows, the growth of domestic debt, the market interest rate,
the monetary base of the economy, and the real and nominal exchange
rates on domestic price levels. We also examine the impact of capital
inflows on the exchange rate volatility. Third, and more importantly, we
consider the possibility of nonlinearities in the relationship between
capital inflows and the other underlying variables with domestic prices.
Fourth, and finally, we examine the differential effects of capital
inflows and the other said variables on the price level during periods
of low (1990-2000) and high (2001-2012) capital inflows.
3. EMPIRICAL MODEL, METHODOLOGY, AND DATA
3.1. The Empirical Model
The impact of foreign capital inflows on domestic prices can be
explained through the following example. Suppose the private sector of
an economy receives a gift of G dollars from abroad. Now government does
not allow the private sector to use these dollars and buys the dollars
from the private sector at the current exchange rate, e, and adds G
dollars to its reserves. Consequently, the aggregate expenditures can be
defined as follows:
E = [bar.M] + eG (1) ... ... ... ... ... ... ... (1)
where E denotes the nominal expenditures on goods and services,
[bar.M] is the pre-gift nominal money stock, and e is the nominal
exchange rate. As expression (1) also represents the demand for money,
the money market equilibrium condition is:
[M.sup.d] = [M.sup.s] = [bar.M] + eG ... ... ... ... ... ... ...
(2)
Considering the quantity theory of demand for money, the nominal
price ([P.sub.N]), in equilibrium is defined as (6)
[P.sub.N] = V x [M.sup.x]/Y = V x [bar.M] + eG/Y ... ... ... ...
... ... ... (3)
Where V is the income velocity of money and Y denotes the aggregate
level of output. Equation (3) describes a positive relationship between
foreign capital inflows and domestic price levels (i.e., [partial
derivative][P.sub.N]/[partial derivative]G > 0) and negative
relationship between the level of output and prices (i.e., [partial
derivative][P.sub.N]/[partial derivative]Y < 0). Thus, as long as the
government adds the gift G to its reserves, and does not allow it to be
absorbed in the economy, it would produce only an inflationary effect.
Different explanatory variables are used in estimation of Equation
(3) to ensure that empirical links between capital inflows and
inflationary dynamics are not spurious. The choice of explanatory
variables in our empirical work is based on availability of data,
previous evidence found in the literature, and aforesaid theoretical
rationale.
3.2. The Empirical Methodology: Nonlinear Cointegration and Granger
Causality Tests
Regarding the linear long- and short-run relationship, we use the
standard Johansen's cointegration test and the Granger causality
test, respectively. As these two tests are very common in the
literature. Below, however, the nonlinear cointegration and causality
tests are explained in detail. We use the Lin and Granger (2004) tests
to explore the nonlinear long-run relationships between foreign capital
inflows and domestic price levels.
As in Lin and Granger (2004), let [x.sub.t] be a linear integrated
process and [y.sub.t] and [x.sub.t] are nonlinearly cointegrated with
function f provided [u.sub.t] = [y.sub.t] - f ([x.sub.t]) has asymptotic
order smaller than those of y and f(x). Lin and Granger (2004) define
the following steps to test the null of nonlinear cointegration against
alternative of no nonlinear cointegration.
(1) Identify the possible nonlinear function for using Alternative
Conditional Expectation (ACE) criterion (i.e., logarithm, exponential,
square root, Box-Cox transformation, etc.).
(2) Apply the Nonlinear Least Square (NLS) method to estimate the
parameters of the specified function.
(3) Obtain the residuals from the estimated model and store.
(4) Apply KPSS test for estimated residual to test the null of
nonlinear cointegration. (7)
To examine the nonlinear short-run causality, we use the
Hristu-Varsakkeis and Kyrtsou (2010) nonlinear Granger causality
test--known as the bivaraile noisy Mackey-Glass (hereafter M-G) model
and is based on a special type of nonlinear structure developed by
Kyrtsou and Labys (2006). The model is given below:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)
where X and Y are a pair of related time series variables, the
[[alpha].sub.tj] and [[delta].sub.ij] are parameters to be estimated,
[[tau].sub.I] are delays, [c.sub.i] are constants.
As mentioned in Kyrtsou and Labys (2006, 2007), Kyrtsou and Vorlow
(2009), and Kyrtsou and Terraza (2010), the principle advantage of Model
(4) over a simple VAR alternative is that the nonlinear M-G terms are
able to capture more complex dependent dynamics in a time series. The
test aims to capture whether past samples of a variable Y have a
significant nonlinear effect (of the type [MATHEMATICAL EXPRESSION NOT
REPRODUCIBLE IN ASCII]) on the current value of variable X.
Testing procedure begins by estimating the parameters of a M-G
model that best fits the given series, using ordinary least squares. To
test reverse causality (i.e., from X to Y), a second M-G model is
estimated, under the constraint [[alpha].sub.22] = 0. Let [[??].sub.It]
and [[??].sub.It] be the residuals produced by the unconstrained and
constrained best-fit M-G models, respectively. Next, we compute the sums
of squared residuals [S.sub.c] = [[summation].sup.N.sub.t=1]
[[??].sub.It] and [S.sub.u] = [[summation].sup.N.sub.t=1] [[??].sub.It].
Let m be the number of free parameters in the M-G model and k is the
number of parameters set to zero when estimating the constrained model,
then the test statistic is defined as:
[S.sub.F] = ([S.sub.c] - [S.sub.u])/k / [S.sub.u]/(N - m - 1)
[approximately equal to] [F.sub.k, N - m - 1]
If the calculated statistics is greater than a specified critical
value, then we reject the null hypothesis that Y does not nonlinearly
caused (in Granger sense).
3.3. The Data
We use monthly data from January 1990 to June 2012. The main source
of data is the IMF's International Financial Statistics database.
The variables are market interest rate (line 60b and denoted by MMR),
the log of nominal exchange rate (linear and denoted by LNER), the log
of real effective exchange rate (line 65um and denoted by LREER), the
log of manufacturing (industrial) production index (line 66ey and
denoted by LMPI), the log of consumer price index (line 64 and denoted
by LCPI), the ratio of net foreign assets to GDP (line 31n divided by
line 90b and denoted by FAR), the ratio of capital account to GDP (line
37a divided by 90b and denoted by CAR), the ratio of foreign reserves to
GDP ratio (line I Id times linear divided by line 90b and denoted by
FRR), the ratio of money supply to GDP (lines 34 plus 35 divided by line
90b and denoted by MSR) and the log of domestic credit (line 32 and
denoted by LDC). (8)
4. EMPIRICAL RESULTS
4.1. Identifying the Order of Integration
We start our investigation of the existence of long-run
relationship between foreign capital inflows and domestic price levels
by testing the order of integration. In particular, to examine whether
variables are integrated of order zero or one, we employ the ADF and the
KPSS [proposed by Kwiatkowski, et al. (1992)] unit root tests. The
results for both sub-periods are presented in Table 3. To find an
appropriate lag length for ADF tests, we use the criterion developed by
Campbell and Perron (1991). Under this procedure, one should start with
a maximum lag length (say k) and sequentially delete insignificant lags
until the last lag appears statistically significant. The ADF results
show that the null hypothesis of non-stationarity cannot be rejected at
any common level of significance for all the series. This implies that
the series at their levels are non-stationary. Said differently, they
have unit roots at their levels. These findings hold for both
sub-periods.
The KPSS test statistics [[eta].sub.u] and [[??].sub.[tau]] are
estimated to test the null hypothesis of stationarity against the
alternative hypothesis that the series contains a unit root with and
without a linear time trend, respectively. Since the estimated test
statistics, [[eta].sub.u] and [[??].sub.[tau]], are greater than the
critical values for all the said series, we reject the null hypothesis
of stationarity in favour of the alternative hypothesis of unit root.
That is, all the series at their levels have unit roots. The KPSS unit
root test results confirm the results of the ADF unit root test. Since
the first differences of the series under study appear stationary, we
conclude that all the series are integrated of order one (i.e. I(1)).
(9)
4.2. The Linear Relationship between Capital Inflows and Domestic
Prices
To examine the short- and long-run relationship between capital
inflows and the price level, we apply cointegration and Granger
causality tests. The results from multivariate Johansen's
cointegration procedure for the first sub-period (January 1990-December
2000) as well as for the second sub-period (January 2001-June 2012) are
given in Tables 4 and 5, respectively.
We use three different measures, namely, the net foreign assets to
GDP ratio, the foreign reserves to GDP ratio, and the capital account
surplus to GDP ratio, as proxies for foreign capital inflows.
Accordingly, the four models are estimated using a set of other control
variables, which vary from model to model, to explore the impact of
capital inflows on the price level. The estimates provide strong
evidence of the existence of, at least one cointegrating vector. The
existence of the long-run relationship holds for all models. This
indicates that the cointegration results that we report here are robust
to different proxies for foreign capital inflows and to different
specifications. The results also suggest that evidence about the
presence of long-run relationship between foreign capital inflows and
domestic prices holds for both sub-periods. This implies that foreign
capital inflows and domestic price levels are integrated (in
cointegration sense) during periods of small as well as massive capital
inflows.
The results given in Tables 4 and 5 suggest that there is a
long-run co-movement between domestic prices and capital inflows. These
findings imply that capital inflows are significant in determining price
levels in the host economy. A possible explanation for the existence of
a significant relationship between foreign capital inflows and domestic
price levels is that the surge in foreign capital inflows not only
increases the monetary base of the economy but also increases the
aggregate expenditures in the recipient economy. Consequently, the price
level would increase in the economy. The capital inflows may also
positively affect domestic prices if they are caused by an exogenous
growth in productivity of domestic capital or/and by a drop in interest
rate in foreign money markets. These findings are in accordance with
several previous empirical studies including Kim and Yang (2009, 2011).
Sayek (2009), Rashid (2010). Bernanke (2010), Nazir. el at. (2012), and
Tillmann (2013) that document a significant association between foreign
capital inflows and prices.
4.3. The Response of Domestic Prices to Capital Inflow Shocks
We estimate impulse response functions (IRFs) as an additional
check of the cointegration test's findings. Order and Fisher
(1993), Cholesk-type of contemporaneous identifying restrictions are
employed to draw a meaningful interpretation. The recursive structure
assumes that variables appearing first contemporaneously influence the
latter variables but not vice versa. It is important to list the
exogenous variables earlier than the endogenous variables.
Impulse response functions for the first and second sub-periods are
presented in Figures 1 and 2 given in the annexure, respectively. The
response is considered significant if confidence intervals do not pass
through zero line. For both the periods, the directions of changes
observed in the impulse responses are according to economic theory. For
the first sub-period, the immediate and permanent effect of a one
standard deviation shock to net foreign reserves on domestic price
levels is positive. The effect of a one standard deviation shock to the
ratio of money supply to GDP on price levels is negative in the
short-run; however, it is positive in the long run. The graphs also
reveal that the money market rate, the nominal exchange rate,
manufacturing output, and the capital account surplus to GDP ratio do
not have any significant long run effect on domestic prices.
For the second sub-period, the net effect on domestic price levels
of a one standard deviation shock to the ratio of foreign assets to GDP,
the ratio of money supply to GDP, and the change in level of domestic
debt is positive in the short run as well as in the long run. One the
other hand, we observe that a one standard deviation shock to the ratio
of capital account surplus to GDP has a positive effect initially but
the permanent effect is negative. Impulse response functions confirm the
findings of cointegration tests that there exists a long-run equilibrium
relationship between foreign capital inflows and domestic price levels.
After confirming the existence of the long-run relationship
(cointegration) between foreign capital inflows and domestic price
levels, we explore the short-run dynamics. Since the variables are
cointegrated, using the Vector Error Correction (VEC) model, we test
whether the variables individually Granger cause domestic price levels
in all the four models. For this, we test for the joint significance of
lagged coefficient of each variable along with the error correction
term. The estimated results for the first sub-period are repotted in
Table 6.
One can see from the fable that the null hypothesis of no short-run
Granger causality cannot be rejected for the net foreign assets-to-GDP
ratio as well as for the foreign reserves-to-GDP ratio. This implies
that neither the net foreign assets nor the amount of foreign reserves
significantly cause (in Granger sense) domestic prices during the period
1990-2000. These findings suggest that the foreign capital inflows do
not have causal linkages with the price level during the periods of low
capital inflows. That is, smooth flows of foreign capital do not create
inflationary pressure in the recipient country. This finding is
consistent with the literature that indicates that only large episodes
of foreign capital inflows do matter for the host economy. The results
regarding our third proxy of foreign capital inflows that is the ratio
of capital account to GDP reveal that domestic prices are significantly
Granger caused by foreign capital inflows via capital account surplus.
(10)
The results do not provide any significant evidence of the
rejection of the null hypothesis that domestic price level is not
Granger caused by the money market rate and the exchange rate (neither
the nominal nor the real one) in any estimated model during the period
1990-2000. These observations indicate that the interest rate and the
exchange rate both do not have any significant short-run causal
relationship with domestic price levels. These findings also suggest
that during the period 1990-2000, both interest rate and exchange rate
policies were not effective in controlling inflation in the economy.
The results given in Table 6 also show that the domestic price
level is significantly influenced (in Granger sense) by domestic credit
and money supply. This implies that increases in monetary base of the
economy during the period 1990-2000 have inflated domestic prices.
Likewise, more credit supply to domestic sector has also significantly
and positively contributed to the price level. We find that there is a
significant Granger causality sunning from manufacturing output to
domestic prices. This implies that the level of industrial output has a
significant short-run impact (in Granger sense) on the level of prices.
On the whole, we observe from the results presented in Table 6 that
during January 1990 to December 2000, the capital account to GDP ratio,
the money supply to GDP ratio, and the level of domestic debt
significantly cause the domestic price level. However, we show that the
net foreign assets to GDP ratio, the foreign reserves to GDP ratio, the
money market rate and both nominal and real effective exchange rates do
not significantly influence the rate of inflation. Thus, we can say that
during the period 1990-2000, domestic prices are significantly caused by
domestic macroeconomic factors, namely money supply, domestic credit,
and manufacturing output, instead of foreign capital inflows in the
short run.
The results for the second sub-period spanning January 2001 to June
2012- a period of large capital surge--are given in Table 7. Contrary to
the period of low capital inflows (1990-2000), yet consistent with our
expectation, foreign capital inflows are significantly related to
short-run dynamics of inflation during surges in capital inflows. In
particular, we find strong evidence to reject the null hypothesis of no
Granger causality for net foreign assets in Model I and Model IV. This
implies that domestic prices are significantly Granger caused by the net
foreign assets-to-GDP ratio. There is also significant evidence of the
presence of the short-run causal relationship between the ratio of
foreign reserves to GDP and the price level (see Model II). These
findings suggest that the impacts of foreign capital inflows that we
reported here are robust to different proxies of foreign exchange rate
and different specifications, and thus, any specific proxy or any
particular specification of the model does not drive them.
It is noteworthy that both the proxies for capital inflows, namely
the net foreign assets to GDP ratio and the foreign reserves to GDP
ratio, do not have any short-run causal relationship with domestic price
levels during an episode of smooth capital inflows (January 1990 to
December 2000). Nonetheless, during the period of large capital inflows
ranging from 2001-2012, both of the proxies have a significant impact
(in Granger sense) on domestic price levels, which is what we expect.
This implies that the higher the level of the foreign capital inflows,
the higher the level of the inflation. These findings suggest that the
abrupt increase in foreign capital inflows would not only undermine
central bank's ability to achieve monetary stability but also
increase monetary base, fuel spending booms, and cause asset market
bubbles without benefiting significantly the real sector of the host
economy. Thus, capital inflow bonanzas generate inflationary pressures
in the recipient country. These findings also suggest that policymakers
can provide nominal anchor to the economy by stabilising the dynamics of
foreign capital inflows. Our findings are consistent with those studies
that document that large and sudden capital inflows significantly fuel
domestic credit growth and price levels in host economies.
It is also important to note that although, during the period
1990-2000, capital account surplus to GDP ratio Granger causes domestic
prices, we do not find any significant evidence of the presence of the
short-run causal relationship between capital account surplus and the
price level during 2001-2012. This finding is contrary to the
preliminary evidence provided by correlation estimates that capital
account surplus is significantly related to the price level during both
the sub-periods. Similarly, there is no evidence of the short-run impact
of market interest rate on the price level.
This finding holds for both sub-periods. Further, the results
reveal that consistent with the first sub-period, neither the nominal
nor the real exchange rate is significantly related to the short-run
dynamic of inflation. Finally, we find that domestic debt, manufacturing
output, and money supply significantly Ganger cause domestic price
levels. These findings also hold for both sub-periods, indicating the
persistent inflationary effect of these variables. These findings
suggest that domestic credit growth and industrial output are
significant for controlling inflationary dynamics in Pakistan. However,
both the exchange rate and the money market rate cannot be effectively
used as policy tools for stabilising short-run price dynamics.
4.3. The Effect of Capital Inflows on the Exchange Rate Volatility
In this section, we examine the impact of capital inflows on the
exchange rate volatility. In particular, we investigate the differential
effect of capital inflows on the nominal and real exchange rate
volatility during periods of low and large capital inflows. The
volatility of nominal exchange rate (VNEX) and real effective exchange
rate (VREER) has been calculated by using the three-period moving
average standard
deviation: [S.D..sub.1] = [(l/m [m.summation over (i=1)]
[([EX.sub.t+1-l] - [EX.sub.t+1-2]).sup.2]].sup.1/2], where m = 3 and EX
denotes the
underlying exchange rate series. Before examining the influence of
capital inflows on the exchange rate volatility, we test the order of
integration of generated volatility series. For this, we apply the ADF
and the KPSS unit root tests. The results for both sub-periods are given
in Table 8. The results indicate that both volatility series are
stationary at their levels.
Since the exchange rate volatility series are stationary at their
levels, we estimate the VAR model for testing the short-run Granger
causality between the exchange rate volatility and the change in foreign
capital inflows. The results summarised in Table 9 provide evidence that
both the nominal and real effective exchange rate volatility is
significantly influenced by the change in net foreign reserves during
1990-2000. This implies that during the first sub-period, capital
inflows are significantly related to the short-run dynamic of both
nominal and real exchange rates. Although during this period, the flows
are relatively small and smooth, they play significant role in
determining exchange rate fluctuations. It should be noted that during
this period, foreign capital inflows not only affect the nominal
exchange rate volatility but also the real effective exchange rate
volatility. Thus, in turn, the inflows affecting foreign competitiveness
increase international trade imbalances and escalate vulnerability to a
financial crisis.
When we observe the Granger causality results for the second
sub-period from January 2001 to June 2012, we find that the change in
capital inflows has a significant impact (in Granger sense) on the
volatility of real effective exchange rate. This finding indicates the
persistent effect of capital inflows on the real exchange rate
volatility during both sub-periods. This implies that the real effective
exchange rate volatility is significantly influenced by the inflows of
foreign capital regardless of whether these flows are smooth or of
bonanza nature. The effects of foreign capital inflows on the real
effective exchange rate that we presented here are consistent with the
findings previously reported in the literature [Calvo, et al. (1993),
Bandara (1995), Edwards (1998), Agenor (1998), Chen and Rogoff (2003),
Lartey (2007, 2008), Cashin, et al. (2004), Lee, et al. (2009),
Saborowski (2009), Rashid (2010), and Combes, et al. (2012)]. (11) These
studies also document significant impacts of foreign capital inflows on
real exchange rates. Our findings are also consistent with the view that
ebb and flow of foreign capital inflows deteriorate macroeconomic and
financial management in the recipient countries and overheat the economy
by causing real appreciation. This set of findings suggests that there
is a critical need to adopt more flexible exchange rate policies that
would be useful in dampening the real exchange rate volatility, which
stem from surges in capital flows.
4.4. The Nonlinear Causation between Capital Inflows and Domestic
Prices
In this sub-section, we comprehensively analyse the existence of
nonlinearity in capital inflows-domestic prices nexus. To test a
long-run nonlinear relationship, we run a bi-variate regression of LCPI
on a constant and BOX-COX transform of the underlying explanatory
variable. Specifically, the function is defined as follows:
[LCPI.sub.t] = (([absolute value of [X.sub.t]]).sup.[theta]] -
1))/[theta] ... (7)
where [X.sub.t] denotes the underlying explanatory variable. We use
the nonlinear least squares (NLS) method to estimate the underlying
parameters ([??]), and then apply the KPSS test to the residual to test
the null hypothesis of nonlinear cointegration against an alternative
hypothesis of no nonlinear cointegration. The estimates are given in
Table 10.
The results provide strong evidence of the presence of nonlinear
cointegration between domestic price levels and the net foreign
assets-to-GDP ratio, the money supply-to-GDP ratio, manufacturing
output, and domestic debt in both the examined periods. On the other
hand, the results reveal that there is no significant nonlinear
association between the price level and both market interest and nominal
exchange rates. In particular, we find that the null hypothesis of
nonlinear cointegration between foreign capital inflows and domestic
prices cannot be rejected when we include a linear time trend in the
KPSS test specification. The existence of the long-run nonlinear
relationship between capital inflows and price levels holds for both
sub-periods. This observation suggests that nonlinearity in the capital
inflows--domestic prices nexus is not attributed to the size of the
waves of capital inflows. Rather, this asymmetric association may be
heritable and stem from economic wellsprings.
To examine the nonlinear short-run causality between domestic
prices and the other underlying variables, we use the Hristu-Varsakkeis
and Kyrtsou (2010) nonlinear Granger causality test--known as the
bi-varaite noisy Mackey-Glass model. The first step is to estimate the
nonlinear VEC model (i.e., Equation (4) is estimated using the first
differences of the variables and error correction term by ordinary least
squares, in a specification [[tau].sub.1] = [[tau].sub.1] = 4 and
[c.sub.1] = [c.sub.2] = 2) selected by the Log Likelihood procedure
without and with restriction on lagged parameters of explanatory
variable. We then obtain the residuals to calculate the test statistics
(says [S.sub.1]) for testing nonlinear Granger causality between the
variables. For each variable, we estimate separately the nonlinear VEC
model to examine the nonlinear causal impact on domestic prices of the
underlying variable. We examine the nonlinear short-run causality during
both subperiods. Specifically, we aim to analyse whether the nonlinear
short-run influence of capital inflows on prices depends on the size of
flow of foreign capital inflows. However, for nonlinear Granger
causality analysis, we utilise only the net foreign assets-to-GDP ratio
as foreign capital inflows proxy. Table 11 presents the estimated
[S.sub.l] for both sub-periods.
We do not find any significant evidence of the existence of the
nonlinear short-run causality between foreign capital inflows (the net
foreign assets-to-GDP ratio) and domestic prices during the first
sub-period when capital inflows are relatively smooth and small in size.
During the second sub-period when there are surges in capital inflows,
however, domestic price levels are significantly nonlinearly Granger
caused by foreign capital inflows. This implies that the nonlinear
short-run association between the price level and foreign capital
inflows is asymmetric, depending on the amount of capital inflows. These
findings are similar to our earlier findings of linear Granger causality
tests--Granger causality running from capital inflows to domestic prices
only for the period of massive capital inflows.
Results regarding other variables indicate that there is a
significant nonlinear Granger causality running from the level of
domestic debt, manufacturing output, and the money supply to GDP ratio
to the rate of inflation. These results hold for both sample periods,
suggesting the persistence in nonlinear short-run inter-linkages across
low and high capital inflow regimes. In other words, ebbs and flows of
foreign capital do not affect the nonlinear association between domestic
prices, domestic debt, manufacturing output, and money supply to GDP
ratio. Finally, we do not find significant evidence of the nonlinear
Granger causality running from the market interest rate as well as the
nominal exchange rate to the level of price in either period.
Several striking findings emerge from the evidence presented here.
First, although the long-run linear and nonlinear association between
foreign capital inflows and domestic price levels is independent of the
size of foreign capital inflows, the short-run linear and nonlinear
Granger causality exists merely during surges in capital inflows.
Second, the causal impact on the level of price of domestic factors,
namely money supply, manufacturing output, and domestic credit growth is
robust regardless of whether foreign capital inflows are in small amount
or of bonanza nature. Third, both the market interest rate and the
exchange rate do not have any causal influence (in Granger sense) on
domestic prices. Fourth, pronounced waves of foreign capital inflows
significantly fuel the real effective exchange rate volatility. The
significant influences of foreign capital inflows on domestic prices and
the exchange rate volatility provide indication of so called
"transfer problem"--which generally refers to the effect of
foreign capital movements on the recipient economy. Our findings suggest
that exchange rate flexibility and effective absorption and
sterilisation of foreign capital inflows are necessary to penalise
destructive capital inflows and lessen inflationary effects of capital
inflows in the host economy. These measures, in turn, would be
significant in dampening financial system vulnerability originating from
surges in capital inflows.
5. CONCLUSIONS AND POLICY IMPLICATIONS
This paper has empirically investigated the inflationary effects of
foreign capital inflows for Pakistan using monthly data covering the
period from January 1990 to June 2012. To provide economic intuition,
the paper has also proposed an empirical model of the equilibrium prices
based on the standard classical quantity theory of demand for money
subject to capital inflows. Further, we have divided the full sample
into two subsamples to study the differential effects of capital inflows
on the price level across the low and high episodes of capital inflows.
Our empirical results suggest that there is a positive and
significant impact of foreign capital inflows (in Granger sense) on
domestic price levels, particularly, during the periods of massive
capital inflows from 2001 to 2012. Our results, however, suggest lack of
causality between capital inflows and domestic price level for the
period 1990-2000. Besides the existence of linear causation between
capital inflows and price levels, we find significant evidence of
nonlinear Granger causality running from capital inflows to the rate of
inflation. This implies that hikes in domestic price levels are not only
linearly but also nonlinearly caused by changes in foreign capital
inflows. The presence of nonlinearity in capital inflows-domestic prices
linkages that we have unfolded in this paper would definitely provide
new insights about the existence of causal links between the price level
and capital inflows. We also show that both the market interest and the
exchange rate do not have any cause-effect relationship with the rate of
inflation in either period. Finally, we find that foreign capital
inflows have significant causal linkages with the exchange rate
volatility. Our analysis suggests that the influence of capital inflows
on the real effective exchange rate volatility holds during both low and
high flow of capital inflows.
From the policy perspective, the findings are of particular
interest to the government authority and the SBP. Since the capital
inflows have played a significant role to push up domestic prices,
particularly during the period of capital inflows surges (2001-2012),
the foreign exchange management policy of SBP is questionable. The
findings suggest that there is a need to absorb the capital inflows in
such a way that they should neither create an inflationary pressure in
the economy nor fuel the exchange rate volatility. More precisely, the
SBP should put the limit to arbitrate in the forex market and should
allow the private sector to use the foreign capital for productive
purposes to increase the production in the economy, rather than just to
add it to government foreign reserves. This policy can prevent the
economy from overheating and dampen financial fragility.
The most effective ways to deal with capital inflows would be to
deepen the financial markets, strengthen financial system supervision
and regulations, where needed, and improve the capacity to design and
implement sound macroeconomic and financial sector policies. These
actions would certainly help increase the absorption capacity and
resilience of the economy and financial systems to the risk associated
with the inflows. The analysis may establish a useful base for future
empirical work in this field and suggest that researchers should also
consider nonlinearity in modelling to test the influence of surges in
capital inflows on inflationary dynamics. We have unambiguously linked
foreign capital inflows to consumer prices and the exchange rate
volatility in both linear and nonlinear causality terms. It would also
be enlightening to know how capital inflows and outflows differently
affect asset price dynamics, in particular, house price inflation.
ANNEXURE
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
REFERENCES
Adams, S. (2009) Foreign Direct Investment, Domestic Investment,
and Economic Growth in Sub-Saharan Africa. Journal of Policy Modeling
31:6, 939-949.
Agenor, P. R. (1998) Capital Inflows, External Shocks, and the Real
Exchange Rate. Journal of International Money and Finance 17:5, 713-40.
Aghion, P., P. Bacchetta, and A. Banerjee (2000) Capital Markets
and the Instability of Open Economies. In Pierre-Richard Agnor, Marcus
Miller, David Vines, and Axel Weber (eds.) The Asian Financial Crisis:
Causes, Contagion and Consequences. Cambridge: Cambridge University
Press. 167-194.
Aizenman, J. and R. Glick (2009) Sterilisation, Monetary Policy,
and Global Financial Integration. Review of International Economics
17:4, 777-801.
Alfaro, L., A. Chanda, S. Kalemli-Ozcan, and S. Sayek (2004) FDI
and Economic Growth: The Role of Local Financial Markets. Journal of
International Economics 64:1, 89-112.
Azman-Saini, W., S. H. Law, and A. H. Ahmad (2010) FDI and Economic
Growth: New Evidence on the Role of Financial Markets. Economics Letters
107:2, 211-213.
Baharumshah, A. Z. and M. A. M. Thanoon (2006) Foreign Capital
Flows and Economic Growth in East Asian Countries. China Economic Review
17:1, 70-83.
Bandara, J. S. (1995) Dutch Disease in a Developing Country: The
Case of Foreign Capital Inflows to Sri Lanka. Seoul Journal of Economics
8, 314-29.
Bernanke, B. S. (2005) The Global Saving Glut and the US Current
Account Deficit. Sandridge Lecture Speech.
Bernanke, B. S. (2010) Monetary Policy and the Housing Bubble.
Speech at the Annual Meeting of the American Economic Association,
Atlanta, January 3.
Buch, C. M., J. Kleinert, A. Lipponer, and F. Toubal (2005)
Determinants and Effects of Foreign Direct Investment: Evidence from
German Firm-level Data. Economic Policy 20:41, 52-110.
Caballero, J. (2012) Do Surges in International Capital Inflows
Influence the Likelihood of Banking Crises? Cross-country Evidence on
Bonanzas in Capital Inflows and Bonanza-boom-bust Cycles. (IDB Working
Paper No. IDB-WP-305).
Caballero, R. J., E. Farhi, and P. O. Gourinchas (2008) Financial
Crash, Commodity Prices and Global Imbalances. National Bureau of
Economic Research, Inc. (Working Paper).
Caballero, R. J. and A. Krishnamurthy (2006) Bubbles and Capital
Flow Volatility: Causes and Risk Management. Journal of Monetary
Economics 53:1, 35-53.
Calvo, G., L. Leiderman, and C. Reinhart (1993) Capital Inflows and
Real Exchange Rate Appreciation in Latin America: The Role of External
Factors. IMF Staff Papers 40:1, 108-51.
Campbell, J. Y. and P. Perron (1991) Pitfalls and Opportunities:
What Macroeconomists Should Know about Unit Roots. NBER Macroeconomics
Annual 6, 141-220. MIT Press.
Cardarelli, R., S. Elekdag, and M. A. Kose (2010) Capital Inflows:
Macroeconomic Implications and Policy Responses. Economic Systems 34:4,
333-356.
Carrion-i-Silvestre, J. L. and A. Sanso (2006) Testing the Null of
Cointegration with Structure Break. Oxford Bulletin of Economics and
Statistics 68:5, 623-646. Cashin, P., L. Cespedes, and R. Sahay (2004)
Commodity Currencies and the Real Exchange Rate. Journal of Development
Economics 75:1, 239-68.
Cecen, A. and L. Xiao (2012) Capital Flows and Current Account
Dynamics in Turkey: A Nonlinear Time Series Analysis. Society of Policy
Modeling. (Working Paper).
Combes, J., T. Kinda, and P. Plane (2012) Capital Inflows, Exchange
Rate Flexibility, and the Real Exchange Rate. Journal of Macroeconomic
34:4, 1034-1043.
Chen, Yu-Chin and K. Rogoff (2003) Commodity Currencies. Journal of
International Economics 60:1, 133-60.
Choong, C. K., S. Y. Lam, and Z. Yusop (2010) Private Capital Flows
to Low Income Countries: The Role of Domestic Financial Sector. Journal
of Business Economics and Management 11:4, 598-612.
De Gregorio, J. (2012) Living with Capital Inflows. Bank for
International Settlements. Speach at the LACEA meetings in Santiago,
Chile.
De Mello, L. R. (1996) Foreign Direct Investment, International
Knowledge Transfers, and Endogenous Growth: Time Series Evidence.
Department of Economics, University of Kent.
De Mello, L. R. (1997) Foreign Direct Investment in Developing
Countries and Growth: A Selective Survey. Journal of Development Studies
34:1, 1-34.
Edwards, S. (2007a) Capital Controls, Capital Flow Contractions,
and Macroeconomic Vulnerability. Journal of International Money and
Finance 26:5, 814-840.
Edwards, S. (2007b) Capital Controls, Sudden Stops, and Current
Account Reversals. Capital Controls and Capital Flows in Emerging
Economies: Policies, Practices and Consequences. 73-120. University of
Chicago Press.
Edwards, S. (1998) Capital Flows, Real Exchange Rates, and Capital
Controls: Some Latin American Experiences. National Bureau of Economic
Research, Inc. (NBER Working Papers 6800).
Eichengreen, B. (2004) Capital Flows and Crises. MIT Press.
Forbes, K. J. and F. E. Wamock (2012) Capital Flow Waves: Surges,
Stops, Flight, and Retrenchment. Journal of International Economics
88:2, 235-251.
Furceri, D., S. Guichard, and E. Rusticelli (2011) Episodes of
Large Capital Inflows and the Likelihood of Banking and Currency Crises
and Sudden Stops. OECD.
Furceri, D., S. Guichard, and E. Rusticelli (2012) The Effect of
Episodes of Large Capital Inflows on Domestic Credit. The North American
Journal of Economics and Finance 23:3, 325-344.
Glick, R. and M. Hutchison (2009) Navigating the Trilemma: Capital
Flows and Monetary Policy in China. Journal of Asian Economics 20:3,
205-224.
Hegerty, S. W. (2009) Capital Inflows, Exchange Market Pressure,
and Credit Growth in Four Transition Economies with Fixed Exchange
Rates. Economic Systems 33:2, 155-167.
Hermes, N. and R. Lensink (2003) Foreign Direct Investment,
Financial Development and Economic Growth. Journal of Development
Studies 40:1, 142-163.
Hristu-Varsakelis, D. and C. Kyrtsou (2010) Testing for Granger
Causality in the Presence of Chaotic Dynamics. Brussels Economic Review
53:2, 323-327.
Hussain, M., A. Berg, and S. Aiyar (2009) The Macroeconomic
Management of Increased Aid: Policy Lessons from Recent Experience.
Review of Development Economics 13:3, 491-509.
Kim, S. and D. Y. Yang (2011) The Impact of Capital Inflows on
Asset Prices in Emerging Asian Economies: Is Too Much Money Chasing Too
Little Good? Open Economies Review 22:2, 293-3 15.
Kim, S. and D. Y. Yang (2009) Do Capital Inflows Matter to Asset
Prices? The Case of Korea. Asian Economic Journal 23:3, 323--348.
Kose, M. A., E. Prasad, K. Rogoff, and S. J. Wei (2009) Financial
Globalisation: A Reappraisal. IMF Staff Papers 56:1, 8-62.
Kwiatkowski, D., P. C. B. Phillips, P. Schmidt, and Y. Shin (1992)
Testing the Null Hypothesis of Stationarity against the Alternative of a
Unit Root: How Sure Are We that Economic Time Series Have a Unit Root?
Journal of Econometrics 54:1, 159-178.
Kyrtsou, C. and W. C. Labys (2006) Evidence for Chaotic Dependence
between US Inflation and Commodity Prices. Journal of Macroeconomics
28:1, 256-266.
Kyrtsou, C. and W. C. Labys (2007) Detecting Positive Feedback in
Multivariate Time Series: The Case of Metal Prices and US Inflation.
Physica A: Statistical Mechanics and its Applications 377:1, 227-229.
Kyrtsou, C. and M. Terraza (2010) Seasonal Mackey-Glass-GARCH
Process and Short-term Dynamics. Empirical Economics 38:2, 325-345.
Kyrtsou, C. and C. Vorlow (2009) Modelling Non-linear Comovements
between Time Series. Journal of Macroeconomics 31:1, 200-211.
Laibson, D. and J. Mollerstrom (2010) Capital Flows, Consumption
Booms and Asset Bubbles: A Behavioural Alternative to the Savings Glut
Hypothesis. The Economic Journal 120:544, 354-374.
Lartey, E. K. K. (2007) Capital Inflows and the Real Exchange Rate:
An Empirical Study of Sub-Saharan Africa. Journal of International Trade
and Economic Development 16:3, 337-57.
Lartey, E. K. K. (2008) Capital Inflows, Dutch Disease Effects and
Monetary Policy in a Small Open Economy. Review of International
Economics 16:5, 971-989.
Lee, J., G. M. Milesi-Ferretti, and L. A. Ricci (2008) Real
Exchange Rates and Fundamentals: A Cross-country Perspective. (IMF
Working Paper 08/13, IFM).
Li, Y. and F. Rowe (2007) Aid Inflows and the Real Effective
Exchange Rate in Tanzania. (World Bank Policy Research Working Paper No.
4456, BW).
Lin, J. and C. Granger (2004) Testing Nonlinear Cointegration. In
COMPSTAT'2004 Symposium in Computational Statistics, ed.
Physica-Verlag/Springer 2004, 1413-1419.
Mendoza, E. G. and M. E. Terrones (2008) An Anatomy of Credit
Booms: Evidence from Macro Aggregates and Micro Data. National Bureau of
Economic Research. (Working Paper).
Mongardini, J. and B. Rayner (2009) Grants, Remittances, and the
Equilibrium Real Exchange Rate in Sub-Saharan African Countries.
(Working Paper 09/75, IMF).
Nazir, S., N. Sarwar, and S. Ullha (2012) Impact of Capital Inflows
on Domestic Inflation: A Case Study of Pakistan. Journal of Asian
Development Studies 1:1, 4-12.
Obstfeld, M. (2009) International Finance and Growth in Developing
Countries: What Have We Learned and Quest. IMF Staff Papers 56:1,
63-111.
Obstfeld, M., J. C. Shambaugh, and A. M. Taylor (2005) The Trilemma
in History: Tradeoffs among Exchange Rates, Monetary Policies, and
Capital Mobility. Review of Economics and Statistics 87:3, 423-438.
Orden, D., and L. A. Fisher (1993) Financial Deregulation and the
Dynamics of Money, Prices, and Output in New Zealand and Australia.
Journal of Money, Credit, and Banking 25:2, 273-292.
Rashid, A. (2013) The Response of Inflation to Its Determinants:
Does It Differ at Short and Long Horisons? International Journal of
Economics and Business Research 6:2, 229-242.
Rashid, A. (2010) Capital Inflows and Exchange Rate Volatility in
Pakistan. International Review of Applied Economic Research 5:1,
109-116.
Rashid, A. and F. Husain (2010) Capital Inflows, Inflation and
Exchange Rate Volatility: An Investigation for Linear and Nonlinear
Causal Linkages. Pakistan Institute of Development Economics, Islamabad,
Pakistan. (Working Paper No. 2010: 63).
Reinhart, C. M. and V. R. Reinhart (2008) Capital Flow Bonanzas: An
Encompassing View of the Past and Present. National Bureau of Economic
Research. (Working Paper).
Reisen, H. and M. Soto (2002) Which Types of Capital Inflows Foster
Developing-country Growth? International Finance 4:1, 1-14.
Rodrik, D. and A. Velasco (1999) Short-term Capital Flows. National
Bureau of Economic Research. (Working Paper).
Saborowski, C. (2009) Capital Inflows and the Real Exchange Rate:
Can Financial Development Cure the Dutch Disease? (IMF Working Paper
09/20, IMF).
Sayek, S. (2009) Foreign Direct Investment and Inflation. Southern
Economic Journal 76:2, 419-443.
Sethi, N. and S. Sucharita (2012) Effect of FDI on Economic Growth
in Bangladesh and India: An Empirical Investigation. Society of Policy
Modeling. (Working Paper).
Tillmann (2013) Capital Inflows and Asset Prices: Evidence from
Emerging Asia. Journal of Banking and Finance 37:3, 717-729.
Ventura, J. (2002) Bubbles and Capital Flows. National Bureau of
Economic Research. (Working Paper).
Wang, M. and M. C. S. Wong (2009) Foreign Direct Investment and
Economic Growth: The Growth Accounting Perspective. Economic Inquiry
47:4, 701-710.
(1) However, Laibson and Mollerstorm (2010) show that instead of an
excessively abundant supply of global savings, mismatch of international
balances is mainly the result of domestic consumption booms and national
asset bubbles.
(2) See Obstfeld, et al. (2005), Reinhart and Reinhart (2008),
Click and Hutchison (2009), Aizenman and Glick (2009), Cardarelli, et
at. (2010), and De Gregorio (2012) for effective policy measures in
response to capital flow bonanzas.
(3) Other things remain constant, capital inflows owing to
"pull" factors will cause an upward pressure on domestic
interest rates, whereas, capital inflows caused by "push"
factors, such as a fall in international interest rates, will have a
tendency to put downward pressure on domestic interest rates on one
hand. On the other hand, it will initially drive up nominal and real
balances, but then, as domestic price level increases, real balances may
decline. See, Rashid and Husain (2010) for the potential differential
effects of capital inflows caused by "pull" and
"push" factors on financial indicators
(4) See data Section 4 of the paper for definition of the
variables.
(5) The breakdown of the whole sample is based on the flow of
foreign capital inflows, as our main objective is to analyse the
differential effect on domestic price levels and the exchange rate
volatility of foreign capital inflows across low and large flows.
However, one should note that the objective of our study is not to test
apparently the presence of structure break in the capital
inflows-domestic prices relationship, for testing the possibility of
structure breaks, a separate comprehensive analysis is required. One may
extend our analysis along these lines by applying sophisticated
econometric techniques such as Carrion-i-Silvestre and Sanso (2006).
(6) We can understand that prices in a country such as Pakistan are
not fully determined by market forces. They are commonly twisted by
providing subsidies and setting ceiling and floor price. However, we do
not consider government distortion in price determination in order to
keep our model simple.
(7) Lin and Granger (2004) argue that if the null hypothesis is
specified as cointegration, then the LOSS test would give the right
distribution under the null hypothesis and power approaching one as
sample size grows under the alternative.
(8) Here, the domestic debt includes claims on general government
(net), claims on non-financial public enterprises, claims on private
sector, and claims on nonblank financial institutions.
(9) The unit root test results for first differences of the
variables are not given here to economise the space. However, are
available from authors.
(10) This differential causal impact across different proxies of
foreign capital inflows suggests that it would be worth exploring the
impact of different components of foreign capital inflows such as
foreign direct investment (FDI), foreign portfolio investment (FPI),
foreign bank borrowing, remittances, etc. on domestic price levels.
Further, it would also be useful to investigate the differential effects
of private versus public foreign inflows on host economies. However, one
should note that we do not extend our analysis along these lines in
order to emphasise more on the objectives of our study.
(11) Our findings regarding the effects on the exchange rate
volatility of capital inflows are, however, inconsistent with Li and
Rowe (2007), Mongardini and Rayner (2009), and Hussain, et al. (2009),
who show that official foreign capital inflows are not significantly
associated with the real effective exchange rate.
Authors' Note: We are grateful to anonymous reviewers for
useful comments and suggestions. We are also grateful to the Editor. The
early version of the paper has been published as PIDE working paper.
Abdul Rashid <abdulrashid@iiu.edu.pk> is Assistant Professor,
International Institute of Islamic Economics (IIIE), International
Islamic University (IIU), Islamabad. Fazal Husain
<fazal@pide.org.pk> is Head, Department of Economics, Pakistan
Institute of Development Economics (PIDE), Islamabad.
Table 1
Correlation Coefficients; Sample Period:
January 1990 to December 2000
Ratio Series
Variables CAR FAR FRR MSR
FAR -0.260#
FRR 0.130 0.648#
MSR -0.322# 0.047 0.125
LCPI 0.316# -0.023 0.130 -0.098
MMR -0.203 0.095 0.040 0.246#
LNER -0.037 -0.066 -0.129 0.022
LDC 0.182 -0.249 -0.040 -0.003
LMPI 0.326# -0.076 0.166# 0.165
First Difference of Series
Variables LCPI MMR LNER LDC
FAR
FRR
MSR
LCPI
MMR -0.089
LNER -0.026 -0.040
LDC 0.011 0.229 0.239#
LMPI 0.395# 0.314# 0.403# 0.415#
Note: Bold values indicate that the correlation is
significantly different from zero at the 5 percent level.
MMR = the market interest rate, LNER = the log of nominal
exchange rate, LCPI = the log of consumer price index, FAR =
net foreign assets-GDP ratio, CAR = capital account-GDP
ratio, FRR = the ratio of foreign reserves to GDP ratio,
LMPI = the log of manufacturing production index, MSR = the
ratio of money supply to GDP, LDC = the log of domestic
credit.
Note: Bold values indicate that the correlation is
significantly different from zero at the 5 percent level
are indicated with #.
Table 2
Correlation Coefficients; Sample Period:
January 2001 to June 2012
Variables Ratio Series
CAR FAR FRR MSR
FAR 0.436#
FRR 0.121 0.963#
MSR 0.763# 0.834# 0.827#
LCPI 0.439# 0.509# 0.483# 0.354#
MMR 0.561# -0.361# -0.531# -0.009
LNER 0.045 -0.128 -0.116 -0.071
LDC 0.283# 0.365# 0.358# 0.523#
LMPI 0.639# 0.708# 0.583# 0.472#
Variables First Difference of Series
LCPI MMR LNER LDC
FAR
FRR
MSR
LCPI
MMR -0.023
LNER -0.124 -0.036
LDC 0.076 0.132 0.007
LMPI 0.677# 0.537# 0.556# 0.693#
Note: Bold values indicate that the correlation is
significantly different from zero at the 5 percent level.
MMR = the market interest rate, LNER = the log of nominal
exchange rate, LCPI = the log of consumer price index, FAR =
net foreign assets-GDP ratio, CAR = capital account-GDP
ratio, FRR = the ratio of foreign reserves to GDP ratio,
LMPI = the log of manufacturing production index, MSR = the
ratio of money supply to GDP, LDC = the log of domestic
credit.
Note: Bold values indicate that the correlation is
significantly different from zero at the 5 percent level
are indicated with #.
Table 3
Unit Root Test Results for Level Series
January 1990 to December 2000
ADF KPSS
[t.sub. [t.sub. [LM.sub. [LM.sub.
Series ADF(c)] ADF(c+1)] KPSS(c)] KPSS(c+1)]
FAR -2.456 -2.708 0.516 0.197
FRR -2.156 -2.205 1.013 0.589
CAR -1.074 -3.152 2.446 0.218
MSR -2.193 -1.179 0.960 0.341
LCPI -2.203 -0.574 2.732 0.542
LMMR -1.668 -1.552 0.517 0.235
LNER 0.205 -3.429 2.720 0.224
LREER -1.137 -3.726 1.932 0.471
LDC -2.251 -0.938 2.679 0.505
LMPI 1.325 -0.796 1.295 0.640
January 2001 to June 2012
ADF KPSS
[t.sub. [t.sub. [LM.sub. [LM.sub.
Series ADF(c)] ADF(c+1)] KPSS(c)] KPSS(c+1)]
FAR -2.570 -1.561 1.254 0.447
FRR -2.037 -1.278 1.013 0.589
CAR -0.071 -2.126 1.815 0.385
MSR -1.399 -3.120 1.682 0.239
LCPI 2.430 -2.076 2.166 0.477
LMMR -1.955 -2.244 0.610 0.404
LNER -2.142 -2.129 0.522 0.532
LREER -1.982 -2.091 0.581 0.407
LDC 2.568 -2.381 2.135 0.451
LMPI -1.087 -1.963 1.982 0.521
Notes: [t.sub.ADF(c)] and [t.sub.ADF(c+1)] are the standard
ADF tcsl statistics for the null of non-stationarity of the
variable in the study without and with a trend,
respectively, in the model for testing. [LM.sub.KPSS-(c)]
and [LM.sub.KPSS(c+1)] are the KPSS test statistics for the
null of stationarity of the variable in the study without
and with a trend, respectively, in the model for testing.
MMR = the market interest rate, LNER = the log of nominal
exchange rate, LREER = the log of real effective exchange
rate, LCPI = the log of consumer price index, FAR = net
foreign assets-GDP ratio, CAR = capital account-ODP ratio,
FRR = the ratio of foreign reserves to GDP ratio, LMPI = the
log of manufacturing production index, MSR = the ratio of
money supply to GDP, LDC -the log of domestic credit
Table 4
Results from Multivariate Johansen's Cointegration Tests
(January 1990 to December 2000)
Model I Model II
[[lambda]. [[lambda]. [[lambda]. [[lambda].
Null Hypothesis sub.max] sub.Trace] sub.max] sub.Trace]
r = 0 31.36 * 66.95 * 39.63 * 104.50 *
r [less than or 21.11 35.59 * 27.31 64.87 *
equal to] 1
r [less than or 9.00 14.48 23.86 17.57
equal to] 2
r [less than or 5.48 5.48 9.53 13.71
equal to] 3
r [less than or -- -- 1.17 4.17
equal to] 4
Model III Model IV
[[lambda]. [[lambda]. [[lambda]. [[lambda].
Null Hypothesis sub.max] sub.Trace] sub.max] sub.Trace]
r = 0 41.50 * 84.93 * 51.94 * 126.12 *
r [less than or 18.80 43.43 * 31.52 * 74.17 *
equal to] 1
r [less than or 17.38 24.62 21.83 42.65 *
equal to] 2
r [less than or 7.25 7.25 1 1.77 20.82
equal to] 3
r [less than or -- -- 9.05 9.05
equal to] 4
Note: * Denotes the rejection of the hypothesis
at the 1 percent level of significance.
Table 5
Results from Multivariate Johansens Cointegration
Tests (January 2001 to June 2012)
Model I Model II
[[lambda]. [[lambda]. [[lambda]. [[lambda].
Null Hypothesis sub.max] sub.Trace] sub.max] sub.Trace]
r = 0 32.58 * 14.62 * 44.16 * 76.43
r [less than or 13.21 15.01 38.48 * 49.62 *
equal to] 1
r [less than or 11.54 13.76 15.14 26.12
equal to] 2
r [less than or 0.02 0.02 10.09 11.80
equal to] 3
r [less than or -- -- 0.06 0.06
equal to] 4
Model III Model IV
[[lambda]. [[lambda]. [[lambda]. [[lambda].
Null Hypothesis sub.max] sub.Trace] sub.max] sub.Trace]
r = 0 37.85 * 65.72 * 46.66 * 95.85 *
r [less than or 17.94 36.14 * 19.57 23.76
equal to] 1
r [less than or 13.63 14.98 13.38 17.49
equal to] 2
r [less than or 0.98 0.98 9.62 10.37
equal to] 3
r [less than or -- -- 0.83 0.83
equal to] 4
Note: * Denotes the rejection of the hypothesis
at the I percent level of significance.
Table 6
Linear Granger Causality Test Results for
January 1990 to December 2000
Number [chi
of square] Decision (at
Null Hypothesis Lags Square the 5% level)
Model 1: LCPI = f (FAR, LMMR, LMPI)
LCPI is not Granger caused by FAR 3 3.089 Do not reject
LCPI is not Granger caused by MMR 3 2.356 Do not reject
LCPI is not Granger caused by LMPI 3 9.178 Reject
Model II: LCPI = f (FRR, LMMR, MSR,
LMPI)
LCPI is not Granger caused by FRR 3 0.129 Do not reject
LCPI is not Granger caused by MMR 3 3.188 Do not reject
LCPI is not Granger caused by MSR 3 10.769 Reject
LCPI is not Granger caused by LMPI 3 12.994 Reject
Model III: LCPI = f (CAR, LDC, LNER)
LCPI is not Granger caused by CAR 3 7.908 Reject
LCPI is not Granger caused by LDC 3 10.232 Reject
LCPI is not Granger caused by LNER 3 1.150 Do not reject
Model IV: LCPI = f (FAR, LMMR, LDC,
LREER)
LCPI is not Granger caused by FAR 3 4.115 Do not reject
LCPI is not Granger caused by LDC 3 21.699 Reject
LCPI is not Granger caused by MMR 3 5.020 Do not reject
LCPI is not Granger caused by 3 1.808 Do not reject
LREER
Note: MMR = the market interest rate, LNER = the log of
nominal exchange rate, LREER = the log of real effective
exchange rate, LCPI = the log of consumer price index, FAR =
net foreign assets-GDP ratio, CAR = capital account-GDP
ratio, FRR = the ratio of foreign reserves to GDP ratio,
LMP1 = the log of manufacturing production index, MSR = the
ratio of money supply to GDP, LDC = the log of domestic
credit.
Table 7
Linear Granger Causality Test Results for January 2001 to June 2012
Number [chi Decision
of square] (at the
Null Hypothesis Lags Square 5% level)
Model 1: LCPI = f (FAR, LMMR, LMPI)
LCPI is not Granger caused by FAR 2 6.726 Reject
LCPI is not Granger caused by LMMR 2 0.638 Do not reject
LCPI is not Granger caused by LMPI 2 8.076 Reject
Model 11: LCPI = f (FRR, LMMR, MSR,
LMPI)
LCPI is not Granger caused by FRR 2 6.326 Reject
LCPI is not Granger caused by LMMR 2 1.175 Do not reject
LCPI is not Granger caused by MSR 2 8.254 Reject
LCPI is not Granger caused by LMPI 2 9.984 Reject
Model III: LCPI = f (CAR, LDC, LNER)
LCPI is not Granger caused by CAR 2 2.637 Do not reject
LCPI is not Granger caused by LDC 2 16.609 Reject
LCPI is not Granger caused by LNER 2 1.487 Do not reject
Model IV: LCPI = f (FAR, LMMR, LDC,
LREER)
LCPI is not Granger caused by FAR 2 13.980 Reject
LCPI is not Granger caused by LDC 2 10.721 Reject
LCPI is not Granger caused by LMMR 2 1.843 Do not reject
LCPI is not Granger caused by LREER 2 1.654 Do not reject
Note: MMR = the market interest rate, LNER = the log of
nominal exchange rate, LREER = the log of real effective
exchange rate, LCPI = the log of consumer price index, FAR =
net foreign assets-GDP ratio, CAR = capital account-GDP
ratio, FRR = the ratio of foreign reserves to GDP ratio.
LMPI = the log of manufacturing production index. MSR = the
ratio of money supply to GDP, LDC = the log of domestic
credit.
Table 8
Unit Root Test Results: The Exchange Rate Volatility
January 1990 to January 2001 to
December 2000 June 2012
Volatility Series ADF K.PSS ADF K.PSS
VNEX -5.469 * 0.484 * -3.654 * 0.312 *
VREER -7.100 * 0.119 * -5.783 * 0.453 *
* Indicates the series is stationary at the 1 percent level.
Table 9
Granger Causality Test Results: Capital Inflows and the
Exchange Rate Volatility
January 1990 to December 2000
[chi square]- Decision
Square (at the 5% level)
Direction of Causality
Afar [right arrow] vnex 7.579 (3) Do not reject
AFAR [right arrow] VREER 8.776 (3) Do not reject
January 2001 to June 2012
[chi square]- Decision
Square (at the 5%
Direction of Causality level)
Afar [right arrow] vnex 0.930 (2) Reject
AFAR [right arrow] VREER 8.546 (2) Do not reject
Note: Here the arrow points out the direction of causality.
Values in parentheses are optimal lag-length selected by the
AIC.
Table 10
Pairwise Nonlinear Cointegration Test Results
Sample Period:
January 1990 to December 2000
Variables included in
Cointegration Equation [LM.sub.KPSS(c)] [LM.sub.KPSS(c+1)]
LCPI and FAR 1.286 0.102 *
LCP1 and LDC 0.107 ** 0.098 *
LCPI and MSR 1.261 0.137 **
LCPI and LMPI 0.187 * 0.121 *
LCPI and MMR 1.412 0.238
LCPI and LNER 1.167 0.546
Sample Period:
January 2001 to June 2012
Variables included in
Cointegration Equation [LM.sub.KPSS(c)] [LM.sub.KPSS(c+1)]
LCPI and FAR 1.329 0.132 *
LCP1 and LDC 0.1 13 * 0.162 *
LCPI and MSR 0.457 * 0.201 *
LCPI and LMPI 0.235 * 0.117 *
LCPI and MMR 1.377 0.275
LCPI and LNER 1.876 0.921
Note: * and ** denote rejection of the null hypothesis at
the 1 percent and 5 percent significant levels,
respectively. FAR = the ration of net foreign reserves to
GDP, LDC = the log of domestic debt, MSR = the ratio of
money supply to GDP, LMPI = the log of manufacturing output
index, MMR = money market rate, and LNER = the log of
nominal exchange rate.
Table 11
Pairwise Nonlinear Granger Causality Test Results
Sample Period:
January 1990 to December 2000
Direction of [S.sub.F]-- Decision
Nonlinear Causality statistic (at the 5% level)
FAR [right arrow] LCPI 0.364 Reject
LDC [right arrow] LCPI 3.283 Do not reject
MSR [right arrow] LCPI 4.247 Do not reject
LMPI [right arrow] LCPI 3.673 Do not reject
LNER [right arrow] LCPI 1.446 Reject
MMR [right arrow] LCPI 1.318 Reject
Sample Period:
Direction of January 2001 to June 2012
[S.sub.F]-- Decision
Nonlinear Causality statistic (at the 5% level)
FAR [right arrow] LCPI 9.454 Do not reject
LDC [right arrow] LCPI 3.987 Do not reject
MSR [right arrow] LCPI 9.545 Do not reject
LMPI [right arrow] LCPI 7.169 Do not reject
LNER [right arrow] LCPI 0.004 Reject
MMR [right arrow] LCPI 0.164 Reject
Note: The arrow points to the direction of nonlinear
causality. FAR = the ration of net foreign reserves to GDP,
LDC = the log of domestic debt, MSR = the ratio of money
supply to GDP, LMPI = the log of manufacturing output index.
MMR = money market rate, and LNER = the log of nominal
exchange rate.
