Indian Financial Integration: Evidence from the NDF Market
Vineet Srivastava^{1}*, Arup Chattopadhyay^{2}
^{1}Research Fellow, Department of Economics, The University of Burdwan,
Golapbag, 713104, West Bengal, India.
2Professsor, Department of Economics, The University of Burdwan, Goplapbag, 713104, West Bengal, India.
*Corresponding Author Email: srivineet13@gmail.com
ABSTRACT:
This study seeks to investigate the extent and causes of financial integration of India with the rest of the world during the sample period of January 2000 to October 2017. The study compares the interest rate which ought to prevail, with respect to the NonDeliverable Forward (NDF) rate, in case of Covered Interest Parity and what actually does prevail. The difference between the Covered Interest Parity implied domestic interest rate and the actual domestic interest rate is our measure of financial integration. The study tries to find out different macroeconomic variables which affect the financial integration of India. The study used ordinary least squares, Johansen’s Cointegration technique along with VECM and Granger causality to unearth the long run and short run relationship of financial integration with plausible macroeconomic variables. We found India to be significantly nonintegrated. Inflation volatility, changes in NEER and GDP along with Bank Rate and CRR were found to unequivocally influence financial integration of India. We also found structural break present in variables under study at June 2008.
KEYWORDS: Financial integration, Covered interest parity, NDF, Structural break, Cointegration, VECM.
INTRODUCTION:
Emerging market economies like India, while undertaking economic reforms made integration of the economy with the world economy an important agenda. Progressive liberalization of the current as well as capital account was also undertaken. The Indian foreign exchange market also witnessed a metamorphosis from a heavily guarded foreign exchange regime to a more open and market oriented one (Misra and Behera, 2006). However in the recent aftermath of subprime lending crisis countries have become a wary of unbridled flow of funds across borders because of the risks associated with it.
Risks include high degree of concentration of capital in certain countries, the procyclical nature of short term capital flows and the risk of abrupt reversals, possibility of contagion effects etc. (Dadush, Dasgupta and Ratha, 2000). For policymakers the issue of financial globalisation also holds a special significance because of its immediate relevance and disparate effects on countries practicing it, specially the emerging ones (Kose, Prasad Rogoff and Wei 2009). The amount of financial integration also has a significant bearing over the conduit of monetary policy in the economy (Higgins and Humpage 2005). It is all the more important after the subprime lending crisis as to see how the countries, in our case India, has fared i.e. whether it has increased or decreased association with the world financial markets.
Against this backdrop this study tries to measure the extent of financial integration of India with the world along with its time behavior and especially in the aftermath of US subprime lending crisis, and establish its economic determinants. In order to measure the amount financial integration there are a number of quantitative and qualitative measures present. We employ a price based approach of establishing Covered Interest Parity (CIP) to measure the extent of integration of the Indian financial market with that of the world. With rapidly increasing cross border trade and evolving currency market, India being circumspect of the open financial borders, has put semipermeable controls like the partial convertibility of Rupee in the capital account. Hence disallowing the hedging opportunities to foreign investors as a result of which offshore markets for NonDeliverable Forwards (NDF) have developed. Due to the growing activity in the NDF market and it being relatively unexplored in the Indian context encouraged us to include it in our study. We also try to explore the possibilities of long run association of the non integration with other macroeconomic variables. Section 2 of the paper presents literature surveyed for this study. Section 3 deals with the data and methodology used in the study. Section 4 presents the results and discusses the same. In Section 5 we provide concluding remarks and also discuss the limitations of the study.
LITERATURE REVIEW:
Among various measures of financial integration Frankel (1992) claimed of the condition ‘that the covered interest differential is zero, is an unalloyed criterion for "capital mobility" in the sense of the degree of financial market integration across national boundaries.’ It is also considered closest thing to a physical law in international finance by Borio, McCauley, McGuire and Sushko V (2016). Covered Interest Parity (CIP) has generally been established by using the usual variables and the onshore forwards rate like (Taylor M. P., 1987), (Baba & Packer, 2009) among many others. Some studies took cognizance of the suitability of the NonDeliverable Forwards (NDF) in markets where credit migration across borders was constricted by decree, and incorporated it in the examination of the CIP. Studies in the Indian context was done by Ma, Ho and McCauley (2004) where a number of other Asian currencies were also examined noting the declining interest rate spread for the Indian economy in the study period of January 1999 to February 2004 whereas Misra and Behera (2006) used NDFs to calculate the onshore/offshore interest rate spread for the Indian economy during the period of November 2004 to January 2007. They reported nonzero spread between the implied yield in the forward and NDF market hence indicating the presence of capital control and segmented financial markets. Hutchison, Pasricha and Singh (2012) also study the Indian capital controls along with that of the Chinese. They find significant reductions in the arbitrage barriers since 2009 indicated by the deviations from the Covered Interest Parity (CIP) also noting the asymmetric nature of barriers on capital inflows and outflows.
A number of studies were consulted to find out what variables influence the covered interest differential. Studies have tried to explain the deviations of CIP equilibrium from its theoretical limits. Frenkel and Levich (1975) in their paper provide a procedure for estimating transaction costs in the market for foreign exchange and securities. He asserts that any deviation from the theoretical limits of the CIP may not be taken as an opportunity for unexploited profits. There may be other factors such as lags in executing arbitrage or differences in demand and supply elasticity in various markets as well as presence of transactions costs. Hence there is band within which the forward premium may deviate without violating the parity theorem. Earlier empirical studies of Branson (1969) has attributed deviations from interest parity to transactions cost while in later studies by Oskooee and Das (1985), Clinton (1988) play down the role of the such costs. Aliber (1973), Dooley and Israd (1980) has stressed the role of political risk for aberrations from parity which was corroborated by Coffey, Hrung and Sarkar (2009) where deviations of Covered interest differential was found to be caused by counterparty risk in the face of crisis in the world financial market. Taylor (1989) observes that market turbulence as well as length of maturity of assets is positively related to the profitable arbitrage opportunities in the markets. Borio et. al (2016) attribute demand to hedge USD forwards and proxies for the risks associated with CIP arbitrage better explains the deviations from CIP than bank credit and liquidity conditions. Fixed income spreads and nominal interest rates were two factors were identified by Du, Tepper, Verdelhan (2018) by again discrediting any role by transactions cost or the credit risk posed for influencing the deviations in the parity. Studies estimating and examining CIP in the Chinese context by Cheung and Qian (2010), Cheung and Herrala (2014) tried to explain the deviations by several macroeconomic and country risk factors. The latter study found credit market tightness indicators and NEER to have an influence on the CID whereas the former got more macroeconomic variables that influence the differential including the capital flight variable. Both studies found the CID to be persistent.
Using onshore data Ferreira (2011) establishes that earlier members of European Union have CIP across all maturities; countries where CIP does not hold are inferred to have incomparable assets or data. Skinner and Mason (2011) found that CIP holds for advanced economies whereas it did not hold for emerging economies for longer maturities. The causes are found to be various aspects of credit risk rather than the transactions cost. Another paper by Casavas (2012) established the no arbitrage condition in CIP by appending a band around which deviations could take place. The study is for developed economies during the year 20082011.
Bhatt and Virmani (2005) studied India’s money market integration with the world by measuring the uncovered as well as the covered interest parity. He found the CIP to exist, albeit weak, and UIP failing to hold during April 1993 to March 2002, ascribing the cause of which to existence of exchange risk premium over and above the expected depreciation of currency. Jain and Bhanumurthy (2005) found that there is a strong integration of the domestic call money market with the London Interbank Offered Rate (LIBOR). Misra and Mahakud (2009) observed violation interest parity in their study spanning April 1993 to March 2003.
DATA AND METHODOLOGY:
In case of Covered Interest Parity (CIP) using the offshore market equation 1 holds:
where is NonDeliverable Forward rate of onemonth maturity given at time t, the symbol _{} represents the interest rate prevalent at home, i_{f} is the foreign interest rate given by the London Interbank Rate (LIBOR) and S_{t} is the spot exchange rate. We transpose the above equation to find out the implied home interest rate given the spot exchange rate, NDF and foreign interest rate LIBOR. Under perfect integration of the markets, such implied interest rate obtained would exactly be equal to the actual interest rate prevailing at home. In other words, if we take a difference between the implied interest rate and the actual interest rate prevalent at home the result would be zero. Hence we rearrange the equation to arrive at the INR covered interest differential, henceforth referred differential as:
In addition to the already explained symbols, i_{t, t+1, ON} is the actual onshore interest rate which is taken to be the Mumbai Inter Bank Rate (MIBOR). The differential so obtained gives the magnitude of aberration found in the ideal interest rate and the actual one. We shall calculate the differential over a period of time to see how it has evolved. We have a sample of monthly data from January 2000 to October 2017 comprising 214 data points each for NDF, LIBOR, MIBOR and Spot USDINR exchange rates.^{1 }The covered interest rate differential, henceforth referred to as differential, calculated as above is known to exist in presence of country and counterparty risk Coffey, Hrung and Sarkar (2009), Dooley and Israd (1980). Following a study by Cheung and Qian (2011) on Chinese covered interest parity we take INR USD Exchange rate Volatility (ExV), Change in Real Gross Domestic Product (dGDP), Inflation Volatility (InfV), Change in Nominal Effective Exchange Rate (NEER) as variables that can explain the differential.^{2} The GDP^{ }at factor cost was available with quarterly frequency, hence it was converted to monthly frequency by the method provided by Chow and Lin (1971). With the help of variables closely related to the GDP and available at monthly frequency, GDP at monthly frequency was interpolated.^{3} Following Jeane (2012) and Cheung and Herrala (2014) we consider more economic determinants for explaining the differential. Change in Volatility Index (VIX), Change in Bank Rate (dBR), Change in Cash Reserve Ratio (dCRR), Change in Exports (dX), Change in Imports (dM) are taken into account.^{4} S & P 500 VIX is also taken into consideration to account for global volatility. Two credit market liquidity indicators are used to see if they have any bearing on the Z_{t}._{ }
On ordinary least squares (OLS) estimation it was inferred from the value of the DubrinWatson statistic that the model suffered from autocorrelation problem. Such autocorrelation might be because of interpolated data used or the nonstationarity of the dependent variable (Gujrati, Porter and Gunasekar, 2012). In order to tackle these problems we perform the CocharanneOrcutt twostep procedure (Cochranne & G.H, 1949). The said procedure not only corrected serial autocorrelation in our case but also rendered the dependent variable stationary. The corrected variable will be depicted by _{}.
However, the above analysis does not account for the possible break in the variables. To that end, we first undertake the QuandtAndrews structural breakpoint test. After settling the breakpoint in the data we go on to estimate the equation taking into account the discontinuity in the regression. Following Stock and Watson (2011) we build an interaction variable by building a dummy and multiplying it with the ‘discontinuous’ variable which takes the value zero before the break and value one after the break. Such interaction variables along with the original variables are regressed on the dependent variable, _{}. The coefficient of the _{} of the respective variable represent the postbreak scenario while other variables are interpreted as usual. We consider first giving a regression to macroeconomic volatility indicators of the Indian economy.
.Eq3
We further use Equation 4 as a general specification to gauge the effects of additional economic determinants along with those mentioned in equation 3, on the Differential
Where _{} is a vector which includes other economic variables in addition to the above macroeconomic stability proxies,_{ }are the vector of relevant interaction variables. Further variables which are integrated of order one are employed to test the long run association among them by the Johansson’s Cointegration test. To see the shortrun dynamics and causation we test VECM along with the Ganger causality test.
EMPIRICAL FINDINGS:
The MIBOR is everywhere greater than the LIBOR. The two variables appear to comove till March 2009 but dissociate with each other afterwards. The differential, calculated by equation 2, remains positive throughout the sample period. Its level however decreases till mid 2006, registers high fluctuations for the next three years and then starts breaking away to increase for a sustained period of time. Such increase coincides with the time when the US subprime lending crisis had hit the financial world. Increased magnitude of the differential around that time might be because of fund reversals from the emerging economies such as India, leading to a high domestic interest rate and hence a greater interest rate cleft leading to greater nonintegration of Indian financial markets with the rest of the world.
Before we start to determine factors affecting the differential we perform an endogenous structural breakpoint test, the QuandtAndrews breakpoint test which establishes the structural breakpoint at June 2008. The estimated Maximum LR Fstatistic (2008M06) has value 6.457538 and the corresponding pvalue of 0.00, furthermore Maximum Wald Fstatistic (2008M06) has the value of 64.57538 and pvalue of 0.00 hence rejecting the null hypothesis of no structural break. The other reported variables also indicate presence of a structural break. The test was conducted with standard 15 per cent trimming.
The differential starts breaking away from the trend to decrease for a couple of quarters and then rise. Table 1 presents the summary statistics of the variables used in explaining the CID, including the differential, where it is also noted that all the variables are stationary and integrated of order zero. To establish the determinants of the interest rate gap we estimate equation 3 and equation 4 by OLS, results of which are presented in Table 2. Among the macroeconomic stability variables presented in Column 1 of Table 2, we see _{} and _{} have positive and significant coefficients implying instability in the economy which leads to rise interest rate cleft depicted by a rise in _{} in the postbreak scenario. _{} also has expected sign of the coefficient indicating strengthening Rupee leads to more financial integration. The second column of Table 2, which considers additional variables, reports _{}and _{} to have the significant positive and negative coefficients post break. We observe that the change in trade variables _{} and _{} show opposite effect on the differential while _{}gives expected negative significant coefficient.
In order to further investigate the causes that drive the difference between the interest rates we resort to Johansen Cointegration test. Results of unit root tests are presented in Table 3, to find out I(1) variables to consider testing three models where variables of different nature are tested to see if they have any long run equilibrium relationship exists. First model is where _{} and _{} are considered, they being a part of macroeconomic stability variables. Model II considers _{} and _{}, the trade variables. Model III sees whether _{}is has a long run relationship with the credit market tightness indicator viz. _{}and _{}.The result for Johansen Cointegration is presented in Table 4.
Table 1: Descriptive statistics of the variables:
Z_{t}^{*} 
dVIX 
ExV 
dNEER 
dGDP 
dBR 
dCRR 
dX 
dM 
dInfV 

Mean 
5.607223 
1.568883 
0.328534 
0.18779 
0.400692 
0.0244 
0.29692 
1.183976 
1.222004 
0.665394 
Median 
5.283508 
1.50851 
0.223879 
0.1977 
1.36711 
0 
0 
1.837356 
0.10223 
0.625196 
Maximum 
11.13251 
134.571 
2.461666 
6.022805 
23.21975 
58.33333 
15 
36.60125 
36.13685 
1.199685 
Minimum 
2.335264 
38.4898 
0.82081 
6.58665 
19.0812 
12.5 
33.3333 
27.8797 
26.6553 
0.54271 
Std. Dev. 
1.419525 
21.29049 
0.339234 
1.740796 
6.376226 
4.81331 
3.868723 
11.99422 
11.44128 
0.112512 
Skewness 
1.06674 
1.755452 
2.095067 
0.28991 
1.03325 
8.971972 
3.97655 
0.09666 
0.415263 
1.886934 
Kurtosis 
4.957893 
10.37899 
11.56882 
4.617669 
4.132974 
106.7783 
34.44111 
2.909882 
3.138906 
6.945105 
JarqueBera Probability 
74.06821 (0) 
589.8544 (0) 
803.6728 (0) 
26.08519 (0.000002) 
49.0608 (0) 
97978.64 (0) 
9290.856 (0) 
0.40186 (0.81797) 
6.26344 (0.043643) 
263.2857 (0) 
ADF 
3.77091 (0.0038) 
16.9369 (0) 
5.45851 (0) 
11.6651 (0) 
4.28511 (0.0006) 
14.1661 (0) 
12.1309 (0) 
3.69696 (0.0048) 
4.1877 (0.0009) 
5.10903 (0) 
PP 
8.979213 (0) 
18.7126 (0) 
9.09732 (0) 
11.5971 (0) 
7.62942 (0) 
14.2584 (0) 
12.2127 (0) 
22.0114 (0) 
16.0059 (0) 
5.05646 (0) 
Observations 
212 
212 
212 
212 
212 
212 
212 
212 
212 
212 
Source: Mentioned in the text.
Note: Figures in parentheses in the last two rows are the pvalues. ADF and PP test are done at level and including intercept term only.
Table 2: Results of the regression on Differential Z_{t}^{* }
(1) 
(2) 

dVIX 
0.00649 (0.006581) 
0.00711 (0.006203) 
DUMMY * dVIX 
0.00561 (0.007826) 
0.00503 (0.007385) 
Ex_V 
0.856623 (0.736496) 
0.573894 (0.738439) 
DUMMY * Ex 
0.574547 (0.789626) 
0.635275 (0.78695) 
dNEER 
0.17658** (0.07381) 
0.13928** (0.071134) 
DUMMY * dNEER 
0.090582 (0.09359) 
0.03741 (0.090732) 
dGDP 
0.002833 (0.013995) 
0.005913 (0.015501) 
DUMMY * dGDP 
0.050822** (0.025479) 
0.036287 (0.026733) 
dBR 

0.043073 (0.046163) 
DUMMY * dBR 

0.021076 (0.048595) 
dCRR 

0.021165 (0.031341) 
DUMMY * dCRR 

0.08176** (0.04018) 
dX 

0.022119** (0.010596) 
DUMMY * dX 

0.001527 (0.015021) 
dM 

0.02024* (0.011252) 
DUMMY * dM 

0.015405 (0.016471) 
InfV 
0.56203 (0.779082) 
0.14858 (0.746183) 
DUMMY * InfV 
1.514537*** (0.356848) 
1.482169*** (0.347836) 
C 
4.972988*** (0.476103) 
4.771792*** (0.452213) 
Adjusted R^{2} 
0.434415 
0.500945 
BreuschGodfrey LM test (H_{0}: No serial correlation) 
0.126172 (0.7228) 
2.607963 (0.108) 
BreuschPaganGodfrey test (H_{0}: Homoskedasticity) 
1.404436 (0.1803) 
1.220132 (0.248) 
No. of observations 
212 
212 
Source: Authors’ estimation.
Note: Standard errors are presented in parentheses. BGLM and BPG test present the Fstatistic with pvalue in the parentheses. ‘***’ denotes 1% level of significance, ‘**’ denotes 5% level of significance and ‘*’ denotes 10% level of significance
Table 3: Result of unit root test on variables
ADF at level 
ADF at 1^{st} difference 
PP at Level 
PP at 1^{st} difference 
Order of integration 

Zt 
2.36045 (0.154348) 
17.2291 (0.00) 
2.15271 (0.224505) 
17.5614 (0) 
I(1) 
GDP 
0.253324 (0.9753) 
4.11456 (0.0012) 
1.42746 (0.5683) 
8.71715 (0) 
I(1) 
X 
1.5662 (0.498) 
3.94325 (0.0021) 
1.9962 (0.2885) 
21.3136 (0) 
I(1) 
M 
2.12069 (0.2368) 
9.32031 (0) 
2.16126 (0.2213) 
9.32031 (0) 
I(1) 
NEER 
0.97601 (0.7618) 
11.4325 (0) 
0.81509 (0.8125) 
11.3166 (0) 
I(1) 
ExV 
5.89458 (0) 
9.82061 (0) 
I(0) 

VIX 
4.01216 (0.0016) 
3.95955 (0.002) 
I(0) 

InfV 
5.25503 (0) 
5.22578 (0) 
I(0) 

CRR 
2.72932 (0.0708) 
12.4967 (0) 
2.64384 (0.0859) 
12.4957 (0) 
I(1) 
BR 
2.09352 (0.2476) 
12.0675 (0) 
1.93112 (0.3176) 
14.1973 (0) 
I(1) 
Source: Authors’ estimation.
Note: ADF Augmented DickyFuller; PP PhillipsPerron. tvalues are reported with pvalues in parentheses.
Table 4: Results for Johansen’s Cointegration Test
Model I* 
Model II** 
Model III 

Hypothesised no. of CE(s) 
Trace statistic 
Max Eigen Statistic 
Trace statistic 
Max Eigen Statistic 
Trace statistic 
Max Eigen Statistic 
None 
59.15962 (0.0000) 
43.82692 (0.0000) 
68.69294 (0.0000) 
50.15565 (0.0000) 
31.76012* (0.0293) 
16.77112 (0.1831) 
At most 1 
15.3327 (0.0529) 
11.97011 (0.0529) 
18.53729 (0.0168) 
15.38864 (0.0331) 
14.989 (0.0595) 
11.57118 (0.1278) 
At most 2 
3.362587 (0.0667) 
3.362587 (0.0667) 
3.148659 (0.076) 
3.148659 (0.076) 
3.417818 (0.0645) 
3.417818 (0.0645) 
Source: Authors’ estimation.
Note: ‘*’ and ‘**’ indicates presence of one and two cointegrating relationships at 5 percent level of significance. Model I and II uses the SIC while Model III uses the AIC criterion for optimal lag length determination.
From Table 4 it is seen that there are one and two cointegrating equations in the models under study. We perform the VEC estimations to find the significance of association between variables. The longrun cointegrating equations obtained from VECM estimations are presented below with their absolute tvalues given below their coefficients.
Model I 

Model II 



Model III 

We see expected signs of long run relations of Z_{t} with GDP and NEER, indicating that as the economy’s output rises and currency strengthens in the long run the differential narrows, implying greater integration. In Model III tightening credit market in the form of increased BR raises the differential but the other indicator has an opposite effect. Higher import is also seen to widen the differential. It must however be noted that even though the error correction term for the first two models are negative, it is statistically significant for the second model’s first equation only with value 0.07954 and absolute tvalue 2.55693. The third model’s error correction term has positive and insignificant coefficient hence ruling out any longterm causality^{5}. In order to delve deeper into causation and short run dynamics, we test for Granger causality.
Table 5: Results for Granger Causality Test

Dependent Variable 


Model I 
Model II 
Model III 
Excluded Variables 
D(Z_{t}) 
D(Z_{t}) 
D(Z_{t}) 
D(GDP) 
1.788802 (0.4089) 
 
 
D(NEER) 
0.559968 (0.7558) 
 
 
X_{t} 
 
0.134332 (0.7140) 
 
M_{t} 
 
3.387375 (0.0657) 

CRR 
 
 
3.670667 (0.1596) 
BR 
 
 
24.54321 (0.0000) 
Source: Authors’ estimation.
Note: Values reported are Chisquares with their pvalues in parentheses. Lag length for Model I and II are decided by AIC criterion and SC for Model III.
From Table 5 we see that imports Granger cause Z_{t} at 10 per cent level while credit market tightness indicator BR causes the differential at 1 per cent level of significance. Rest of the variables does not have a statistically significant causation.
CONCLUSION AND LIMITATIONS:
The evidence in the analysis suggests Indian economy is sparingly integrated with the rest of the world as the value of the covered interest differential is significantly large. We also see that the differential worsened midway the sample period and increased around the time of global financial crisis. We infer that low inflation volatility, strong Rupee and growth in GDP make a more conducive environment for greater financial integration of India with the rest of the world. We see that high macroeconomic stability leads to lessening of the differential, hence more financial integration in the country. Tighter credit market indicated by higher bank rate and higher imports worsen integration in the long run as well as in the short run. Hence stable economy and easier credit market must go a long way in integrating the Indian economy to the world financial markets. (Ma, Ho, & McCauley, 2004)
In order to understand financial integration of a country more comprehensively, one must consider including more segments of the financial market viz. stock, bond market and others. We must however be cautious while interpreting the coefficients of changes in GDP, exports and imports since they use the data of GDP which is interpolated and at best an approximation of real values.
Notes:
1. Data on NonDeliverable Forward and LIBOR is obtained from Bloomberg. MIBOR and Spot exchange rates are sourced from the Database on Indian Economy, Reserve Bank of India.
2. The data for NEER, real GDP (nominal GDP deflated by WPI) and Inflation volatility has been adjusted have the same base year of 200405. Inflation Volatility is calculated as the standard deviation of the GARCH (1,1) of the Wholesale Price Index. Above mentioned variables along with the USD INR exchange rate have been sourced from Database on Indian Economy, Reserve Bank of India.
3. IIP, credit disbursement by all scheduled commercial banks and L2 measure of liquidity was used to interpolate monthly GDP. The variables have year 200405 base and sourced from Handbook of Indian Economy, Reserve Bank of India.
4. Import and export data are deflated by their respective monthly GDP. The change in each variable implies monthly percentage change of the respective data. The data on VIX is from Chicago Board Options Exchange (CBOE) website.
5. Full results are not presented in the article to economize the use of words and space.
REFERENCES:
1. Aliber, R. Z. (1973). The interest rate parity theorem: A reinterpretation. Journal of Political Economy, 81(6), 145159.
2. BaamaniOskooee, M., & Das, S. P. (1985). Transaction Costs and the Interest Parity Theorem. Journal of Political Economy , 793799.
3. Baba, N., & Packer, F. (2009). Interpreting deviations from covered interest parity during the financial. Journal of Banking & Finance , 19531962.
4. Bhatt, V., & Virmani, A. (2005). Global integration of India's money market: Interest rate parity in India, Working Paper, No. 164, Indian Council for Research on International Economic Relations (ICRIER), New Delhi
5. Borio, C., McCauley, R. N., McGuire P., & Sushko V. (2016). Covered interest parity lost: understanding the crosscurrency basis. BIS Quarterly Review
6. Branson, W. H. (1969). The minimum covered interest differential needed for international arbitrage activity. Journal of Political Economy, 77(6), 102835.
7. Cheung, Y.W., & Herrala, R. (2014). Chinas’s capital controls: Through the prism of covered interest differentials. Pacific Review Economic , 112134.
8. Chow, G. C., & Lin, A.l. (1971). Best Linear Unbiased Interpolation, Distribution, and Extrapolation of Time Series by Related Series. The Review of Economics and Statistics , 28 (3), 372375.
9. Clinton, K. (1988). Transactions costs and covered interest arbitrage: Theory and evidence, Journal of Political Economy, 96(2), 358–370.
10. BIBLIOGRAPHY\l 1033 Cochranne, D., & G.H, O. (1949). Application of Least Squares Regression to Relationships Containing AutoCorrelated Error Terms. Journal of the American Statistical Association , 44 (245), 3261.
11. Coffey, N., Hrung. W., & Sarkar, A. (2009). Capital constraints, counterparty risk, and deviations from covered interest rate parity,” Staff Report 393, Federal Reserve Bank of New York.
12. (2016). Covered interest parity with default risk. The European Journal of Finance, 22(12),11301144
13. Dadush, U., Dasgupta, U & Ratha, D. (2000). The Role of Shortterm Debt in Recent Crises. Finance and Development, 37(4), 5457.
14. Dooley, M. P., & Peter, I. (1980). Capital Controls, Political Risk, and Deviations from InterestRate Parity. Journal of Political Economy , 370384.
15. Du, W., Tepper, A., & Verdelhan, A. (2018). Deviations from Covered Interest Rate Parity. The Journal of Finance , 915957.
16. Ferreira, P. (2011). Monetary Integration in the European Union. Journal of Emerging Market Finance , 93120.
17. Frankel, J. (1992). Measuring international capital mobility: A review. American Economic Review, 82(2),197–202.
18. Frenkel, J.A., and Levich, R.M. (1975). Covered interest arbitrage: unexploited profits?. Journal of Political Economy, 83( 2), 325–38.
19. BIBLIOGRAPHY\l 1033 Gujrati, D. N., Porter, D. C., & Gunasekar, S. (2012). Basic Econometrics (5th Edition ed.). McGraw Hill.
20. Greene, W. H. (2003). Econometric Analysis. New Jersy: Prentice Hall.
21. Hansen, B. E. (1997). Approximate Asymptotic P Values for StructuraSChange Tests. Journal of Business and Economic Statstics , 15 (1), 6567.
22. Higgins, P., & Humpage, O.F. (2005). Nondeliverable forwards: can we tell where the renminbi is headed?. Federal Reserve Bank of Cleveland, September
23. Hutchison, M.M., Pasricha, G.K., & Singh, N. (2012). Effectiveness of Capital Controls in India: Evidence from the Offshore N Market. IMF Economic Review, 60(3), 395 438
24. Jain, S., & Bhanumurthy, N. (2005). Financial Markrts Integration in India. AsiaPacific Development Journal , 1532.
25. Jeanne, O. (2012). Capital account policies and the real exchange rate. NBER Working Paper
26. Series # 18404
27. Kose, M.A, Prasad, E, Rogoff, K & Wei, SJ. (2009). Financial globalization: a reappraisal. IMF Staff Papers, 56(1), 862.
28. Misra,S and Behera, H. (2006). NonDeliverable Forward Exchange Market: An Overview. Reserve Bank of India Occasional Papers, 27(3) 25–55.
29. Misra, A. K., & Mahakud, J. (2009). Emerging trends in financial markets integration: the Indian experience. International Journal of Emerging Markets , 235251.
30. Ma, G., Ho, C., & McCauley, R. N. (2004). The markets for nondeliverable forwards in Asian Currencies. BIS Quarterly Review.
31. Qian, X. (2011). Deviations from Covered Interest Parity: The Case of China. In Frontiers of Economics and Globalization.
32. Skinner, F. S., & Mason, A. (2011). Covered interest rate parity in emerging markets. International Review of Financial Analysis , 355366.
33. Stock, J. H., & Watson, M. W. (2011). Introduction to Econometrics. AddisonWesley.
34. Taylor, M. P. (1989). Covered Interest Arbitrage and Market Turbulence. The Economic Journal , 376391.
35. Taylor, M. P. (1987). Covered Interest Parity: A HighFrequency, HighQuality Data Study. Economica , 429438.
Received on 14.06.2019 Modified on 07.07.2019
Accepted on 27.07.2019 ©A&V Publications All right reserved
Res. J. Humanities and Social Sciences. 2019; 10(3):840846.
DOI: 10.5958/23215828.2019.00138.4