“Dynamic relationship between equity, bond, commodity, forex and foreign institutional investments: Evidence from India”

The interrelationship between equity, bond, commodity and forex movements can provide investors with abundant trading opportunities regardless of whether one market is trending upward or downward. Hence, to understand the interlinkage between markets, this study examines the long-run and causal linkage between forex, G-sec bonds, oil prices, gold rates, foreign institutional investment (FII) flows, and equity market and sectoral index returns. Daily time-series data from August 2012 to August 2021 were considered for empirical analysis. Johansen’s cointegration test revealed that foreign exchanges like USD, Euro, GBP and Yen, oil and gold rates, G-bond returns and FII flows were significantly cointegrated with the stock market and sectoral indices in the long run. Further, Granger causality found a uni-directional relationship between forex rates (i.e., USD, Euro, Yen) and the market, as well as sectoral indices, except Nifty 50 and Nifty IT indices. Oil price movements were found to effectively predict future price changes of Nifty consumer durables, auto, IT indices. Gold prices are useful to predict Nifty-Auto, Bank, Financial Services, Oil & Gas and PSU. The study also found a bi-directional relationship from FII inflows to the stock market and sectoral indices. The findings suggest that forex rates, oil prices and FII flows significantly affect India’s stock market and sectoral performance. The study contributes to the existing literature by comprehensively examining the interlinkage between commodities such as oil and gold, foreign exchanges like USD, Euro, GBP and Yen, G-bond, FII flows and the stock market, and fourteen sectoral indices in the Indian context.


INTRODUCTION
The sustained rise in the interdependence of commodity, bond, forex and equity markets accelerated the financialization process and witnessed increased volatility (Parab & Reddy, 2020).A volatile capital market has large implications for financial and economic stability.The interdependence between the market and economies plays a pivotal role in pricing securities in the market.The most traded commodities such as oil and gold, foreign exchanges like USD, euro, GBP and Yen, bond rates and foreign institutional inflows and outflows, and equity markets are intercorrelated.Therefore, price changes in one market can trigger price fluctuations in other markets.It is necessary to understand the long-run and causal relationship between the popular macro-economic indicators in the short run and long run, because economic indicators have been considered risk factors that significantly influence the financial market.The direction of causality between

LITERATURE REVIEW
The studies on the degree of interdependency and co-movements between markets have captured the attention of researchers.The arbitrage pricing theory proposed by Ross (1976) argues that macroeconomic indicators significantly affect the performance of the stock market.Similarly, 'Portfolio balance approach' proposed by Frankel and Rodriguez (1975) established the interlinkage between the markets.The theory argues that a dynamic stock market attracts a large pool of foreign investment to the market and increases the demand for the domestic currency (Aravind, 2017, p. 2).Subsequently, market hypotheses such as Frenkel's asset market hypothesis and 'Goods and market approach' (Dornbusch & Fischer, 1980) also explain the association between forex and stock markets.Furthermore, Friedman (1988) explained the relationship between stock market performance and money supply.Stable market conditions and the economy attract foreign investments to the market.Thus, foreign institutional investment (FII) is one of the key indicators for understanding equity return movements (Chandra, 2012).The most traded commodities, such as oil and gold, and government bonds can become good portfolio diversifiers by channelising funds during market crashes (Arafoui & Rejeb, 2017).
In recent years, global markets have been vulnerable to unforeseen financial crises, including unexpected equity and foreign exchange price oscillations.The relationship between stock market return, price movements, and forex price variations has been widely explored (Patel, 2017).Aravind (2017) analyzed the effect of volatility in Forex rates of the Indian rupee with other major global currencies like US Dollar, Euro, Japanese Yen, and GBP on the stock price and return volatility in the Indian stock market using the granger causality test.The study exhibited a weak causal relationship between the exchange rate of USD, EUR and JPY to stock returns but a positive relationship with GBP.It concluded that an increase in the exchange rate of GBP causes an increase in the stock return; in contrast, any variations in other exchange rates of currencies will not influence stock returns significantly.Tudor and Dutaa (2012) examined the causal relationship between equity returns and thirteen countries' exchange rates.They found a significant bi-directional relationship only in South Korea and the uni-directional impact of the exchange rate on returns in other countries.Parsva and Tang (2017) examined the causal association between equity prices and forex rates of middle east economies.The study established the bi-directional causality between equity prices and exchange rates in middle east economies such as Iran, Oman, Saudi Arabia, and non-causality in Kuwait.At the same time, Farooq and Keung (2004) claimed the non-stationarity and non-cointegration among the exchange rate and stock prices, as contrary to other studies.It is also argued that uni-directional causality from stock price to exchange rates are possible only in the short run; it could not exist in the long term.Akram (2009) concluded that exchange rates and interest rates are negatively correlated with commodity prices; Commodity prices tend to increase when interest rates and foreign exchange rates are low (Mitra & Bhattacharjee, 2015;Bataineh, 2022).Adjasi et al. (2008) explored the effect of exchange rates and macroeconomic variables on the stock market using EGARCH models in the Gahan stock exchange.It is evident that the movement in the exchange rate influences volatility in the stock market (Saxena & Bhaduriya, 2012).
Further studies have extended the scope of research by exploring relative market indicators such as oil prices, gold rates and government securities to the existing causal relationship of forex and stock prices.Arfaoui and Rejeb (2017) explored the interlinkage between oil, gold, forex and stock market prices.The study found that oil and stock price are negatively correlated.Nevertheless, oil price positively influences gold prices and US exchange rates; USD is negatively correlated with oil, gold, and stock prices.Ingalhalli et al. (2016) found the uni-directional relationship between oil, gold, forex, and security prices.Results exhibited that oil price effectively predicts exchange rates and gold prices.It implies that any fluctuations in oil prices cause variations in the stock prices.Exchange rates and oil prices have a substantial influence on each other.In support of this argument, the relationship between U.S. exchange rates and oil prices in major oil-importing and exporting countries through incorporating the MS-VECM model was analyzed; Results argued that the U.S. exchange rate movement has significantly influenced the oil prices (Beckmann & Czudaj, 2013;Kumar et al., 2021).
Additionally, Singh and Sharma (2018) explored the correlation between gold rates, oil prices, USD exchange rates, and Sensex in the pre-crisis and post-crisis periods using Johansen's cointegration test to verify the long-run relationship.The study revealed a significant long-term relationship between gold, oil, USD, and Sensex.The studies by Arouri et 2021) found that the oil rate has a significant positive impact on the equity return.In contrast, Burian and Brcak (2012) argued that the variation in oil prices has no impact on the stock market return.
Macroeconomic performance and its indicators are essential to predict the price movement in the market.The macro-economic indicators such as global crude oil rates, forex rates and gold prices are highly correlated and influence the stock, bond yields and other market-related indices.Singhal et al. (2019) explored the interrelationship between crude oil rates, gold prices, forex rates and equity returns in Mexico.International gold prices have a significant positive effect on stock prices, whereas it has no significant influence on forex rates.Oil prices were found to be more volatile than other variables and have a considerable negative impact on both equity and forex prices.This relationship exhibits an increase in oil prices, causing a decrease in both stock prices and forex rates.Ratanapakorn and Sharma (2007) investigated the short-run and long-run relationship between equity prices and macroeconomic indicators.The research found a positive relationship between money supply, industrial production, forex, and short-term interest rates.Granger causality revealed that all the macro-economic variables cause the fluctuation in stock prices in the long run, not in the short run.Krchniva (2016) analyzed the causal relationship between economic activity and the stock market in seven countries and found a strong correlation.The study opined that the stock index could be a prominent predictor of economic activities.Diebold et al. (2009) found that returns on bonds, stocks, and gold are not correlated and will not significantly influence each other.Similarly, Sumner et al. (2010) found a weak correlation between the gold, equity and bond returns.On the contrary, studies on the linkages between return on gold, stock, and U.S. bonds found a significant correlation between gold, stock and bond returns.And no correlation with other macro-economic variables (Lawrence, 2003).Furthermore, Kolluri et al. (2015) explored cointegration between the Indian stock and bond market and other major foreign equity markets.Stock and bond market returns were cointegrated in the Indian and the other five foreign stock markets.In the Indian context, the association between bond and equity market is less explored along with other macro-economic variables.Therefore, the current study hypothesis as follows: External factors like foreign institutional investments (FII) play a significant role in determining the stock prices and returns in the stock market.If the market condition is stable, it attracts a massive amount of funds in the form of foreign investments.Chandra (2012)  Initially, a natural logarithm is used to decrease the skewness in the data; the returns are calculated using the following formula: where ln = Natural log; P t is the price of the current period; and P t-1 is the price of the previous period.Further, stationarity of the time series was verified using the Augmented Dickey-Fuller (ADF) test (Dickey & Fuller, 1979).The stationarity of the time series data implies that statistical properties will remain unchanged in the future.Unit root test is essential in econometrics forecasting and models.Granger-causality and Johansen cointegration tests assume the stationarity of time series data.The mathematical expression of ADF is as follows: where y t denotes the time series to be tested; β = coefficient on a time trend; p = lag order of the autoregressive process; a = constant, ε t = error term.
Further, the Johansen cointegration test (Johansen, 1988) was applied to examine the cointegrating vectors in the data series.It evaluates the cointegrated vectors in two forms, i.e., Trace test and the Max-Eigen test.In time series results, trace tests determine the number of linear combinations, i.e., K is equal to the K 0 value, and the hypothesis that K is more significant than K 0 .It is expressed as In addition, the Granger causality test was used to examine the short-run causal relationship between forex, gold, oil, equity flow of FII, Nifty 50 and other sectoral indices.The Granger causality test evaluates the ability to predict the variable's potential to predict the future movements of the time series using prior actions of another time series data.The regression equation for the granger causality test is where Y t and X t are variables to be tested; e 1t and e 2t are error terms, t is time period, k is no of lags.The null hypothesis is rejected if the ADF test values are less than the critical value at 5% (i.e., the p-value is less than 0.05).ADF-test results show that all the variables are stationary at levels.Therefore, the null hypothesis (H 0 : There is no stationarity in the data) has been rejected.Hence, data is stationary for all the selected variables or does not have a unit root (H 1 ) by accepting the alternative hypothesis.

Johansen cointegration analysis
Johansen's cointegration test exhibits the long-run relationship between variables.This test is proposed by Johansen (1988); it examines the cointegrating vectors in the data series.Johansen cointegration test examines the cointegrated vectors in two forms, i.e., Trace test and the Max-Eigen test.In time series results, trace tests determine the number of linear combinations, i.e., K is equal to the K 0 value, and the hypothesis that K is more significant than K 0 .It is expressed as H 0 : K = K 0, H 1 : K > K 0 .

Granger causality test
The Granger causality test evaluates the ability to predict the variable's potential to predict the future movements of the time series using prior actions of another time series data.It is essential to understand the usefulness of the time-series data to predict another time series' movement.The null hypothesis (H 0 : There is no causality between variables) was tested to study the causal relationship of forex rates (H The complete results of the Granger causality test are presented in Appendix A, Table A1.

Table 4. Summary of hypothesis testing results
Source: Authors' calculation.

CONCLUSION
This paper examined the long-run and causal relationship between commodities such as oil and gold, foreign exchanges like USD, euro, GBP and Yen, G bond returns and FII inflows and outflows, and the equity market sectoral indices.Johansen's cointegration test revealed that macro-economic indicators commodities such as oil and gold, foreign exchanges like USD, euro, GBP and Yen, G bond index returns and FII inflows and outflows were significantly cointegrated with the stock market and sectoral indices in the long run.Taken together, these results suggest that economic indicators such as oil and gold rates, forex rates, and financial indicators like G bond index returns and FII flow are intercorrelated and crucial in predicting the stock market and sectoral performance in the long term.The current findings highlight the importance of the long-run relationship between the variables.Hence, this study provides insights for regulatory bodies, policymakers and investors in long-term investment decisions.
Further, the Granger causality test establishes a uni-directional relationship between forex rates (i.e., USD, Euro, Yen) and the market as well as sectoral indices.Additionally, the study found a uni-directional relationship between GBP and NSE sectoral indices such as Nifty -Consumer durables, Pharma, PSU, Pvt.Bank indices, and a bi-directional relationship from GBP to Nifty-Bank, FMCG, Financial services, Metal, Oil & Gas and Reality sectoral indices movements.Taken together, these results suggest that forex rates significantly impact the stock market performance.Currency crises may adversely impact the stock market and sectoral prices.Therefore, regulatory bodies and policymakers could timely implement the policies as a preventive measure, emphasizing transparent pricing by preventing price volatility.
The current study revealed that commodities such as oil price movements were found to be efficient in predicting future price changes of stock markets and sectors, as oil is considered a key input material in industrial production.Regulatory bodies could moderate prices through fiscal and monetary policy amendments to prevent adverse price volatility.The study also found that FII inflow is meaningful in forecasting the movements of all the market and sectoral indices.Hence, FII inflows and outflows are significant predictors for the equity market and sectoral indices price movement, which is considered a significant financial indicator by institutional and retail investors for their investment decisions.Therefore, policymakers should regulate foreign capital flows and prevent imbalances through credible investment policies.
Investors may use the findings to forecast the expected returns from a given investment avenue explored in the study.The result of the study will be helpful for retail investors and financial institutions to understand the causal relationship between various macroeconomic variables, market indicators and sectoral performance indicators.This paper has contributed to the body of literature in terms of understanding the combined effect of variables on the stock market and is beneficial to analyze the inter-relation between forex, commodity, gold and capital markets.This study also provides insights on the influence of macro-economic variables such as forex rates, gold, oil, and foreign institutional investments on the Nifty 50 market indicator and other relative sectoral indices.
The current studt analyzed the relationship between forex, oil prices, gold rates, G-sec bonds, FII purchases and FII sales on the market and sectoral indices.However, the Consumer Price Index (CPI), Wholesale price index (WPI), GDP and other macro-economic variables can be explored along with the existing independent variables.Future research can be expanded by considering the companies listed in the Nifty 50 or Sensex indices.In addition, cross-country comparative studies can be undertaken to enhance the generalizability of the results.

Further, the study
found a bi-directional relationship from FII inflows to the stock market and sectoral indices except for Nifty -Media & Pharma.On the other hand, FII outflows are meaningful to forecast the movements of all the market and sectoral indices except Nifty -Bank, financial services and PSU.These results align with those of previous studies by Chandra (2012), Murthy and Singh (2013), Goyal (2013), Vohra (2016), Dhingra et al. (2016), Arora (2016), Agarwal (2016), and Parab and Reddy (2020).Hence, financial indicators such as foreign institutional investment inflows and outflows are significant predictors for the equity market and sectoral indices price movement.This implies that FII inflows significantly forecast the indices' future price movements.
The summary of descriptive statistics is demonstrated in Table1.The Nifty 50, Nifty CD, Nifty Auto, Nifty Bank, Nifty FS, Nifty FMCG, Nifty IT, Nifty O&G, Nifty PH, and Nifty P Bank sectors showed positive returns with the highest mean returns for Nifty CD and Nifty IT (0.05%).While Nifty Metal, Nifty PSU, and Nifty REA sectors exhibited negative returns, with the lowest mean returns for Nifty PSU (-0.05%).Further, positive mean returns were observed for US Dollar, Euro, Japanese Yen, GBP, Gold, and Government Bonds except for oil (-0.04%).Standard deviation explained the variation from the actual mean and showed the highest deviations for FII Purchase and FII Sales with 44.26% and 42.29%, respectively.This deviation is because of high fluctuations in foreign institutional investment inflow and outflows.The skewness of the data depicts positive skewness only for USD, Euro, Yen, and gold.The kurtosis measured the flatness of the data and found that the data was too peaked.Further, the Jarque-Bera test is applied to determine the data normality, and the null hypothesis is H 0 : The data are normally distributed.The results showed that the p-value of the test statistics is insignificant at a 5% significance level and thus, data is not normally distributed (H 1 ).

Table 2
Dickey and Fuller (1979)ts of the ADF proposed byDickey and Fuller (1979), which tests the stationarity of the time-series data.Unit root test is essential in econometrics forecasting and models.The stationarity of the time series data implies that the statistical properties such as mean and variances should be constant over time.The Granger-causality test assumes the stationarity of time series data.The unit root test has been done using the ADF test to confirm the stationarity.This test postulates null hypothesis H 0 : There is no stationarity in the data or a unit root in the data.

Table 3
shows the Trace and Max-Eigen values of all the dependent variables with the associated critical values.Forex rates, oil prices, gold rates, G-sec bond returns, FII purchases and sales are considered as independent variables in the test analysis.The null hypothesis (H 0 : No cointegration between the variables) has been tested at the critical value of 5%.The long-run relationship among variables is proved only when the Trace and Max-Eigen values are higher than the respective critical values.The trace's critical value at 5% is 239.2354, and Max-Eigen is 64.5047.Hence, Johansen cointegration test result values of Trace and Max-Eigen are consistently higher than their respective critical values.Therefore, the alternative hypothesis is accepted (H 1 : There is cointegration between the variables.).At a 5% significance level, at least ten cointegrating pairs among the variables are visible in the Johansen cointegration test results.Two or more cointegrated pairs in the time series are verified and exhibit bi-directional or uni-directional relationships in the Granger causality test.

Table 3 .
Results of Johansen's cointegration test Source: Authors' own calculation.
Oil price movements were found to be efficient in predicting the future price changes of Nifty -consumer durables, auto, & IT indices.In addition, the uni-directional relationship of oil with Nifty-50, Bank, Financial services, Metal, Oil & Gas, Pvt.Bank suggests that future prices can be forecasted by each other.Uni-directional relationship between Nifty 50 returns and oil supported the results of(Arouri etal., 2012; Ingalhalli et al., 2016; Aydogan et al., 2017; Singh & Sharma, 2018).The bi-directional relationship has been noticed between Gold and other sectoral indices.Gold prices are useful to predict Nifty-Auto, Bank, Financial Services, Oil & Gas and PSU.The results are not supported by Diebold et al. (2009) and Sumner et al. (2010) and argued that gold prices might not be the best predictor for stock returns.G-sec bond index returns, and Nifty-IT index can be predicted by each other.It is also significant to forecast the price movements of Nifty 50, Nifty-Auto, Bank, Financial Services, FMCG, and Pvt.Banks and realty indices movements.Results of Lawrence (2002), Diebold et al.

Table A2 .
Granger causality test results of NIFTY -Auto, Bank, Metal, Oil and Gas indices

Table A3 .
Granger causality test results of NIFTY -Financial services, FMCG, Pharma, PSU indices