Dynamic Nexuses between Macroeconomic Variables and Sectoral Stock Indices: Reflection from Indian Manufacturing Industry

Abstract

Determination of significant sector specific macroeconomic factors under the board manufacturing industry is an important task. In Indian context, using the monthly data on five major manufacturing sector specific indices (such as BSE-Basic Materials, BSE-Consumer Discretionary Goods and Services, BSE-Fast Moving Consumer Goods, BSE-Health Care and BSE-Industrials) and the macroeconomic variables (gold price, index of industrial production, wholesale price index, money supply, foreign portfolio investment ratio (FPIR), rate of interest, real effective exchange rate and crude oil price and economic policy uncertainty) for the period September, 2005 to November, 2016, the present study attempted to explore the significant sector specific macroeconomic variables in long run as well as short run. The empirical results obtained by applying the ARDL-UECM model suggested that economic policy uncertainty, FPIR and price factor were observed to be the most important determinants of all the five sectoral stock indices for the study period.

Keywords

Manufacturing sector stock indices economic policy uncertainty foreign portfolio investment ratio price factor

Introduction

Stock market development is one of the most important prerequisites of industrial development (Sultana & Reddy, 2017) and economic growth of a country as it acts as an important medium in the financial system that helps in channelizing fund from the surplus sector to the deficit sector in an economy (Mohammad et al., 2009). In the Efficient Market Hypothesis, it is argued that stock price absorbs every sort of information (historical, public and even private). For a long period of time efforts have been made by the scholars across the world to ascertain the different factors that affect the stock market. Share prices of the companies in the stock market get affected due to a number of factors. These are micro-economic, i.e., firm specific factors, industry specific factors and macroeconomic factors. Amongst all, with reference to a globalized and liberalized economy, perhaps the macroeconomic factors are the most important determinants of share prices of the different companies. Thus, the issue on the inter-relationship between macroeconomic fundamentals and the stock market movement has received considerable attention from the scholars across the globe.

The share price of a particular company gets determined at the equilibrium point of the supply and demand situation in the market. The firm specific, industry specific and the macroeconomic factors have a direct bearing on the demand and supply situation of the shares in the stock market leading to the movement of share prices in either direction. In Figure 1, the inter-relation between the various macroeconomic variables and share price is presented.

Figure 1.

Source: The authors.

For example, reduction of the cost due to the implementation of the new technology will have positive impact on the operating earnings of the company and this information acts as a catalyst in creating a positive sentiment about the share of that particular company, which increases its demand in the market leading to increase in the share price. Similarly, entry of a new but strong market player enhances the competition in the same industry or imposition of additional indirect tax by the government on a particular industry results in creation of a negative sentiment amongst the investors, which ultimately reduces the demand of the shares in the market and consequently the prices of the shares of the companies belonging to that industry get negatively affected. In a similar fashion macroeconomic fundamentals such as growth rate of the GDP or trend of Index of Industrial Production (IIP), inflation, money supply (MS), interest rate, openness of the economy, political risk, foreign fund flow, etc., have a direct bearing on the stock market. For example, if the foreign investment increases that will create more demand of the share in the stock market and such a rising demand will also positively affect the overall stock market. It is not possible to track the shares of each and every company and therefore stock market index is generally used as an indicator of the overall health of the stock market.

Review of Existing Literature

The issue associated with the impact of the macroeconomic factors on the share price discovery is a debatable issue. In line with the Random Walk Hypothesis, Malkiel and Fama (1970) propagated that the effect of changes in the macroeconomic factors is supposed to be factored by the investor community and thus modelling the stock price return for making an abnormal profit is not possible. However, this proposition was challenged by Fama and Scwartz (1977) and Ross (1976). The Arbitrage Price Theory suggests that the macroeconomic variable and the stock market are interlinked with each other. Later on, to test this hypothesis, many scholars across the globe have tried to investigate into the linkages between the movement of the stock market index and different macroeconomic variables. Some of the notable contributions in this direction are made by Fama and Scwartz (1977), Chen et al. (1986), Fama and French (1989), Mukherjee and Naka (1995), Easterly and Rebelo (1993), Thornton (1993), Kaneko and Lee (1995), Cheung and Ng (1998), Darrat and Dichens (1999), Nishat et al. (2004), Al-Sharkas (2004), Herve et al. (2011), Sultana and Pardhasaradhi (2012), and Antonakakis et al. (2013). The genesis of the effect of the macroeconomic variables on the stock market can be traced from the simple dividend discounting model of Share Price Valuation.¹

There are many macroeconomic variables that bear long-run relationship with the stock prices such as MS (Al-Sharkas, 2004; Kraft & Kraft, 1977), inflation (Ammer, 1994), government consumption (Grier & Tullock, 1989), government policy (Croce et al., 2012), growth of GDP (Barro, 1996; Durham, 2002), FDI and FIIs (Clark and Berko, 1997; Froot et al., 2001), policy uncertainty (Antonakakis et al., 2013), oil prices (Hondroyiannis & Papapetrou, 2001), rate of interest (Malkiel, 1982), exchange rate (Granger et al., 2000), gold price (GP), investment in housing sector, industrial production index, domestic interest rate (Al-Sharkas, 2004), and foreign exchange reserve (Chen et al., 1986; Nishat et al., 2004), international crude oil price (COP) (Gay Jr., 2011; Hosseini et al., 2011), integration with the other international stock markets (Sheng and Tu, 2000). In Indian scenario, in the light of the existing studies it can also be concluded that the important stock indices in India such as Nifty and Sensex have also a strong relationship with the various macroeconomic fundamentals mentioned earlier (Agrawalla & Tuteja, 2008; Ferdows & Roy, 2012; Hassan & Sangmi, 2013; Mukhopadhyay & Sarkar, 2003; Naik & Padhi, 2012; Patel, 2012; Pethe & Karnik, 2000; Rao & Bhole, 1990; Srinivasan, 2011; Sultana & Pardhasaradhi, 2012). One lacuna that can be pointed out in the context of the existing literature is that, generally, when any stock market index (such as Sensex or Nifty or S&P 500 or Nikkei) is considered, it represents the market sentiments because the indices are constructed with the help of market capitalization of the companies across different sectors. Therefore, one pertinent question that may be raised is that whether all the sectors or industries or the companies belonging to the different sectors or industries get equally affected due to the dynamic nature of the macroeconomic variables. Are all the macroeconomic factors equally relevant to the different industries? It is obvious that the answer would disagree with the propositions already made by the scholars. It is generally accepted that since the different industries are different in respect of nature, customer base, demand situation, foreign market operational exposure, cost structure and financing pattern, the effect of the dynamic and diverse macroeconomic variables on the stock prices of the companies belonging to the different industries would also be different.

In this prelude, the focus of the present study is primarily concentrated on exploring the significant sector or industry specific macroeconomic variables and also on the assessment of the long-run as well as short-run relationships between those macroeconomic factors and the sectoral stock indices pertaining to the Indian manufacturing industry. Careful review of the existing literature reveals that the studies so far carried out to examine the relationship between the different macroeconomic variables and the sector specific stock indices especially with reference to the Indian manufacturing industry are scanty.

Research Objective

The review of the existing literature suggested that cross country empirical studies have attempted to establish the long-run as well as short-run relationships between the macroeconomic variables and the stock market using broad market indices (such as Sensex or Nifty in Indian context, Shanghai Stock Exchange Composite Index in respect of China, Kuala Lumpur Stock Exchange Index for Malaysia, etc.). However, there is a dearth in the literature that has tried to examine such relationship with regard to sectoral stock indices especially in the context of Indian manufacturing industry. Keeping this aspect in focus, the present study attempts to bridge this research gap and contribute to the existing literature by examining the interrelation between the different macroeconomic fundamentals and the sectoral stock indices belonging to the Indian manufacturing industry and thereby identifying the crucial macroeconomic variables affecting the manufacturing sector specific indices.

Research Design

Data Sources

The review of the existing literature indicates that there are various macroeconomic variables having significant influence on the stock market. Taking cue from the existing literature GP, IIP, wholesale price index (WPI), MS (as proxied by M3, i.e., broad money), foreign portfolio investment ratio (FPIR), rate of interest (as indicated by 91 days Treasury Bill Rate), real effective exchange rate (REER) and COP were taken into account while conducting this study. Inclusion of these variables will serve the following purposes.

In all the studies so far carried out, the relationship between these macroeconomic variables and broad market indices was assessed. However, the present study attempted to examine whether there is a notable relationship between the macroeconomic variables and the sectoral manufacturing indices.

Unlike other studies, economic policy related variable was also introduced in this study as economic policy uncertainty (EPU) plays an important role in shaping the movement of the stock markets.

The present study attempted to identify the important macroeconomic variables having impact on overall stock market indices out of all the macroeconomic variables that were identified by the previous researchers. This will also help in checking whether the outcomes of the study are consistent with those derived from other studies.

There are altogether nine constituent sectors of S&P BSE India Manufacturing Index² that was obtained from the official website of Asia Index Private Limited.³ These are Basic Materials, Consumer Discretionary Goods and Services, Fast Moving Consumer Goods, Health Care, BSE-Industrials, Energy, Telecom, Information Technology and Diversified. Under these nine BSE-manufacturing sectors there are 69 BSE-industry subgroups, out of which 18 subgroups belong to Basic Materials, 14 belong to Consumer Discretionary Goods and Services, 12 belong to Fast Moving Consumer Goods, 4 belong to Health Care, 12 belong to Industrials, 4 belong to Information Technology, 2 belong to Telecom, 1 belong to Diversified and 2 belong to Energy industry subgroups. For the purpose of this study, five major manufacturing sector specific indices, such as BSE-Basic Materials (BSE-BM), BSE-Consumer Discretionary Goods and Services (BSE-CDGS), BSE-Fast Moving Consumer Goods (BSE-FMCG), BSE-Health Care (BSE-HC) and BSE-Industrials (BSE-IND), were selected. There are a number of reasons for selecting these five manufacturing BSE indices.

Firstly, these five sectors together represent a total number of 60 (86.96%) industry sub-groups. Therefore, it can be deduced that the selected five BSE-Sectors are dominant within the overall manufacturing sector. The rest of the BSE-Sectors such as Information Technology, Telecom, Diversified and Energy together only account for 13.04% of the total industry subgroups.⁴

Secondly, the information and Technology sector includes firms belonging to ‘Networking Equipment’, ‘Computer Hardware’, ‘Storage Media & Peripherals’ and ‘Electronic Components, Equipment & Instruments’ industry subgroups. However, the BSE-Information & Technology Index⁵ is constituted taking into consideration the software development companies (such as HCL Technologies Ltd, Tech Mahindra Ltd, Wipro Ltd, Mind Tree Ltd, Tata Consultancy Services Ltd, etc). It can be observed that there is no such information and technology manufacturing index in BSE. Thus, inclusion of BSE-Information & Technology Index for the purpose of this study was not possible.

Thirdly, telecom manufacturing sector includes the firms operating in ‘telecommunications equipment’ and ‘telecom cables’ industry sub-groups. However, the available BSE-Telecom index⁶ is constructed taking into account telecom services companies (such as Bharti Airtel Ltd, Reliance Communications Ltd, Vodafone Idea Ltd, etc.). Thus, BSE-Telecom index was also not taken into account for the purpose of this study.

Fourthly, the fate of the Information Technology and Telecom sector is largely associated with the development and trend of the service sector as a whole. So the macroeconomic variables that affect the BSE-Telecom and BSE-Information & Technology will also be dominating the stock price of the telecom and information & technology manufacturing sub-industries.

Fifthly, the ‘diversified’ manufacturing firms could not be included in the study as there is no BSE index in respect of the ‘diversified’ manufacturing firms.

Sixthly, although BSE-Energy index considers the firms belonging to the ‘refineries & petroleum products’ and ‘integrated oil & gas’, the growth and development of this sector gets significantly affected by the oil production and price at the international level. Therefore, it is expected that COP will be the dominating macroeconomic variable affecting BSE-Energy Index. Moreover, the representation of BSE-Energy in terms of industry sub-group is very small (2.89%) relative to the total number (69) of industry sub-groups under BSE-Sector.

In this study, the period September, 2005 to October, 2016 was chosen as the reference time frame with respect to which the monthly data for all the macroeconomic factors and sectoral indices were available. The monthly sectoral stock indices data were collected from the official website of Bombay Stock Exchange,⁷ Mumbai. The monthly data of GP, IIP, MS, FPIR, WPI, REER and 91-days Treasury Bill Rate were obtained from the Database of Indian Economy website of Reserve Bank of India,⁸ whereas the data of COP and EPU were collected from the Petroleum Planning and Analysis Cell of Ministry of Petroleum and Natural Gas, Government of India⁹ and Baker et al. (2016),¹⁰ respectively. It must be noted that all the variables were converted into its natural logarithmic values, which not only helped in reducing the sharpness of the data but also provided the measurement of elasticities (Shahbaz et al., 2016). Trend of all these macroeconomic variables is presented in Figure A.1 of the Appendix.

Principal Component Analysis (PCA)

Factor analysis is one of the most popular mathematical factor models for reduction of dimensions of the data that immensely helps in getting rid of the multicollinearity problem (Brooks, 2014). Tripathi and Seth (2014) in their study found multicollinearity problem due to the presence of high degree of correlation between the pairs of WPI, IIP, MS and oil price. Therefore, PCA was used by them to construct a comprehensive variable for incorporating the basic features of the four closely related variables. In this study also it was observed that the pairs of different combination of the four macroeconomic variables such as WPI, IIP, MS and GP possessed high degree of correlation.¹¹ Thus, in line with Tripathi and Seth (2014), these four factors were analyzed using PCA to form a new variable named ‘Price Factor’ (PF). However, there is the requirement of giving the economic logic of considering these variables together to get the clubbed variable PF. In the first instance, it can be argued that the pair wise correlations of the variables were found to be positive during the period of study. Secondly, the positive relationship between the MS and WPI is well documented in the literature. In fact, the Quantity Theory of Money propagates that, as the supply of money increases, the inflation also moves in the same direction due to demand side pressure (Lucas, 1980; Siddique, 1975). In the pioneering attempt, Friedman and Schwartz (1963) opined that any increase/decrease in the supply of money was followed by business expansion/contraction respectively. In fact, there is also a vast literature supporting the argument of the presence of a positive relationship between the stock of money and the economic activity (Ahmed, 1993; Ogunmuyiwa & Ekone, 2010; Sims, 1972; Thornton, 1993). Therefore, following the positive linkage between the MS and economic activity and Quantity Theory of Money, it is argued that as the MS increases, the economic activity also increases and with the increase in MS the inflation also enhances. Moreover, the positive relationship between inflation and real economic activity is well accepted in the literature in the light of Philips Curve Theory¹² (Ram & Spencer, 1983). Similarly, the people tend to invest in gold in anticipation of inflation. Therefore, inflation and GP move in the same direction (Mahdavi & Zhou, 1997; Tkacz, 2007; Worthington & Pahlavani, 2007). Considering these arguments the four factors (in presence of high degree of correlation) were clubbed together to form the new variable PF.

For four macroeconomic variables (which were highly correlated), PCA technique can yield four principal components that are independent of the linear combination of the original variables (Brooks, 2014) in the following manner:

\ln ​ P F = β_{i j} \ln ​ G P + β_{i j} \ln ​ M S + β_{i j} \ln ​ I I P + β_{i j} \ln ​ W P I,

where β_ij are the respective coefficients or factor loadings to be estimated for the ‘jth’ variable in the ‘ith’ principal component. The factor loadings with respect to the different variables were first determined using PCA. Later, the factor loadings were normalized by scaling to unity sum as per the methodology suggested by Research Centre-European Commission (2008). The weights¹³ that were finally used to construct PF were 0.25 for GP, 0.26 for MS, 0.24 for IIP and 0.26 for WPI. The observed value of KMO χ² was also found to be statistically significant indicating adequacy of the sample size in the study.

Stationarity Test

In financial economics, it is observed that most of the financial or macroeconomic data are found to be non-stationary with no tendency to revert back to an underlying trend value.¹⁴ In this context, the most desirable characteristics of macroeconomic and financial time series data are the property of stationary.¹⁵ For the purpose of investigating into the stationarity of the variables two popular methods such as Augmented Dickey-Fuller test (ADF test) and Philips and Perron test (PP test) are used.

In this study to test the stationary property of the time series data, ADF test was used. ADF test is an augmented version of the Dickey-Fuller test. For a series y_t the ADF test consists of a regression of the first difference of the series against the series lagged k times as follows:

Δ y_{t} = α + δ y_{t - 1} + \sum^{​} ​ β_{s} Δ y_{t - k} + ε_{t},

where ∆y_t denotes the change in the value of y and ε_t is the white noise error term.

The following hypotheses are tested under ADF test:

H_{0} : δ = 0 (U n i t r o o t e x i s t s),

H_{1} : δ < 0 (N o u n i t r o o t e x i s t s) .

To test H₀, the value of test statistic $τ = \frac{(\hat{δ} - 1)}{S E (\hat{δ})}$ is computed. Then, the computed value of τ is compared with the Mac Kinnon critical value. If the computed value of τ is more than the Mac Kinnon critical value, then the null hypothesis of δ = 0 gets rejected and it can be concluded that no unit root is present and therefore the series is stationary.

The assumption underlying the ADF test is that the error term is serially independent with a constant variance (Bhaumik, 2015). However, it may not be valid in reality in many cases. To overcome this problem Philips and Perron (1988) suggested the use of non-parametric method to tackle the serial correlation in the error terms (Gujarati, 2009).

PP test considers estimation of the following equation:

Δ y_{t} = \emptyset_{0} + δ y_{t - 1} + α_{t} .

Like ADF test, similar hypotheses are also formulated in PP test.

Under PP test a non-parametric correction is introduced in the t-statistics to make correction of any serial correlation and heteroscedasticity in the errors a_t of the test regression by directly modifying the test statistics.

Autoregressive Distributed Lag—Error Correction Model

There are a number of ways to examine the short-run and long-run relations between the different macroeconomic factors and the sector specific stock indices. One of the popular techniques for determining long-run inter-relation is the cointegration test as developed by Johansen (1988). This method is used to test the restrictions imposed by cointegration on the unrestricted vector auto regression (VAR) involving the series. The estimation procedure used in the cointegration test is based on the error-correction representation of the VAR model with Gaussian errors. In this method, if it is established that the variables possess long-run relationship, then the short-run adjustment process is represented by the vector error correction model. However, Johansen test of cointegration is only applicable if all the variables are of same order (more specifically integrated of order one). Generally, the time series data should be non-stationary at level but stationary at first difference. But in reality, most of the financial time series data do not follow this property (Nkoro & Uko, 2016). In different studies it is found that variables are integrated of different orders, i.e., I(0), I(1) or I(2). This type of problem can be easily tackled by the autoregressive distributed lag (ARDL) model as suggested by Pesaran and Shin (1998) and Pesaran et al. (2001). The ARDL model is capable of capturing the long-run and short-run relationships among the time series variables that are either I(0) or I(1) but not I(2) (Bhattacharya et al., 2019; Gokmenoglu & Fazlollahi, 2015; Khan et al., 2017; Lima et al., 2016; Pesaran et al., 2001; Shahbaz et al., 2016; Tursoy, 2019; Wada, 2017). Moreover, the ARDL bound test for cointegration technique¹⁶ is also employed in deriving the unrestricted error correction model (UECM) through a linear transformation of the ARDL model. The error correction term (ECT) derived from such reparameterization of the ARDL model yields the short-run adjustment speed to restore long-run equilibrium (Nkoro & Uko, 2016) without losing the longrun information.

The ARDL¹⁷ specification is expressed in the form of the following equations:

\begin{array}{l} \begin{matrix} Δ \ln ​ B S E_{t}^{k} = α_{0} + \sum_{i = 1}^{n} α_{1} Δ \ln ​ B S E_{t - i}^{k} + \sum_{i = 1}^{n} α_{2} Δ \ln ​ E P U_{t - i} \\ + \sum_{i = 1}^{n} α_{3} Δ \ln ​ F P I R_{t - i} + \sum_{i = 1}^{n} α_{4} Δ \ln ​ I N T_{t - i} \\ + \sum_{i = 1}^{n} α_{5} Δ \ln ​ R E E R_{t - i} + \sum_{i = 1}^{n} α_{6} Δ \ln ​ P F_{t - i} \end{matrix} \\ \begin{matrix} + \sum_{i = 1}^{n} α_{7} Δ \ln ​ C O P_{t - i} + δ_{1} \ln ​ B S E_{t - 1}^{i} + δ_{2} \ln ​ E P U_{t - 1} \\ + δ_{3} \ln ​ F P I R_{t - 1} + δ_{4} \ln ​ I N T_{t - 1} + δ_{5} \ln ​ R E E R_{t - 1} \\ + δ_{6} \ln ​ P F_{t - 1} + δ_{7} \ln ​ C O P_{t - 1} + ε_{t} \end{matrix} \end{array}

(1)

where k = BM, FMCG, HC, IND and CDGS.

In Equation (1), ∆ is the first difference operators α₁ to α₇ are the short-run dynamics and δ₁ to δ₇ are the long-run dynamics.

The long-run relationship is explored by estimating the bound F-statistics (Nkoro & Uko, 2016). For this purpose, the following hypothesis has to be tested with reference to the critical value range within I(0) and I(1) as suggested by Pesaran et al. (2001).

$H_{0} : δ_{1} = δ_{2} = δ_{3} = δ_{4} = δ_{5} = δ_{6} = δ_{7} = 0 (No long-run relationship),$

$H_{1} : δ_{1} \neq δ_{2} \neq δ_{3} \neq δ_{4} \neq δ_{5} \neq δ_{6} \neq δ_{7} \neq 0 (Long-run inter-relation)$ .

In the ARDL approach, the selection of the appropriate lag length is very crucial in order to ensure that the Gaussian error term is free from non-normality, autocorrelation and heteroskedasticity problems (Omoniyi & Olawale, 2015). The optimum lag order¹⁸ was selected based on appropriate model selection criteria such as Akaike information criterion (AIC), and Schwarz Bayesian criterion (SBC) (Nkoro & Uko, 2016). For the purpose of this study, the trial and error process was used in case of the selection of maximum lag also so that the best ARDL model can be selected. For example initially for both the dependent and independent variables the maximum lag was defined as (1,1) within which the optimal lag length to be selected, in the next instance maximum lag of (1,2) then (1,3) was tested and the process was continued for all the possible combination up to (5,5). For each such combination the AIC values were noted and finally the model with the lowest possible of value of AIC was finally selected as the appropriate ARDL model. Thus, different variables under the ARDL model had different optimal lag length. Once the long-run relationship was established, in the second step, the UECM was estimated to find out the short-run dynamics. This was easily done by reparameterization of the ARDL model to estimate the ECM.

The ARDL-UECM is expressed as follows:

\begin{matrix} Δ \ln ​ B S E_{t}^{k} = α_{0} + \sum_{i = 1}^{n} α_{1} Δ \ln B S E_{t - i}^{k} + \sum_{i = 1}^{n} α_{2} Δ \ln E P U_{t - i} + \sum_{i = 1}^{n} α_{3} Δ \ln F P I R_{t - i} \\ + \sum_{i = 1}^{n} α_{4} Δ \ln I N T_{t - i} + \sum_{i = 1}^{n} α_{5} Δ \ln R E E R_{t - i} + \sum_{i = 1}^{n} α_{6} Δ \ln P F_{t - i} \\ + \sum_{i = 1}^{n} α_{7} Δ \ln C O P_{t - i} + λ E C T_{t - 1} + ϕ_{t} \end{matrix}

(3)

where k = BM, FMCG, HC, CDGS and IND.

In Equation (3), ECT is the residuals of the cointegrating models estimated through solving Equation (2). λ is the speed of adjustment. If the value of λ is positive it implies that the variables under consideration are drifting apart in the short run and therefore there will be no long-run equilibrium. On the other hand, negative value of λ indicates that the variables are coming close to each other at the speed of λ in each period to converge in the long run. Thus, a negative and statistically significant value of λ is imperative to argue in favour of both short-run and long-run relationships among the variables.

Finally, some diagnostic and stability tests were also performed in order to ensure that the results derived from the ARDL-ECM were reliable as well as stable. Breusch–Godfrey serial correlation Lagrange multiplier test (LM) was performed to examine whether the serial correlation in the residuals was present. Autoregressive Conditional Heteroscedasticity (ARCH) test was applied in examining whether there was no heteroscedasticity problem in the error terms and the error terms were independent of the explanatory variables. The normality of the residuals was checked by using Jarque–Bera (J–B) test and the correctness of the model specification was examined using Ramsey’s RESET test. Moreover, cumulative sum of recursive residuals (CUSUM) and cumulative sum of squares of recursive residuals (CUSUM of squares) plots were used to ensure the stability of the ARDL-ECM and the long-run and short-run coefficients were correctly derived.

Structural Break

To examine the effect of financial crisis of 2008–09 on the manufacturing stock indices, the existence of structural breaks in the relationship between the sectoral stock indices and the different explanatory macroeconomic variables was assessed. In order to do so first the multiple break points in the relationship were tested by using Bai–Perron test,¹⁹ on the basis of the structure of the data. Once the multiple break points were identified with respect to the different indices (in ARDL framework), in the next stage Chow test (with known break points) was used to determine the significance of the already determined break points. Existence of statistically significant break point within the period 2008–09, in fact, validates the effect of global financial crisis on the sector specific indices.

Bai–Perron Test of Multiple Break Points

Multiple break point methodology is quite helpful in a situation where the break points are unknown. Under this approach the following multiple break equations are considered assuming ‘n’ number of breaks:

y_{t} = γ {\overset{´}{x}}_{t} + δ_{1} {\overset{´}{z}}_{t} + μ_{t}, w h e r e t = 1, ​ 2, \dots \dots, T_{1}

(A)

y_{t} = γ {\overset{´}{x}}_{t} + δ_{2} {\overset{´}{z}}_{t} + μ_{t}, w h e r e t = T_{i} + 1, T_{i} + 2, \dots \dots, T_{2}

(B)

y_{t} = γ {\overset{´}{x}}_{t} + δ_{n + 1} {\overset{´}{z}}_{t} + μ_{t}, w h e r e t = T_{n} + 1, T_{n} + 2, \dots \dots, T

(C)

where y_t is the dependent time series variable, ${\overset{´}{x}}_{t}$ is the $(q ×1)$ , γ and δ_j( j = 1, 2, …., n + 1) are the respective vector coefficients. μ_t is error term at time t. The unknown break points (T_1, T₂, …., T_n) are estimated along with the unknown parameters (γ and δ₁, … … .,δ_n₊₁) when T observations are available under this approach. Once the break points are indentified, the entire time frame gets divided into ‘n’ number of break points and specific model for each of such segments is estimated. It is to be noted at this point that in the general framework of the model ${\overset{´}{x}}_{t}$ is assumed to be variable that does not have any break (Carlson et al., 2000). This type of theoretical assumptions may be taken up in special cases and therefore it can be seen that in the general specifications from model (A) to (C) the vector of the coefficients associated with the vector of the covariate ${\overset{´}{x}}_{t}$ has remained same (γ). In the case of this study no such theoretical assumptions were made relating to the non-breaking covariate. Thus, the general form can be finally written as follows:

y_{t}^{k} = δ_{n + 1} {\overset{´}{z}}_{t} + μ_{t}, w h e r e t = T_{n} + 1, T_{n} + 2, \dots \dots, T,

where $y_{t}^{k}$ represents the vector of independent variables under ARDL framework, δ_n₊₁ indicates the respective vector of coefficients and μ_t implies the error term.

Chow Test

In the Chow test procedure,²⁰ F-test chooses the specification of the equation that better suits the empirical data: single regression equation or two different regression equations.

To explain it let us assume the following equations:

y_{t} = β_{0} + β_{1} x_{1} + ε_{t}

(a)

y_{t} = γ_{0} + γ_{1} x_{1} + ε_{1 t}

(b)

y_{t} = δ_{0} + δ_{1} x_{1} + ε_{2 t} ​

(c)

Equation (a) represents a single regression equation with no consideration of structural break. But if structural break appears at time t then the single regression line will have a kink at t period. Therefore, before and after t, the nature of the relationship may alter, and thereby suggest the presence of a structural break. Equations (b) and (c) actually represent two regression equations that are derived by segregating (a) based on the structural change at period t. If the parameters are equal (γ₀ = δ₀ and γ₁ = δ₁) then there is no structural break and therefore, Equation (a) is suitable. However, if γ₀ ≠ δ₀ and γ₀ ≠ δ₀, there exists structural break and two separate regression equations (b and c) are to be estimated instead of a single Equation (a). F-test is used to test the null hypothesis (H₀) that there exists no structural break:

F s t a t i s t i c s = \frac{[R S S_{p} - (R S S_{1} + R S S_{2})] / k}{(R S S_{1} + R S S_{2}) / (n_{1} + n_{2} - 2 k)} ~ F, (k), (n_{1} + n_{2} - 2 k) d . f .,

where RSS_p is the sum of residuals for pooled data, RSS₁ is the sum of residuals for the data before the break and RSS₂ is the sum of residuals for the data after the break, k is the total number of parameters to be estimated (three in this case) and n₁ and n₂ are the respective number of observations in the pre-break and post-break periods, respectively.

Results and Discussion

In Figure 2, the trends of the sectoral manufacturing stock indices are presented. The evaluation of the trend lines clearly showed that BSE-FMCG and BSE-HC followed upward rising movement during the period of study. The prime reason may be the huge domestic market demand for food products and pharmaceutical products, which enabled these two sectors to become resilient as compared to other sectors in the manufacturing industry. Other indices such as BSE-BM, BSE-IND and BSE-CDGS depicted slight upward rising movement along with a very erratic movement during the period under study. It can be seen that all the sectoral indices declined sharply during 2008–09 as during that period there was huge drainage of FPIs from the Indian stock market. However, it is also observed that the recovery of the sectoral indices took place from the end of 2009. The rate of recovery was noticed to be very fast in respect of BSE-FMCG and BSE-HC.

Figure 2.

Source: The authors.

The visual inspection of the trend line of the different indices apparently discloses that the volatility of BSE-HC and that of BSE-FMCG were less as compared to those of the other indices. However, the analysis of the S.D. and C.V. from the descriptive statistics²¹ of the indices revealed that the volatility²² of BSE-HC and that of BSE-FMCG were found to be higher as compared to the other indices. The genesis of this difference lies in the range of the different indices. The ranges of BSE-FMCG (7.235–9.085) and BSE-HC (7.862–9.801) were much higher than the respective ranges of BSE-BM (6.473–7.844), BSE-CDGS (6.389–8.091) and BSE-IND (6.797–8.127).

One of the prerequisites of applying the time series data is to identify the order of integration for avoiding any sort of spurious results to be derived from the time series econometric techniques. ADF and PP tests were used in order to check whether the variables were stationary or not at their level form and at their first difference form under the assumptions that the data had ‘constant and no trend’ and ‘constant and linear trend’.

The analysis of Table 1 reveals that BSE-BM, BSE-IND, EPU and PF were observed to be stationary at their level form indicating that these variables were integrated of order zero, i.e., I(0), whereas all other sectoral stock indices (BSE-CDGS, BSE-FMCG and BSE-HC) and rest of the macroeconomic variables (INT, COP, REER and FPIR) were found to be stationary at first difference, i.e., integrated of order one.

Table 1.

Test of Stationarity

Variables	Test	Constant and No Trend		Order of Integration	Constant and Linear Trend		Order of Integration
Variables	Test	Level	First Difference	Order of Integration	Level	First Difference	Order of Integration
lnBSE^BM	ADF	–3.06^$	–9.63^#	I(0)	–3.24*	–9.60^#	I(0)
	PP	–2.75*	–9.72^#	I(0)	–3.05	–9.69^#	I(0)
lnBSE^CDGS	ADF	–0.85	–9.93^#	I(1)	–1.83	–9.89^#	I(1)
	PP	–1.25	–10.07^#	I(1)	–2.36	–10.04^#	I(1)
lnBSE^FMCG	ADF	–0.78	–12.54^#	I(1)	–2.56	–12.51^#	I(1)
	PP	–0.75	–12.54^#	I(1)	–2.50	–12.51^#	I(1)
lnBSE^HC	ADF	–0.24	–11.71^#	I(1)	–2.25	–11.66^#	I(1)
	PP	–0.24	–11.70^#	I(1)	–2.37	–11.65^#	I(1)
lnBSE^IND	ADF	–3.02^$	–8.85^#	I(0)	–3.32*	–8.83^#	I(0)
	PP	–2.63*	–8.82^#	I(0)	–2.96	–8.81^#	I(1)
lnEPU	ADF	–2.28	–12.76^#	I(1)	–1.90	–12.92^#	I(1)
	PP	–3.41	–18.03^#	I(0)	–3.22*	–21.12^#	I(0)
lnFPIR	ADF	–7.87^#	–9.16^#	I(0)	–7.84^#	–7.84^#	I(0)
	PP	–7.83^#	–53.26^#	I(0)	–7.80^#	–70.15^#	I(0)
lnINT	ADF	–1.98	–15.22^#	I(1)	–1.95	–15.18^#	I(1)
	PP	–2.42	–15.21^#	I(1)	–2.50	–15.17^#	I(1)
lnREER	ADF	–1.74	–10.05^#	I(1)	–2.27	–10.02^#	I(1)
	PP	–1.92	–10.00^#	I(1)	–2.44	–9.96^#	I(1)
lnPF	ADF	–3.54*	–16.15^#	I(0)	–0.86	–17.15^#	I(1)
	PP	–3.40^#	–15.56^#	I(0)	–1.30	–17.22^#	I(1)
	ADF	–2.42	–7.20^#	I(1)	–2.44	–7.23^#	I(1)
lnCOP	PP	–1.96	–7.19^#	I(1)	–1.93	–7.21^#	I(1)

Source: The authors.

Note: *, $ and # indicate statistically significant at 10 per cent, 5 per cent and 1 per cent, respectively.

It is very much evident from Table 2 that the different variables were integrated of different order, i.e., I(0) or I(1) but not I(2). Therefore, in order to establish the long-run and short-run inter-relations between the sectoral stock indices and the different macroeconomic variables, Johansen’s test of cointegration and VECM could not be applied. As an alternative to these methods ARDL model and the UECM were used in determining the long-run and short-run relationships among the sectoral indices and the different macroeconomic variables as discussed in the methodology section.

ARDL bound test approach as suggested by Pesaran et al. (2001) is presented in Table 2. In this test procedure, F-statistics were compared with the ARDL critical bound values at I(0) and I(1). It is observed from Table 2 that the computed values of F-statistic exceeded the upper bound, i.e., I(1). It shows that a long-run relationship between the different sector specific manufacturing stock BSE-indices and the macroeconomic variables was noticed.

Table 2.

Result of Bound Test for Cointegration

Equation	F-statistics	ARDL Critical Bounds at 1 per cent Level of Significance
Equation	F-statistics	I(0)	I(1)
$F_{\ln B S E^{B M}} (\ln B S E^{B M} \| \ln E P U, \ln F P I R, \ln I N T, \ln R E E R, \ln P F, \ln C O P)$	7.20	2.88	3.99
$F_{\ln B S E^{C D G S}} (\ln B S E^{C D G S} \| \ln E P U, \ln F P I R, \ln I N T, \ln R E E R, \ln P F, \ln C O P)$	12.90	2.88	3.99
$F_{\ln B S E^{F M C G}} (\ln B S E^{F M C G} \| \ln E P U, \ln F P I R, \ln I N T, \ln R E E R, \ln P F, \ln C O P)$	4.64	2.88	3.99
$F_{\ln B S E^{H C}} (\ln B S E^{H C} \| \ln E P U, \ln F P I R, \ln I N T, \ln R E E R, \ln P F, \ln C O P)$	5.98	2.88	3.99
$F_{\ln B S E^{I N D}} (\ln B S E^{I N D} \| \ln E P U, \ln F P I R, \ln I N T, \ln R E E R, \ln P F, \ln C O P)$	10.65	2.88	3.99

Source: The authors.

Long-run Elasticity

Once the prevalence of the long-run relationship was established between the different macroeconomic variables and the sectoral stock indices, then the next step was to determine the long-run elasticities of the different independent macroeconomic variables. To put it simply, how the different stock indices react to the dynamism in the different macroeconomic variables is a matter of utmost interest in this regard. In Table 3 the long-run elasticities of the different macroeconomic factors with respect to different sectors are presented. The appropriate lag lengths of the ARDL model for the different sectoral stock market indices (which were derived applying trial and error process as elaborated in the methodology section) is presented in the last row of Table 3. In Table 3, the long-run elasticities of the different macroeconomic factors with respect to the five sectoral stock indices belonging to the manufacturing sector are presented.

Table 3.

Estimation of Long-run Elasticities

Variables	lnBSE^BM	lnBSE^CDGS	lnBSE^FMCG	lnBSE^HC	lnBSE^IND
lnEPU	–0.3897*** (–2.6975)	–0.5026*** (–3.7142)	–0.1099 (–0.9720)	–0.4637* (–1.8851)	–0.42387** (–2.0190)
lnFPIR	1.5898*** (2.7016)	2.0730*** (3.4027)	1.5046** (2.2645)	2.4721 (1.4981)	2.8144*** (2.5079)
lnINT	–0.3004* (–1.6623)	0.1359 (0.7837)	0.1067 (0.7070)	–0.0233 (–0.0597)	–0.11470 (–0.44484)
lnREER	1.7655* (1.6296)	0.5585. (0.5502)	0.6762 (0.6272)	–1.8592 (–0.8017)	1.1582 (0.69859)
lnPF	0.5320** (2.4081)	1.0654*** (4.9557)	1.5150*** (7.2638)	2.1624*** (4.8830)	0.55451* (1.7740)
lnCOP	0.0405 (0.2153)	–0.4440** (–2.2323)	–0.2136 (–1.2885)	–0.4128 (–1.7774)	–0.36661 (–1.1500)
Constant	–5.8347 (–1.3720)	–5.3020 (–1.3448)	–13.8549*** (3.3838)	–6.7568 (–0.7731)	–1.3701 (–0.19665)
ARDL (p, q) model	(1,4,1,0,0,0,2)	(1,4,0,0,0,2,3)	(1,0,2,0,0,1,0)	(1,3,0,2,0,2,2)	(1,4,0,0,0,2,3)

Source: The authors.

Note: *, ** and *** indicate statistically significant at 10 per cent, 5 per cent and 1 per cent, respectively.

Any stock market index basically reflects the sentiments of the investors which in turn get affected by a large number of factors. Amongst these, the economic state of a particular country is expected to exacerbate significant effect on the stock market. Strong economic situation, political stability, rule of law, strong governance systems, etc. are of paramount importance to have a resilient stock market. In the absence of these conditions the stock market becomes fragile in nature and susceptible to the vulnerability of the external shocks. Prosperity or misery of a country’s economic fate is largely determined by the political situation prevailing in a nation (Bilson et al., 2002). Different political parties subscribe to different ideological schools. When a particular political party comes to the power with people’s mandate, the ideology of that political party gets reflected through the socio-economic policies adopted by the government. The success and failure of the firms to a large extent are dependent on such policy decisions adopted by the government (Hillman & Hitt, 1999). The stability of the government, industrial unrest, divestment policy, foreign investment policies, prevalence of corruption, high rate of taxes, import and export restrictions, red tape, etc. are the components of political risk, which is of great significance in the operation of the firms in an economy (Cherian & Perotti, 2001; Hillman et al., 1999).The presence of coalition government creates frequent impediments towards taking strong policy decision and eventually shoots up the policy uncertainty. However, measurement of such political risk and policy uncertainty is not an easy task. EPU is one such comprehensive index as propounded by Baker et al. (2015).²³ The results obtained from Table 3 reveal that the coefficients associated with the EPU with respect to the BSE-BM, BSE-CDGS, BSE-HC and BSE-IND were found to be negative but not statistically significant. The EPU elasticity of BSE-FMCG was also observed to be negative but statistically not noticeable. Thus, it can be said that the EPU had significant adverse effect on the sectoral indices belonging to the Indian manufacturing industry. This finding is in tune with the observations made by Antonakakis et al. (2013), Liu and Zhang (2015), Ko and Lee (2015) and Arouri and Roubaud (2016).

Perhaps, one of the most important factors that shapes the stock price movement is the foreign portfolio investment especially in the context of emerging economies like India. After opening up of the investment opportunity to the foreign institutional investors in September, 1992, Indian stock markets have been able to attract huge volume of foreign institutional investment on account of higher return overtaking other global stock markets (Bhargava & Malhotra, 2015). The FPI of around $15 billion during the year 2009 shows the confidence that the foreign institutional investors have on Indian equity market (Bhargava & Malhotra, 2015). Therefore, Indian stock markets heavily rely upon the foreign institutional investment, which makes the Indian stock market highly integrated to the other global stock markets (Vardhan & Sinha, 2016).²⁴ To identify the impact of foreign institutional investment on the sectoral stock indices, it becomes very important to consider the investment and withdrawal simultaneously to ascertain the real effect. Thus, following Vardhan and Sinha (2014), the ratio of purchase and sale by the foreign institutional investors as denoted by FPIR was considered. Increase in FPIR is a sign of more investment by the foreign institutional investors. The analysis made in Table 3 shows that the elasticities associated with FPIR in respect of the sectoral indices such as BSE-BM, BSE-CDGS, BSE-FMCG and BSE-IND were found to be positive and statistically significant whereas the same in respect of BSE-HC was found to be positive but not statistically noticeable. Therefore, it can be said that the firms belonging to the Indian manufacturing sector had been able to attract the foreign portfolio investment during the period of study and FPIR was found to be one of the major determinants of the stock prices for this set of sectors. The findings of the study are consistent with the outcome of the studies conducted by Suganthi and Dharshanaa (2014), Srivastava and Behl (2015), Acharya et al. (2016), Sultana and Reddy (2017) and Mishra (2018).The outcome of the study carried out by Mukherjee and Roy (2016) also showed that the foreign investment had a significant positive effect on the Indian stock market and such effect became stronger in the post 2008 regime.

Table 3 shows that the elasticity associated with rate of interest (INT) with respect to BSE-BM was found to be negative and statistically significant, which is in line with the accepted theoretical proposition that the rate of interest and the stock price movements are negatively associated. A hike in the rate of interest, on one hand, enhances the cost of borrowings for the firms (Ibrahim & Musah, 2014), while on the other hand, it induces more savings leading to lesser consumption and siphoning the funds from the stock to fixed deposits (Mukherjee & Roy, 2016). This finding is very much consistent with the outcome of the studies conducted by Mukherjee and Naka (1995), Liu and Shrestha (2008), Khan et al. (2017) and Shabbir (2018). In the context of India, Mukherjee and Roy (2016) suggested that stock market return in India gets negatively affected by the increase in the rate of interest internationally and since the interest rate in India is higher than the global rates such impact gets more intensified in the event of change in the domestic rate of interest. However, the effect of the rate of interest was found to be statistically insignificant in the cases of BSE-FMCG, BSE-CDGS, BSE-HC and BSE-IND (except BSE-BM) during the period of study. It indicates that the rate of interest was observed to be an insignificant determinant of the stock price for the rest of the sector. This outcome is not unusual because many other researchers such as Talla (2013), Ibrahim and Musah (2014) and Msindo (2016) also found that interest rate does not affect the stock price. Similar observations were also found in the studies carried out by Naik and Padhi (2012) and Tripathi and Seth (2014) in the context of India in respect of Sensex and Nifty.

Effect of rate of exchange on the stock prices is always considered as a subject for the purpose of making empirical investigation (Cenedese et al., 2015). The analysis of the results as shown in Table 3 indicates that the exchange rate was also not found to be a significant determinant of the stock price except for BSE-BM. The long-run elasticities of REER were not found to be statistically significant in case of BSE-CDGS, BSE-FMCG, BSE-HC and BSE-IND. Thus, the share prices of companies pertaining to these sectors remained insulated from the volatility of the rate of exchange during the period of study. Only in case of BSE-BM the effect of the rate of exchange was found to be positive and statistically significant. The possible explanation may be that the effect of rate of exchange can only be positive if the firms belonging to this sector are export oriented. The positive effect of the exchange rate was observed by Mukherjee and Naka (1995), Naik and Padhi (2012) and Khan et al. (2017), which supports the sign of REER in respect of BSE-BM.

From the analysis made in Table 3 it is evident that PF was found to be the most important factor having bearing on all the sectoral indices within the broad manufacturing industry. It must be kept in mind that PF was considered in the model as a principal factor and was constituted of four important macroeconomic factors such as IIP, WPI, M2 and GP. None of the studies carried out so far considered these four variables together as one factor. Thus, the outcome derived from it cannot directly be compared with the findings of the existing studies. However, there are plenty of evidences of their individual effect on the stock prices, which can be used partially to compare this outcome. IIP is an indicator of the real economic activity in the economy. Increase in IIP is expected to affect the cash flows and earnings positively, which in turn will impact favourably the stock price (Mukherjee & Naka, 1995). Moreover, the dividend growth is also largely conditioned upon the economic growth as indicated by the increasing IIP that affects the stock prices positively (Liu & Shrestha, 2008). The positive effect of IIP on the stock price is observed in the studies carried out by Mukhopadhyay and Sarkar (2003), Maysami et al. (2004), Mohammad et al. (2009), Naik and Padhi (2012) and Shahbaz et al. (2016). The effect of MS on the stock prices is subject to empirical investigation. Naik and Padhi (2012) suggested that increase in the MS will lead to a shift in the portfolio, i.e., non-interest bearing money asset to financial assets such as stocks and thereby stock price will increase. In addition to that, economic stimulus brought in by the increased MS will also affect the stock prices positively (Mukherjee & Naka, 1995). On the other hand, enhanced MS may have adverse effect on the rate of discount by creating an inflationary pressure, which will negatively affect the stock prices (Nishat et al., 2004). However, it is not right to claim that inflation always affects the stock prices negatively. There are empirical evidences where it was found that the inflation exacerbated favourable effect on the stock prices too. If the general level of price steps up, the rate of discount in the dividend discounting equity valuation model will also augment as the nominal risk-free rate enhances in account of inflation (Liu & Shrestha, 2008). Moreover, Naik and Padhi (2012) justified the negative effect of inflation on the stock price using Fama’s proxy effect argument which states that the stock return and inflation are negatively associated.²⁵ On the contrary, in the event of inflation, the demand for money reduces along with the expected return to money and therefore the stock will be positively affected as the demand for equity increases (Marshall, 1992). In addition to that, mild inflation will create opportunity to enhance the corporate profitability, which ultimately affects the stock prices favourably and thereby helps in neutralizing the negative effect on the discounting rate (Khan et al., 2017; Mukherjee and Naka, 1995; Shahbaz et al., 2016). It suggests that investment in stocks can be considered as a hedging mechanism to beat inflation and concluded that inflation and stock prices are therefore positively associated. In the context of India, it can be said that gold is considered to be the best hedge against the stock market volatility (Hasanzadeh & Kianvand, 2012). Therefore, the GP is negatively associated with the stock prices. Thus, amongst the constituting factors of PF, the effect of IIP on stock price is positive, the effects of MS and inflation on stock price can either be positive or negative and the effect of GP on stock price is negative. The long-run elasticity of PF for BSE-BM, BSE-CDGS, BSE-FMCG, BSE-HC and BSE-IND were found to be positive and statistically notable during the study period. It shows that the effect of IIP was found to be positive as expected, the stimulus brought in by the increased MS and such a stimulus negated any negative effect of inflation on the rate of discount and finally on stock prices. Since three out of the four variables (accounting for about 75% of the principal factor ‘PF’) had highly positive effect on the sectoral stock prices, the overall effect of PF was also observed to be favourable in all the selected sectoral stock indices.

It is also evident from the analysis of Table 3 that COP was also found to be an insignificant determinant of the stock prices as the long-run elasticity of COP was found to be negative and statistically significant only in case of BSE-CDGS. The same for all other sectors were not found to be statistically noticeable. The reason behind the negative effect of COP on BSE-CDGS may be due to the fact that Automobiles and Auto Parts and Equipment Companies and Consumer Durable companies are the major constituents of BSE-CDGS index. Any increase in the oil price has a negative effect on the auto sector and consumer durable sector and thereby on the BSE-CDGS index as a whole. Such a negative effect is consistent with the findings of the studies carried out by Chancharat et al. (2008) and Basher et al. (2012).

Short-run Elasticity

Table 4 exhibits the short-run elasticities and the equilibrium relationship between the sector-specific BSE manufacturing indices and the select macroeconomic factors. In long run there may be a stable relationship while in short run there may be disequilibrium. ECT_t–1 is the error correction term in the model. To establish long-run equilibrium, it is imperative that the value of the error correction term is negative and found to be statistically significant. Moreover, negative and statistically significant ECT_t–1 also indicates long-run causal relationship running from the macroeconomic variables to the respective sectoral stock indices and the speed of adjustment taking place in short run to restore long-run equilibrium. From the analysis, it is observed that the values of ECT_t–1 corresponding to all the sectoral indices were negative and found to be statistically significant. It reinforced the proposition made earlier that there is a long-run equilibrium relationship between the manufacturing sector-specific BSE-indices and the selected macroeconomic variables, which is consistent with the findings of the studies conducted by Liu and Shrestha (2008), Shahbaz et al. (2016) and Khan et al. (2017). It can also be concluded that the speed of adjustment to correct the previous month’s disequilibrium took place at a magnitude of 16.84 per cent, 14.56 per cent, 10.88 per cent, 6.37 per cent and 11.81 per cent for BSE-BM, BSE-CDGS, BSE-FMCG, BSE-HC and BSE-IND, respectively, to restore long-run equilibrium.

Table 4.

Short-run Elasticities

Variables	ΔlnBSE^BM	ΔlnBSE^CDGS	ΔlnBSE^FMCG	ΔlnBSE^HC	ΔlnBSE^IND
ΔlnEPU_t	–0.030987 (–1.3293)	–0.031951 (–1.5649)	–0.011963 (–0.82938)	–0.028085 (–1.5694)	–0.034539 (–1.4005)
ΔlnEPU_t–1	0.069749*** (2.5088)	0.097509*** (3.9450)	–	0.057346*** (3.0128)	0.084713*** (2.9421)
ΔlnEPU_t–2	0.058661** (2.2660)	0.062071*** (2.8499)	–	0.030115* (1.7884)	0.051818** (1.9952)
ΔlnEPU_t–3	0.032781 (1.4136)	0.048747*** (2.4458)	–	–	0.039350* (1.6504)
ΔlnFPIR_t	0.34159*** (6.9764)	0.30200*** (7.7013)	0.16072*** (5.1356)	0.15747*** (4.6483)	0.33258*** (6.7635)
ΔlnFPIR_t–1	–	–	–0.063076** (–1.9795)	–	–
ΔlnINTR_t	–0.050591* (–1.6771)	0.019809 (0.74735)	0.032293 (1.4158)	0.069017* (1.6126)	–0.013554 (–0.44114)
ΔlnINT_t–1	–	–	–	0.086118** (2.0359)	–
ΔlnREER_t	0.29735 (1.3141)	0.081377 (0.52151)	0.073592 (0.67777)	–0.11843 (–0.98507)	0.13687 (0.61534)
ΔlnPF_t	0.089593*** (2.5014)	–0.42711 (–1.1580)	–0.36256 (–1.3808)	–0.52294* (–1.6422)	–0.71442 (–1.5837)
ΔlnPF_t–1	–	–0.72676** (–2.0225)	–	–0.57819* (–1.8658)	–0.71257* (–1.6260)
ΔlnCOP_t	0.17199** (2.0997)	–0.077468 (–1.1167)	–0.023246 (–1.2621)	–0.034962 (–0.59156)	0.057440 (0.67926)
ΔlnCOP_t–1	0.19245** (2.3137)	0.21683*** (2.9986)	–	0.13649** (2.2894)	0.16058* (1.8213)
ΔlnCOP_t–2	–	0.11502* (1.6414)	–	–	0.11900 (1.3999)
ECT_t–1	–0.16842*** (–3.7700)	–0.14569*** (–3.9750)	–0.10883*** (–2.6264)	–0.063698* (–1.6941)	–0.11817*** (–2.9823)
R ²	0.54485	0.54584	0.27867	0.38406	0.52137
${\bar{R}}^{2}$	0.48944	0.48153	0.21905	0.29761	0.45360
F-statistics	12.5149***	10.4469***	5.8431***	5.9236***	9.4686***
DW statistics	1.9614	1.7830	1.9261	2.0161	1.7522

Source: The authors.

Note: *, ** and *** indicate statistically significant at 10 per cent, 5 per cent and 1 per cent, respectively.

The results of the different diagnostic test²⁶ indicated the robustness of the models for different sectors. The results of ARCH test revealed that there was no heteroscedasticity problem in the error terms and the error terms were independent of the predictors in case of all the models. The null hypothesis under J–B test could not be rejected in case of any of the models, which indicates that the residuals in all the models were found to be normally distributed. Statistically insignificant values of χ² and F in respect of Ramsey’s RESET test showed that the all the models were correctly specified.

Finally, the analysis of the CUSUM and CUSUM of squares plots²⁷ revealed that CUSUM and CUSUM of squares lines were lying well within the upper and lower bounds of 5 per cent level. It signified the stability of the ARDL-UEC models. Moreover, it also confirms that the long-run coefficient and short-run dynamics were correctly derived.

Structural Break

In Table 5, the outcomes of the Bai–Perron test of multiple break points and Chow test are presented. From the results of Bai–Perron test, it is observed that one break point in case of BSE-BM (December, 2008), three break points in case of BSE-CDGS (December, 2007; July, 2009; May, 2014), one break point in case of BSE-FMCG (November, 2008), three break points in case of BSE-HC (December, 2007; February, 2010; June, 2014) and also three break points in case of BSE-IND ( December, 2009; February, 2014;October, 2007) were present. However, the result of the Chow test clearly showed that all the structural breaks were not statistically significant. During 2008–09, many important economic (contagious effect of US subprime crisis), political (re-election of UPA govt.) and social events (Mumbai terror attack²⁸) took place, which had affected the stock indices across all sectors. A closer look at Figure 1 also signifies that either the indices fell heavily or recovered fast earmarking change in the structure during the break points. It must be kept in mind that in respect of BSE-BM and BSE-FMCG, the effect of 2008–09 financial crisis was very prominent, which was further intensified by the Mumbai terror attacks. On the other hand, BSE-CDGS and BSE-IND increased significantly since late 2009, which may be because of the re-election of the UPA Government. Out of all these events, perhaps the US sub-prime crisis had most deepening and contagious effects across stock indices of the countries of the world. The outcome of the Chow test clearly showed that all the selected stock indices exhibited structural break in their relationship with the macroeconomic variables during 2008–09 except health care indices (BSE-HC). As far as the BSE-HC is concerned, it can be seen that structural break was found in February 2010. This can be attributed to the government decision of huge investment of Rs. 3000 crore (venture capital fund) in the pharmaceutical sector for research and development for new drug discovery.²⁹ Structural breaks (in upward direction) of BSE-HC in 2014 can be explained by the formation of NDA government with full majority, which signalled the political stability and upcoming economic reforms.

Table 5.

Determination of Structural Break Points

ARDL Models	Bai–Perron Multiple Break Points Test	Chow Test (F-statistics)
BSE-BM	December, 2008	3.1265**
BSE-CDGS	December, 2007	0.6980
	July, 2009	2.2330**
	May, 2014	1.403930
BSE-FMCG	November, 2008	3.112913**
BSE-HC	December, 2007	0.884471
	February, 2010	3.1478**
	June, 2014	1.872797*
BSE-IND	October, 2007	1.176426
	December, 2009	2.73251**
	February, 2014	1.423905

Source: The authors.

Note: ** indicates significant at 1 per cent level and * shows significant at 5 per cent level.

Conclusion

Vibrant stock market has always been considered important for industrial growth as well as economic growth. Effective and efficient stock market helps in channelizing funds from the surplus sector to the deficit sector in the financial system. Exploring the determinants of the stock prices has been a matter of great interest amongst the finance scholars across the globe for a long period of time. Obviously the share prices of the corporate houses get influenced by the demand and supply situation in the stock market that in turn gets driven by the company specific factors, industry specific factors and the macroeconomic factors. Since late 1980s researcher community was primarily interested in understanding the role of the fundamentals of the companies so to say the company specific factors in an isolated fashion in modelling the stock price predictors. Limited number of empirical studies has been conducted then in order to identify the significant macroeconomic factors influencing the stock market development. However, with the emergence of the free market economy in early 1990s the macroeconomic factors begun to play pivotal role in shaping the stock market movement across countries. The countries that adopted the new form of market economy became more vulnerable to the externalities. The higher the extent of liberalization of an economy, the greater is the chance of getting affected by the macroeconomic factors that are beyond the control of the corporate sector. Naturally, the direction of the research studies has also changed according to the changed factor dynamics. In tune with this development in the stock market of different countries the focus of the researchers shifted towards incorporating the various macroeconomic factors in the existing models of stock price determination.

In this context, the present study evaluated the short-run and long-run relationships between the different sectoral stock market indices (BSE-BM, BSE-CDGS, BSE-FMCG, BSE-HC and BSE-IND) belonging to the Indian manufacturing sectors and the different macroeconomic variables (EPU, FPIR, INT, REER, PF and COP). While testing the stationarity of all the variables involved it is found that some of the variables were integrated of different orders, which necessitated the application of ARDL Bound test approach to explore the prevalence of long-run equilibrium relationship between the different sectoral indices and the select macroeconomic variables. The outcome of the ARDL bound test revealed the existence of the stable long-run interrelation between all the stock indices and the selected macroeconomic variables. The long-run causality was established via UECM. The error correction term associated with each of the models corresponding to different indices were observed to be negative and statistically significant. It showed that the errors existent in short run getting corrected at the speed ranging from approximately 6 per cent to 16 per cent for different stock indices to establish long-run equilibrium. The analysis of the long-run elasticities indicated that EPU, FPIR, INT, REER and PF were the important macroeconomic determinants in case of BSE-BM whereas EPU, FPIR, PF and COP were recognized as the factors affecting the BSE-CDGS index significantly during the period under study. Moreover, the results of the study showed that BSE-FMCG sector got influenced by only two macroeconomic factors such as FPIR and PF. In case of BSE-HC, EPU, PF were observed to be the major determining factors during the study period, whereas EPU, FPIR and PF were established themselves as the significant macroeconomic factors affecting the movement of the BSE-IND during the same period. Overall, it can be said that EPU, FPIR and PF were the most important determinants of all the five sectoral stock indices during the study period. COP was considered as a notable factor only in case of BSE-CDGS, whereas REER and COP were also found to be significant only in case of BSE-BM.

Moreover, the outcomes derived from the different diagnostic and stability tests indicated that the results obtained were robust and the specified models were stable. Thus, the result of the study established that all the macroeconomic variables were not equally significant for all the industries although each and every sector is operating in the context of identical business environment either in respect of any time horizon. The structural break analysis also revealed that the effect of crisis of 2008–09 was found to be prominent in case of BSE-BM, BSE-CDGS, BSE-IND and BSE-FMCG while it was not at all noticeable in case of BSE-HC. Thus, BSE-HC remained more or less insulated from the effect of financial crisis of 2008–09. However, some other forms of structural changes on account of important domestic events were also observed in case of BSE-HC.

Policy Implications of the Study

The outcome of the study has several policy implications which are as follows:

Firstly, the outcome of the study is expected to be useful to the policy makers and researchers to get an insight into the nature of interrelationn between the select macroeconomic indicators and stock price behaviour belonging to the diverse sectors within the Indian manufacturing industry and also in understanding the behaviour of the stock market investors in the light of changing macroeconomic fundamentals.

Secondly, the findings of the study will enable the prospective investors to make their investment decisions prudently in the different sectors belonging to the Indian manufacturing industry. Moreover, the outcome of the study will also enable the existing investors to take appropriate investment decisions with respect to different sectors based on the changes in the macroeconomic environment as the sector specific responses of the stock price to the dynamism in macroeconomic variables are precisely ascertained from such outcome in both long run and short run. For example, with an increase in EPU, the investors have to take a call on the firms pertaining to BSE-BM, BSE-CDGS, BSE-IND, BSE-HC, etc. but they can maintain status quo in respect of the shares belonging to BSE-FMCG. Similarly, changes in the rate of interest and exchange rate are only expected to exert influence on the BSE-BM.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Appendix

Table A4.

Diagnostic Test Result

Diagnostic Test	Test Statistics	lnBSE^BM	lnBSE^CDGS	lnBSE^FMCG	lnBSE^HC	lnBSE^IND
Breusch–Godfrey LM test	x ²	0.65551 [0.985]	6.2276 [0.285]	1.9002 [0.863]	2.3407 [0.800]	7.0792 [0.215]
Breusch–Godfrey LM test	F	0.11149 [0.990]	1.0868 [0.372]	0.33886 [0.888]	0.39661 [0.850]	1.2440 [0.294]
Ramsey’s RESET test	x ²	0.01133 [0.915]	0.00694 [.934]	0.0182 [0.893]	1.1528 [0.283]	0.97691 [0.323]
Ramsey’s RESET test	F	0.009941 [0.921]	0.00598 [.938]	0.0166 [0.898]	0.8298 [0.365]	0.84802 [0.359]
Jarque–Bera normality test	J–B	1.0156 [0.602]	1.2757 [0.528]	1.9666 [0.374]	1.9218 [0.382]	2.7959 [0.247]
ARCH test of heteroscedasticity	x ²	7.6523 [0.176]	2.4466 [.785]	3.0444 [0.693]	3.3835 [0.641]	1.3909 [0.925]
ARCH test of heteroscedasticity	F	1.3760 [0.239]	0.41431 [0.838]	0.54771 [0.740]	0.57799 [0.717]	0.23361 [0.947]

Source: The authors.

Note: Values in bracket represent P values.

Notes

References

Acharya

V. V.

, Anshuman

V. R.

, & Kumar

K. K.

(2016). Foreign fund flows and asset prices: Evidence from the Indian Stock Market. http://pages.stern.nyu.edu/~sternfin/vacharya/public_html/pdfs/Foreign%20Fund%20Flows%20May_2016.pdf

Agrawalla

R. K.

, & Tuteja

S. K.

(2008). Share prices and macroeconomic variables in India. Journal of Management Research, 8(3), 1–12.

Al-Sharkas

(2004). The dynamic relationship between macroeconomic factors and the Jordanian stock market. International Journal of Applied Econometrics and Quantitative Studies, 1, 97–114.

Ammer

(1994). Inflation, inflation risk, and stock returns (International Finance Discussion Paper No. 464). Board of Governors of the Federal Reserve System. https://www.federalreserve.gov/pubs/ifdp/1994/464/ifdp464.pdf

Antonakakis

, Chatziantoniou

, & Filis

(2013). Dynamic co-movements of stock market returns, implied volatility and policy uncertainty. Economics Letters, 120(1), 87–92.

Arouri

, & Roubaud

(2016). On the determinants of stock market dynamics in emerging countries: The role of economic policy uncertainty in China and India. Economics Bulletin, 36(2), 760–770.

Bai

, & Perron

(2003). Computation and analysis of multiple structural change models. Journal of Applied Econometrics, 18(1), 1–22.

Baker

S. R.

, Bloom

, & Davis

S. J.

(2016). Measuring economic policy uncertainty. The Quarterly Journal of Economics, 131(4), 1593–1636.

Barro

R. J.

(1996). Determinants of economic growth: A cross-country empirical study (National Bureau of Economic Research Working Paper No. w5698). https://www.nber.org/papers/w5698.pdf.

10.

Basher

S. A.

, Haug

A. A.

, & Sadorsky

(2012). Oil prices, exchange rates and emerging stock markets. Energy Economics, 34(1), 227–240.

11.

Bhargava

, & Malhotra

D. K.

(2015). Foreign institutional investment and the Indian stock market. The Journal of Wealth Management, 17(4), 101–116.

12.

Bhattacharya

S. N.

, Bhattacharya

, & Basu

(2019). Stock market and its liquidity: Evidence from ARDL bound testing approach in the Indian context. Cogent Economics & Finance, 7(1), 1–12.

13.

Bhaumik

(2015). Principles of econometrics: A modern approach using eviews. OUP India.

14.

Bilson

C. M.

, Brailsford

T. J.

, & Hooper

V. C.

(2002). The explanatory power of political risk in emerging markets. International Review of Financial Analysis, 11(1), 1–27.

15.

Brooks

2014, Introductory econometrics for finance (2nd ed.). Cambridge University Press.

16.

Carlson

J. B.

, Craig

, & Schwarz

J. C.

(2000). Structural uncertainty and breakpoint tests: An application to equilibrium velocity. Journal of Economics and Business, 52(1–2), 101–115.

17.

Cenedese

, Payne

, Sarno

, & Valente

(2015). What do stock markets tell us about exchange rates? Review of Finance, 20(3), 1045–1080.

18.

Chancharat

, Valadkhani

, & Havie

(2008). The influence of international stock markets and macroeconomic variables on the Thai stock market. https://ro.uow.edu.au/cgi/viewcontent.cgi?referer=https://scholar.google.com/&httpsredir=1&article=1397&context=commpapers.

19.

Chen

N. F.

, Roll

, & Ross

S. A.

(1986). Economic forces and the stock market. Journal of Business, 59(3), 383–403.

20.

Cherian

J. A.

, & Perotti

(2001). Option pricing and foreign investment under political risk. Journal of International Economics, 55(2), 359–377.

21.

Cheung

Y.W.

, and Ng

L.K.

(1998). International evidence on the stock exchange and aggregate economic activity. Journal of Empirical Finance, 5(3), 281–296.

22.

Chow

G. C.

(1960). Tests of equality between sets of coefficients in two linear regressions. Econometrica: Journal of the Econometric Society, 28(3), 591–605.

23.

Clark

, & Berko

(1997). Foreign investment fluctuations and emerging market stock returns: The case of Mexico (Staff Report No. 24). Federal Reserve Bank of New York. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=993813.

24.

Croce

M. M.

, Kung

, Nguyen

T. T.

, & Schmid

(2012). Fiscal policies and asset prices. Review of Financial Studies, 25(9), 2635–2672.

25.

Darrat

A. F.

, & Dichens

R. N.

(1999). On the inter-relationship among real monetary and financial indicators. Applied Financial Economics, 9(3), 289–293.

26.

Durham

J. B.

(2002). The effects of stock market development on growth and private investment in lower-income countries. Emerging Markets Review, 3(3), 211–232.

27.

Easterly

, & Rebelo

(1993). Fiscal policy and economic growth. Journal of Monetary Economics, 32(3), 417–458.

28.

Fama

E. F.

, & Schwert

W. G.

(1977). Asset returns and inflation. Journal of Financial Economics, 5(2), 115–146.

29.

Fama

E. F.

(1981). Stock returns, real activity, inflation, and money. The American Economic Review, 71(4), 545–565.

30.

Fama

E. F.

, & French

K. R.

(1989). Business conditions and expected returns on stocks and bonds. Journal of Financial Economics, 25(1), 23–49.

31.

Ferdows

S. S.

, & Roy

(2012). A study on the international diversification in the emerging equity market and its effect on the Indian capital market. International Journal of Contemporary Business Studies, 3(4), 79–96.

32.

Froot

K. A.

, O’Connel

P. G.

, Seasholes

M. S.

(2001). The portfolio flows of international investors. Journal of Financial Economics, 59(2), 151–194.

33.

Gay Jr

R. D.

(2011). Effect of macroeconomic variables on stock market returns for four emerging economies: Brazil, Russia, India, and China. International Business & Economics Research Journal, 7(3), 1–8.

34.

Gokmenoglu

K. K.

, & Fazlollahi

(2015). The interactions among gold, oil, and stock market: Evidence from S&P500. Procedia Economics and Finance, 25, 478–488.

35.

Granger

C. W. J.

, Huang

B. N.

, Yang

C. W.

(2000). A bivariate causality between stock prices and exchange rate: Evidence from recent Asian Flu. Quarterly Review of Economics and Finance, 40(3), 337–354.

36.

Grier

K.B.

& Tullock

(1989). An empirical analysis of cross-national economies 1951–1980. Journal of Monetary economics, 24(2), 259–276.

37.

Gujarati

D. N.

(2009). Basic econometrics. Tata McGraw-Hill Education.

38.

Hasanzadeh

, & Kianvand

(2012) The impact of macroeconomic variables on stock prices: The case of Tehran stock exchange. Money and Economy, 6(2), 171–190.

39.

Hassan

M. M. S.

, & Sangmi

M. U. D. G.

(2013). Macro-economic variables and stock prices in India [Doctoral dissertation]. http://dspaces.uok.edu.in:8080/jspui/bitstream/1/1375/1/Mohd.%20Mubashar%20Hassan.pdf

40.

Herve

D. B. G.

, Chanmalai

, & Shen

(2011). The study of causal relationship between stock market indices and macroeconomic variables in Cote d’Ivoire: Evidence from error-correction models and granger causality test. International Journal of Business & Management, 6(12), 146–167.

41.

Hillman

A. J.

, & Hitt

M. A.

(1999). Corporate political strategy formulation: A model of approach, participation, and strategy decisions. Academy of Management Review, 24(4), 825–842.

42.

Hillman

A. J.

, Zardkoohi

, & Bierman

(1999). Corporate political strategies and firm performance: Indications of firm-specific benefits from personal service in the US government. Strategic Management Journal, 20(1), 67–81.

43.

Hondroyiannis

, & Papapetrou

(2001). Macroeconomic influences on the stock market. Journal of Economics and Finance, 25(1), 33–49.

44.

Hosseini

S. M.

, Ahmad

, & Lai

Y. W.

(2011). The role of macroeconomic variables on stock market index in China and India. International Journal of Economics and Finance, 3(6), 233–243.

45.

Ibrahim

, & Musah

(2014). An econometric analysis of the impact of macroeconomic fundamentals on stock market returns in Ghana. Research in Applied Economics, 6(2), 47–72.

46.

Johansen

(1988). Statistical analysis of cointegration vectors. Journal of Economic Dynamics and Control, 12(2–3), 231–254.

47.

Joint Research Centre-European Commission (2008). Handbook on constructing composite indicators: Methodology and user, guide (OECD). https://www.oecd.org/sdd/42495745.pdf.

48.

Kaneko

, & Lee

B. S.

(1995). Relative importance of economic factors in the US and Japanese stock markets. Journal of the Japanese and International Economies, 9(3), 290–307.

49.

Khan

M. K.

, Teng

J. Z.

, Parvaiz

, & Chaudhary

S. K.

(2017). Nexuses between economic factors and stock returns in China. International Journal of Economics and Finance, 9(9), 182–191.

50.

J. H.

, & Lee

C. M.

(2015). International economic policy uncertainty and stock prices: Wavelet approach. Economics Letters, 134, 118–122.

51.

Kraft

, & Kraft

(1977). Determinants of common stock prices: A time series analysis. Journal of Finance, 32(2), 417–425.

52.

Lima

, Vasconcelos

C. F.

, Simão

, & de Mendonça

H. F.

(2016). The quantitative easing effect on the stock market of the USA, the UK and Japan: An ARDL approach for the crisis period. Journal of Economic Studies, 43(6), 1006–1021.

53.

Liu

M. H.

, & Shrestha

K. M.

(2008). Analysis of the long-term relationship between macro-economic variables and the Chinese stock market using heteroscedastic cointegration. Managerial Finance, 34(11), 744–755.

54.

Liu

, & Zhang

(2015). Economic policy uncertainty and stock market volatility. Finance Research Letters, 15, 99–105.

55.

Lucas

R. E.

(1980). Two illustrations of the quantity theory of money. The American Economic Review, 70(5), 1005–1014.

56.

Mahdavi

, & Zhou

(1997). Gold and commodity prices as leading indicators of inflation: Tests of long-run relationship and predictive performance. Journal of Economics and Business, 49(5), 475–489.

57.

Malkiel

B. G.

(1982). Risk and return: A new look (National Bureau of Economic Research Working Paper No. 700). https://www.nber.org/papers/w0700.pdf.

58.

Malkiel

B. G.

, & Fama

E. F.

(1970). Efficient capital markets: A review of theory and empirical work. The Journal of Finance, 25(2), 383–417.

59.

Marshall

D. A.

(1992). Inflation and asset returns in a monetary economy. The Journal of Finance, 47(4), 1315–1342.

60.

Maysami

R. C.

, Howe

L. C.

, & Hamzah

M. A.

(2004). Relationship between macroeconomic variables and stock market indices: Cointegration evidence from stock exchange of Singapore’s All-S sector indices. Jurnal Pengurusan, 24, 47–77.

61.

Milton

, & Schwartz

A. J.

(1975). Money and business cycles. In The state of monetary economics (pp. 32–78). National Bureau for Economic Research.

62.

Misra

(2018). An Investigation of the macroeconomic factors affecting the Indian stock market. Australasian Accounting, Business and Finance Journal, 12(2), 71–86.

63.

Mohammad

S. D.

, Hussain

, & Ali

(2009). Impact of macroeconomics variables on stock prices: Empirical evidence in case of KSE (Karachi Stock Exchange). European Journal of Scientific Research, 38(1), 96–103.

64.

Msindo

Z. H.

(2016). The impact of interest rates on stock returns: Empirical evidence from the JSE securities exchange [Doctoral dissertation]. http://wiredspace.wits.ac.za/bitstream/handle/10539/22148/Zethu’sFinalpercent20Researchpaper14Sept.pdf?sequence=1&isAllowed=y.

65.

Mukherjee

, & Roy

(2016). What drives the stock market return in India? An exploration with dynamic factor model. Journal of Emerging Market Finance, 15(1), 119–145.

66.

Mukherjee

T. K.

, & Naka

(1995). Dynamic relations between macroeconomic variables and the Japanese stock market: An application of a vector error correction model. Journal of Financial Research, 18(2), 223–237.

67.

Mukhopadhyay

, & Sarkar

(2003). Stock return and macroeconomic fundamentals in model specification framework: Evidence from Indian stock market. https://www.isical.ac.in/~eru/erudp/2003–05.pdf

68.

Naik

P. K.

, & Padhi

(2012). The impact of macroeconomic fundamentals on stock prices revisited: Evidence from Indian data. Eurasian Journal of Business and Economics, 5(10), 25–44.

69.

Nishat

, Shaheen

, & Hijazi

S. T.

(2004). Macroeconomic factors and the Pakistani equity market. The Pakistan Development Review, 43(4), 619–637.

70.

Nkoro

, & Uko

A. K.

(2016). Autoregressive Distributed Lag (ARDL) cointegration technique: Application and interpretation. Journal of Statistical and Econometric Methods, 5(4), 63–91.

71.

Ogunmuyiwa

M. S.

, & Ekone

A. F.

(2010). Money supply-economic growth nexus in Nigeria. Journal of Social Sciences, 22(3), 199–204.

72.

Omoniyi

L. G.

, & Olawale

A. N.

(2015). An application of ARDL bounds testing procedure to the estimation of level relationship between exchange rate, crude oil price and inflation rate in Nigeria. International Journal of Statistics and Applications, 5(2), 81–90.

73.

Patel

(2012). The effect of macroeconomic determinants on the performance of the Indian stock market. NMIMS Management Review, 22(1), 117–127.

74.

Pesaran

M. H.

, & Shin

(1998). An autoregressive distributed-lag modelling approach to cointegration analysis. Econometric Society Monographs, 31, 371–413.

75.

Pesaran

M. H.

, Shin

, & Smith

R. J.

(2001). Bounds testing approaches to the analysis of level relationships. Journal of applied econometrics, 16(3), 289–326.

76.

Pethe

, & Karnik

(2000). Do Indian stock markets matter? Stock market indices and macro-economic variables. Economic and Political Weekly, 35(5), 349–356.

77.

Philips

, & Perron

(1988). Mean Reversion in Stock Prices. Journal of Financial Economics, 22, 27–59.

78.

Rad

A. A.

, & Eslami

M. R.

(2012). Long run and short run effects of monetary and exchange variables on stock prices in Iran. International Journal of Economics and Finance Studies, 4(1), 1–10.

79.

Ram

, & Spencer

D. E.

(1983). Stock returns, real activity, inflation, and money: Comment. The American Economic Review, 73(3), 463–470.

80.

Rao

K. N.

, & Bhole

L. M.

(1990). Inflation and equity returns. Economic and Political Weekly, 25(21),91–96.

81.

Ross

S. A.

(1976). The arbitrage theory of capital asset pricing. Journal of Economic Theory, 13(3), 341–360.

82.

Shabbir

M. S.

(2018). The impact of foreign portfolio investment on stock prices of Pakistan. https://www.researchgate.net/publication/325257541_M_P_RA_The_Impact_of_Foreign_Portfolio_Investment_on_Stock_Prices_of_Pakistan_The_Impact_of_Foreign_Portfolio_Investment_on_Stock_Prices_of_Pakistan

83.

Shahbaz

, Rehman

I. U.

, & Afza

(2016). Macroeconomic determinants of stock market capitalization in an emerging market: Fresh evidence from cointegration with unknown structural breaks. Macroeconomics and Finance in Emerging Market Economies, 9(1), 75–99.

84.

Sheng

H. C.

, & Tu

A. H.

(2000). A study of cointegration and variance decomposition among national equity indices before and during the period of the Asian financial crisis. Journal of Multinational Financial Management, 10(3), 345–365.

85.

Siddique

A. K. M.

(1975). Money and inflation in Bangladesh. The Bangladesh Development Studies, 3(1), 27–42.

86.

Sims

C. A.

(1972). Money, income, and causality. The American Economic Review, 62(4), 540–552.

87.

Srinivasan

(2011). Causal nexus between stock market return and selected macroeconomic variables in India: Evidence from the National Stock Exchange (NSE). IUP Journal of Financial Risk Management, 8(4), 7.

88.

Srivastava

, & Behl

(2015). Macroeconomic variables on Indian stock market interactions: An analytical approach. JIM Quest: Journal of Management and Technology, 11(2), 1–8. http://jaipuria.edu.in/jim/wp-content/uploads/2012/06/jim_quest_july_dec_2015a.pdf#page=5

89.

Suganthi

, & Dharshanaa

(2014). Interrelationship between FII and stock market and their causal relationship with selected macroeconomic variables in India. TSM Business Review, 2(1), 29–46.

90.

Sultana

S. T.

, & Pardhasaradhi

(2012). Impact of flow of FDI & FII on Indian stock market. Finance Research, 1(3), 4–10.

91.

Sultana

S. T.

, & Reddy

K. S.

(2017). The Effect of macroeconomic factors on Indian stock market: An empirical evidence. FIIB Business Review, 6(1), 68–76.

92.

Talla

J. T.

(2013). Impact of macroeconomic variables on the stock market prices of the Stockholm stock exchange (Master’s thesis). Jonkoping International Business School. https://pdfs.semanticscholar.org/531c/af1a5eded883d4a663b9f200bb9b666abcb5.pdf

93.

Thornton

(1993). Money, output and stock prices in the UK: Evidence on some (non) relationships. Applied Financial Economics, 3(4), 335–338.

94.

Tkacz

(2007). Gold prices and inflation (Bank of Canada Working Paper No. 2007, 35). https://www.econstor.eu/bitstream/10419/53831/1/534786065.pdf

95.

Tripathi

, & Seth

(2014). Stock market performance and macroeconomic factors: The study of Indian equity market. Global Business Review, 15(2), 291–316.

96.

Tursoy

(2019). The interaction between stock prices and interest rates in Turkey: Empirical evidence from ARDL bounds test cointegration. Financial Innovation, 5(1), 1–12.

97.

Vardhan

, & Sinha

(2014). Influence of foreign institutional investments (FIIs) on the Indian stock market. https://mpra.ub.uni-muenchen.de/53611/1/MPRA_paper_53611.pdf

98.

Vardhan

, & Sinha

(2016). Influence of foreign institutional investments (FIIs) on the Indian stock market: An insight by VAR models. Journal of Emerging Market Finance, 15(1), 49–83.

99.

Wada

(2017). Dynamic causality in energy production and output growth in Nigeria revisited: ARDL bounds test approach. Energy Sources, Part B: Economics, Planning, and Policy, 12(11), 945–951.

100.

Worthington

A. C.

, & Pahlavani

(2007). Gold investment as an inflationary hedge: Cointegration evidence with allowance for endogenous structural breaks. Applied Financial Economics Letters, 3(4), 259–262.