January 2001

 

January 2001

ECONOMIC REFORMS, CAPITAL INFLOWS

AND MACRO ECONOMIC IMPACT IN INDIA

Indrani Chakraborty

Working Paper No. 311

2

ECONOMIC REFORMS, CAPITAL INFLOWS AND

MACRO ECONOMIC IMPACT IN INDIA

Indrani Chakraborty

Centre for Development Studies

Thiruvananthapuram

January 2001

This is a revised version of the paper presented in an open seminar at the

Centre for Development Studies. I am grateful to Prof. K.N.Raj and Dr.

D. Narayana for helpful comments. Thanks are due to Mr.K. A. Anilkumar

3

ABSTRACT

The study attempts to explain the effects of inflows of private

foreign capital on some major macroeconomic variables in India using

quarterly data for the period 1993-99. The analyses of trends in private

foreign capital inflows and some other variables indicate instability.

Whereas net inflows of private foreign capital (FINV), foreign currency

assets, wholesale price index, money supply, real and nominal effective

exchange rates and exports follow an I(1) process, current account deficit

is the only series that follows I(0). Cointegration test confirms the

presence of long-run equilibrium relationships between a few pairs of

variables. But the dependence of each I(1) variable on FINV invalidates

such cointegration except in two cases: cointegration exists between

foreign currency assets and money supply and between nominal effective

exchange rate and exports, even after controlling for FINV. The Granger

Causality Test shows unidirectional causality from FINV to nominal

effective exchange rates- both trade-based and export-based-, which raises

concern about the RBI strategy in the foreign exchange market. Finally,

instability in the trend of foreign currency assets could be partially

explained by the instability in FINV with some lagged effect.

JEL Classification: F21, F41, and C22

Keywords: Private foreign capital, economic reforms, instability, India

4

Introduction

Deregulation of private foreign investment in India started in 1993

in the form of partial liberalization of the capital account. Outflows of

capital by the Indian residents remained strictly controlled, whereas

inflows and outflows of capital by non-residents were partially

deregulated. These changes in policy framework not only led to a surge

in inflows of private foreign capital but also contributed to a significant

change in the form in which private capital was coming in. External

commercial borrowing which was the major source of private foreign

capital inflows during the eighties became less important during the

nineties. In the nineties, the predominant role of the portfolio investment

followed by the foreign direct investment (FDI) has been envisaged1 .

Estimates of portfolio investment through foreign institutional investors

(FIIs) and global depository receipts (GDRs) are $9 billion and $6 billion,

respectively, covering the period 1993-98 (Ahluwalia, 1999). Distinct

changes in the guidelines for portfolio investment and FDI vis-à-vis

external commercial borrowing during August and September of 1992

certainly encouraged the former categories (R.B.I, 1992). A look at some

of the items in the new guidelines will make it clear:

a) Use of external commercial borrowing would be prioritised to

the infrastructure and core sector export-oriented and importsubstitution

units and also medium-sized / small-scale units.

Infrastructure or core sector included power, oil exploration and

refining, telecommunication, fertilizer and transport.

5

b) External commercial loans could be raised only for meeting

foreign exchange cost of capital investment. Expenditure on

working capital could not be met by external commercial

borrowing. In addition, all external commercial borrowing should

have a minimum final maturity period of five years.

c) Priority will be given to those proposals for external commercial

borrowing which will be used for total export productions, which

will be self-liquidating i.e. the principal installments and interest

would be entirely serviced out of export-earning and for which

no security would be provided by a commercial bank/financial

institution in India.

d) Foreign institutional investors (FII) would be welcome to invest

in all the securities tradable in the primary and the secondary

markets. There will be no restriction on the volume of investment

by the FIIs. Moreover, there would be no lock-in period for the

proposed investments by the FIIs.

e) FIIs would be given tax-benefits. Concessional tax rate of 20 per

cent was proposed for the dividend and interest income. In

addition, a tax rate of 10 per cent on long-term capital gains (more

than one year) and 30 per cent on short-term capital gain were

proposed.

f) In connection with the FDI, the limit of 40 per cent of foreign

shareholdings imposed by the Foreign Exchange Regulation Act

(FERA) of 1973 on the Indian companies was raised to 51 per

cent in July 1991. The approval of foreign direct investment in

the priority industries where high technology was thought to be

needed was made automatic. Criteria for approval were also made

liberalised, in general.

6

The trend of private foreign capital depicted in Fig.1 shows

that portfolio investment exceeded the other two sources

between 1993-94 and 1996-97. However, the trend of

portfolio investment exhibits instability. On the other hand,

the flow of FDI appears to be increasing steadily over the

years. External commercial borrowing, the only major

source of private foreign capital prior to reform, reached

its all time trough in 1992-93 when it became negative.

Afterwards it followed an upward trend and it exceeded

the other two forms of private foreign capital during the

last two years i.e., 1997-98 and 1998-99.

Fig. 1. Net Inflows of Foreign Capital

Year

Net Inflows of foreign capital in Rs. crores

ComB

Total

PORTF

FDI

.

Foreign direct investment (FDI)

Portfolio investment (PORTF)

External commercial borrowing (CoMB)

Total

7

A brief account of the net annual capital inflow in India since the

beginning of the 90’s is presented in Table 1. It reveals that net inflow of

private foreign capital increased from Rs.605.2 crores (0.86% of GDP)

in 1992-93, the year just prior to the deregulation of private foreign

investment, to Rs.15187.5 crores (1.87% of GDP) in 1993-94. Easing of

restrictions on inflows of private foreign capital has also led to its

increasing share in gross domestic capital formation from 2.85% in 1990-

91 to 8.05% in 1993-94 and 9.3% in 1997-98. In terms of net capital

account, private inflows of foreign capital accounted for 53.58% and

79.01% in 1993-94 and 1998-99, respectively, whereas the same stood

at only 21.16% in 1985-86.

Some important changes may also be observed in the exchange

rate policy since July 1991. A significant downward adjustment in the

exchange rate took place in July 1991. On July 1, 1991, exchange rate of

rupee per unit of dollar was devalued from Rs.21.40 to Rs.23.25 and on

July 3 to Rs.26. Since March 1992, dual exchange rate system was

instituted. It was characterised by the coexistence of the official exchange

rate determined by the RBI and the market rate determined in the interbank

foreign exchange market. Since March 1993, exchange rate of rupee

was left to be determined by the market forces. In March 1993, the

exchange rate of rupee per unit of dollar became Rs.31.40 and it remained

steady for over two years at that level. But since the middle of September

1995 there were periodic speculative pressures on the exchange rate which

called for active intervention by the Reserve Bank of India in the foreign

exchange market. In the floating exchange rate regime, the predominant

objective of India’s exchange rate policy was to maintain a stable REER

in order to prevent an erosion in the incentives available to exporters. To

meet this objective, it was announced that, "RBI stands ready to intervene

to maintain orderly market condition and to curb excessive speculation"

(G.O.I, 1995-96). The period after 1993, therefore, witnessed

8

interventions by the Reserve Bank of India several times to reduce the

excess volatility of the exchange rates.

Given these aspects of the reform programme, it is pertinent to

analyse the macroeconomic responses in the post-1993 period in India.

Importance of this issue can be traced back to different macroeconomic

consequences of liberalization of foreign capital in the countries of South-

East and East Asia vis-a-vis Latin America. Various studies observed

that in the post liberalization period in countries viz., Thailand, Indonesia,

Malaysia and Chile both investment and exports grew without any

substantial appreciation in real exchange rate. On the other hand,

Argentina, Mexico, Brazil, Columbia, Korea and Philippines experienced

a strong appreciation of their real exchange rate. In contrast to this

difference, it was further noted that in all these countries in both the

regions current account deficits and inflation increased (Khan and

Reinhart, 1995, World Bank, 1995 and Corbo and Hernandez, 1996). A

comparative picture of these two regions in terms of selected

macroeconomic indicators in the period of increasing inflow of foreign

capital is presented in Table 2.

The objective of this study is to observe and analyse the dynamics

of some selected macroeconomic indicators in relation to the inflows of

private foreign capital as a consequence of economic reforms in India.

The paper is organized as follows: section I discusses the trends in some

macroeconomic indicators and tries to explain them. Section II reports

the findings of some econometric analyses based on quarterly data from

1993.II to 1999.II. Finally we conclude in section III.

I. Trend behaviour of some Macroeconomic Indicators in relation to

Inflows of Private Foreign Capital in India with an Analytical Overview:

This section begins with the question: how might liberalization of

capital inflows affect macroeconomic aggregates in an open economy?

9

An economy seeking to attract foreign capital can experience different

macroeconomic consequences under different exchange rate regimes.

In what follows we consider the macroeconomic consequences under a

floating exchange rate regime and the discussion follows the arguments

available in the existing literature.

In a floating exchange rate regime, an increase in capital inflows

will lead to an appreciation of nominal exchange rate because of an excess

supply of foreign exchange. This appreciation under a floating exchange

rate regime generally leads to overshooting of nominal as well as real

exchange rate. This happens because as the financial market adjusts at a

faster rate than the goods market, surge of capital inflows may lead to

excessive appreciation of nominal and real exchange rates above their

equilibrium levels (Rangarajan, 1998). Overshooting of nominal

exchange rate will, in effect, ration the sale of foreign exchange for current

transactions. This will affect exports adversely because the consequence

will be similar to imposing an implicit tax on exports (Sachs, 1989).

Such a rationing, on the other hand, is likely to lead to the development

of a black market for foreign exchange. The rationing of foreign exchange

will lower the relative price of exports and bias production away from

exports. At the same time, imports may have increased due to the

appreciation of real exchange rate which, in turn, will have adverse effect

on the current account balance. In consequence expected depreciation

will generate pressure which will lead to excessive depreciation of

nominal exchange rate with some lag. This tendency of instability in

nominal exchange rate may result in a loss of confidence on the part of

the foreign institutional investors. Such adverse effects due to the

instability of the exchange rate may lead to a system of "managed float".

Under this system, the central bank intervenes in the foreign exchange

market to reduce the excess volatility in the exchange rate so that the

equilibrium is restored. There may be two types of central bank

intervention.

10

In the first type the central bank purchases foreign exchange against

the domestic currency. This will help control further appreciation of the

nominal exchange rate. On the other side, net foreign assets being one

component of reserve money, such intervention leads to the growth of

high-powered money and consequently increases the money supply in

the economy. With no change in the demand for money this will lead to

an increase in domestic absorption. Increased domestic absorption may

come through increased spending on either investment or consumption

or both. This increased spending will go to both the categories of goods

viz., tradables and nontradables. Increased expenditure on tradables will

increase the size of trade deficit. On the other hand, increased spending

on nontradables will increase the relative price of nontradables to

tradables. This, in effect, will have two consequences. One is the

reallocation of factors of production towards nontradable sector due to

the increase in its relative price. So it is interesting to observe that while

a large nontradable sector emerges, the consumer expenditure switches

from nontradables to tradables (Corbo and Henandez, 1996). The other

consequence is the effect on the real exchange rate. However, the direction

of change will depend on the rate of inflation relative to the initial

depreciation of the nominal exchange rate. If real exchange rate

appreciates there will be further deterioration of the current account

balance which will require further intervention by the central bank in

the form of buying foreign exchange. On the other hand, if real exchange

rate depreciates that may help improve the current account balance. But

depreciation of real exchange rate may also fail to produce all the desirable

results. Because too much depreciation may lead to reversal of capital

inflows simultaneously with the other effect. Hence, that will require

further intervention by the central bank but in the form of selling foreign

exchange so that nominal exchange rate appreciates to some extent. Such

intervention by the central bank will continue till a new "equilibrium" is

reached for the exchange rate.

11

The second type of central bank intervention is known as "sterilized

intervention". In this process the central bank buys foreign exchange in

exchange of government securities. It helps to curb the growth of money

supply in the economy and hence there will be no increase in domestic

absorption. In consequence, there will be no increase in the current

account deficit too. But it creates an upward pressure on the domestic

interest rate and hence increase fiscal deficits (Joshi and Little, 1994).

Further an increase in the interest rate may attract more foreign capital

which would aggravate the problem of management of capital inflows.

Another limitation of sterilized intervention is that the central bank has

to incur some costs while using this instrument. The associated cost will

be equivalent to the difference between the interest to be paid by the

central bank on government securities and the return enjoyed by the

central bank on holding foreign reserves. This will happen because

sterilized intervention by the central bank in the context of liberalization

of private foreign capital will lead to an increase in interest rate for

government securities2 .

In the light of the above discussion, let us now analyse the trend

behaviour of some macroeconomic indicators in India. For the trend

Fig. 2. Foreign Exchange Reserve

Year

Foreign exchange reserve in US $ million

12

analysis we have chosen the period 1990-91 to 1998-99 which will help

to form a comparative picture of the post-reform period and the period

just prior to the reform.

The foreign exchange reserve in India has increased considerably

since the initiation of economic reforms (Fig.2). The reserve has gone

up by US $27286 million between 1990-91 and 1998-99. Although it

was steadily increasing since 1991-92, the volume was not quite high

during the first three years. The reserve had started to peak up since

1993-94. It happened mainly because of intervention by RBI against the

surge of capital inflows. Initially RBI followed "sterilized intervention".

But since the treasury bill market was not properly developed this process

could not be continued for a longer time. During the later period

intervention by RBI was mostly through purchase of foreign currency

for domestic currency (Reddy, 1999, Rangarajan, 1998). Consequent

upon the increase in foreign currency assets due to RBI intervention in

the foreign exchange market, there was a rapid growth in money supply

in the post-reform period (Fig.3 and Table 4).

Fig. 3. Money Supply (M3)

Year

Money supply in Rs. crores

13

Let us now consider if the increase in money supply led to an

increased domestic spending in the post-1993 period. Figure 4 shows

that investment as a percentage of GDP follows an upward trend since

1993-94 and continued till 1997-98 when the level of investment was

much higher than that during the two years prior to 1993-94. The same

figure also reveals one of the important features in the post-reform period,

that is, a sharp increase in the private sector investment. However, this

achievement should not be attributed entirely to the liberalization of

capital inflows because some other reform measures have also contributed

towards this end. These reform measures include relaxation of restrictions

on industrial licensing and reduction of tariff on imports of capital goods

(Athukorala and Sen, 1995). Consumption expenditure as a percentage

of GDP, on the other hand, does not exhibit any upward rising trend in

general, but sharp increases are observed during the last two years (Fig.5).

Total consumption expenditure, combining both private and public,

Fig. 4. Savings and investment

Savings and investment as a % of GDP

Year

.

Gross domestic savings (GDMSAV)

(Total) Gross domestic cap form (GDMC)

(Private) Gross domestic cap form (GDMC)

(Public) Gross domestic cap form (GDMC)

Total GDMC

GDMSAV

Private GDMC

Private GDMC

14

increased sharply in 1997-98 to 75.8% of GDP from 67.9% of GDP in

1996-97. Interestingly enough, private consumption expenditure steadily

declined after economic reform until 1996-97 and then increased during

the last two years (Table 3). While disaggregated data on consumption

expenditure are not available, data on different categories of imports

suggest that the rising expenditure on consumption was not heavily driven

by manufactured imports (Table 3). In contrast to this it may be noted

that the Latin American countries experienced a consumption boom

mainly driven by imports of consumer durables in the post-liberalization

period. The period since 1993-94 also witnessed an increase in real GDP.

Although the growth in real GDP has reached its highest value in 1996-

97, it was varying between 7.2% and 7.5% during the period 1994-95 to

1996-97 (Table 3). However, it started to decline sharply since 1997-98.

Fig. 5. Consumption Expenditures

Year

Consumption expenditure as a % of GDP

.

(Total) Consumption Exp. As % of GDP

(Private) Consumption Exp. As % of GDP

(Public) Consumption Exp. As % of GDP Public

Total

Private

Public

15

Trend behaviour of savings as a percentage of GDP was almost similar

to that in investment except the fact that the gap between savings and

investment was widening since 1993-94 (Fig.4).

Trend behaviour of inflation is reported in Fig.6. Increased money

supply led to a rise in the inflationary pressure during the initial years

after reform viz. between 1990-91 and 1992-93. Due to the policy of

targetting inflation stabilization, however, it had been possible to keep

inflation below double digit level since 1993-94 excepting the year 1994-

95. Nominal effective exchange rate (NEER) and real effective exchange

rate (REER), (both export-based and trade-based) reveal a declining trend

since 1990-91 and it appears that the nominal effective exchange rate

depreciated at a faster rate than the real effective exchange rate (Fig.7).

Prior to 1993-94, it had been made possible by the devaluation of nominal

exchange rate of the rupee twice in July 1991 and by containing inflation

within a stable limit. Even after the liberalization of capital inflows since

1993-94, nominal effective exchange rate did not appreciate mainly

because of the direct intervention by the Reserve Bank of India in the

Fig. 6. Inflation

Year

Inflation as a percentage change in WPI

16

foreign exchange market, as discussed earlier. The real effective exchange

rate (REER), however, appreciated between 1993-94 and 1994-95

because of a rise in inflation in India in that period at a faster rate

compared to her trading partners. On the whole, movement of both the

Fig. 7. Nominal and Real Effective Exchange Rates (36 countrybased

weight)

Fig. 8. Current Account Deficits, Exports and Imports

Year

NEER, REER ( In SDR)

Year

Current account deficits, exports and imports as a % of GDP

REER (trd)

REER (ex)

NEER (trd)

NEER (ex)

. REER (ex) REER (trd) NEER (ex) NEER (trd)

Imports

Exports

CAD

CAD

Exp

Imp

17

indices of exchange rate in the downward direction during the entire

period 1990-91 to 1998-99 indicates that the policy of targetting REER

by adjusting NEER seems to have been partly successful during the 1990s.

Depreciation of REER seems to have improved the external trade

competitiveness because, it appears from Fig.8 that exports follow an

increasing trend since 1990-91. Imports as a percentage of GDP, on the

other hand, declined initially to 7.7% in 1991-92 from 8.1% in 1990-91

and started to increase thereafter. In relative terms, imports increased at

a faster rate than exports particularly in two years viz., from 1991-92 to

1992-93 and from 1994-95 to 1995-96. Hence, there were sharp increases

in the current account deficits during these two periods. If we look at

Fig.4 it is visible that, during these two periods there were sharp increases

in investment relative to savings too. Despite some fluctuations, the

overall trend in the current account deficit appears to be declining (Fig.8).

Thus, unlike some of the Latin American countries, the surge in the

inflows of capital, did not result in an increase in the current account

deficits in India.

II. Findings and Analysis

This section empirically analyses the effects of inflows of private

foreign capital on some of the major macroeconomic variables in India

using the quarterly data for the period 1993.II-1999.II. We try to

understand if the observed fluctuations in the time-series of some

macroeconomic variables viz., foreign currency assets3 , wholesale price

index, money supply, real and nominal effective exchange rates, exports

and current account deficit, as reported in the earlier section, can be

explained in relation to the fluctuations in the time series of inflows of

private foreign capital. Research done over the past decades shows that

before indulging in any econometric modelling using time-series data,

one should be concerned about the problem of non-stationarity or unit

18

root problem. Results from a regression exercise involving nonstationary

data is observed to be spurious (Granger and Newbold, 1974 and Granger,

1981). Therefore, the following empirical analysis is carried out in the

light of the recent developments in the time-series analysis.

In the first stage, stationarity of the series on each variable is

examined using both the Dickey-Fuller (DF) test and Augmented Dickey-

Fuller (ADF) test. The DF test is based on the following regression:

Yt = C + át + äYt-1 + åt…………….(1)

where C is a constant and t is the trend component.

The null hypothesis of unit root in Yt or nonstationarity of Yt is

rejected if ä is negative and statistically significant. If C and á failed

to be statistically significant we run the above regression dropping the

constant and trend. Critical values for ä in such a situation are noted to

be different from the one in equation (1) above.

For ADF test we include the lagged difference terms as regressors

in the above equation i.e.

k

Yt = C + át + äYt-1 + Ó âi Yt-i + åt ………………(2)

i =i

Following Enders (1995), to select the number of lagged

differenced terms we started with a relatively long lag length namely 15

in this case. Since the t -value for â at lag 15 was statistically insignificant

we estimated equation (2) with 14 lagged differenced terms and again

we tested for statistical significance of the t-value corresponding to â at

lag 14. This process is repeated until a lag is found which is statistically

significant. The number of lags chosen is reported below each equation

under ADF test.

19

The results are reported in Table 5. All the variables are transformed

to natural logarithm, except CAD which includes negative values. DF

test shows, for none of the variables at the level, the hypothesis of

nonstationarity can be rejected at 1% level of significance. However, it

is rejected at 5% level only for some of the variables viz., FINV, FNCA,

WPI, and CAD. But the DF test at the first difference of the series shows

that stationarity condition is uniformly supported at the 1% level of

significance by all the series except CAD. On the other hand, it follows

from the ADF test that none of the series, except CAD and NEERX are

stationary at the level. However, all the series except REERT, EXP and

CAD appear to be stationary at their first difference following ADF test.

Therefore, from these results it follows that all the series have unit roots

except CAD which is only stationary at the level.

The fact that all the series except CAD are I(1), or nonstationary

at the level, is important. What follows from the nonstationarity or the

presence of unit root in a time-series variable is that the time-path of the

variable is diverging from its equilibrium. The idea of convergence

towards equilibrium represents "stability" in the context of difference

equation, where lies the conceptual origin of the term "unit root". Thus

the presence of unit root indicates instability. However, if a set of

nonstationary variables is observed to be cointegrated then it follows

that the variables will come back to equilibrium in the long-run, even if

they drift away from equilibrium in the short-run. Hence it is necessary

to examine if there exists any cointegrating relationship between the set

of variables observed to have I(1) process before drawing any inference

regarding their instability.

Before going to the second stage, some diagnostic checking was

carried out to verify if the number of differenced lags was selected for

ADF test appropriately. The residual analysis for ADF regressions at the

20

first difference of each variable is reported in Table 6. The appropriate

number of lags in ADF regression should not reveal any significant

autocorrelation among the residuals or heteroscedasticity. Presence of

autocorrelation can be verified using Ljung-Box Q-statistic and Box-

Pierce Q-statistic whereas the presence of heteroscedasticity can be

verified using the ARCH test. Varying the number of lags in the residuals,

considering for example 2, 4 and 6 we observe that residuals exhibit no

autocorrelation for all the I(1) series except REERT and REERX. On

the other hand, ARCH test supports the assumption of homoscedasticity

for residuals of all the I(1) series4 .

In the second stage, tests for cointegration are applied to examine

if there exists any long-run equilibrium relationship between any pair of

I(1) variables. A number of series are said to be cointegrated if they are

nonstationary at the level and have same order of integration but there is

at least a linear combination of these variables which is stationary. We

have carried out cointegration test for each pair of variables having I(1)

series by making use of the methodology suggested by Engle and Granger

(1987). The results are reported in Table 7. We find, following either the

DF or ADF test results, that all the I(1) variables individually have

cointegrating relationship with FINV. In addition, cointegration is

observed between the following pairs of variables: FNCA and M3, M3

and WPI, WPI and REERX, WPI and REERT, REERX and EXP, REERT

and EXP, NEERX and EXP, NEERT and EXP. The results of

cointegration test in the latter sequence of relations suggest that the longrun

equilibrium relationship is restored between the following pairs of

variables viz, foreign currency assets and money supply, money supply

and inflation, inflation and real exchange rate, real exchange rate and

exports during the period 1993-99. These long-run relationships, based

on the observed data, reflect that the covariate fluctuations for the

variables in each pair are correlated over time. These relationships,

21

however, need to be analysed carefully, because such cointegration

relationship between variables in each pair breaks down in most of the

cases when we include FINV as a third variable. The results of the test

of cointegration, reported in Table 8, reveal that we fail to reject the null

hypothesis of no cointegration in all the cases but with two exceptions.

These two exceptional cases are (M3, FINV, FNCA) and (NEERX, FINV,

EXP) where these two sets are observed to be cointegrated following

the DF and ADF tests, respectively. The above results suggest that if we

control for the variable FINV, no long-run equilibrium relationship holds

between the variables for most of the above mentioned pairs of variables.

These findings are indicative of the fact that the increased inflows of

foreign capital in India since 1993 might account for the disturbances in

the equilibrium relationship between a number of macroeconomic

variables with a few exceptions. Exceptions, which follow from our study,

are between foreign currency assets and money supply and between

nominal effective exchange rate and exports.

The test of cointegration ignores the effect of the past values of

one variable on the current value of the other variable. So, finally, we

tried the Granger causality test to examine such possibilities. Since the

reliability of results of the Granger causality test depends on whether

the variables are stationary or not, we applied this test on the first

difference of the log transformed series which are reported to be

stationary. It is well-known that Granger causality test is sensitive to the

choice of lag length. To avoid this problem, as noted in Enders (1995)

we have applied Akaike information criterion to choose the optimum

lag length5 .

The results are reported in Table 9. Major observations are

discussed here. The most important observation is that FINV Granger

causes NEERT and NEERX. This has relevance for the exchange rate

22

policy. What it implies is that the past information on FINV improves

the predictability of NEERT and NEERX. As discussed earlier, RBI

intervened in the foreign exchange market with certain objectives since

1993, one of which was to "curb excessive speculation". The above

finding, however, challenges this objective. The direction of Granger

causality from FINV to NEERX and NEERT indicates that even if RBI

does not disclose its strategy of intervention a priori, it is possible to

speculate about the nominal exchange rate given the past information

on the inflows of private foreign capital. We further observe that FINV

Granger causes FNCA. This result suggests that, in the post reform period,

instability in the trend behaviour of foreign currency assets can be

explained partly by the instability in the trend behaviour of the inflows

of private foreign capital with some lagged effect. However, no causality

is observed between FINV and other variables having I(1) process.

III. Conclusion

A large volume of recent literature, while analysing the experiences

of Asian and Latin American countries, reveals that financial liberalization

led to severe macro-economic instability in several of those countries

and no unique pattern emerged in this respect. This study, therefore,

made a modest attempt to analyse the dynamics of some major

macroeconomic variables during the post-reform period in India. The

main focus of this study lies in analysing the behaviour of some selected

macro-economic indicators in relation to the surge in inflows of private

foreign capital in India since 1993, the year in which several major reform

programmes were initiated. A review of the analytical literature shows

that macroeconomic consequences of financial liberalization are the

results of the combined effect of monetary, fiscal as well as trade and

exchange rate policies followed by the government of a country. So,

there is no straightforward way of predicting the resulting macroeconomic

effects of financial liberalization in any country.

23

Major observations from the trend analysis, covering the period

1990-91 to 1998-99, are as follows: (a) Inflows of private foreign capital,

measured as the aggregates of foreign direct investment, portfolio

investment and external commercial borrowing, increased sharply since

1993-94. Although the volume of portfolio investment increased

enormously its trend exhibits instability. Flow of foreign direct

investment, on the other hand, increased steadily after the reforms. (b)

Foreign exchange reserve increased by a considerable amount which

indicates intervention of RBI against the surge of capital inflows.

Consequently, a rapid growth was observed in money supply. (c)

Investment as a percentage of GDP, private investment in particular,

followed an upward rising trend. Total consumption expenditure as a

percentage of GDP, on the other hand, did not reveal any clear pattern.

Private consumption expenditure, however, steadily declined. Real GDP

followed an increasing trend. (d) Inflationary pressure was mostly under

control except in the three years prior to 1993-94. Except in the two

years 1993-94 and 1994-95, real effective exchange rate (both exportbased

and trade-based) declined sharply. It had been made possible by

the downward adjustment of nominal effective exchange rate as well as

containment of inflation within a stable limit. (e) Unlike the Latin

American countries, current account deficits as a percentage of GDP did

not increase sharply excepting the two years 1992-93 and 1994-95.

Some econometric analyses based on quarterly data for the period

1993.II-1999.II reveal a number of interesting observations. It is found

that each of the following series viz., inflows of private foreign capital,

foreign currency assets, wholesale price index, money supply, real and

nominal effective exchange rates and exports follows an I(1) process

whereas the only series which follows an I(0) process is the current

account deficit. I(1) process indicates instability in the trend behaviour

of the variable under consideration. Tests for cointegration are applied

24

to examine if there exists any long-run equilibrium relationship between

any pair of I(1) variables. All the I(1) variables individually have

cointegrating relationship with FINV. In addition, cointegration is

observed between the following pairs of variables: FNCA and M3, M3

and WPI, WPI and REERX, WPI and REERT, REERX and EXP, REERT

and EXP, NEERX and EXP, NEERT and EXP. Further tests shows that

such cointegration relationship between variables in each pair breaks

down in most of the cases when we include FINV as a third variable.

The two exceptional cases where the cointegration relationship exists

even after controlling for FINV are foreign currency assets and money

supply and between nominal effective exchange rate and exports. Results

of the cointegration test are indicative of the fact that the increased inflows

of private foreign capital in India since 1993 might account for the

disturbances in the equilibrium relationship between a number of

macroeconomic variables with a few exceptions.

Finally, the Granger causality test is applied to examine if there is

any lagged effect of inflows of private foreign capital on the

macroeconomic variables under consideration. The direction of causality

from FINV to NEERX and NEERT has some relevance for the exchange

rate policy. It raises concern about the strategy of RBI intervention in

the foreign exchange market, one objective of which is to curb

speculation. Another finding from the causality test, that FINV Granger

causes FNCA, suggests that in the post reform period, instability in the

trend behaviour of foreign currency assets can be explained partly by

the instability in the trend behaviour of the inflows of private foreign

capital with some lagged effect.

25

Notes:

1. Private foreign capital may be classified into three categories viz., (a)

foreign direct investment (b) portfolio investment and (c) external

commercial borrowing. With the opening of the Indian stock market to

foreign institutional investors (FII) and allowing the private corporate

sector to issue global depository receipts (GDRs) in 1993, portfolio

investment entered as a new category into the private foreign investment

in India in the nineties. Nevertheless, liberalization of foreign capital in

the form of foreign direct investment can be traced back to the beginning

of 1980’s with a distinct change in the country’s foreign investment policy

(RBI, 1991). In 1980 a scheme was introduced to attract investments

from the oil-exporting developing countries. Under this scheme investors

from the Gulf region were allowed to invest in the equity capital of Indian

companies upto 40% of the total paid-up capital. Liberalised investment

facilities to non-resident Indians was another important policy decisions

taken during early 1980’s.

2. Prior to financial liberalization, interest rate on government securities

may be deliberately kept at a low level. But financial liberalization will

lead to a market-determined interest rate on government securities which

will be definitely higher than the earlier level.

3. Since foreign currency assets form the major component of foreign

exchange reserve that influence money supply in an open economy, we

have included foreign currency assets in our empirical analysis.

4. Absence of serial autocorrelation and heteroscedasticity in the estimated

residuals confirm that the power of ADF test is reliable. Absence of

heteoscedasticity is evident from the ARCH test results reported in Table

7. It justifies that Phillips-Perron test is not required under this

circumstance.

5. Some empirical studies which have applied this criterion include Samanta

and Mitra(1993),Masih and Masih(1994), Ghosh(1995).

26

References

Ahluwalia, Montek Singh (1999), "Reforming India’s Financial Sector:

An Overview, in James A. Hanson and Sanjay Kathuria (ed),

India: A Financial Sector for the Twenty-first Century, Oxford

University Press.

Athukorala, P. and Sen, K.(1995), "Economic Reforms and the Rate of

Savings in India", Economic and Political Weekly, 30 (35),

September 2.

Corbo Vittorio and Hernandez Leonardo (1996), "Macroeconomic

Adjustment to Capital Inflows: Lessons from Recent Latin

American and East Asian Experience", The World Bank Research

Observer, 11 (1), February.

Enders, Walter (1995), Applied Econometric Time Series, John Wiley &

Sons.

Engle, R.F. and Granger, C.W.J. (1987), "Cointegration and Error-

Correction: Representation, Estimation and Testing",

Econometrica, 55.

Ghosh, A.R. (1995), International Capital Mobility among the Major

Industrialised Countries: Too Little or Too Much?, Economic

Journal, 105 (428), January.

G.O.I. (1993-94 & 1995-96), Economic Survey.

Granger, C.W.J. (1981), "Some Properties of Time-Series Data and Their

Use in Econometric Model Specification", Journal of

Econometrics, 16.

Granger, C.W.J. and Newbold, P. (1974), "Spurious Regressions in

Econometrics", Journal of Econometrics, 2.

Joshi, Vijay and Little, I.M.D. (1994), India: Macroeconomics and

Political Economy, 1964-1991, Oxford University Press.

27

Joshi, Vijay and Little, I.M.D. (1998), India’s Economic Reforms: 1991-

2001, Oxford University Press.

Khan, Mohsin S. and Reinhart, Carmen M. (1995), "Macroeconomic

Management in APEC Economies: The Response to Capital Inflows"

in Khan, M.S. and Reinhart, C.M. (ed), Capital Flows in

the APEC Region, IMF Occasional Paper No. 122, March.

Masih, A.M. and Masih, R. (1994), "Temporal Causality between Money

and Prices in LDCs and the Error-Correction Approach: New

Evidence from India, Indian Economic Review, 29 (1).

Rangarajan, C. (1998), Indian Economy- Essays on Money and Finance,

UBSPD, New Delhi.

Reddy, Y.V. (1999), "Managing Capital Flows", Reserve Bank of India

Bulletin, January.

Reserve Bank of India (1991), Reserve Bank of India Bulletin, April.

Reserve Bank of India (1992), Reserve Bank of India Bulletin, November.

Sachs, Jeffrey D.(ed), (1989), Developing Country Debt and Economic

Performance, vol.2, University of Chicago Press.

Samanta, G.P. and Mitra, S. (1998), "Recent Divergence between Wholesale

and Consumer Prices in India – A Statistical Exploration",

RBI Occasional Papers, 19(4), December.

Sen, Kunal and Vaidya, Rajendra R. (1998)," India", in Fanelli, J.M. and

Medhora, Rohinton (ed), Financial reform in Developing Countries,

International Development Research Centre, Canada.

World Bank (1995), India: Recent Economic Developments and Prospects,

Washington, D.C.

28

Table 1. Net Annual Capital Flows (in Rs. Crores)

Year

FDI

Portfolio

External

Total Private

External

% share of net

Total

Total private

(1)

investment

Commercial

inflows of

Assistan

private capital

Capital

inflows of

(2)

borrowing

capital

(4)

inflows in

Account

capital as a %

(3)

(1) + (2) + (3)

GDP

(6)

of GDCF

(5)

(7)

1990-91

173.6

9.9

4034.4

4217.9

3964.9

0.78

12660.8

2.85

1991-92

329.8

9.9

3806.6

4146.3

7394.5

0.67

9812.5

2.86

1992-93

958.7

741.2

-1094.7

605.2

5749.7

0.86

12208.9

0.36

1993-94

1837.8

11444.8

1904.9

15187.5

5963.9

1.87

28341.9

8.05

1994-95

4216

11233.4

3237.8

18687.2

4798.3

1.93

27683

7.21

1995-96

7176.5

9097.1

4548

20821.6

3355.8

1.86

14271.1

6.85

1996-97

10094

11735.2

10003.6

31832.8

3998.3

2.5

39269.4

9.13

1997-98

13193.6

6766.6

14557.4

34517.6

3430.3

2.4

44532.8

9.3

1998-99

10387.7

-219.4

18557

28725.3

3485

1.85

36354

8.06

Notes:

1.

GDP at current price for the year 1997-98 is converted to 1980-81 prices. Basic data source is RBI (1999), Handbook of Statistics on Indian

Economy. Figures for 1988-99 in this column is obtained from C.S.O (1999), National Accounts Statistics.

2.

GDCF at current price is similarly computed and the source till 1997-98 is R.B.I (1999), as stated above. The same figure for 1998-99 is

computed on the basis that it is 23.4% of GDP, which information is collected from CMIE (June 2000), Monthly Review of the Indian

Economy.

3.

Source of "External Assistance" and "Total Capital Account" is R.B.I (1999), Handbook of Statistics on Indian Economy.

4.

Source of cols.(1), (2) and (3) is RBI Bulletin, various issues.

29

Table 2. Selected Macroeconomic Indicators

Country

Annual average from first year of inflows to 1994

Year in

% change

% change in

Capital

Current account

Real effective

which the

in real

prices

account

deficit as a

exchange

capital inflows

GDP

balance as

% of GDP

rate (%

began

a % of GDP

change)

AsiaIndonesia

1990

6.8

8.7

5.3

2.5

-6.2

Malaysia

1989

8.7

3.6

10.1

4.8

-3.9

Thailand

1988

10.0

5.0

9.4

6.0

1.9

Philippines

1992

2.3

8.5

8.3

4.2

20.9

Latin AmericaArgentina

1991

7.7

52.8

4.4

3.1

20.1

Mexico

1989

3.0

16.1

5.7

6.8

23.4

Brazil

1992

3.0

1941.9

2.0

0.2

57.9

Colombia

1991

4.1

25.6

2.8

4.2

37.1

Chile

1990

6.4

17.5

5.5

1.8

13.5

Source: Corbo and Hernandez (1996), International Financial Statistics (IMF) and World Economic Outlook (IMF)

30

Table 3 : Consumption Expenditure and Imports

Growth rate of Consumption Expenditure as a Manufacturing

RGDP % of GDP imports in total

imports(%)

Total Private Public

1990-91 5.4 73.6 62.1 11.5 12.88

1991-92 0.8 73.8 62.5 1.3 13.11

1992-93 5.3 72.8 61.7 11.1 12.64

1993-94 6 72.7 61.6 11.1 16.56

1994-95 7.2 70.2 59.7 10.5 17.75

1995-96 7.2 68.4 58 10.4 19.49

1996-97 7.5 67.9 57.5 10.3 15.83

1997-98 5 75.8 64.5 11.3 17.61

1998-99 3.8 75.9 63.6 12.3 15.02

Notes :

(i) Growth rate of real GDP in col.(1) represents percentage change

in GDP at factor cost (at constant prices) and the base is 1980-

81 =100

(ii) Source of cols. (1) to (4) is Monthly Review of the Indian

Economy, CMIE, various issues

(iii) Source of col.(5) is Foreign Trade Statistics of India, CMIE,

various issues

31

Table 4. Money Supply, Inflation and Exchange Rates

Money supply

Inflation

REER

REER

NEER

NEER

(M3)

(export based)

(trade-based)

(export-based)

(trade-based)

(Rs. Crores)

1990-91

265828

10.3

74.54

76.59

68.32

69.26

1991-92

277603

13.7

64.55

67.13

55.08

56.29

1992-93

366825

10

60.53

64.47

47.02

49.23

1993-94

378878

8.4

57.86

60.23

43.3

44.47

1994-95

452185

10.9

61.82

64.51

42.88

44.08

1995-96

530802

7.7

60.78

63.44

39.78

40.83

1996-97

696012

6.3

59.45

62.05

37.72

38.6

1997-98

821332

4.4

63.38

66.45

39.05

40.07

1998-99

972204

5.9

61.57

64.88

35.25

37.29

Notes:

(i)

Source of col.(1) is RBI Bulletin, RBI, various issues

(ii)

Source of col.(2) is Monthly Review of the Indian Economy, CMIE, various issues

(iii)

Source of cols. (3) to (6) is RBI Bulletin, RBI, various issues

(iv)

Inflation is measured as the % change in WPI and for WPI series base is 1980-81=100

(v)

36 country bilatateral weights are used for REER and NEER and the base is 1985 =100

32

Table 5. Test for Unit Roots

Variables (Xt)

DF test

ADF test

Levels

First Difference

Levels

First Difference

FINV

-3.1930**

-4.6582 *

-2.0690

-1.6255***

(with C)

(no C & T)

(15 lags, with C & T)

(14 lags, no C &T)

M3

-3.2267

-6.3117*

-2.3074

-2.8578***

(with C & T)

(with C)

(15 lags, with C)

(6 lags, with C )

WPI

-3.0611 **

-4.3817*

-3.1862

-4.5889*

(with C)

(with C & T)

(15 lags, with C &T)

( 9 lags, with C & T)

REERT

-2.5224

-7.5214*

1..5836

-2.8267

(with C)

(no C & T)

(15 lags, with C &

T)

(12 lags, no C &T)

REERX

-2.4996

-6.5821*

3.4028

-2.0187**

(with C)

(no C & T)

(15 lags, no C &T)

(13 lags, no C &T)

NEERT

-1.9669

-6.1035**

-1.8982***

-4.0576*

(with C & T)

(no C & T)

(11 lags no C &T)

(11 lags, with C &T)

NEERX

-1.9683

-5.3986*

-1.7740

-2.7765*

(with C & T)

(no C & T)

(11 lags, no C &T)

(10 lags, no C &T)

EXP

-2.8085

-8.0557*

1.7272

-1.2418

(with C & T)

(with C)

(15 lags ,no C & T)

(12 lags, no C & T)

cont'd

33

CAD

-4.6727 **

-5.5617 *

(with C & T)

(15 lags, with C & T)

FNCA

-3.3121***

-4.7181*

-5.5617*

3.2984

(with C & T)

(with C)

(15 lags, with C & T)

(10 lags, with C)

Note:

k

i.

ADF test is based on the regression Yt = C + át + äYt-1 +

Óâi Y

t-i + åt. The DF test is based on the same equation

i = 1

without the summation of the lagged difference terms on the right hand side. The figures reported in the table are tvalues

of ä.

ii.

‘C’ stands for constant and ‘T’ stands for trend

iii. * signifies statistically significant at 1 % level

iv.

** signifies statistically significant at 5 % level

v.

*** signifies statistically significant at 10 % level

Table 5. Cont'd....

Variables (Xt)

DF test

ADF test

Levels

First Difference

Levels

First Difference

34

Table 6: Residual Analysis for ADF Regression at First Difference

ADF first

Ljung-Box Q

Box-Pierce Q

ARCH Test

differenced

F-statistic

nR2

FINV

Q(2) = 1.52(0.46)

Q (2) = 1.26(0.53)

F(2) = 0.38(0.68)

Z(2) = 0.87(0.64)

Q(4) =2.69(0.61)

Q)(4) =2.17(0.70)

F(4) = 0.21(0.92)

Z(4) = 1.14(0.88)

Q(6) = 4.49(0.61)

Q(6)= 3.38(0.76)

F(6) = 0.64(0.69)

Z(6) = 4.88(0.56)

M3

Q(2) = 0.00(0.99)

Q(2) = 0.00(0.99)

F(2) = 0.71(0.50)

Z(2) =1.54(0.46)

Q(4) =0.37(0.98)

Q(4) = 0.28(0.99)

F(4) = 0.79(0.54)

Z(4) = 3.53(0.47)

Q(6)= 0.94(0.98)

Q(6) =0.68(0.99)

F(6) = 0.66 (0.68)

Z(6) = 4.82(0.56)

WPI

Q(2) = 2.93(0.23)

Q(2) = 2.46(0.29)

F(2) = 0.30(0.74)

Z(2) = 0.68(0.71)

Q(4) =4.66(0.32)

Q(4) = 3.84(0.43)

F(4) =0.18(0.94)

Z(4) =0.97(0.91)

Q(6) =6.48(0.37)

Q(6) =5.16(0.52)

F(6)= 0.68(0.66)

Z(6) = 4.94(0.55)

FNCA

Q(2) =1.58(0.45)

Q(2) = 1.40(0.49)

F(2) = 0.38(0.69)

Z(2) = 0.84(0.65)

Q(4) =1.97(0.74)

Q(4) = 1.69(0.79)

F(4) = 0.66(0.62)

Z(4) =3.02(0.55)

Q(6) = 2.29(0.89)

Q(6) = 1.92(0.93)

F(6) = 0.37(0.89)

Z(6) = 3.05(0.80)

REERT

Q(2) = 3.17(0.20)

Q(2) =2.69(0.26)

F(2)= 1.79(0.19)

Z(2) = 3.49(0.17)

Q(4) =4.09(0.39)

Q(4)= 3.42(0.49)

F(4) =0.75(0.57)

Z(4) = 3.36(0.49)

Q(6) =13.17(0.04)*

Q(6) =9.76(0.13)

F(6) =1.15(0.39)

Z(6) = 6.96(0.32)

REERX

Q(2) = 3.36(0.18)

Q(2) = 2.94(0.23)

F(2) = 0.35(0.71)

Z(2) = 0.78(0.67)

Q(4) = 4.37(0.36)

Q(4) = 3.72(0.44)

F(4) = 0.16(0.95)

Z(4) = 0.86(0.93)

Q(6) = 10.70(0.09)**

Q(6) = 8.14(0.23)

F(6) = 0.28(0.93)

Z(6) =2.46(0.87)

(cont'd)

35

NEERT

Q(2) = 1.76(0.41)

Q(2) = 1.52(0.46)

F(2) = 0.006(0.99)

Z(2) = 0.013(0.99)

Q(4) =3.68(0.45)

Q(4) = 3.01(0.55)

F(4) = 0.35(0.84)

Z(4) = 1.69 (0.79)

Q(6) = 4.80(0.57)

Q(6) = 3.81(0.70)

F(6) =0.38(0.87)

Z(6) = 3.07(0.79)

NEERX

Q(2) = 0.08(0.96)

Q(2) =0.06(0.96)

F(2) = 1.62(0.22)

Z(2) = 3.19(0.20)

Q(4) = 1.17(0.88)

Q(4) =0.92(0.92)

F(4) =1.68(0.20)

Z(4) = 6.20(0.18)

Q(6) =4.46(0.61)

Q(6) =3.20( 0.78)

F(6) =0.84(0.56)

Z(6) =5.68(0.46)

EXP

Q(2) = 0.65(0.72)

Q(2) = 0.56(0.75)

F(2) = 0.18(0.83)

Z(2) =0.42(0.80)

Q(4) = 1.43(0.84)

Q(4) =1.16(0.88)

F(4) = 0.10(0.97)

Z(4) = 0.54(0.97)

Q(6) = 3.14(0.79)

Q(6) = 2.37(0.88)

F(6) = 0.09(0.99)

Z(6) =0.85(0.99)

Notes :

(i)

Q(n) reports Ljung-Box Q/ Box-Pierce Q statistic for the autocorrelations of the n residuals of the estimated

model. With 24 observations, T/4 is equal to 6. Significance levels are in parentheses.

(ii)

F-statistic and nR2

provide ARCH test for the heteroscedasticity in the estimated residuals. ARCH test is

based on the specification that the squared residuals from the estimated model is related to the lagged

squared residuals. Each statistic is estimated using three different lags viz., 2, 4 and 6. nR2 statistic has a chisquare

distribution with degrees of freedom equal to the number of lagged squared residuals and here n

refers to the number of observations.

(iii)

* indicates significant at 1% level and ** indicates significant at 5% level.

Table 6. Cont'd.......

ADF first

Ljung-Box Q

Box-Pierce Q

ARCH Test

differenced

F-statistic

nR2

36

Table 7. Test for pairwise cointegration

Equations: Xt on Yt

µ

ã

DF

ADF

LFINV on LM3

6.45

0.12

-2.90*

-3.27*

(no C & T)

(1 lag, no C& T)

LM3 on LFINV

13.14

0.024

-2.26

-0.9952

(with C & T)

(3 lags, no C& T)

LFINV on LWPI

5.87

0.39

-2.89*

-3.2655*

(no C & T)

(1 lag, no C& T)

LWPI on LFINV

5.59

0.014

-1.38

-5.4491

(with C )

(1lag, with C& T)

LFINV on LFNCA

5.29

0.254

-2.6125

-2.1417**

(no C & T)

(5 lags, no C & T)

LFNCA on LFINV

9.80

0.16

-1.6114

-3.2052

(no C & T)

(3 lags, with C &T)

LFINV on LREERT

-13.73

5.26

-2.7195**

-2.16

(no C & T)

(5 lags, no C& T)

LREERT on LFINV

4.03

0.015

-2.0771**

12.8368*

(no C & T)

(10 lags, with C& T)

LFINV on LREERX

-17.78

6.30

-2.7253*

-6.1864*

(no C & T)

(10 lags, with C&T)

cont'd

37

LREERX on LFINV

3.97

0.016

-2.1369**

-28.5731*

(no C & T)

(10 lags, with C& T)

LFINV on LNEERT

6.49

0.442

-3.0257*

-2.0660**

(no C & T)

(5lags, no C& T)

LNEERT on LFINV

3.64

0.006

-0.6929

-0.7996

(no C & T)

(8 lags, no C& T)

LFINV on LNEERX

6.56

0.428

-3.0268*

-2.0667**

(no C & T)

(5 lags, no C& T)

LNEERX on LFINV

3.62

0.006

-0.6599

-3.3511

(no C & T)

(3 lags, with C& T)

LFINV on LEXP

3.87

0.418

-2.7207**

-2.1502**

(no C & T)

(5 lags, no C & T)

LEXP on LFINV

9.62

0.068

-2.4868

-2.6298

(with C & T)

(4 lags, no C& T)

LFNCA on LM3

-11.24

1.67

-3.0379*

-3.3395**

(no C & T)

(9 lags, with C & T )

LM3 on LFNCA

7.53

0.52

-3.3889***

-3.3449

(with C & T)

(9 lags, with C & T)

Table 7. Cont' d

Equations: Xt on Yt

µ

ã

DF

ADF

con'td

38

Equations: Xt on Yt

µã

DF

ADF

LM3 on LWPI

-0.016

2.34

-1.4261

-3.1806*

(no C & T)

(8 lags, no C& T)

LWPI on LM3

0.139

0.41

1.6485***

-3.2935*

(no C & T)

(8 lags, no C& T)

LWPI on LREERX

0.75

1.206

-1.9027

3.3791***

(with C & T)

(3 lags, with C& T)

LREERX on LWPI

3.60

0.088

-2.0074**

-7.7701*

(no C & T)

(10 lags, with C& T)

LWPI on LREERT

-0.38

1.468

-1.8706

-3.3334***

(with C & T)

(3 lags, with C& T)

LREERT on LWPI

3.46

0.12

-1.9661**

-3.4772*

(no C & T)

(7 lags, with C& T)

LREERX on LEXP

3.83

0.027

-2.1446**

-19.2301*

(no C & T)

(10 lags, with C & T)

LEXP on LREERX

3.08

1.72

-2.4502

-15.8815*

(with C & T)

(9 lags, with C& T)

LREERT on LEXP

3.74

0.04

-2.0952**

-3.7998**

(no C & T)

(7 lags, with C& T)

Table 7. Cont' d

Cont'd

39

LEXP on LREERT

0.68

2.28

-2.2895

-10.8381*

(with C & T)

(9 lags, with C& T)

LNEERX on LEXP

6.31

-0.26

-2.9133*

-3.3111*

(no C & T)

(6 lags, with C& T)

LEXP on LNEERX

21.95

-3.21

-3.4578*

-5.1963*

(no C & T)

(6 lags, with C& T)

LNEERT on LEXP

6.32

-0.25

-3.0249*

-0.8514

(no C & T)

(7 lags, no C& T)

LEXP on LNEERT

22.11

-3.23

-3.5563*

-4.2851*

(no C & T)

(6 lags, with C& T)

Note:

i.

Cointegration regression for two variables Xt and Yt is given by Xt = µ + ãYt + Zt where µ and ã are constant

and cointegrating parameter, respectively.

ii.

DF and ADF tests are carried out using regressions similar to that in Table 5.

iii.

* indicates significant at 1% level

iv.

** indicates significant at 5% level

v.

*** indicates significant at 10% level

Table 7. Cont'd

Equations: Xt on Y

t

µ

ã

DF

ADF

40

Table. 8 Test for Cointegration

Variables DF test ADF test

LFINVQ LM3Q LFNCA -4.8650** -0.7400

LFINVQ LM3Q LWPI -3.4596 -3.8497

LFINVQ LWPI LREERX -1.9165 -1.0948

LFINVQ LWPI LREERT -1.8686 -1.0545

LFINVQ LEXP LREERT -1.9000 -0.0209

LFINVQ LEXP LREERX -1.8856 0.0389

LFINVQ LEXP LNEERX -2.3841 -6.2334**

LFINVQ LEXP LNEERT -2.4095 -0.5647

Notes:

(i) Reported results are based on regressions including a constant

and a trend .

(ii) Reported results for ADF test correspond to 11 lags, the highest

possible number of lags that can be chosen for the given number

of observations. ADF test is also tried with other lags going

down to the least possible number of lags. The null hypothesis

of no cointegration is rejected in none of these case.

(iii) ** indicates significant at 5% level.

41

Table 9: Pairwise Granger Causality Test

Dependent

Explanatory

m

n

F- Statistic

p-value

Remarks

Variable

Variables

FINV

FINV, M3

1

1

1.09

0.31

No causality from M3 ¨ FINV

M3

M3, FINV

1

1

0.009

0.92

No causality from FINV ¨ M3

FINV

FINV, WPI

1

1

57

0.23

No causality from WPI ¨ FINV

WPI

WPI, FINV

2

1

0.38

0.54

No causality from FINV ¨ WPI

FINV

FINV,FNCA

1

1

0.075

0.78

No causality from FNCA¨ FINV

FNCA

FNCA, FINV

1

1

6.79

0.02

Causality from FINV ¨ FNCA

FINV

FINV, REERT

1

1

2.28

0.15

No causality from REERT ¨ FINV

REERT

REERT,FINV

1

1

2.01

0.17

No causality from FINV ¨ REERT

FINV

FINV, REERX

1

1

2.40

0.14

No causality from REERX ¨ FINV

REERX

REERX, FINV

1

1

2.09

0.16

No causality from FINV ¨ REERX

FINV

FINV, NEERX

1

1

2.04

0.17

No causality from NEERX ¨ FINV

NEERX

NEERX, FINV

1

1

3.74

0.06

Causality from FINV ¨ NEERX

FINV

FINV, NEERT

1

1

2.14

0.16

No causality from NEERT ¨ FINVcont'd

42

NEERT

NEERT, FINV

1

1

3.005

0.10

Causality from FINV ¨ NEERT

FINV

FINV, EXP

1

1

0.29

0.59

No causality from EXP ¨ FINV

EXP

EXP, FINV

4

1

0.13

0.72

No causality from FINV ¨ EXP

FNCA

FNCA, M3

1

1

0.026

0.87

No causality from M3 ¨ FNCA

M3

M3, FNCA

1

1

0.41

0.53

No causality from FNCA ¨ M3

M3

M3, WPI

1

1

0.15

0.70

No causality from WPI ¨ M3

WPI

WPI, M3

2

1

0.41

0.53

No causality from M3 ¨ WPI

REERX

REERX,WPI

1

1

2.15

0.16

No causality from WPI ¨ REERX

WPI

WPI, REERX

2

1

1.16

0.29

No causality from REERX ¨ WPI

REERT

REERT, WPI

1

1

2.33

0.14

No causality from WPI ¨ REERT

WPI

WPI, REERT

2

1

0.99

0.33

No causality from REERT¨ WPI

REERX

REERX, EXP

1

1

0.49

0.49

No causality from EXP ¨ REERX

EXP

EXP, REERX

4

4

2.12

0.15

No causality from REERX ¨ EXP

REERT

REERT, EXP

1

2

1.06

0.37

No causality from EXP ¨ REERT

EXP

EXP, REERT

4

4

2.27

0.12

No causality from REERT ¨ EXP

Notes: Optimum lag lengths (m, n) are determined by minimizing the Akaike Information Criteria.

Table 9 Cont'd

43

CENTRE FOR DEVELOPMENT STUDIES

LIST OF WORKING PAPERS

(From 1991 onwards)

MRIDUL EAPEN Hantex: An Economic Appraisal.

September, 1991, W.P.242

SUNIL MANI Government Intervention in Commercial Crop Development:

A Case of Flue Cured Virginia Tobacco.

November, 1991, W.P.243

K. PUSHPANGADAN Wage Determination in a Casual Labour Market: The

Case Study of Paddy Field Labour in Kerala.

January, 1992, W.P.244

K.N. NAIR & S.P. PADHI Dynamics of Land Distribution: An Alternative

Approach and Analysis with Reference to Kerala.

January, 1992, W.P.245

THOMAS ISAAC Estimates of External Trade Flows of Kerala - 1975-76 and

1980-81.

March, 1992, W.P.246

THOMAS ISAAC, RAM MANOHAR REDDY, NATA DUVVURRY Regional

Terms of Trade for the State of Kerala.

March, 1992, W.P.247

P. MOHANAN PILLAI Constraints on the Diffusion of Innovations in Kerala:

A Case Study of Smokeless Chulas.

March, 1992, W.P.248

R. ANANDRAJ Cyclicality in Industrial Growth in India: An Exploratory

Analysis.

April, 1992, W.P.249

T.M. THOMAS ISAAC, RAM MANOHAR REDDY, NATA DUVVURY

Balance of Trade, Remittance and Net Capital Flows: An Analysis of

Economic Development in Kerala since independence.

October, 1992, W.P.250

M. KABIR, T.N. KRISHNAN Social Intermediation and Health Transition:

Lessons from Kerala,

October, 1992, W.P.251

44

SUNIL MANI, P. NANDAKUMAR Aggregate Net Financial Flows to India:

The Relative Importance of Private Loan vis-a-vis Foreign Direct Investments.

August, 1993, W.P.252

PULAPRE BALAKRISHNAN Rationale and the Result of the Current

Stabilisation Programme.

November, 1993, W.P.253

K.K. SUBRAHMANIAN, P. MOHANAN PILLAI Modern Small Industry

in Kerala: A Review of Structural Change and Growth Performance.

January, 1994, W.P.254

DILIP M.MENON Becoming Hindu and Muslim : Identity and Conflict in

Malabar 1900-1936.

January, 1994, W.P.255

D. NARAYANA Government Intervention in Commodity Trade: An Analysis

of the Coffee Trade in India.

January, 1994, W.P.256

K.J. JOSEPH, P. NANDAKUMAR On the Determinants of Current Account

Deficits: A Comparative Analysis of India, China and South Korea.

January, 1994, W.P.257

K.K. SUBRAHMANIAN, K.J. JOSEPH Foreign Control and Export Intensity

of Firms in Indian Industry.

February, 1994, W.P.258

PULAPRE BALAKRISHNAN, K. PUSHPANGADAN Total Factor Productivity

Growth in Indian Manufacturing - A Fresh Look.

April 1994, W.P.259

D. NARAYANA, K.N. NAIR Role of the Leading Input in Shaping Institutions:

Tendency in the Context of Irrigation Uncertainty.

May, 1994, W.P.260

G. MURUGAN, K. PUSHPANGADAN Pricing of Drinking Water: An Application

of Coase Two-part Tariff.

December, 1994 W.P.261

MOHANAN PILLAI On the Mexican Crisis.

December, 1995, W.P.262

SUNIL MANI Financing Domestic Technology Development through the Venture

Capital Route.

December, 1995, W.P.263

45

T.T. SREEKUMAR Peasants and Formal Credit in Thiruvithamcore: The

State Institutions and Social Structure 1914-1940.

December, 1995 W.P.264

AMITABH Estimation of the Affordability of Land for Housing Purposes in

Lucknow City, Uttar Pradesh (India): 1970-1990.

March, 1996. W.P.265

K. PUSHPANGADAN, G. MURUGAN, K. NAVANEETHAM Travel Time,

User Rate & Cost of Supply: Drinking Water in Kerala, India:

June 1996. W.P.266

K.J. JOSEPH Structural Adjustment in India: A Survey of Recent Studies &

Issues for Further Research,

June 1996 W.P.267

D. NARAYANA Asian Fertility Transition: Is Gender Equity in Formal Occupations

an Explanatory Factor?

October, 1996 W.P.268

D. NARAYANA, SAIKAT SINHAROY Import and Domestic Production of

Capital Goods from Substitution to Complementarity,

October 1996. W.P.269

NEW SERIES

W.P. 270 ACHIN CHAKRABORTY On the Possibility of a Weighting System

for Functionings December 1996

W.P. 271 SRIJIT MISHRA Production and Grain Drain in two inland Regions

of Orissa December 1996

W.P. 272 SUNIL MANI Divestment and Public Sector Enterprise Reforms,

Indian Experience Since 1991 February 1997

W.P. 273 ROBERT E. EVENSON, K.J. JOSEPH Foreign Technology Licensing

in Indian Industry : An econometric analysis of the choice

of partners, terms of contract and the effect on licensees’ performance

March 1997

W.P. 274 K. PUSHPANGADAN, G. MURUGAN User Financing & Collective

action: Relevance sustainable Rural water supply in India.

March 1997.

W.P. 275 G. OMKARNATH Capabilities and the process of Development

March 1997

W. P. 276 V. SANTHAKUMAR Institutional Lock-in in Natural Resource

Management: The Case of Water Resources in Kerala, April 1997.

W. P. 277 PRADEEP KUMAR PANDA Living Arrangements of the Elderly

in Rural Orissa, May 1997.

46

W. P. 278 PRADEEP KUMAR PANDA The Effects of Safe Drinking Water

and Sanitation on Diarrhoeal Diseases Among Children in Rural

Orissa, May 1997.

W.P. 279 U.S. MISRA, MALA RAMANATHAN, S. IRUDAYA RAJAN

Induced Abortion Potential Among Indian Women, August 1997.

W.P. 280 PRADEEP KUMAR PANDA Female Headship, Poverty and

Child Welfare : A Study of Rural Orissa, India, August 1997.

W.P. 281 SUNIL MANI Government Intervention in Industrial R & D, Some

Lessons from the International Experience for India, August 1997.

W.P. 282 S. IRUDAYA RAJAN, K. C. ZACHARIAH Long Term Implications

of Low Fertility in Kerala, October 1997.

W.P. 283 INDRANI CHAKRABORTY Living Standard and Economic

Growth: A fresh Look at the Relationship Through the Non- Parametric

Approach, October 1997.

W.P. 284 K. P. KANNAN Political Economy of Labour and Development in

Kerala, January 1998.

W.P. 285 V. SANTHAKUMAR Inefficiency and Institutional Issues in the

Provision of Merit Goods, February 1998.

W.P. 286 ACHIN CHAKRABORTY The Irrelevance of Methodology and

the Art of the Possible : Reading Sen and Hirschman, February 1998.

W.P. 287 K. PUSHPANGADAN, G. MURUGAN Pricing with Changing

Welfare Criterion: An Application of Ramsey- Wilson Model to Urban

Water Supply, March 1998.

W.P. 288 S. SUDHA, S. IRUDAYA RAJAN Intensifying Masculinity of Sex

Ratios in India : New Evidence 1981-1991, May 1998.

W.P. 289 JOHN KURIEN Small Scale Fisheries in the Context of

Globalisation, October 1998.

W.P. 290 CHRISTOPHE Z. GUILMOTO, S. IRUDAYA RAJAN Regional

Heterogeneity and Fertility Behaviour in India, November 1998.

W.P. 291 P. K. MICHAEL THARAKAN Coffee, Tea or Pepper? Factors

Affecting Choice of Crops by Agro-Entrepreneurs in Nineteenth

Century South-West India, November 1998

W.P. 292 PRADEEP KUMAR PANDA Poverty and young Women's Employment:

Linkages in Kerala, February, 1999.

W.P. 293 MRIDUL EAPEN Economic Diversification In Kerala : A Spatial

Analysis, April, 1999.

W.P. 294 K. P. KANNAN Poverty Alleviation as Advancing Basic Human

Capabilities: Kerala's Achievements Compared, May, 1999.

47

W.P. 295 N. SHANTA AND J. DENNIS RAJA KUMAR Corporate Statistics:

The Missing Numbers, May, 1999.

W.P. 296 P.K. MICHAEL THARAKAN AND K. NAVANEETHAM

Population Projection and Policy Implications for Education:A

Discussion with Reference to Kerala, July, 1999.

W.P. 297 K.C. ZACHARIAH, E. T. MATHEW, S. IRUDAYA RAJAN

Impact of Migration on Kerala's Economy and Society, July, 1999.

W.P. 298 D. NARAYANA, K. K. HARI KURUP, Decentralisation of the

Health Care Sector in Kerala : Some Issues, January, 2000.

W.P. 299 JOHN KURIEN Factoring Social and Cultural Dimensions into

Food and Livelihood Security Issues of Marine Fisheries; A Case

Study of Kerala State, India, February, 2000.

W.P. 300 D. NARAYANA Banking Sector Reforms and the Emerging

Inequalities in Commercial Credit Deployment in India, March, 2000.

W.P. 301 P. L. BEENA An Analysis of Mergers in the Private Corporate

Sector in India, March, 2000.

W.P. 302 K. PUSHPANGADAN, G. MURUGAN, Gender Bias in a

Marginalised Community: A Study of Fisherfolk in Coastal Kerala,

May 2000.

W.P. 303 K. C. ZACHARIAH, E. T. MATHEW, S. IRUDAYA RAJAN ,

Socio-Economic and Demographic Consequenes of Migration in

Kerala, May 2000.

W.P. 304 K. P. KANNAN, Food Security in a Regional Perspective; A View

from 'Food Deficit' Kerala, July 2000.

W.P. 305 K. N. HARILAL, K.J. JOSEPH, Stagnation and Revival of Kerala

Economy: An Open Economy Perspective, August 2000.

W.P. 306 S. IRUDAYA RAJAN, Home Away From Home: A Survey of Oldage

Homes and inmates in Kerala, August 2000.

W.P. 307 K. NAVANEETHAM, A. DHARMALINGAM, Utilization of

Maternal Health Care Services in South India, October 2000.

W.P. 308 K. P. KANNAN, N . VIJAYAMOHANAN PILLAI, Plight of the

Power Sector in India : SEBs and their Saga of Inefficiency

November 2000.

W.P. 309 V. SANTHAKUMAR AND ACHIN CHAKRABORTY,

Environmental Valuation and its Implications on the Costs and

Benefits of a Hydroelectric Project in Kerala, India, November 2000.

W.P. 310 K. K. SUBRAHMANIAN, E. ABDUL AZEEZ, Industrial Growth

In Kerala: Trends And Explanations November 2000