IMPACT OF OIL TRADE ON EXCHANGE RATE OF INDIA

Introduction
India in the 21st century is one of the fastest growing countries of the world. Oil being the bloodline of the growing economy, is a necessary commodity and has a very inelastic demand, steadily growing with time. In 2011, India was the fourth largest energy consumer in the world after the United States, China, and Russia. India's economy grew at an annual rate of approximately 7 percent since 2000 and proved relatively resilient to the 2008 global financial crisis. India was the 10th largest economy in the world in 2011, as measured by nominal gross domestic product (GDP). In the International Energy Outlook 2011, EIA projects India and China to account for the biggest share of Asian energy demand growth through 2035. India is heavily dependent on crude oil imported from the Middle East and imports more than 70% of its domestic demand. Due to a stagnation of domestic production, the import of crude has gone up from 11.68 million tons (mt) in 1970–1971 to 196 mt in 2007–2008. Oil import bill for India in 2007–2008 was $144.93 billion. With the high demand of oil and other petroleum, and their fluctuating price in the global markets, we are at a very high risk of foreign exchange risk. With so much purchase of energy imports, it might lead to exchange rate movements. And the volatility in the exchange rates (caused by the oil price volatility) may have severe effects on the economy, especially on infrastructural projects and FDI. There have been new studies which are even extending the linkage of the oil rate and exchange rate to the stock markets. Oil is the second largest source of primary commercial energy in India after coal. Share of oil in India’s total energy consumption is 31% (source: EIA). India was under fixed exchange regime till March 1992, where exchange rate of rupee was determined and adjusted by Reserve Bank of India (RBI), the central bank. Following economic reforms and liberalization, rupee was initially made partially convertible and finally went ahead with full convertibility on the current account. RBI, however, intervenes in foreign exchange market from time to time to prevent extreme volatility. This study probes oil price-exchange rate nexus for India, an oil importing developing country, within a generalized autoregressive conditional heteroscedasticity (GARCH) framework, following the work of Bollerslev [9] and Nelson [10]. The paper uses daily data in accordance with the studies of Narayan and Narayan [4] and Hammoudeh and Yuan [11] for the span 1 July, 2010 to 1 December, 2011.

Acknowledgement
I would like to sincerely thank Prof. Archana Srivastava (Dept. of Economics, BITS-Pilani K.K Birla Goa Campus) for her constant support and guidance on this project. Her knowledge and work constantly inspired me to work ahead in the project.

Literature Review
Amano and Norden (1998) find a stable linkage exists between oil price shocks and the US real effective exchange rate over the longer horizon. Their findings indicate that oil prices have been the dominant source of persistent shocks on real exchange rate. Chaudhuri and Daniel (1998) obtain similar results, asserting that the main source of US real exchange rate fluctuations comes from the real price of oil. Other studies confirming the significant impacts on real exchange rates in developed countries from oil price shocks include Dibooglu and Koray (2001) and Zhou (1995). The ability of real oil prices to explain movements in the US real effective exchange rate. The potential importance of the price of oil for exchange rate movements has been noted by, inter alia, Krugman (1983) and Golub (1983). They construct multi-country model to explain the transfer of wealth due to oil trade and other exports, to explain the fluctuation of exchange rates. Narayan and Narayan examine the relationship between oil price and the Fiji-US exchange rate using daily data for the period 2000–2006. They use the generalised autoregressive conditional heteroskedasticity (GARCH) and exponential GARCH (EGARCH) models to estimate the impact of oil price on the nominal exchange rate. In favor of the presence of cointegration between the real exchange rates and the real oil prices, Chen and Chen (2006) performed a rigorous analysis. They examined the long-run relationship between real oil prices and real exchange rates for the G7 countries. They showed that real oil prices may have been the dominant source of real exchange rate movements and found evidence of a cointegrating relationship between real oil prices and real exchange rates. Tiwari et al. find causality using a wavelet based analysis for India between the rupee exchange rate and oil prices is frequency dependent. At lower time scales (high frequency), no causal relationship is found; but at higher scales (low frequency) they find causality. In particular, evidence of unidirectional causality from exchange rates to oil prices was found.

Methodology
Daily data on Brent Crude oil and rupee-dollar exchange rate have been collected from the websites of Energy Information Administration (EIA) (www.eia.doe.gov) and Reserve Bank of India (www.rbi.org.in). Present study uses nominal data because of unavailability of daily consumer price index. According Narayan et al., tracking the daily behavior of oil price and exchange rate does not require knowledge of real values. Daily returns on oil price and exchange rate are calculated based on the following formula: ( )

where yt and yt-1 are price of oil or exchange rate for the periods t and (t -1). Let grext and groilt be the daily returns on oil price and exchange rate on tth date. The mean equation can be written as

grext  c   groilt   t
The corresponding GARCH-M model is grext  c   groilt   t2   t

The variance equation for GARCH (p, q) has the following form

∑

∑

The mean of the volatility equation is denoted by w. represents the size effect, which indicates how much volatility increases irrespective of the direction of the shock. Both series display volatility and volatility clustering, although volatility clustering seems to be more in magnitude in the case of the oil price series. As shown in Table 1, the statistics relating to Skewness, Kurtosis and Jarque–Bera reveal that the variables are non-normal in nature. The descriptive statistics for the two series are presented in Table 1. TABLE 1: Variable groilt grext Max. 0.164137 0.024903 Min. -0.12827 -0.03007 Std. dev 0.029218 0.005261 Skewness 0.003861 0.049767 Kurtosis 7.509083 7.838241 J-B 310.9(0.0) 358.107 (0.0)
*values in brackets are probability values

Empirical Analysis
At the first stage, stationarity of the variables has been examined. Table 2 presents the results of unit root tests based on augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) statistics on the level of the variables. In ADF and PP tests, the null hypothesis is the series has a unit root against the alternative of stationarity. The tests reveal that both the series are stationary in nature. Absence of non-stationarity of individual series rules out the possibility of a cointegrating relationship. TABLE 2: Level(constant and trend) Variable groilt grext ADF -9.240709 -4.333661 PP -20.07850 -17.90929

In the next stage, Eq. (2) has been estimated using ordinary least square (OLS) technique, results of which is shown in Table 3. The variable groil is found to be statistically significant at 5% level in grex equation. Diagnostic tests reveal that the residuals are free from serial correlation up to 36 lags based on Ljung-Box Q-statistics though ARCH-LM test up to 36 lags appears to be statically significant, indicating the presence of ARCH effect. To deal with the ARCH effect present in the residual series, GARCH (1, 1) and GARCH (1, 1)M models have been estimated using maximum likelihood estimation procedure assuming normally distributed errors. Optimal orders of the GARCH models are determined based on SBC (Schwartz Bayesian Criteria). We find p=1 and q=1. As shown in Table 3, the mean equation of GARCH (1, 1) model reveals that an increase in oil price has a negative impact on nominal exchange rate. A 10% increase in the oil price return leads to 0.12% depreciation of Indian currency vis-à-vis US dollar. The residual series is found to be free from autocorrelation and ARCH effects. In GARCH (1, 1)-M equation,  is found to be statistically insignificant as shown in Table 3. This implies that exchange rate volatility has no impact on the exchange rate itself.

TABLE 3: Estimation of the models Variable/parameter OLS 0.0005 c (0.0002) -0.047  (0.009)  II. Variance Equation w

GARCH(1,1) 7.36E-05* (0.0001) -0.012 (0.005) -

GARCH-M(1,1) -0.0001* (0.0002) -0.012 (0.005) 19.94* (13.44) 7.59E-07 (1.92E-07) 0.30 (0.049) 0.71 (0.032) 3.12 [0.86] 18.75 [0.76] 31.83 [0.66] 3.90 [0.68] 22.60 [0.54] 31.52 [0.68]

-

7.49E-07 (1.92E-07) 0.30 (0.05) 0.722 (0.033) 2.79 [0.83] 17.64 [0.82] 30.69 [0.71] 3.905 [0.68] 23.33 [0.50] 32.24 [0.64]

III. Diagnostics Q-statistics(6) Q-statistics(24) Q-statistics(36) Arch-LM(6) Arch-LM(24) Arch-LM(36)

4.46 [0.61] 32.47 [0.11] 42.55 [0.21] 35.44 [0.00] 89.30 [0.00] 89.46 [0.00]

* Statistically insignificant at 5% level. Figures in () are standard errors and that in [ ] are probability values.

Conclusion
This paper probes oil price and exchange rate relationship for India using daily data for a period of extreme oil price volatility. The study reveals that the return of oil price – exchange rate relationship exhibits time-varying volatility. GARCH and EGARCH models have been employed to examine the impact of oil price shocks on nominal exchange rate returns. The conclusions of this study are as follows. An increase in the oil price return leads depreciation of Indian currency vis-à-vis US dollar. Being an oil importing country, such oil price – exchange rate relationship is in accordance with theoretical consistency. So, an increase or expected increase in international oil price obligates Indian refineries to procure excess dollars to pay for costlier oil import leading to a depreciation of Indian currency. The results are consistent with many such studies, which imply an important relationship between oil price and exchange rates. It shows us how oil price changes will have an impact on exchange rates and thus various other markets based on imports. We believe that our modeling exercises can be interpreted as shedding light on the observation that the shocks from real aspects do have an important and significant bearing on the determination of both long- and short-run exchange rates in India. Such oil price – exchange rate nexus should also have significant impacts on Indian Stock Market. Taking a cue from Markowitz, for foreign investors, a depreciation of the Indian currency can lead to a portfolio switch from domestic assets, such as stocks, to foreign assets since depreciation reduces returns when these funds are translated to the home currency. For the internationally- diversified domestic investor, the depreciation of the Indian currency would cause foreign stocks to be more expensive. The investor would substitute foreign assets by domestic assets and hence domestic stock price would increase due to increased demand. Future study, thus, should be directed to examine dynamic relationship between international oil price, exchange rate and India Stock Market.

References
1. Amano, Robert A., and Simon Van Norden. "Oil prices and the rise and fall of the US real exchange rate." Journal of international Money and finance 17.2 (1998): 299-316. 2. Dawson, Jennifer C. "The effect of oil prices on exchange rates: a case study of the Dominican Republic." The Park Place Economist 14 (2006): 1-9. 3. Al‐mulali, Usama, Che Sab, and Che Normee Binti. "The impact of oil prices on the real exchange rate of the dirham: a case study of the United Arab Emirates (UAE)." OPEC Energy Review 35.4 (2011): 384-399. 4. Narayan, Paresh Kumar, Seema Narayan, and Arti Prasad. "Understanding the oil priceexchange rate nexus for the Fiji islands." Energy Economics 30.5 (2008): 2686-2696. 5. Tiwari, Aviral Kumar, Arif Billah Dar, and Niyati Bhanja. "Oil price and exchange rates: A wavelet based analysis for India." Economic Modelling 31 (2013): 414-422. 6. Golub, Stephen S. "Oil prices and exchange rates." Economic Journal 93.371 (1983): 576-93. 7. Chaudhuri, Kausik, and Betty C. Daniel. "Long-run equilibrium real exchange rates and oil prices." Economics Letters 58.2 (1998): 231-238. 8. Akram, Q. Farooq. "Oil prices and exchange rates: Norwegian evidence." the econometrics Journal 7.2 (2004): 476-504. 9. Bollerslev, Tim. "A conditionally heteroskedastic time series model for speculative prices and rates of return." The review of economics and statistics(1987): 542-547. 10. Nelson, Daniel B. "Stationarity and persistence in the GARCH (1, 1) model."Econometric theory 6.03 (1990): 318-334. 11. Hammoudeh, Shawkat, and Yuan Yuan. "Metal volatility in presence of oil and interest rate shocks." Energy Economics 30.2 (2008): 606-620.