Dynamic Relationship between Crude Oil, Gold and Exchange Rates: The Case of India

 

Dr. Manu K S¹*, Harshita Agarwal²

¹Assistant Professor, Department of Management Studies, Christ University, Bangalore, Karnataka

²Student, BBA, Department of Management Studies, Christ University, Bangalore, Karnataka

 

ABSTRACT:

Gold and Crude Oil are the most trading commodities in the international market. EURO, USD, JPY and GBP are the four most trading currencies available for the market participants to trade. There are many factors which determine the changes in the Exchange rates. Similarly, change in the Gold prices and Crude Oil prices might impact the changes in the determination of Exchange rates. The study considered US Crude Oil (WTI- West Texas Intermediate) Prices,IB Oil(Indian Crude Basket)and Gold prices to assess the relationship. The study used Multiple Linear Regression Analysis and Granger causality test to analyze the relationship between gold, crude oil and exchange rates. The study found that only gold returns has positive impact on EURO/INR returns. Further granger causality test results revealsuni directional causality from Euro returns to Gold returns, Gold returns to JPY returns, IB Oil returns to JPY Returns and IB Oil returns to USD returns. The market participants are suggested to consider the effect of gold returns and IB Oil performance to take investment or trading decision in forex market (JPY/INR, USD/INR and EURO/INR exchange rates).

 

 

 

INTRODUCTION:

The Indian Commodities market has been facing continuous fluctuations in the prices of Crude Oil, Gold Prices and exchange rates. Due to these changes in the prices of the most traded commodities on the markets, it becomes an important area of investigation to find out the effects of the changes and to determine the relationship between these variables.

 

The main reasons to analyze the relationship are: first, from the view point of the investors who want to invest in the diversified portfolios, to understand better the consequences of selecting each commodity and secondly, the nature of crude oil, gold prices and exchange rates, the relation between them and these major financial assets dependencies on each other.

 

The total world oil production as per the records of 2016 was 80,622,000 barrels per day (source: U.S. Energy Information Administration), 66% of the production resulted from Non- OPEC and 44% of the production came from the fourteen current members of the OPEC. The top three producers, since 2014 have been producing around 9 to 11 million barrels of oil per day (source: Wikipedia).  The Oil industry and the gas industry are considered among the top 10 industries worldwide based on their revenue and are considered as the most powerful sector worldwide. Among top 10 wealthiest companies, six belong to the oil industry (source: statista.com). The oil prices have been constantly on a rise till 2014 but from 2015 the benchmarks of crude oil prices–WTI, Brent Crude Oil have been significantly decreasing.  As per the statistics of March 2017, The OPEC Crude Oil prices were on the average of 50.32 USD per barrel which have reduced from the average annual price of 52.03 USD per barrel.  It has been expected by many top oil producing companies to increase their production till 2020.  As of 2008, the total gold mine production in 2008 was approximately 2,280 metric tones and it increased to 3,000 metric tons by 2015. The maximum gold reserves worldwide cannot be more than 56,000 metric tons. The size of the Gold market is approximately 2.5 trillion dollars as per World Gold Council. The forex market is the world’s most traded market. The size of the market comprises of almost 5 trillion USD value transactions per day.  U.S. Dollar is considered to be the strongest currency and it is the International currency available for trading.  As it is considered the strongest currency that is the reason it is paired with every other currency. (Source: dailyfx.com).

 

Crude Oil and Gold Prices have seen a drastic fluctuation in their prices. The change in one variable prices cause the effect on the other. For example, if the crude oil prices are increasing, there is a higher demand on the USD which appreciates the dollars against the home currency. Thereby, causing effect on the exchange rates. Similarly, gold can be said to be somewhat dependent on the exchange rates. It does not have a greater impact on the exchange rates. After the fall of the Bretton Woods system after 1971 and the entry of the Flexible exchange rates regime, there have been major fluctuations in the gold prices. Gold performs a very unique function in the world as it performs store of value and the medium of exchange functions. The main trading center for gold is London and it accounts for approximately 70% of the world wide trading of gold followed by US markets and Shanghai Stock Exchange. Any changes in the gold prices simultaneously affect the exchange rates of where they are traded. The fluctuations in the gold prices are due to many reasons like supply and demand, international trade, monetary policies, value of US Dollar, interest rates etc. The relationship between gold and exchange rates is very a discussed topic as gold is used to hedge against the dollars and inflation. So, whenever dollar exchange value is decreasing, gold is required to buy more dollars which in turn increases the value of gold. Thus, there exists a relationship between exchange rates and gold and exchange rate sand crude oil.

 

Crude Oil and gold prices are said to be positively related and it has been seen also from the past data that gold prices and crude oil prices move in the same pattern. During the period of 1010to 2016, both are seen falling at the same time and also depicting inverse relationship with the exchange rates. From investor’s point of view buying and selling of crude oil, gains significance to determine the impact of changes in the crude oil prices. When countries import crude oil, say Brent Crude Oil then the importing country’s purchasing power is affected by the price of crude oil. So, the country’s exchange rate may depreciate or appreciate depending on the price of WTI or Brent Crude Oil or Indian Crude Oil. Thus, there comes a problem as to what is the impact of crude oil prices which result in change (depreciate or appreciate) in the exchange rates of the country.

 

LITERATURE REVIEW:

Several researchers attempted to find a link between different macroeconomic factors and bond market with the use of varying tools for analysis, for different time periods and also with relevance to different countries. Kumar and Onkar  (2016) identified that that all the selected variables including fiscal deficit, inflation rate, etc., except GDP and Trade openness, meaningfully reflected the volumes of corporate bonds with the use of regression and unit root test. Using a different approach, Sensarma and Indranil (2016) carried out research on the result of monetary policy on the shape of the corporate yield curve and credit spread using a macro finance perspective with the variables of output, price and exchange rate. Anurag Agnihotri (2015) similarly used the findings of regression test to prove that bonds of varying maturities usually move together with macroeconomic variables. On the other hand, Kapingura and Makhetha (2014), through the use of Grangular causalty and Grangular cointegration, reflected that policy initiatives for bond market development can help in domestic resource mobilization and promote economic growth. While, Fredrick (2014) used a causal research to establish that exchange rate, interest rate, GDP per capita and GDP at purchasing power parity have an impact on the bond market development. Even Chowdhury et al (2013) used regression for studying bond market spreads. The research undertaken by Nair and Thenmozhi (2012) found that the predictive power of the macroeconomic factors is higher in case of India and Brazil as compared to the developed bond markets. In similar studies, Baele et al (2010) interpreted that macro factors are useful for understanding bond return volatility using the tool of correlation. In contrast to this, Bekaert, et al (2010) studied both stock and bond return in the same research. Sydney C. Ludvigson and Serena (2009) conducted a research study to analyze if there are any prominent cyclical fluctuations in the bond market premiums and identify with which macroeconomic factors these premiums fluctuate. The study was conducted in the US bond market context.  In contrast, Adelegan et al (2009) conducted a study in sub Saharan Africa to find factors behind bond market development. Rasheed Ameer (2007) also conducted a similar study specific to the Asian countries of Malaysia and Korea. Furthermore, John D. Burger and Francis E. Warnock (2007) studied the reasons behind the local bond market development using tool of regression. While Diebold et al (2006) used chi square to find a strong impact on the yield curve. Goeij et al (2006) used univariate analysis for research on bond market. In other studies, again based in US, Fair (2003) tried finding impact of macro related events on the stock, bond prices and exchange rate where it was found that change in an exchange rate compared to the change in the bond price is lesser for price events than for monetary and real events. Various other approaches were used where Andrew Ang and Piazzesi (2003), who used Vector Auto regression model and found 80% variation in bond yields. Lingfeng (2002) also undertook correlation analysis but by using an asset pricing model for G7 nations. Pierre Collin, Robert Goldstein and Spencer Martin (2001) used correlation and regression analysis to study credit spread changes. While Balduzzi et al (2001) concluded that there was increase in volatility of US treasury bonds after macroeconomic announcements. In contrast to this, Roland Beck (2001) highlighted that the developing market Eurobond spreads after the Asian crisis can be studied through market expectations and looked at the Asian context. On the other hand, Bollerslev, et al (2000) studied the return volatility in US Treasury bond futures contracts. Further analysis carried out for the US government bonds, but using GARCH model, was by Charlotte Christiansen (2000) and Ernst Konrad for Europe. Even Jonesa et al (1998) examined the response of Treasury bond prices to macroeconomic news in the US, but using ARCH and regression analysis.

 

MATERIAL AND METHODS:

OBJECTIVES:

1.       To Assess the Granger Causality Relationship Between Forex Market, Crude Oil and Gold Returns. 

2.       To Analyze the Impact of Global Crude Oil and Gold Returns on Exchange rates Returns.

 

HYPOTHESIS:

H_1a - The return series of the Gold, Crude Oil and Exchange Rates are non-stationary.

H_0a - There is no impact of Crude Oil Returns and Gold Returns on selected exchange rate returns

H₀₃-The daily return series of Crude Oil Returns and Gold Returnsdo not granger cause the daily return series of selected Exchange rates

 

DATA AND METHODOLOGY:

Data:

The study collected monthly data Crude Oil prices (Indian Crude Basket and WTI Crude Oil) Exchangerates (USD/INR, GBP/INR, YEN/INR, EUR/INR) and GOLD Prices.

 

Period of the study:

The data collected for the period of 10 years from April 1^st 2007 to 31^st March 2017.

 

Sources of data:

Data has been collected monthly Crude Oil prices, Gold Prices and Exchange rates from Investing.com, worldstopexports.com, www.eia.gov., oilprice.com, marketrealist.com, ppac.org, indexmundi.com, GJEPC India, STC ltd., goldprice.org, world gold council etc.

 

Tools for Analysis:

Augmented Dickey–Fuller test (ADF):

The study used ADF to test the presence of unit root in the selected variables during sample period.

 

Granger Causality Tests:

The study used Granger Causality Tests to test the Granger causality between selected variables

 

  (1)

 

  (2)

 

Equation (1) and (2) shows granger causality equations. Where the residual errors (e1t and e_2t) are not correlated. It shows that y variables depends on its own past values and past values of x. Similarly, x is related to past values of x and past values of y.

 

Regression Analysis:

Regression analysis used to analyze the impact of crude oil prices on exchange rate return. The study considered US Crude Oil (WTI- West Texas Intermediate) Prices, Indian Crude Basket and Gold Prices as independent variables. Further, the study considered USD/INR, GBP/INR, YEN/INR, and EURO/INR as dependent variables.

 

(USD/INR) = β₀+ β₁ (GOLD) + β₂ (Indian Crude Basket) + β₃ (WTI Crude)

(GBP/INR) = β₀+ β₁ (GOLD) + β₂ (Indian Crude Basket) + β₃ (WTI Crude)

 (JPY/INR) = β₀+ β₁ (GOLD) + β₂ (Indian Crude Basket) + β₃ (WTI Crude)

 (EURO/INR) = β₀+ β₁ (GOLD) + β₂ (Indian Crude Basket) + β₃ (WTI Crude)

 

 

 

 

 

EMPIRICAL RESULTS AND INTERPRETATION:

Table 4. 1: Descriptive statistics of Exchange rates, crude oil and Gold returns

Particulars	REURO	RGOLD	RIBOIL	RJPY	RUSD	RWTI	RGBP
Mean	0.1764	0.8753	-0.2023	0.4431	0.3845	-0.2185	-0.0064
Median	0.1996	1.0692	1.2776	0.2828	0.1724	1.0234	0.1232
Maximum	7.6646	13.1099	19.1694	12.6428	7.6681	21.3866	9.6323
Minimum	-8.5522	-11.0525	-33.6914	-8.7545	-6.8299	-33.1981	-7.9664
Std. Dev.	2.9292	4.0569	9.4873	3.8496	2.5337	9.6739	2.9069
Skewness	-0.0959	0.1647	-1.0459	0.2761	0.3488	-0.8447	-0.0040
Kurtosis	3.4470	3.4839	4.8358	3.1492	3.9459	4.5401	3.4261
Jarque-Bera	1.1734	1.6990	38.4059	1.6225	6.8493	25.9125	0.9005
Probability	0.5562	0.4276	0.0000	0.4443	0.0326	0.0000	0.6375
Observations	119	119	119	119	119	119	119

Source: Researcher’s Own Calculation

 

Table (4.1) shows the descriptive analysis of Exchange rates, Gold and Crude Oil Prices, it can be depicted that the standard deviation between the exchange rates is moderate which shows that the variables are moderately dispersed whereas the standard deviation between the gold prices and the crude oil prices are high. It is also observed from the above table that the skewness is positive for Gold prices, JPY and USD which signifies that the data is skewed right. The positive skewness also means that the right tail is longer than the left tail whereas the skewness value of EURO, Oil Prices and GBP is negative. On the other side, the value of kurtosis over a period of time remains constant and is not much fluctuating. If the value of kurtosis is more than three, like in the means that the tails of the data set are more flat and long as compared to the normal distribution. If the kurtosis value is less than three, it means that as compared to the normal distribution, the tails of data set are shorter and thinner. So from the above table, it shows that all the values are above 3 i.e. leptokurtic.

 

Table 4. 2: Augmented Dickey Fuller (ADF) Test results

At Level
Index	t- Statistic	p	Conclusion
EURO RETURN	-11.0712 ***	0.0000	I(0)
GBP RETURN	-10.2612 ***	0.0000	I(0)
USD RETURN	-4.4602 ***	0.0000	I(0)
JPY RETURN	-9.2292 ***	0.0000	I(0)
GOLD RETURN	-9.7722 ***	0.0000	I(0)
WTI RETURN	-5.7734 ***	0.0000	I(0)
IB OIL RETURN	-6.9533 ***	0.0000	I(0)

(Source: Researcher’s Own Calculation)  (*** indicates significant at 1% level)

 

 

 

Table (4.2) shows Augmented Dickey Fuller (ADF) Test results. It clearly indicates that there is no presence of any unit root in any of the selected dependent and independent variables. It shows that all the exchange rates, gold returns and crude oil returns are stationary at level.

 

 

 

 

Table 4.3: Multiple regression analysis results

Variables	N	β0	β1	β2	β3	F- statistic	Adj R2
USD	119	0.282	0.1034	0.0139	-0.0742	3.2455	0.054
		(-0.2252)	(0.0687)	(0.8133)	(0.2182)	(-0.0245)
EURO	119	0.047306	0.1582 ***	0.103	-0.0648	2.9448 ***	0.4711
		(-0.8604)	(0.0169)	(0.1337)	(0.3534)	(0.0359)
GBP	119	-0.0814	0.1052	0.0795	-0.0019	3.6493 ***	0.0631
		(0.7581)	(0.1043)	(0.2373)	(0.9772)	(0.0147)
JPY	119	0.5477	-0.0008	-0.0015	0.0009	0.2989	-0.0181
		(0.0000)	(0.6950)	(0.4837)	(0.6697)	(0.8261)

(Source: Researcher’s Own Calculation)  (*** indicates significant at 1% level)

 

 

 

Table (4.3) shows consolidated multiple regression analysis. It indicates that all the probabilities of coefficients of the independent variables are more than 0.05 except the coefficient of gold returns (β₁) which has a value of 0.0169. It’s significant at 5% level. Thus, it clearly indicates that the Gold returns have positive impact on the Euro/INR Exchange rates. Further the regression results shows no impact of other independent variables (WTI crude oil and IB crude oil returns) on the selected exchange rates.

 

 

 

 

 

Table 4. 4: Granger Causality Test results

Null Hypothesis:	Obs	F-Statistic	Prob.
GOLD  RETURN  does not granger cause EURO RETURN	117	1.86297	0.16
EURO RETURN does not granger cause GOLD  RETURN		4.49471 ***	0.0133
IB OIL  RETURN EURO RETURN	117	1.13138	0.3262
EURO RETURN and IB OIL  RETURN		0.11668	0.89
WTI RETURN and EURO RETURN	117	1.71211	0.1852
EURO RETURN and WTI RETURN		0.19146	0.826
JPY RETURN and GOLD  RETURN	117	0.53140	0.5893
GOLD  RETURN and JPY RETURN		3.40495 ***	0.0367
USD RETURN and GOLD  RETURN	117	1.32335	0.2704
GOLD  RETURN and USD RETURN		1.14232	0.3228
WTI RETURN and GOLD  RETURN	117	0.96290	0.3849
GOLD  RETURN and WTI RETURN		0.68702	0.5052
GBP RETURN and GOLD  RETURN	117	1.83186	0.1649
GOLD  RETURN and GBP RETURN		1.52113	0.2229
JPY RETURN and IB OIL  RETURN	117	0.09950	0.9054
IB OIL  RETURN and JPY RETURN		3.36436 ***	0.0381
USD RETURN and IB OIL  RETURN	117	0.40478	0.6681
IB OIL  RETURN and USD RETURN		5.21007 ***	0.0069
WTI RETURN and USD RETURN	117	1.86387	0.1598
USD RETURN and WTI RETURN		0.46055	0.6321
GBP RETURN and WTI RETURN	117	0.74046	0.4792
WTI RETURN and GBP RETURN		1.09151	0.3393
WTI RETURN and JPY RETURN	117	1.89090	0.1557
JPY RETURN and WTI RETURN		0.02915	0.9713

(Source: Researcher’s Own Calculation)  (*** indicates significant at 1% level)

 

 

Table 4.4a Consolidated Granger Causality Table Showing unidirectional and Bidirectional Effect

Uni- Directional Granger Causality Effect from one variable to other variable		Bi - Directional Granger Causality Effect
Euro Return	Gold Return	No Variables have a bi- directional Granger CausalityEffect
Gold Return	JPY Return
IB Oil Return	JPY Return
	USD Return

(Source: Table 4. 4)

 

 

Table (4.4) and table (4.4a) shows the results of paired Granger Causality test and consolidated Granger Causality test results. The results clearly indicate uni directional causality exists from Euro returns to Gold returns, Gold returns to JPY returns, and IB Oil returns to JPY Returns and IB Oil returns to USD returns. The results shows no bi directional causality exists between the selected variables.

 

CONCLUSION:

The study pertains to analyze the impact of crude oil and gold returns on exchange rates returns. The study used multiple linear regression analysis and Granger casualty test to analyze the relationship. The study found that only gold returns has positive impact on EURO/INR returns. Further granger causality test results reveals uni directional causality Euro returns to Gold returns, Gold returns to JPY returns, and IB Oil returns to JPY Returns and IB Oil returns to USD returns. The market participants are suggested to consider the effect of gold returns and IB Oil performance to take investment or trading decision in JPY/INR, USD/INR and EURO/INR exchange rates.

 

 

 

REFERENCES:

1.     Adelegan, O. J., and Bak, B. R. (2009). What Determines Bond Market Development in sub-Saharan Africa? IMF Working Paper.

2.     Agnihotri, A. (2015). Study of various factors affecting return in bond market - A case of India. AIMA Journal of Management and Research, Vol. 9(2).

3.     Ameer, R. (2007). What moves the primary stock and bond markets? Influence of macroeconomic factors on bond and equity issues in Malaysia and Korea. Asian Academy of Management Journal of Accounting and Finance, Vol. 3(1), pp. 93-116.

4.     Ang, A., and Piazzesi, M. (2003). A no-arbitrage vector autoregression of term structure dynamics with macroeconomic and latent variables. Journal of Monetary Economics, Vol. 50, pp. 745–787

5.     Asia Index Pvt. Ltd. (2018). Asiaindex.co.in. Retrieved from http://www.asiaindex.co.in/

6.     Baele, L., Bekaert, G., and Inghelbrecht, K. (2010). The Determinants of Stock and Bond Return Co-movements. The Review of Financial Studies, Vol 23(6), pp. 2375-2428.

7.     Balduzzi, P., Elton, E. J., and Green, T. C. (2001). Economic News and Bond Prices: Evidence from the U.S. Treasury Market. Journal of Financial and Quantitative Analysis, Vol. 36(4), pp. 523-541.

8.     Beck, R. (2001). Do Country Fundamentals Explain Emerging Market Bond Spreads? Center for Financial Studies Working Paper.

9.     Bekaert, G., Engstrom, E., and Grenadier, S. R. (2010). Stock and Bond returns with Moody Investors. Journal of Empirical Finance, Vol. 17, pp. 867-894

10.  Bollerslev, T., Cai, J., and Song, F. M. (2000). Intraday periodicity, long memory volatility, and macroeconomic announcement effects in the US Treasury bond market. Journal of Empirical Finance, Vol 7, pp. 37-55.

11.  Burger, J. D., and Warnock, F. E. (2007). Foreign participation in ocal currency bond markets. Review of Financial Economics, Vol. 16, pp. 291–304.

12.  Chowdhury, I., Bayar, Y., and Kilic, C. (2013). Effect of major Macroeconomic Indicators on Emerging Markets Bond Index. Afyon Kocatepe Üniversitesi, İİBF Dergisi, pp.15-30.

13.  Christiansen, C. (2000). Macroeconomic announcement effects on the covariance structure of government bond returns. Journal of Empirical Finance, pp. 479–507.

14.  Collin, P., Goldstein, R., and Martin, S. (2001). The Determinants of Credit Spread Changes. The Journal of Finance, Vol. 56(6), pp.2177-2207.

15.  Diebold, F. X., Rudebusch, G. D., and Aruoba, S. B. (2006). Macro economy and the yield curve: a dynamic latent factor approach. Journal of Econometrics, Vol. 131, pp. 309-338.

16.  F, K., and P, M. K. (2014). The Causal Relationship between the Bond Market Development and Economic Growth in Africa: Case Study of South Africa. Mediterranean Journal of Social Sciences, Vol. 5(3), pp. 126-131.

17.  Fair, R. C. (2003). Shock effects on stocks, bonds, and exchange rates. Journal of International and Monetary Finance. Vol. 22(3), pp.307-341.

18.  Fredrick, O. (2014). The Effect of Selected Macro-Economic Variables on Bond Market Development in Kenya. DBA Africa Management Review, Vol 4(2), pp. 54-62.

19.  Goeij, P. D., and Marquering, W. (2006). Macroeconomic announcements and asymmetric volatility in bond returns. Journal of Banking and Finance, Vol. 30(10), pp. 2659-2680.

20.  Indicators (2018). Data.worldbank.org. Retrieved from https://data.worldbank.org/

21.  Jonesa, C. M., Lamontb, O., and Lumsdainec, R. L. (1998). Macroeconomic news and bond market volatility. Journal of Financial Economics, Vol. 47, pp. 315-337.

22.  Konrad, E. (2009). The impact of monetary policy surprises on asset return volatility: the case of Germany. Swiss Society for Financial Market Research, Vol. 23, pp. 111-135.

23.  Li, L. (2002). Macroeconomic Factors and the Correlation of Stock and Bond Return. Yale International Center for Finance.

24.  Ludvigson, S. C., and Ng, S. (2009). Macro Factors in Bond Risk Premia. The Review of Financial Studies, Vol. 22, pp. 5028-5067.

25.  Maurya, A. K., and Mishra, O. N. (2016). The Macro-Economic Determinants of Corporate Bond Market in India. Pacific Business Review International, Vol 1(1), pp. 163-176.

26.  Nair, K. S., and Thenmozhi, M. (2012). Macroeconomic factors and conditional bond volatility: evidence from emerging and developed bond markets. American J. Finance and Accounting. Vol. 2(4), pp. 326-345

27.  NSE - National Stock Exchange of India Ltd (2018). Nseindia.com. Retrieved from https://www.nseindia.com/

28.  Reserve Bank of India. (2018). Rbi.org.in. Retrieved from https://www.rbi.org.in/

29.  Sensarma, R., and Bhattacharyya, I. (2016). The impact of monetary policy on corporate bonds in India. Journal of Policy modeling, Vol. 38(3), pp. 587-602.

 

Received on 15.02.2018          Modified on 27.03.2018

Accepted on 23.04.2018           ©A&V Publications All right reserved