1 Introduction

Stock indices are considered as significant indicators of the well being of an economy. An Industry specific stock index is nowadays assessed to track the performance and growth of a sector, since it is considered to be a true reflector of the behaviour of a particular industry . Economists and financial researchers have always been very keen on the interaction of stock prices with the macroeconomic variables . Among various macroeconomic variables , inflation rate is considered to be a prominent one, as it plays vital role in the monetary policy decisions of a country.

Inflation is the continuing upward movement of general price level of various goods and services in a country over a period of time. Inflation rate in India was initially represented by Wholesale Price Index of selected basket of wholesale commodities. Later in April, 2014 Reserve Bank of India considered Consumer Price Index (combined) as the key measure of inflation (The Hindu 2014). CPI is the measure of change in the general level of retail prices of selected consumer goods and services. CPIs have been widely used as a macroeconomic indicator of inflation , and also as a tool by Government and Central Bank for targeting inflation and monitoring price stability. (Consumer Price Index : Changes in the Revised Series 2015).

The relationship between stock price and inflation rate is widely discussed and investigated topic both in developed and developing countries. There are many studies inland and abroad, examining the liaison between these two variables. But studies specifically investigating the relationship between inflation and stock indices of different industries are limited. Every industry need not behave in a common way to the change in inflation rate.

The current study aims to establish the long run relationship between inflation and industry specific stock indices. Since industry specific stock indices exhibit the behaviour of sectors, the effect of inflation on the chosen sector may be inferred. Investors and business decision makers may use the findings of the study to gauge the performance of a particular industry during the time of rise or fall in inflation rate and plan the strategies accordingly.

1.1 Stock Returns and Inflation

As per random walk theory, stock prices equip the information in the economy and promptly react to it. Therefore the stock market may become agog towards the fluctuation in macroeconomic variables . Being a major economic indicator, the variations in inflation rate may be an omen to stock market to either tank or surge prices.

Further, Arbitrage Pricing Theory developed by Stephen Ross, assumes return on any asset is linearly related to a set of systematic risk factors. The risk factors were again classified into macroeconomic and microeconomic risk factors. The macroeconomic variables ; the spread between long and short interest rate, high and low grade bonds, Industrial Production and expected and unexpected inflation are significant risk factors that systematically affect the stock market (Chen et al. 1986). Hence the stock return can be linearly related to the major macroeconomic variable, inflation .

Fisher (1930) proposed that the expected nominal interest rate is the sum total of expected real interest rate and the expected inflation rate. From the Fisher’s theory, it is inferred that there exists a positive relation between the stock returns and inflation . Fama and Schwert (1977) generalized Fisher’s hypothesis on to various assets like real estate, government debt instruments, short term bills and stock returns, of which stock returns and inflation showed a negative relationship.

1.2 Selected Industries and Inflation

The study had chosen four major industries of India ; Fast Moving Consumer Goods, Pharmaceutical, Information technology and Banking.

Fast moving consumer goods industry is one of the largest industries in India . It comprises of three main sectors; food and beverages, healthcare and household and personal care. The performance of this industry to a great extent depends on the purchasing power and demographic layout of the country. The purchasing power is explained by the CPI of a country. At the time of inflation the companies’ costs goes up, sales get stagnated and companies may be in doldrums. The FMCG stocks are cyclical and are expected to react to macroeconomic variables .

Indian pharmaceutical industry is one of the fastest growing industries and plays a prominent role in the global pharmaceutical sector. Inflation may have an effect on cost of production and thereby profitability of the companies. But the Pharmaceutical stocks are defensive stocks that are expected to outperform even when the economy slows down, in view of the fact that the sector deals with drugs that are always essential for life.

The fast absorption of digital technology in India has contributed for the tremendous growth of IT industry . The cost competitiveness in providing IT services has given India a remarkable outlook in the global sourcing market. Since IT stocks are growth stocks, they are expected to withstand the changes in inflation rate.

Indian banking industry has always retained the confidence among the investors. The industry has remained stable in many of the global upheavals. The monetary policies largely depend upon the inflation rate prevailing in the country. At the time of rising inflation , RBI announces monetary tightening by raising the interest rate and/or reducing money supply. This may have an adverse effect on the banking activities which may in turn encumber companies’ growth. Like FMCG stocks, Banking stocks are cyclical and are expected to change in accordance with inflation .

2 Literature Review

A large number of studies were carried out for investigating the relationship between stock prices and inflation rates. The survey on the literature exhibits varied results of relationship such as negative, positive and even no relationship between the variables. The studies of Fama (1981) and Schwert (1981) are the two major studies that confirmed negative relationship. According to Fama, there exists a positive relationship between stock returns and real activity and a negative relationship between inflation and real activity. As such he analysed the proxy hypothesis of negative relationship between stock returns and inflation and the study established the negative relationship between expected and unexpected components of inflation and stock returns. Schwert proved that the daily returns of S&P composite portfolio is negatively related to the announcement of unexpected inflation in the CPI.

Feldstein (1983) and Kaul (1987) intrigued the reason for inflation —stock relation. Feldstein found that the inverse relation between inflation and stock price is due to the basic features of US tax laws of that time. From the post war evidence, Kaul found that the negative inflation —stock relation in US, UK, Canada and Germany is due to the money demand and also due to some effect of counter cyclical money supply.

More recently, Bordo et al. (2009) found that the inflation shocks negatively impact the real stock price index (S&P 500) as well as stock market conditions. Geetha et al. (2011) established long run relationship between stock returns and both expected and unexpected inflation in Malaysia, US and China. Both expected and unexpected inflation had a negative impact on the stock market in Malaysia and a positive impact in US. Interestingly in China the expected inflation had a negative impact and unexpected inflation had a positive impact. However short run relationship was established only in China.

Ibrahim and Agbaje (2013) established positive and significant long run and short run relationship between inflation and Nigerian stock market. Later Uwubanmwen and Eghosa (2015) attempted to find the impact of inflation on the Nigerian Stock Market, but confirmed a negative but weak impact. They made a concluding remark that inflation is not strong predictor of Nigerian stock returns.

Amonhaemanon et al. (2014) found that Thai stock returns are negatively related to both unexpected and expected inflation and Vietnamese stock returns are negatively related to unexpected inflation only. But quite the reverse, the Vietnamese stock returns showed a positive relation with the expected inflation . However, the coefficients of these variables were not statistically significant. Afterwards, Hau (2017) found that 12 out 24 Vietnamese stocks have positive and significant relation with ex post inflation . But, he also found that all stocks are not related to the actual inflation .

A negative impact of inflation on stock prices was confirmed by Mahmood et al. (2014) in Pakistan stock market and Silva (2016) in Srilankan stock market. On the contrary, Floros (2004) could not find long run association between Greek Consumer Price Index , which was considered as a proxy for inflation , and Athens Stock Exchange Price Index.

In the Indian context Patel (2012) found long run equilibrium between macroeconomic variables including inflation and stock market indices such as SENSEX and S&P CNX NIFTY. Also he stated that inflation is highly positively correlated with the indices. Ray (2012) tested the granger causality between stock prices and macroeconomic variables and found a unidirectional causality running from stock price to inflation , but not from inflation to stock price . Reddy (2012) opined that interest and inflation rate jointly have a negative influence on Indian stock price . Tripathi and Kumar (2014) conducted a study on BRICS stock markets and could not find any significant long term equilibrium relationship between inflation and Indian stock market. Similarly Gurloveleen and Bhatia (2015) found no significant relationship between inflation and Indian stock.

There are a few studies that concentrated on sectoral stock prices while examining the relationship. Lajeri and Dermine (1999) found that in France during the period of high volatility , the unexpected interest rate and inflation rate jointly showed a negative impact on the bank stock returns. Abadi (2006) confirmed negative and significant impact of expected inflation on the bank stock returns. Umashankar and Himabindu (2015) investigated the impact of inflation on the performance of FMCG stocks. They found a significant negative relation in four years and positive relation in three years. Nurhakim et al. (2016) investigated the influence of profitability and inflation on Pharmaceutical stock prices. The result shows that pharmaceutical stocks are not significantly influenced by the inflation rate.

3 Data and Methodology

The study is confined to four major industry specific stock indices of Bombay Stock Exchange such as S&P BSE Fast Moving Consumer Goods, S&P BSE Healthcare, S&P BSE Information Technology and S&P BSE Bankex. Inflation rate measured in terms of Consumer Price Index (CPI) is chosen for the study. The period of the study is from January 2012 to October 2017. The study has used EViews 9 for econometric analysis.

3.1 Objectives of the Study

The objective of the study is to examine the long run relationship between inflation rate and industry specific stock indices. More specifically the study investigates the long run causality running from Inflation to Industry specific sock indices.

3.2 Hypothesis

H0: There is no long run causality running from Inflation rate to Industry specific Indices.

3.3 Data Required and Source

The data required for the study are monthly inflation rates measured in terms of CPI and monthly index values of four selected indices from January 2012 to October 2017. Inflation rates are obtained from Database of Indian Economy website maintained by RBI and Index Values from BSE website.

3.4 Methodology

Various econometric tools are applied in this study. Initially the time series data are checked for stationarity using Augmented Dickey Fuller (ADF) Test. Since the whole set of data are stationary only at first difference; the study proceeds with the Johansen Cointegration Test to understand the existence and number of cointegrating vectors. Optimal number of lags for the cointegration test was arrived at using Akaike Information Criterion (AIC), Final Prediction Error (FPE) and Hannan-Quinn information criterion (HQ). Further when cointegration exists Vector Error Correction Model is used to examine and confirm the long run causality . If Johansen cointegration test reveals no cointegration between the variables, Vector Autoregressive Model is used to examine the significance of dynamic relationship. As the study is limited to examine the influence of inflation on stock return, the equations considering Stock indices as exogenous variable and inflation rate as endogenous variable are only taken.

4 Empirical Results and Analysis

4.1 Unit Root Test

The study deals with five set of time series data and hence to proceed further the stationarity needs to be checked. ADF test is used to examine the stationarity of all five variables: Inflation rate, S&P BSE FMCG, S&P BSE Healthcare, S&P BSE IT and S&P BSE Bankex. The stock index values are converted to natural log values, before testing for unit root.

  • H0: The time series has a unit root (Time series are non-stationary)

The study failed to reject null Hypothesis at level since, the critical values are less than the t-statistic at 1, 5 and 10% level of significance and the p-value is greater than 0.05. But it got rejected at the first difference for all the five variables. That means all the variables are stationary only at their first difference. The results of ADF test at first difference are shown in Table 4.1.

Table 4.1 Results of ADF test
Full size table

4.2 Lag Selection

Before proceeding with further econometric tools optimal lag length needs to be decided. There are many information criterions available for lag selection. In the current study, AIC, FPE and HQ are chosen. For each sector index and inflation rate set of data, two has been chosen as the optimum lag length.

4.3 Results of Johansen Cointegration Test

Since the data are stationery at first difference, Johansen cointegration test is applied to investigate whether the variables are cointegrated. Two variables are cointegrated if they have a long term equilibrium relationship (Gujarati et al. 2012). Johansen cointegration test uses \( \lambda_{trace} \) and \( \lambda_{max} \) statistics for finding the number of cointegrating vectors. The present study uses \( \lambda_{trace} \) statistic for testing the hypothesis.

S&P BSE FMCG Index and Inflation: The Johansen Cointegration test rejects the null hypothesis of no cointegration between S&P BSE FMCG and Inflation , at 5% level of significance as the trace statistic is greater than the critical value. The test indicates the existence of at least one cointegrating equation, which means the variables are cointegrated. Therefore VECM can be applied. The results are shown in Table 4.2.

Table 4.2 Results of Johansen cointegration test between S&P BSE FMCG and inflation
Full size table

S&P BSE Healthcare index and Inflation : The test failed to reject the null hypothesis of no cointegration between S&P BSE Healthcare and Inflation as the trace statistic is less than the critical value. Since there is no long run equilibrium between the variables, VECM cannot be applied. As a result the study continues with VAR model. The results are shown in Table 4.3.

Table 4.3 Results of Johansen cointegration test between S&P BSE healthcare and inflation
Full size table

S&P BSE IT Index and Inflation: As per the results shown in Table 4.4, the test has failed to reject the null hypothesis of no cointegration at 5% level of significance. There exists no long run equilibrium between S&P BSE IT and Inflation and therefore VAR model will be used.

Table 4.4 Results of Johansen cointegration test between S&P BSE IT and inflation
Full size table

S&P BSE Bankex and Inflation : The results shown in Table 4.5 imply that the null hypothesis of no cointegration between S&P BSE Bankex and Inflation is rejected and there exists at least one cointegrating vector. VECM model will be used as there is long run equilibrium between the variables.

Table 4.5 Results of Johansen cointegration test between S&P BSE bankex and inflation
Full size table

4.4 Results of Vector Error Correction Model

VECM is a restricted VAR designed for use when variables are cointegrated. With the existence of cointegrating vector that exhibits the long run relationship, long run causality must be analysed using VECM. VECM is applied for FMCG and Bankex indices, since they are found to be cointegrated with inflation .

Model 1: VECM equation for dependent variable S&P BSE FMCG Index

$$ \begin{aligned} D\left( {LNFMCG} \right) & = C\left( 1 \right) * \left( {LNFMCG\left( { - 1} \right) + 0.0648759788411 * INF\left( { - 1} \right) - 9.31444727087 } \right) \\ & \quad + C\left( 2 \right) * D\left( {LNFMCG\left( { - 1} \right)} \right) + C\left( 3 \right) * D\left( {LNFMCG\left( { - 2} \right)} \right) \\ & \quad + C\left( 4 \right) * D\left( {INF\left( { - 1} \right)} \right) + C\left( 5 \right) * D\left( {INF\left( { - 2} \right)} \right) + C\left( 6 \right) \\ \end{aligned} $$

In the equation LNFMCG is the log value of FMCG index and INF is the Inflation rate. C(2) and C(3) are the coefficients of lagged values of dependent variable. C(4) and C(5) are the coefficients of lagged values of independent variable and C(6) is the constant. C(1) is the coefficient of cointegrating equation (LNFMCG(−1) + 0.0648759788411 * INF(−1) − 9.31444727087). It is the error correction term and is understood as the speed of adjustment towards long run equilibrium. The results of VECM estimates are shown in the Table 4.6. C(1) is negative and also significant at 5% level. This implies that there is long run causality running from Inflation rate to FMCG Index. Also the cointegrating equation: ‘LnFMCG = 9.31 – 0.065 Inflation ’, shows that the variables are negatively associated in the long run.

Table 4.6 Results of VECM estimates—BSE FMCG index and inflation
Full size table

Model 2: VECM equation for dependent variable S&P BSE Bankex

$$ \begin{aligned} D\left( {LNBANK} \right) & = C\left( 1 \right) * \left( {LNBANK\left( { - 1} \right) + 0.103617920649 * INF\left( { - 1} \right) - 10.4424952318} \right) \\ & \quad + C\left( 2 \right) * D\left( {LNBANK\left( { - 1} \right)} \right) + C\left( 3 \right) * D\left( {LNBANK\left( { - 2} \right)} \right) \\ & \quad + C\left( 4 \right) * D\left( {INF\left( { - 1} \right)} \right) + C\left( 5 \right) * D\left( {INF\left( { - 2} \right)} \right) + C\left( 6 \right) \\ \end{aligned} $$

From the Table 4.7, it is inferred that there is long run causality running from Inflation rate to BSE Bankex as C(1) is negative and significant at 5%. The cointegrating equation: ‘LnBANK = 10.44 − 0.104 Inflation ’, shows that there is a long run negative association between the variables.

Table 4.7 Results of VECM estimates—BSE bankex and inflation
Full size table

4.5 Results of Vector Auto Regression Estimate

VAR model is used to estimate the dynamic relationship when the variables are not cointegrated. VAR model is applied for BSE Healthcare index and BSE IT index as they are not cointegrated with Inflation rate.

Model 3: VAR Model with BSE Healthcare Index (LNHEALTH = log values) as dependent variable

$$ \begin{aligned} D\left( {LNHEALTH} \right) & = C\left( 1 \right) * D\left( {LNHEALTH \left( { - 1} \right)} \right) + C\left( 2 \right) * D\left( {LNHEALTH \left( { - 2} \right)} \right) \\ & \quad + C\left( 3 \right) * D\left( {INF\left( { - 1} \right)} \right) + C\left( 4 \right) * D\left( {INF\left( { - 2} \right)} \right) + C\left( 5 \right) \\ \end{aligned} $$

The VAR estimates (Table 4.8) shows that none of the coefficients are significant at 5%, which means that BSE Healthcare Index is not influenced by Inflation rate.

Table 4.8 Results of VAR estimates—BSE healthcare index and inflation rate
Full size table

Model 4: VAR Model with BSE IT Index (LNIT = log values) as dependent variable

$$ \begin{aligned} D\left( {LNIT} \right) & = C\left( 1 \right) * D\left( {LNIT\left( { - 1} \right)} \right) + C\left( 2 \right) * D\left( {LNIT\left( { - 2} \right)} \right) + C\left( 3 \right) * D\left( {INF\left( { - 1} \right)} \right) \\ & \quad + C\left( 4 \right) * D\left( {INF\left( { - 2} \right)} \right) + C\left( 5 \right) \\ \end{aligned} $$

From Table 4.9, it can be seen that only C(1), which is the coefficient of LNIT(−1) is significant at 5%. It means that the present BSE IT index value associates with the previous month index value. All the other coefficients are insignificant at 5% level; therefore BSE IT Index is not influenced by Inflation rate.

Table 4.9 Results of VAR estimates—BSE IT index and inflation rate
Full size table

5 Conclusion

The study explores the influence of inflation rate on the industry specific stock indices. FMCG index, Healthcare Index, IT Index and Banking Index of Bombay Stock Exchange were considered for the study. The study confirmed the long run causality running from inflation rate to FMCG index and Banking Index. Also the inflation rate is having a negative association with these indices. Healthcare and IT indices were found to have no association with the inflation rate. The results are an implication for the investment decisions in these sector stocks at the time of higher inflation .

FMCG and Banking Stocks are cyclical stocks and are expected to change in accordance with the inflation . Also the Pharmaceutical stocks being defensive stocks and IT stocks being growth stocks are expected not to have a causal relationship with the inflation especially in the long run. Therefore the finding of the present study is in confirmation with the existing knowledge base.

The study may be replicated with indices of various other sectors, to understand the impact of inflation on the performance of other sectors. Also Nifty sector indices may be included and compared to have a clear picture on the inflation impact. Studies investigating the influence of various macroeconomic variables on sector indices will be useful for the investors to make rational investment decisions.