This study investigates the dynamic links between key macroeconomic variables and the performance of stock market in India. With 420 monthly observations and a comprehensive econometric framework, the analysis attempts to assess the impact of Consumer Price Index (CPI), Exchange Rate (EXRATE), Foreign Direct Investment (FDI), and Gross Domestic Product (GDP) on BSE stock returns. Descriptive statistics reveal significant non-normality and volatility, justifying the use of GARCH models to capture market fluctuations. Granger causality and Johansen cointegration tests indicate a unidirectional and long-term influence of macroeconomic factors — particularly FDI, exchange rate, and GDP — on stock returns. Inflation and exchange rate have a positive impact, whereas FDI and GDP show negative associations, highlighting market sensitivity to capital flows and policy conditions. The GARCH (1,1) model accurately describes time-varying stock return volatility. The large ARCH and GARCH coefficients confirm the effects of previous shocks on volatility persistence. Impulse response functions support these conclusions, whereas error correction estimates emphasize the stock market’s role as a shock absorber for the economy. Diagnostics confirm the robustness and structural stability of the model, reflecting the post-liberalization resilience. The research underscores the significant and predominantly unilateral influence of macroeconomic variables on Indian stock markets. It highlights the critical role of a stable macroeconomic framework and investor sentiment in determining market trends.

Macroeconomic variables, stock return, BSE, India, volatility.

There is a widely held belief that well-developed stock markets contribute to the organization of liquidity, encourage the use of risk management tools, reduce information asymmetries, reward performance and efficiency, and ultimately lead to accelerated overall economic growth. For an economy, the role of capital markets is essential to maintain the aggressive desire to compete in today’s environment, which is characterized by increased intercontinental competition, rapid technological evolution, and the increased importance of innovation for economic growth. According to Galbraith (1955), the stock market can be seen as a mirror that reflects the fundamental economic situation of a country. This vibrant relationship between macroeconomic variables and market volatility has been a topic of discussion among academics, researchers, and strategists. Many economic indicators, commonly referred to as macroeconomic variables, play a crucial role in understanding stock market movements and behaviors. These indicators broadly reveal the state of the economy and are directly linked with key factors such as exchange rates, GDP, inflation and foreign direct investment (FDI). Economists and financial experts have conducted various studies on the dynamic causal relationships in different countries and time periods. They have concluded that the types and strength of the relationships differ depending on a country’s financial structure and policies. They argue that it is difficult to predict the response of market returns to macroeconomic policy changes in advance, as it varies across countries. In recent times, the global economic developments have been seen as more important in explaining returns across markets than the domestic developments. Zakaria and Shamsuddin (2012) note that the causal relationships and dynamic interactions between macroeconomic variables and stock prices are significant in the development of a country’s economic policy. Any restrictions on policy regarding the economic environment can have a cascading effect on other markets. Economic theories suggest that corporate profits are a reflection of the current financial performance of a company. At the same time, the stock price serves as an indicator of the company’s anticipated future performance. If stock prices accurately reflect the underlying fundamentals, they should be used as leading indicators of future economic activity, rather than the other way around. The improved performance of the stock market is primarily dependent on the initiatives taken by the government to create a more attractive investment climate and promote economic growth.

The value of corporate equity at an aggregate level depends on the state of economic activity. Any change in the level of uncertainty about future economic conditions is likely to cause a change in stock return volatility. This is observed in relation to some basic macroeconomic factors, such as the exchange rate, gold price, gross domestic product (GDP), oil price, and inflation, among others. To investors, excessive stock price volatility, or “noise”, undermines the usefulness of stock prices as a yardstick for assessing risk. This is because stock prices act as an indicator of the true intrinsic value of a company (Karolyi, 2001). Officer (1973) was the first to notice a link between stock price fluctuations and economic indicators, pointing out high volatility during the Great Depression of the 1930s. Schwert (1988), while examining the relationship between stock returns and macroeconomic uncertainty, found that market volatility exhibited counter-cyclical behavior, with the direction of causality generally flowing from the stock market to macroeconomic variables. Corradi, Distaso, and Mele (2013) also found that volatility moves counter-cyclically in relation to GDP. Karolyi (2001) argued that a decrease in stock prices leads to increased uncertainty in the stock market. Campbell and Hentschel (1992) noted that favourable and unfavourable information could have different effects on volatility. A similar line of research observed a significant relationship between stock returns and the unpredictability of macroeconomic factors. (Morana & Beltratti, 2002, Chowdhury et al., 2006, Saryal, 2007) .

Although there are various perspectives on this relationship in developed and emerging economies, the current study attempts to investigate the dynamic connections between certain key macroeconomic indicators (Gross Domestic Product, Exchange Rate, Foreign Direct Investment, and Inflation) and the volatility of stock returns in India from January 1990 to December 2024. The time series data set comprises the monthly observations of the Bombay Stock Exchange (BSE), Gross Domestic Product (GDP) and Exchange Rate (EX). Inflation Rate (INFLA) and Foreign Direct Investment (FDI) are also included in the data set. These data are used for empirical observations in this phase, which is marked by a series of global and national events with significant impacts on India’s economy. The hypothesis suggests that the selected macroeconomic variables may have an impact on the stability of market returns in India. Today, India is an increasingly popular intercontinental destination for investment, so a better understanding of stock market indicators has become extremely important for both investors and policy makers. The rest of the paper is organized as follows: Section 2 provides a literature review. Section 3 presents the data, models, and methodology used, as well as the empirical results. Concluding remarks are presented in Section 5.

The relationship between economic fundamentals and stock prices is a long-standing topic of discussion among researchers and practitioners. According to the Efficient Market Hypothesis (Fama, 1970), stock prices already reflect all available information, including macroeconomic factors, thus ruling out the possibility of earning abnormal profits. However, this view has been challenged by subsequent studies (Fama & Schwert, 1977; Nelson, 1976), which found that macroeconomic variables do influence stock returns. The Arbitrage Pricing Theory (APT) proposed by Ross (1976) and extended by Chen et al. (1986) also provides a theoretical framework connecting macroeconomic factors and stock prices.

There is no shortage of literature on the study of the causal relationship between the stock market and macroeconomic variables. A great deal of research has been conducted to observe the effects of macroeconomic variables on the stock markets of industrialised economies. Thus, Fama (1999) examined the relationship between real output and stock prices and showed that there was well-built association between gross national product and stock prices. Chen, Roll and Ross (1986) tested the multifactor model in the USA by employing seven macroeconomic variables. They found that the financial market does not price consumption, oil prices or the market index. However, changes in the risk premium, industrial production and the shape of the yield curve were found to be significant in explaining stock returns.

Mukherjee and Naka (1995) explored the relationship between industrial production and stock prices in Japan, finding a positive correlation. Mookerjee and Yu (1997) found that foreign exchange reserves and money supply, as independent variables, were associated with stock prices in Singapore in the long run.

Niarchos and Alexakis (2000) investigated the effect of macroeconomic variables, such as inflation, money supply and the exchange rate, on the Athens Stock Exchange, using monthly data from January 1984 to December 1994. Their findings suggest that monthly stock prices at the Athens Stock Exchange are directly related to these variables.

Bhattacharya and Mukherjee (2002) found a causal relationship between macroeconomic factors and stock prices in the Indian stock market. They applied the Toda-Yamamoto methodology to the period between 1992 and 2001, stating that changes in industrial production affect stock prices.

Wongbangpo and Sharma (2002) showed that high inflation in Indonesia and the Philippines resulted in a negative long-term relationship between money supply and stock prices, whereas money growth in the other ASEAN-5 countries (Malaysia, Singapore and Thailand) had a positive effect on their stock market indices.

In his study, Kim (2003) found that the S&P 500 stock price had a positive correlation with industrial production, but a negative relationship with the real exchange rate, inflation and interest rate. Nishat and Shaheen (2004), in the context of the Pakistan stock market, found that industrial production had the largest positive relationship with stock prices. Using a Granger causality test, Chakravarty (2005) examined the association between industrial production and stock prices, finding unidirectionality from industrial production (IIP) to stock prices (SP) in India.

Tan, Loh and Zainudin (2006) examined the dynamic relationship between various macroeconomic indicators and the Malaysian stock indices (Kuala Lumpur Composite Index) between 1996 and 2005. They discovered that inflation, industrial production, the price of crude oil and the rate of treasury bills were all long-term factors influencing the Malaysian stock market. The results indicate that the Kuala Lumpur Composite Index is significantly and negatively connected with the consumer price index, the industrial production index, the crude oil price and the treasury bills over a longer time frame. However, the industrial production index is coupled with a positive coefficient.

Ahmed and Osman (2007) examined the long-run equilibrium and short-term dynamics between the DSE stock index and a set of macroeconomic variables, such as the money supply, the 91-day T-bill rate, the interest rate, GDP and the industrial production index. The cointegration results demonstrated two cointegrating vectors, one of which was statistically significant. In the VECM test, they found that the lagged stock index adjusted to long-run equilibrium by 43.82% due to the combined lagged influence of all the selected macroeconomic variables. The Granger causality test revealed unidirectional causality from interest rate changes to stock market returns.

Using monthly data on stock price indices, foreign exchange rates, the consumer price index and the industrial production index between January 1993 and December 2002, Mahmood and Mohd Dinniah (2009) observed the strong relationship between stock prices and several macroeconomic indicators in six Asian-Pacific countries, including Malaysia, Hong Kong, South Korea, Japan, Thailand and Australia. The results indicate the continuation of a long-term equilibrium relationship between stock prices and macroeconomic variables in four countries: Korea, Australia, Japan and Hong Kong. A short-run association exists in all countries except Thailand and Hong Kong. In the case of Hong Kong, the results show a correlation between the exchange rate and stock prices. In Thailand, however, there is a clear relationship between production and stock prices.

Mohammad, Hussain and Ali (2009) examine the relationship between macroeconomic indicators and the Karachi Stock Exchange in Pakistan, taking into account quarterly data on the foreign exchange rate, foreign exchange reserves, gross fixed capital formation, money supply, industrial production, interest rates and the wholesale price index. The analysis shows that the exchange rate and foreign exchange reserves have a significant impact on stock prices.

Asaolu and Ogunmuyiwa (2010) studied the impact of macroeconomic variables on the share prices of the Nigerian stock market, with the latter acting as the dependent variable, and external debt, inflation rate, fiscal deficit, exchange rate, foreign capital inflow, investment and industrial output acting as the independent variables. The Granger causality test revealed that the average share price (ASP) did not Granger-cause any of the nine macroeconomic variables in Nigeria during the sample period. Only the exchange rate Granger caused the ASP when considered in pairs. The Johansen co-integration test revealed a long-term relationship between share price and macroeconomic factors considered in the study. The error correction method also revealed a weak relationship between share price and the aforementioned macroeconomic variables, indicating that stock price is not a leading indicator of macroeconomic variables in Nigeria.

Ali M. B. (2011) examined the effect of changes in select macroeconomic indicators upon the stock returns of the Dhaka Stock Exchange (DSE) using a multivariate regression model. The results suggest that inflation and foreign remittances have a pessimistic effect, while market price-to-earnings ratios, the industrial production index, and the monthly percentage average growth in market capitalisation have an optimistic effect on stock returns. No unidirectional Granger causality was found between stock prices and all the predictor variables, except for one unidirectional causal relationship from stock price to market P/Es.

Patel (2012) examined the influence of eight macroeconomic variables — interest rates, inflation, exchange rates, oil prices, the index of industrial production, silver prices and gold prices — on the BSE Sensex and the S&P CNX Nifty, using VECM and Granger causality tests for the period from January 2001 to December 2011. The results showed that the exchange rate, money supply, industrial production and inflation provide vital information for directing a stock exchange. Commodity prices, such as those for gold, silver and oil, are noteworthy predictors of stock market performance.

Kumar (2013) investigated the effect of industrial performance, the macro-environment, and policy rates on the Indian stock market. The results showed that industrial production had a significant impact on stock market performance. Although the influence of the policy rate has no lasting impact on the stock market, its influence should not be dismissed.

Tripathi and Seth (2014) investigated the relationship between the real economy and the stock market in India for the period from July 1997 to July 2011. Their results show that there is a significant correlation between macroeconomic indicators and the stock market. According to Ahmad and Sinha (2015), who studied the relationship between the BSE Sensex, GDP and the exchange rate, GDP was a notable predictor of the BSE Sensex, whereas the exchange rate was not.

Tripathi, Singh & Singh (2016) investigated the relationship between the Indian stock market (BSE) and key macroeconomic variables such as the index of industrial production (IIP), foreign direct investment (FDI) and the wholesale price index (WPI) of the Indian economy, using quarterly data from 2002–03 to 2012–13. The results showed that the IIP is a significant predictor of the BSE Sensex, while the FDI and WPI were not found to be essential predictors of the BSE Sensex. While examining the connection between India’s BSE index and macroeconomic markers from 1999 to 2017, Misra (2018) found an enduring relationship between inflation, IPI, gold price, money supply, FII, interest rates, exchange rate and BSE index. The study also established a temporary connection between the money supply and stock market index, and between inflation and BSE index.

Kaur and Singh (2019) examined the relationship between macroeconomic indicators such as the cash reserve ratio (CRR), the reverse repo rate, commodity prices, inflation rates, gold prices and oil prices, and the Sensex in India from January 2001 to December 2009. They found that the gold price, exchange rate and inflation rates were the most decisive variables for forecasting the Sensex.

In their 2020 paper, Norehan and Ridzuan used an ARDL model to analyse the impact of inflation rates, domestic savings, the money supply and exchange rates on the Malaysian stock market between 1981 and 2017. The results showed that inflation and exchange rates had a positive and significant effect on the stock market index. Additionally, the money supply and domestic savings had a modest but significant influence on the stock market.

Bhattacharjee and Das (2021) examined the impact of macroeconomic indicators on India’s stock market. Their findings showed that the relationship between the money supply and exchange rates is irrelevant in the long term. Money supply, inflation and the foreign exchange rate have a short-term connection with the stock market. Pandey (2022) examined the strong correlation between the movement of Indian stock market sectoral indices and a few macroeconomic variables. The analysis showed that the oil price, exchange rate (though this variable was significantly negative) and gold price all have a notable impact on sectoral indices in the Indian stock market.

To explore the influence of the price of Brent crude oil on the Chinese stock market and selected industries, Hashmi et al. (2022) used a Var-DCC-GARCH model. The results revealed that the Shanghai Composite Index and the chemical, steel, non-ferrous metals and mining industries are considerably affected by Brent crude oil prices.

Ali et al. (2023) investigated the association between the returns of sectoral indices and inflation. The study applied the Pearson correlation method and observed that all the sectoral indices had a statistically significant affiliation with inflation, except the Metal Index, Energy Index and IT Index.

Suriani et al. (2024) examined the relationship between economic factors such as inflation and exchange rates and the Indonesian sectoral stock market in the consumer goods (CGI) sector, the basic industrial and chemical (BIC) sector, and the miscellaneous industries (MSI) sector. The study found that all three industrial sectors reacted positively to changes in inflation and exchange rates.

Mandal and Datta (2024) examined the impact of oil prices on the BSE Sensex and eight sectoral indices (BSE Energy, BSE Carbon, BSE GreenEx, BSE Fast-Moving Consumer Goods, BSE Industry, BSE Health, BSE Information Technology and BSE Metal) during the pre-, post- and pandemic periods. The results suggest that during a period of a COVID outbreak, all the sectoral indices display a high positive correlation over a long-term investment horizon, and a negative correlation over a medium-term investment horizon.

According to Sharma (2024), in a bear market, most economic sectors, such as financial services, utilities, commodities and industrials, as well as consumer discretionary sectors, are pessimistically affected. The information technology industry is unaffected by domestic EPU shocks and negatively influenced by foreign ones.

Despite the extensive body of literature on the correlation between macroeconomic indicators and stock returns in developed markets, research into this relationship in major developing markets is scarce. The extant literature suggests the existence of a significant correlation between macroeconomic indicators and stock prices in developed countries. However, this relationship is not observed in developing economies.

Furthermore, while the results are predominantly conclusive for advanced economies, there is a paucity of consensus for emerging economies due to a dearth of research and the emergence of conflicting results. Several studies on developing markets focus on a limited number of macroeconomic variables and a small number of countries, but there is no comprehensive research that considers major developing markets such as India and a range of important macroeconomic variables at the same time. However, the current trends indicate a shift in focus towards the analysis of stock markets in emerging economies, such as India, owing to their significant profit potential.

Despite extensive research conducted over several decades, there is still no consensus in relation to Indian financial markets due to the variations in variables, frequency, and methodologies used. While monthly data is preferred because of their higher frequency, there are limitations caused by the lack of monthly GDP figures. However, monthly GDP can be estimated using average approximations.

The present study, based on monthly data, aims to evaluate the impact of several macroeconomic variables - GDP growth, exchange rate, inflation, and foreign direct investment (FDI) - on stock market returns in India. Specifically, it considers the BSE Sensex returns for the period from January 1990 to December 2024.

This section outlines the methodology employed to evaluate the relationship between various macroeconomic indicators and stock market fluctuations. It covers the data sources used, the research variables selected, and the formulation of an empirical model for analysing the data.

Following a thorough review of the literature, we selected several key macroeconomic variables, including inflation, exchange rates, GDP and foreign direct investment (FDI), to determine their association with the Indian stock market. Inflation is assessed using CPI index numbers, while the exchange rate is represented by the monthly percentage change in the value of the US dollar against the Indian rupee. The BSE Sensex serves as an approximate measure of market performance.

In this study, we used GDP growth as a proxy indicator of the economic growth that affects the stock market. We collected quarterly GDP data from the Handbook of Statistics on the Indian Economy (several issues covering our study period) and converted it into a monthly series, although the conversion process is not without its flaws. This is because a developing economy like India’s has agriculture as one of the major contributors to its GDP growth. The nature of agricultural products means they do not mature on a monthly or quarterly basis. Most crops either yield once a year or take a full year to produce.

In each case, quarterly data will be divided by 3 to get monthly figure and, wherever quarterly data is not available, yearly GDP data has been converted to monthly data by dividing those by 12.

Dependent Variable: Stock Market Return of BSE SENSEX

Independent Variables: (Macro-economic Variables) Gross Domestic Product, Exchange Rate, Inflation, Foreign Direct Investment.

Hypothesis

H_01: GDP growth has no impact on the stock market return.

H₀₂: Exchange rate has no impact on the stock market return.

H₀₃: Inflation has no impact on the stock market return.

H₀₄: Foreign direct investment has no impact on the stock market return.

All the variables, both dependent and independent, are based on monthly observations covering the period from January 1990 to December 2024. The data has been sourced from the following institutions and publications: BSE SENSEX, World Bank data, WTRG Economics and Handbooks of Statistics. The BSE SENSEX monthly closing prices have been taken into consideration as a proxy for Indian stock prices, and the return has been calculated using the following formula:

R_t = LnP_t – LnP_{t – 1}

R_t = Stock Return for’ t’ time-period

P_t = Price at t time-period

Ln = Natural logarithm

Along with the BSE stock return, all other variables have been converted to their logarithmic form.

The regression equation is expressed as follows:

LnR_t = α + β₁LnCPI + β₂LnEXRATE + β₃LnFDI + β₄LnGDP + ε

A GARCH-OLS model has been employed for a comprehensive analytical study of the impact of macroeconomic factors on the unpredictability of the Indian stock market. The GARCH model (Bollerslev, 1986), a modified version of ARCH(p), is the most widely used model for judging volatility clustering (Ray & Saha, 2016). The OLS method is a linear regression technique used to estimate the relationship between a dependent variable and one or more independent variables by minimising the sum of squared residuals (errors). To assess the impact of the time-varying variance of India’s stock returns, this study adopts the GARCH-OLS model. This combines OLS (Ordinary Least Squares) for estimating the mean equation with GARCH (Generalised Autoregressive Conditional Heteroskedasticity) for modelling the variance of the residuals from the OLS equation. The GARCH-OLS test is as follows:

Mean Equation (OLS): r_t = μ + ϵ_t

Variance Equation (GARCH) ${ϵ}_t=z_t\sqrt{h_t};z_t\sim N(0,1)$

$h_t=\omega +\alpha {ϵ}_{t-1}^2+\beta h_{t-1}$

Where, μ is the constant mean.

ϵ_t is the error term or residual for period t

h_t is the conditional variance

ω > 0, α ≥ 0 and β ≥ 0 are parameters to be estimated

α effect of past squared residuals (ARCH term)

β effect of past variances (GARCH term)

Cointegration techniques address the issue of spurious regression in time series data, including when the data are non-stationary. Many economic theories suggest that a linear combination of variables is stationary, even if the variables themselves are not. If such a stable linear combination exists among the variables, they are said to be cointegrated. Therefore, when working with time series, it is essential to check for stationarity and cointegration in order to avoid spurious regressions. When the variables are non-stationary at levels but are difference stationary, cointegration methodology allows researchers to test for the subsistence of long run equilibrium connection among the variables. If the separate economic time series are stationary after differencing, but a linear combination of them is stationary, then the series are said to be cointegrated.

Johansen (1988) uses the maximum likelihood principle to identify cointegrating vectors in non-stationary time series. The Johansen procedure (1988) relies heavily on the relationship between the rank of a matrix and its characteristic roots. It is essentially a multivariate extension of the Dickey-Fuller test.

To ensure the robustness of the study, Johansen’s cointegration test was applied to reveal whether there is a long-term relationship between the variables.

This model captures both the long-term equilibrium relationship through $\alpha {\beta }^{\prime }{Y}_{t-1}$ and the short-term dynamics through the lagged differenced terms ${\Gamma }_i\Delta Y_{t-i}$.

$\Delta Y_t=\alpha {\beta }^{\prime }{Y}_{t-1}+\sum _{i=1}^{k-1}{\Gamma }_i\Delta Y_{t-i}+{\epsilon }_t$

When performing the Johansen test, the variables must be in the same order of integration, and the appropriate lag length must be selected using a VAR model at level, employing the multivariate generalizations of the AIC or SBC.

Even if y_t and x_t variables are cointegrated, that is, there is a long run equilibrium relationship between them, there may be disequilibrium in the shorter time frame. Thus, the error term ut = y_t – β₁ β₂ xt in the regression equation y_t = –β₁ + β₂x_t + u_t is called the equilibrium error. This error term can be used to tie the short-run behaviour of Y to its long-run value. The error correction models (ECM) first used by Sargan and later popularized by Engle and Granger corrects for disequilibrium. The Granger Representation Theorem says that if two variables y_t and x_t are cointegrated, then the relationship between the two can be expressed as Error Correction Model by: ∆y_t = α + α₁∆x_t + α₂u_{t – 1} + ε_t

Where, ∆ = first difference operator,

ε_t = a white noise error term,

u_{t – 1} = one period lagged value of the error term from the cointegrating regression.

If the error term is non-zero, the model is out of equilibrium. Here the value of α₂ decides how quickly the equilibrium is restored.

In order to explore directional causality, Granger causality tests are conducted.

This test is based on estimating the following two equations:

Equation 1: Restricted Model (Univariate AR Model of Y). This model predicts Y_t using only its own past values.

$Y_t={\alpha }_0+\sum _{i=1}^p{\alpha }_i{Y}_{t-i}+{\epsilon }_{1t}$

Equation 2: Unrestricted Model (Includes Lagged Values of X). This model predicts Y_t using both its own lags and the lags of X_t

$Y_t={\beta }_0+\sum _{i=1}^p{\beta }_i{Y}_{t-i}+\sum _{j=1}^q{\gamma }_j{X}_{t-j}+{\epsilon }_{2t}$

The robustness of the model is diagnosed using the following tests:

• The Breusch-Godfrey LM test confirms serial correlation.
• The Ramsey RESET test confirms functional form misspecification.
• The CUSUM and CUSUMSQ tests confirm structural stability of coefficients over the sample period.

The characteristic of our sample variables—BSE returns (BSE_RT), Consumer Price Index (CPI), Exchange Rate (EXRATE), Foreign Direct Investment (FDI), and GDP—across 420 monthly observations is well captured by descriptive statistics.

The average monthly return of the BSE is 1.11%, but the high standard deviation of 7.75 indicates considerable variability in the Indian stock market. Although the average exchange rate is about 3.82, implying relative currency stability and low volatility, the CPI has a moderate mean of 0.79%, indicating typical inflation behaviour. While FDI is more variable than GDP, both depict steady economic inflows and growth, with respective averages of 9.02% and 13.82%. According to Mandelbrot (1963), skewness and kurtosis measures suggest deviations from normality for all the variables. BSE returns are nearly symmetrical but are heavy-tailed, which points to the presence of outliers or extreme market events/shocks. This leptokurtic behaviour indicates the volatile and unpredictable nature of stock markets (Fama, 1970; Bollerslev, 1986; Mandelbrot, 1963). CPI is right-skewed (Cheng & Tan, 2002), whereas EXRATE, FDI and GDP are left-skewed. This indicates asymmetry in their distributions (Alfaro et al., 2004). The high kurtosis measures, particularly for BSE_RT and FDI, also provide supporting evidence for the presence of outliers or fat tails. The Jarque-Bera test confirms that none of the series are normally distributed since all the p-values are zero. These findings are supported by Cont (2001) and Hamilton (1994), who state that financial and macroeconomic data tend to be non-Gaussian. Overall, the figures reflect the volatility of financial returns in relation to less volatile macroeconomic aggregates, which has significant implications for policy modelling and analysis. These characteristics make models such as GARCH, which are capable of accommodating heteroskedasticity and non-Gaussian distributions, suitable for use (Bollerslev, 1986; Engle, 1982).

H₀: Variables have Unit root

H₀: Residual has Unit root

The results in Table 2(b) show that the residual series is stationary at the conventional 1 percent level of significance. This confirms the existence of long-run relationship among BSE (Returns), Inflation (CPI), Exchange Rate, FDI and real GDP of India.

The probability associated with the ADF statistics shows that ADF values of BSE (Returns), Inflation (CPI), Exchange Rate, FDI and GDP are stationary at first difference. They are also supported by the value of t statistic that is higher than the 5 percent critical value. Thus BSE (Returns), Inflation (CPI), Exchange Rate, FDI and GDP are stationary at I (1).

Since all variables are stationary at first difference and the error terms are stationary at the level, we are moving towards modelling the short-run regression presented in Table 3 below.

As can be seen in Table 3, the exchange rate has a significant short-term effect on the stock market, as evidenced by the statistically significant coefficient on REER. The exchange rate coefficient is -37.09431%, implying that a 1% increase in the exchange rate would depress stock returns by 37.09431% over a shorter time frame. However, inflation, FDI and GDP have no short-term impact on stock market returns.

Since all the variables are stationary at first difference and the error term is stationary at the level, this indicates the possibility of cointegration or a long-run relationship. The ECT is used to detect the long-run causal relationships. The study found that the error correction variable has a negative sign and is statistically significant, which substantiates the existence of a long-term relationship between the variables. In short-run regression models, the ECT coefficient is negative and significant, indicating that the model will adjust towards long-run equilibrium at a rate of approximately 91.16% per unit per month.

The Granger causality test revealed a sequence of one-way causal connections between the variables in question. Notably, foreign direct investment (FDI) Granger-caused stock returns, indicating that foreign investment activity can predict equity market movement, which supports the findings of Kaur & Singh (2019). The absence of reverse causality serves to refute the forward-looking information theory (Chen et al., 1986; Bekaert et al., 2013) which posits that stock prices are a reflection of macroeconomic factors. Stock returns, however, did not Granger-cause any of the macro variables, implying a weak feedback effect. The exchange rate was found to Granger-cause CPI and FDI, which supports its pivotal role in macroeconomic adjustment. This is consistent with the findings of Goldberg and Kolstad (1995), who identified the exchange rate’s role in determining inflation and the direction of investment. Furthermore, it was found that GDP Granger-causes both FDI and CPI, suggesting that economic growth leads to inflationary pressure and foreign investment flows. These results highlight the significance of macroeconomic stability and growth in shaping capital market behaviour and investment flows. The present study finds support for unidirectional causal links between GDP growth, inflation, and foreign direct investment, as posited by Chakrabarti (2001) and Barro (1995). Their findings show that output plays a role in relation to both inflation and capital flows, and economic growth can be a cause of both inflationary pressures and foreign investment.

On the other hand, inflation (CPI) can be a predictor of exchange rates (EXRATE) as it is a relevant factor of their fluctuations. High levels of inflation can erode the purchasing power of a country’s currency, which may lead to its devaluation against other currencies. Investors’ self-confidence may be at risk in a country where inflation is increasing, prompting them to sell that country’s currency, which will have a significant impact on the exchange rate.

The findings of this study also indicate that inflation (CPI) Granger caused foreign direct investment (FDI), which was contrary to the initial hypothesis. In most cases, the evidence is inconclusive and the causal relationship between inflation (CPI) and foreign direct investment (FDI) is not consistently demonstrated. It depends on a variety of factors like the host country’s economic conditions and the specific nature of inflation. In practice, high inflation can lead to a currency being downgraded, making investments in that country riskier for foreign investors and potentially reducing the value of their assets. This is because speedy inflation can also increase business costs, making it difficult for companies to operate profitably and thus discouraging FDI. However, moderate inflation can sometimes be beneficial for FDI, as it can stimulate economic growth, boost exports and reduce the real value of debts for domestic firms.

Stock returns, however, did not Granger cause any of the macro variables, showing weak feedback effect. Interestingly, this phenomenon indicates that, while macroeconomic fundamentals drive equity markets, the reverse is weak or non-existent. In the Indian context, stock prices have limited predictive ability with regard to real economic activity, possibly due to inefficiencies, institutional restrictions, or the relatively low integration of financial markets and the real economy. This result is consistent with Mukherjee and Naka (1995), who showed that in the Japanese market, macro variables affect stock prices, but stock prices do not affect macro variables significantly. Similarly, Raza, Jawaid, and Afshan (2015) reported a lack of feedback of stock markets to macroeconomic variables in South Asia, attributing it to underdeveloped capital markets and limited investor base. The poor feedback may also be explained by market noise, speculative trades, or policy-making processes more responsive to macroeconomic objectives than short-run market movements.

This is in contrast to evidence from developed economies, where stock returns tend to precede economic indicators thanks to high informational efficiency. For instance, Chen, Roll, and Ross (1986) asserted that stock prices reveal expectations of future macroeconomic performance, suggesting strong feedback. Such bidirectional causality, however, may not be present in developing economies with less developed financial systems and macroeconomic policies not immediately responsive to market signals (Fama, 1990). These findings highlight the importance of macroeconomic stability and growth in shaping the dynamics of the capital markets and investment flows.

The Johansson cointegration test, presented in Table 5, used the stock index as the dependent variable and inflation (CPI), exchange rate, foreign direct investment (FDI) and India’s real gross domestic product (GDP) as the independent variables. Cointegration exists if the trace statistic value or maximum eigenvalue is greater than the critical value and the probability is significant, or if the error correction variable in the error correction model is negative and significant. The results of the Johansen cointegration test in Table 5 show that there are two cointegration equations, as the trace statistic value or maximum eigenvalue is greater than the critical value, and the probability values are significant. This indicates the existence of a long-run relationship between stock returns, inflation (CPI), exchange rates, foreign direct investment (FDI) and real gross domestic product (GDP) in India.

[Both the Trace Test (Unrestricted Co-Integration Rank) and Unrestricted Maximum Eigenvalue Test (Co-integration Rank) tests indicated the existence of two co-integration equation; the critical values for the test are those suggested by MacKinnon-Haug-Michelis (1999) p-values.]

The Johansen cointegration test confirms the existence of two long-term equilibrium relationships between stock returns, inflation, the exchange rate, foreign direct investment and GDP between 1990 and 2022. The first vector (normalized cointegrating coefficients; BSE_rt = 0.461CPI + 3.316ExRate – 2.484FDI – 1.049GDP + 0.0114trend) indicates that stock returns are positively correlated with inflation and exchange rate variations, but inversely with FDI and GDP, implying a combination of macro and financial factors in the long term. This suggests that currency devaluation and inflation can boost stock market performance by signalling growth or improved competitiveness of exports. Negative correlations between FDI and GDP may imply investor pessimism or monetary contraction due to growth. These long-term relationships are consistent with the findings of Mukherjee and Naka (1995), Aggarwal (1981) and Barro (1995), but contradict the conclusions of studies such as those by Fama (1990) and Chakrabarti (2001), which suggest a positive effect of GDP and FDI on equity markets.

The second vector (CPI = –35.815 × ExRate + 9.201 × FDI + 3.959 × GDP + 0.00963 × TREND) also suggests that inflation itself is influenced by depreciation in the exchange rate and is positively correlated with foreign investment and output growth. This corroborates the pass-through effect of currency depreciation and the inflationary consequences of growth (Bahmani-Oskooee & Hegerty, 2009; Barro, 1995).

Of all the variables, only stock returns adjust quickly enough to return to long-run equilibrium, highlighting the stock market’s function as an economic shock absorber. These results are consistent with macro-financial transmission mechanisms and support the use of nominal and real variables in long-run asset pricing models. Furthermore, this behaviour in the context of macro-financial transmission mechanisms demonstrates that financial markets can easily incorporate new information and respond more quickly than real macroeconomic variables such as GDP, CPI, or FDI. As stock markets are forward-looking, they capture investors’ expectations regarding future economic conditions. Consequently, they react to policy changes and economic shocks more quickly (Fama, 1990; Chen et al., 1986). This responsiveness lends support to the theoretical framework of asset pricing models, which incorporate both real and nominal variables in order to understand the interaction between financial markets and macroeconomic fundamentals (Campbell & Shiller, 1988; Bekaert et al., 2013). As demonstrated by Schwert (1988), the volatility observed during times of economic uncertainty further emphasises the market’s role in absorbing shocks and serving as a barometer for economic sentiment. The rapid adjustment mechanism of the stock market validates its role in stabilising the broader economy by absorbing and reflecting shocks before they are fully reflected in macroeconomic indicators.

Figure 1. 

Impulse Response Function (IRF). Source: Authors’ own computation

The Impulse Response Function (IRF) graph shows the cumulative effect of single standard deviation shock on important macroeconomic variables - CPI, Exchange Rate, Foreign Direct Investment, and GDP - on BSE stock return (BSERT) over a 20period horizon. The responses were estimated using the Choleski decomposition method considering orthogonal innovations.

The impulse response functions (IRFs) offer dynamic insights into the effect of one-time shocks:

• Inflation shocks lead to a persistent, positive response in stock returns, supporting growth-linked inflation theories (Fama, 1990), although opposing the uncertainty hypothesis by(Geske & Roll, 1983).
• Exchange rate shocks initially reduce stock returns but turn positive in the long run, consistent with the export-competitiveness channel discussed by Phylaktis & Ravazzolo (2005).
• FDI shocks produce a sharp and sustained negative effect on BSE returns, contradicting the findings of Chakrabarti (2001) and Dritsaki et al. (2004), but aligning with Claessens et al. (2002), who observed volatility due to unstable capital flows.
• GDP shocks also result in a negative but minor impact, aligning with studies that associate growth with monetary tightening (Barro, 1995; Chen et al., 1986), yet contradicting Bhattacharya & Mukherjee (2002), who found GDP growth to positively influence the Indian equity returns.

The Impulse Response Function (IRF) analysis reveals that macroeconomic shocks exert persistent influences on Indian stock returns, supplementing the Granger causality and Johansen cointegration outcomes. The IRF results confirm that, in the long run, inflation and exchange rate movements stimulate stock returns, while FDI and GDP shocks have a dampening effect. This emphasises the macroeconomic determination of the structural dependence of Indian equity markets and the weak feedback channel from the financial sector to the real economy (Kaur et al., 2019). These findings highlight the complex transmission mechanisms of macroeconomic shocks to capital markets and demonstrate that not all growth indicators positively impact investor confidence in the short to medium term.

In order to analyse the conditional volatility of stock returns, a GARCH (1,1) model was estimated based on the residuals of an OLS regression with stock return as the dependent variable and GDP, CPI, Exchange Rate and FDI as independent variables. The GARCH model was specified with a student’s t-distribution in order to incarcerate possible fat tails in the return distribution.

The results of the variance equation show that both the ARCH term and the GARCH term are statically significant at 5% levels, measuring 0.0751 and 0.9071 respectively. The sum of these coefficients (0.9822) is less than one, which strengthens the stationarity of the process for variance and indicates high persistence in volatility, a typical characteristic in financial time series data.

The approximated degrees of freedom of the student’s t-distribution are 10.11, which is statistically significant, suggesting the presence of leptokurtosis and supporting the use of the t-distribution instead of the normal distribution. The constant term of the mean equation is statistically insignificant as would be expected because the model is defined in terms of the residuals of the original OLS regression. The Durbin-Watson statistic (1.8168) suggests there is no serial correlation in the residuals, confirming the validity of the model. In general, the GARCH (1,1) model accurately describes the time-varying volatility of stock returns. The large and significant ARCH and GARCH terms confirm the effect of previous shocks and volatility persistence, respectively, and the application of a t-distribution enhances model stability under heavy tails. This aligns well with the works of Bollerslev (1986) and Chittedi (2014), while accounting for financial market reactions to crisis periods such as 1991 reforms, the 2008 global financial crisis and theCOVID-19 pandemic.

The volatility clustering and persistence can be seen in the conditional variance (Figure 2) obtained from the GARCH- OLS model. The time-series evolution of conditional variance gives valuable insights into market response and behaviour towards significant economic events. Conditional variance increased remarkably during the early 1990s, particularly between 1992 and 1994. This increase is likely to have coincided with India’s economic liberalisation in 1991 and global market uncertainties, and is indicative of increased volatility in returns during a period of transition. This is consistent with the findings of Bekaert and Harvey (1997), who observed that liberalisation tends to increase volatility as markets adapt to new flows of information. From the late 1990s to the early 2000s, volatility gradually declined and stabilised. This was a period of adjustment involving moderate economic fluctuations, as the Indian economy adapted to its post-liberalisation regime. This is consistent with evidence from Pandey (2005), who recorded lower volatility following structural economic integration.

Figure 2. 

Residual Diagnostic of GARCH -OLS Model. Source: Authors’ own computation

There was a sharp rise in variance during the 2008–2009 Global Financial Crisis, which is a sign of a record market reaction to systemic risk. The peak in volatility is well captured by the GARCH model, illustrating the responsiveness of financial markets to global shocks. This is supported by the evidence presented by Schwert (2011) and Nelson (1991), who both demonstrated that GARCH-type models accurately predict market responses to global shocks. In the post-crisis era of the 2010s, there has been a consistent decline in conditional variance, which suggests a decade of relatively stable markets and low uncertainty. However, there was a smaller spike around 2020 following the pandemic and subsequent lockdowns. Notably, the period’s shock of volatility was less severe than that of 2008. This can be attributed to timely monetary interventions, improved policy arrangements, and digital market resilience, as evidenced by Zhang, Hu, and Ji (2020). They argued that a quick policy response and infrastructural technology played a crucial role in mitigating volatility during the pandemic.

However, the fall in conditional variance after 2021 supports the view that financial markets ultimately stabilise following systemic shocks, thus confirming the mean-reverting volatility hypothesis proposed by Andersen et al. (2001). Overall, the trajectory of conditional variance validates the GARCH model’s ability to describe evolving market dynamics, particularly in a vulnerable, opening economy such as India’s.

The test checks for model specification errors, BSE shows no significant omitted variable bias.

This test examines whether residuals are auto correlated, Serial correlation is present at 5% significance level.

The model diagnostic tests via the Ramsey RESET test (Table 8(a)) reveal no functional form misspecification. However, the Breusch-Godfrey LM test in Table 8(b) indicates mild serial correlation in the residuals. Nevertheless, structural stability is confirmed by Figures 3(a) and 3(b) of the CUSUM and CUSUMSQ tests, respectively, both of which remain within the 5% critical bounds. This indicates stable parameters and variance throughout the sample period and reflects the maturity and regulatory improvements of the Indian market following liberalisation (Hamilton, 1994; Lamoureux & Lastrapes, 1990).

Figure 3(a). 

CUSUM

Figure 3(b). 

CUSUMSQ

The present research endeavour constitutes a wide-ranging empirical investigation into the dynamic interactions between macroeconomic indicators and Indian stock market returns. The investigation emphasises the critical role played by both long-run equilibrium relationships and short-term volatility patterns. Using 420 monthly observations of key macroeconomic variables — the Consumer Price Index (CPI), the Exchange Rate (EXRATE), Foreign Direct Investment (FDI) and Gross Domestic Product (GDP) — alongside BSE stock returns, we highlight distinctive statistical characteristics and structural relationships that emphasise the complexity of macro-financial linkages in emerging economies.

The descriptive statistics convey non-normality in all variables, with BSE returns exhibiting a leptokurtic distribution and high volatility. This affirms the presence of outliers and market shocks and justifies the use of GARCH models. Granger causality tests show that macroeconomic variables such as foreign direct investment (FDI), the exchange rate and gross domestic product (GDP) influence stock returns, but not vice versa, highlighting a unidirectional transmission mechanism. Johansen cointegration analysis reveals that inflation and the exchange rate have a positive impact on stock returns in the long term. Meanwhile, foreign direct investment (FDI) and gross domestic product (GDP) are negatively associated, suggesting market sensitivity to capital flows and policy expectations. Impulse Response analysis further supports these dynamics, showing persistent positive effects from inflation and exchange rate shocks, and negative effects from FDI and GDP. Importantly, it is only stock returns that exhibit a significant speed of adjustment to long-run equilibrium. This indicates that the stock market acts as an economic shock absorber. This behaviour reflects the forward-looking character of financial markets and their capacity to respond to macroeconomic shocks. However, the weak feedback loop between financial markets and macroeconomic variables means that, until now, policy interventions in India have largely been reactive to macroeconomic variables rather than to market signals.

Finally, model diagnostics confirm the robustness of our econometric approach. Although minor serial correlation is evident, structural stability is reinforced, indicating the resilience of India’s financial ecosystem following liberalisation.

The connection between macroeconomic variables and stock returns, which is highly relevant in the Indian stock market, may be consistent across the other BRICS countries, although sensitivities to specific variables may differ. India’s macroeconomic variables affect the other BRICS economies by generating risk and investment prospects within the bloc, despite the impact varying across different macroeconomic factors. Unlike in Anglosphere markets, India’s inflation is often correlated with its stock market, which can increase costs for other BRICS countries or fuel inflationary expectations globally. Therefore, while higher inflation could benefit India’s stock market, it could also have an adverse effect on those BRICS countries that are net oil importers due to increased costs. Conversely, a strengthening of the Indian rupee, driven by capital inflows, has the potential to bolster the currencies of the other BRICS members. Changes in interest rates, exchange rates and global commodity prices, particularly oil, also may have spillover effects on the other BRICS stock markets. The Indian stock market’s performance can influence the exchange rates of other BRICS countries, as capital often flows in tandem between them. Political instability and trade disputes among the BRICS countries, like those between India and China, can lead to increased uncertainty and risk for investors across the bloc.

Therefore, the empirical findings of this research are of significant relevance in the context of economic development in the BRICS countries. Policymakers in BRICS can use the dynamic nature of the studied relationships to design appropriate economic policies and mitigate risks associated with volatility in financial markets.

This research highlights the asymmetric, complex and largely unidirectional relationship between macroeconomic fundamentals and the Indian stock market. It confirms that stock market dynamics are critically governed by macroeconomic stability and investor expectations shaped by inflation, currency movements, and capital flows. The limitations of the study lie in its scope: the study does not consider a wide range of macroeconomic variables, such as oil and gold prices, corruption, money supply, industrial production indexes, interest rates, business freedom, institutional efficiency, government effectiveness, economic freedom and the ease with which businesses can operate. Besides, the NSE and BSE indices are not taken into account simultaneously. Future studies should further explore how structural reforms and technological advancements influence these parameters, especially in conditions of financial globalization and geopolitical uncertainty. They may also extend their analysis to cross-country exploration using panel data sets that take into account other exogenous factors affecting stock markets, which are absent from this study.

The authors would like to thank the honourable anonymous referees for their valuable comments and helpful suggestions on an earlier draft of this article. Nevertheless, if any pitfalls remain in the article after careful revisions, the responsibility lies solely with the authors themselves.

The authors declare no conflict of interest.