The Relationship of Inflation and Stock Return: An Empirical Study of the Hedging or Wealth Effects in Bangladesh
Md. Sharif Ullah Mazumder, David Aadland
Department of Economics and Finance, University of Wyoming, Laramie, Wyoming, USA
Email address:
To cite this article:
Md. Sharif Ullah Mazumder, Dr. David Aadland. The Relationship of Inflation and Stock Return: An Empirical Study of the Hedging or Wealth Effects in Bangladesh. International Journal of Finance and Banking Research. Vol. 1, No. 2, 2015, pp. 24-31. doi: 10.11648/j.ijfbr.20150102.12
Abstract: The relationship between stock market returns and inflation is a well studied area in economics and financial economics. Theoretical and empirical researches on this topic establish that inflation causes stock market return. In this paper we study on Bangladesh stock markets return and inflation and find interesting results that stock market returns cause inflation. This empirical study has been conducted on basis of the monthly stock market index and CPI data. The result of the study can be a roadmap to the policymaker of Bangladesh to control the inflation within the tolerance level by adopting well regulation on stock market.
Keywords: Stock Market, Inflation, DSE Index, Granger Causality Effect
1. Introduction
Does inflation affect the stock market returns or do the stock market returns affect the inflation rate in Bangladesh? The research of E.F Fama and G.W. Schwert ^{1} show that there is a negative correlation between inflation and stock market returns. Warner Kramer says that the equity return is negatively correlated with inflation and that investors want to be more risk averse in high inflationary situations. This is because of the possibility of negative equity return ^{2}. Another paper, written by Jacob Boudukh and Matthew Richardson, find that there is a long run positive relationship between inflation and stock market returns for the United States and the United Kingdom ^{3}. The lack of the institutional investors and the significant presence of short term investors make the stock market of Bangladesh more impulsive. So, we think the relationship between stock market returns and inflation in Bangladesh is thought provoking.
An abundance of research has been conducted from both an empirical and theoretical standpoint, specifically, whether there is any relationship between inflation and stock market returns. There are a few studies conducted on Bangladesh stock market returns and macroeconomic indicators, like interest rates and GDP. But literature on the relationship between inflation and stock market returns can be expanded on.
The stock market return is a major economic indicator to the policy makers of all developed and developing countries. In Bangladesh, stock market returns become quite volatile with the change of political wave. Also, people living in the city areas that are familiar with the stock market, but the marginal village people are out of access due to lack of market infrastructure and acquaintance. For the remainder of this paper, we will review the results and findings of some empirical and theoretical research and discuss the model of the paper. In the remaining sections, we cover methodology, the economic significance of the results and provide some recommendations from my conclusions.
2. Literature Review
Both in the context of developed and developing countries, there have been extensive theoretical and empirical studies conducted on the relationship between stock market return and inflation. Inflation and financial sector correlation: The case of Bangladesh, by Abu N M Wahid, Muhammad Shahbaz and Pervez Azim (2011), show that inflation tends to induce volatility of the stock returns and so does the returns of savings. Inflation and economic growth in Bangladesh: 1981 to 2005, a working paper conducted by Shamim Ahmed and Golam Mortaza (2005) show that there is a long-run and short-run negative relationship between inflation and economic growth. Owusu Frimpong (2001) find that the stock market return is negatively correlated with inflation. In the study, he conducted a survey and found that people are unwilling to trade in a high-inflation market because they will get less return. Finally, a paper by Nwokoma (2005) find that there is a positive relationship between stock market return and inflation in Nigeria, but that the relationship is inconclusive in the case of Kenya and South Africa^{4}.
3. Model
To examine the relationship between inflation and stock market returns, we will test the causality or direction of influence. Since, the future cannot predict the past, if independent variables cause the dependent variable, then changes in independent variables would precede changes in the dependent variable. Therefore, in a regression of dependent variable on independent variable (including its own past values) if we include past or lagged values of independent variable, then we can significantly improve the prediction of the dependent variables^{5}. The structural models to examine the causality are given below:
(1)
(2)
Here, DGEN_{t} is DSE General Index (proxy of the stock market returns) and INF_{t} is the rate of inflation ^{6}.
Hypothesis for equation (1)
Null Hypothesis: : = 0; Inflation does notGranger cause stock return.
Hypothesis for equation (2)
Null Hypothesis: = 0; Stock return does not Granger cause inflation.
4. Data and Methodology
We now define the variables and provide the data sources. The DSE General Index (DGEN) has been used as the measure of the stock market return. Rate of inflation is computed from the monthly CPI data [6]. Data sources include "Monthly Economic Trends’ published by Bangladesh Bank, Dhaka Stock Exchange website (dsebd. org) and website of Trading Economics. Data are monthly of 144 observation months from January 2001 to December 2012. We omit the data after the period of December 2012 to avoid miss-measurement as the market index calculation has been changed from January 28, 2013.
Brief overview of Dhaka Stock Exchange
Dhaka Stock Exchange (DSE) was established in 1954 as the first stock exchange in the country, and began operating in 1956. In August 2014, there were 307 listed companies in DSE and the market capitalization of Taka was 3030.3 billion. The market performance has been measured by DGEN index (changed the parameters as DSEX from January 2013) and the DSE20 index (changed as DSE30 from January 2013).
Summary statistics of the two variables
DGEN Index | Inflation rate | |
Mean | 2685.32 | .0058 |
Median | 1839.79 | .0037 |
Maximum | 8602.44 | .042 |
Minimum | 607 | -.013 |
Std Deviation | 1992.40 | .009 |
Skewness | 1.01 | .75 |
Kurtosis | 3.03 | 4.56 |
Observation | 144 | 144 |
Methodology
Stationary data
As the data are time series data, the first step involves checking to see if the data is stationary. We have plotted the DGEN variables against time and to see whether there is any trend in the graph. In appendix section 1, the time series data of the DSE general index has been shown graphically. We use the Augmented Dickey Fuller (ADF) test, developed by Dickey and Fuller (1979, 1981), to see whether there is any trend in the variables. We found there is trend in stock market returns. To de-trended the stock market return we use three methods described below and ADF test has done with the three methods and found the trend has been removed successfully.
Lag selection
Final Predictor Error (FPE), Akaike information criterion (AIC), Hannan-Quinn information criterion (HQIC) and Schwarz information criterion (SBIC) criterion have been followed to select the number of lags. All the methods confirm the first lag of both the variables.
Causality effect
The last section of my analysis is to test the causality between stock market return and inflation in Bangladesh. If we reject (accept) the null hypothesis in equation (1), this suggests that stock return does (does not) cause inflation. Alternatively, rejecting (accepting) the null hypothesis in equation (2) suggests that Inflation does (does not) cause stock return.
5. Results and Economic Significance
The result of the ADF model shows whether the variables are stationary or not and determines how many differences of the variables needs to be taken to make the variables stationary. The hypothesis for checking for stationary data is as follows:
Null hypothesis: : Variables have unit root or are not stationary.
From the result, the t statistic of DGEN_{t} is -1.942, but the critical value for the 5% level is -3.44.This means that we cannot reject the null hypothesis. It means that the variable DGEN is non- stationary. But in the case of the inflation variable, the t- statistics is -7.535 and the critical value is -3.44 at 5% significance level. We can reject the null hypothesis (results are shown in appendix section 1). So, from the test we can say that the stock market index is not stationary. To make the data stationary we will use three methods:
1. Growth rates of DSE general index, Δln(DGEN_{t}).
2. Residuals: I will run a regression on ln(DGEN) and save the residuals.
3. Hedrick- Prescott (HP) filters method.
First, we take the first differences of ln(DGEN_{t}) and test for stationary data using ADF method. The result of the ADF has shown that we can reject the null hypothesis as the test value is-11.76 and the critical value for 5% confidence interval is -3.44, so the variable has been de-trended successfully. We can use the de-trended variable in my further analysis. Secondly, the predicted residual of the second model graphically shows in appendix section 1 that the data has been also de-trended successfully. Therefore we will use this de-trended data to run the regression. Finally, the HP filter method is used for smoothed-curve representation of time series data. The adjustment of the sensitivity of the trend to short term fluctuations is achieved by modifying a multiplier λ. A Graph of the filtered data shows that (appendix section 1) it has been de-trended successfully.
To determine the lag length we have used FPE, AIC, HQIC and SBIC parameters. All the parameters are the lowest in one lag, so we can use one lag for the Vector Auto-regressive Model and Granger causality analysis. The results of the lag selection have been shown in appendix section-2.
Multicollinearity test
The test of multicollinearity using VIF test gave a value 1 which is less than 5. In another test of multicollinearity, correlation matrix has shown that the correlation between the two variables is 0.04 which is very insignificant. So we can say that there is no multicollinearity between the variables of lag inflation and first difference of lag lnDGEN. The other regression variables of residuals and inflation has also showed that they are not multicollinearity because the correlation matrix between this variables is also 0.04. The results of the multicollinearity test have been shown in appendix section-5.
Autocorrelation test
Null hypothesis:: There is no auto correlation.
The result of Durbin Watson test (h test), to test autocorrelation has shown that the number of the test is 1.37 for lag 1 and the p-value is 0.24. Therefore, we failed to reject the null hypothesis. Another test for auto correlation named Breusch-Godfrey Test using 8 lags gives the result of 8.87 with p-value 0.3532. So in this case we also failed to reject the null hypothesis. But in the case of lag 12, the test statistics is statistically significant. To measure the seasonality effect we have tested up to 12 lags.
In the other regression, the variables of inflation and HP-filtered DGEN, the Durbin Watson test (h-test) shows that there is no serial correlation in the 1 lag between the residuals. The test statistics is .65 for lag 1 and the p-value is 0.42, so we failed to reject the null hypothesis. The result of Breusch-Godfrey Test with 8 lags gave the result of 10.882 with p-value 0.20. We also failed to reject the null hypothesis. But in the case of lag 12, the test statistics is statistically significant. This means that there is some seasonality effect in the variables. The AIC and SIC lag selection criterion gave one lag to test the Breush-Godfrey test. Therefore, we found that there is no autocorrelation between the residuals. The results of the autocorrelationtest have been shown in appendix section-6.
Model Specification
Null hypothesis: : There is no misspecification.
The result of Ramsey’s RESET test, to test whether the model is correctly specified has shown that the number of the test statistics is .93 and the p-value is 0.43. Therefore, we failed to reject the null hypothesis. The results of the model specification test have been shown in appendix section-7.
Granger causality tests:
The Granger causality test of the first difference in lnDGEN_{t} and inflation shows a very interesting result. It shows that the stock market return causes inflation. From previous literature, we saw that inflation had a negative impact on stock market return, but in the case of Bangladesh, the result tells us that as the stock market return increases, it causes inflation. The test statistics of , the inflation (lag 1) is .11 and the p-value is .88, which means that it is statistically insignificant. We cannot reject the null hypothesis that inflation does not cause stock market returns. Most interestingly, the test statistics of explaining how stock market return causes inflation is statistically significant. This means that the stock market return causes inflation. The results of the Granger causality test is shown in appendix section 3.
The results of the vector autoregressive model are below:
Data Range | 2001m1 to 2012m12 | No of Observation | 143 | ||
1 Regression of Inflation and ΔlnDGEN | |||||
Variables | Inflation | ΔlnDGEN | |||
Co-efficient | t- value | Co-efficient | t- value | ||
Inflation Lag1 | 0.3626 | 4.68*** | 0.1100 | 0.14 | |
ΔlnDGEN lag1 | 0.0200 | 2.30* | 0.0094 | 0.11 | |
Constant | 0.0036 | 4.29*** | 0.0130 | 1.55 | |
R^{2} | 0.1600 | 0.0002 | |||
Adjusted R^{2} | 0.1500 | -0.0142 | |||
2 Regression of Inflation and Residuals | |||||
Variables | Inflation | Residuals | |||
Co-efficient | t- value | Co-efficient | t- value | ||
Inflation Lag1 | .3562 | 4.52*** | 0.1702 | 0.23 | |
Residuals lag1 | .0019 | 0.60 | 0.9501 | 31.09*** | |
Constant | .0037 | 4.44*** | -0.0040 | -0.60 | |
R^{2} | 0.13 | .8737 | |||
Adjusted R^{2} | 0.12 | .8719 | |||
3 Regression of Inflation and HP filtered DGEN | |||||
Variables | Inflation | HP filtered DGEN | |||
Co-efficient | t- value | Co-efficient | t- value | ||
Inflation Lag1 | .3503 | 4.5201*** | 1847.5701 | .6104 | |
HP Filtered DGEN lag1 | .0000 | .9301 | .9302 | 28.3901*** | |
Constant | .0036 | .0008 | -22.1703 | -.6802 | |
R^{2} | .85 | ||||
Adjusted R^{2} | .85 | ||||
**t-value is statistically significant at the 5% level, *** t-value is statistically significant at the 1% level.
OLS estimators and analysis
From the Granger causality effect we can see that the stock market return causes the inflation in Bangladesh in the growth rate model. But other two models do not show this causation. Table-2 shows the coefficients, t-values and adjusted R^{2} of the models. From the Granger causality result, we found that the stock market causes inflation, we are interested in the coefficients of the first dependent variable (inflation). If inflation in the t-1 month increases by 1% then the inflation in the month of t will increase by 0.36, other thing remaining constant. If the growth rate of DGEN (stock market return) in the t-1 month increases by 1% then the inflation will increase by 0.02% in the month of t, other thing remaining constant. The adjusted R^{2} is very insignificant. At 15.00%, the co-relation is very low between these variables and, that means the variables in lag 1(t-1 month) is not highly correlated to the variable of inflation in the month of t.
On the other hand, for the regression of inflation and the linearly de-trended model ( residuals method), we find that if the inflation of the month t-1 increase by 1% then the inflation in the month of t will increase by 0.35, other thing remaining constant. This is very similar to the previous regression result. If the residuals of stock return growth rate (residuals of regression of lnDGEN against time) in the t-1 month increases by 1% then the inflation will increase by 0.002% in the month of t, other thing remaining constant. The adjusted R^{2} dramatically rises to 87%. But here the granger causality result shows that neither of the variables causes each other.
In the other regression model of inflation and HP-filtered DGEN variable, we find that if the inflation of the month t-1 increase by 1% then the inflation in the month of t will increase by 0.35,other thing remaining constant. This result is similar to the previous findings. The lag filtered stock return has almost no impact on the current stock return.
Economic significance of the coefficients
A significant number of coefficients are statistically significant and are economically significant too. Because the previous month’s (t-1) inflation need not affect the current month’s inflation, it is also true for the previous month’s stock return growth. To control the inflation rate the government usually uses intensive macro-economic tools. So, if the month t’s inflation follows the previous month’s inflation, that means the mechanisms are not working properly. It is difficult to comment about the monetary policy of the government to control the inflation with these two factors model because the inflation is dependent on lots of issues other than previous year’s inflation and the stock price returns, which is beyond this paper.
The previous year’s growth of stock market return effect on inflation can be taken more seriously. From my study, we found that stock market returns in Bangladesh positively affect inflation. Inflation is a phenomenon when the money supply increases in the market those results to an artificial demand in the commodity market. The investors who invest in the stock market reap up the short term capital gain from the market, money supply increases in the market, and consequently, it pressures the commodity market.
6. Conclusion
This paper empirically shows the relationship of inflation and stock market return in the context of Bangladesh. In this paper, the granger causality and vector auto regressive models have been applied. We found a very economically significant result from the analysis. It is well known that the inflation affects stock market negatively. However, in case of Bangladesh, the stock market returns affect the inflation, which means that there is a pressure on commodity market of the stock market returns. So, it is an indicator to the policy makers of Bangladesh to focus on the stock market volatility. The government of Bangladesh tries to control the inflation to fulfill their political manifesto, and they adjust several market mechanisms to keep the inflation within a tolerable level. So, a well-established and regulated stock market will decrease the stock market volatility in Bangladesh which will lessen the extra pressure in the commodity market.
Appendix
1. ADF tests and results:
Test Statistics | 1% critical value | 5% critical value |
-1.942 | -4.026 | -3.44 |
P-value .63 |
Test Statistics | 1% critical value | 5% critical value | 10% critical value |
-11.74 | -3.50 | -2.89 | -2.57 |
P-value 0.00 |
De-trended data of DGEN:
Test Statistics | 1% critical value | 5% critical value | 10% critical value |
-7.53 | -4.026 | -3.44 | -3.14 |
P-value 0.00 |
2. Table 2.1: Lag selection:
Selection-order criteria
lag | LL | LR | df | p | FPE | AIC | HQIC | SBIC |
0 | 586.797 | 5.9e-07 | -8.6665 | -8.646 | -8.620 | |||
1 | 598.316 | 23.040* | 4 | 0.000 | 5.3e-07* | -8.775* | -8.722* | -8.644* |
2 | 599.313 | 1.994 | 4 | 0.737 | 5.5e-07 | -8.730 | -8.643 | -8.515 |
3 | 601.754 | 4.882 | 4 | 0.300 | 5.7e-07 | -8.707 | -8.58503 | -8.406 |
4 | 603.292 | 3.076 | 4 | 0.545 | 5.9e-07 | -8.671 | -8.51358 | -8.283 |
5 | 605.394 | 4.203 | 4 | 0.379 | 6.1e-07 | -8.642 | -8.45048 | -8.169 |
6 | 608.149 | 5.509 | 4 | 0.239 | 6.2e-07 | -8.624 | -8.39705 | -8.064 |
7 | 609.105 | 1.912 | 4 | 0.752 | 6.5e-07 | -8.579 | -8.31697 | -7.933 |
8 | 610.724 | 3.239 | 4 | 0.519 | 6.7e-07 | -8.544 | -8.24672 | -7.812 |
3. Granger causality effect:
Null Hypothesis | Chi ^{2} | Df | P-value of chi^{2} |
Stock return does not Granger cause inflation | 5.40 | 1 | .02 |
Inflation does not Granger cause stock return. | .02 | 1 | .88 |
Null Hypothesis | Chi ^{2} | Df | P-value of chi^{2} |
Stock return does not Granger cause inflation | 0.37 | 1 | .54 |
Inflation does not Granger cause stock return. | .053 | 1 | .81 |
Null Hypothesis | Chi ^{2} | Df | P-value of chi^{2} |
Stock return does not Granger cause inflation | 0.85 | 1 | .35 |
Inflation does not Granger cause stock return. | .38 | 1 | .54 |
4. Regression result:
Dependent Variable | Time | |||||
Data | 2001m1 to 2012m12 | No of Observation | 144 | |||
Co-efficient | Std. Error | Z value | P-value | 95% confidence Interval | ||
Time | .0170 | 0.0004 | 37.8610 | 0.0000 | .01631 | .01810 |
Constant | -2.1201 | 0.2600 | -8.2310 | 0.0000 | -2.6331 | -1.6100 |
R^{2} | .9102 | |||||
Adjusted R^{2} | .9101 | |||||
Sum of Squared Estimate | 74.4010 | MSE | 74.392 | F(1,142) | 1433.3600 | |
Sum of Squared residuals | 7.3720 | MSR | .051 | P-value(F) | 0.0001 |
Dependent Variable | Inflation | |||||
Data | 2001m1 to 2012m12 | No of Observation | 142 | |||
Co-efficient | Std. Error | Z value | P-value | 95% confidence Interval | ||
Inflation Lag1 (γ_{2}) | .3601 | .0830 | 4.74 | .0000 | .2100 | .5100 |
1^{st} diff. lnDGEN lag1 (γ_{3}) | .0200 | .0080 | 2.33 | .0200 | .0027 | .0370 |
Constant (γ_{1}) | .00361 | .0008 | 4.33 | .0000 | .0019 | .0053 |
R^{2} | .1601 | |||||
Adjusted R^{2} | .1502 | |||||
Sum of Squared Estimate | .00180 | MSE | .0018 | F(2,139) | 13.19 | |
Sum of Squared residuals | .0100 | MSR | .0096 | P-value(F) | 0.000 |
Dependent Variable | First difference of lnDGEN | |||||
Data | 2001m1 to 2012m12 | No of Observation | 142 | |||
Co-efficient | Std. Error | t- value | P-value | 95% confidence Interval | ||
1^{st} diff. lnDGEN index Lag1 (β_{2}) | .0094 | .0840 | .1100 | .9100 | -.15801 | .17702 |
Inflation Lag1 (β_{3}) | .1100 | .7600 | .1400 | .8800 | -1.4001 | 1.6101 |
Constant (β_{1}) | .0131 | .0080 | 1.5500 | .1241 | -.0036 | .0291 |
R^{2} | .0002 | |||||
Adjusted R^{2} | -0.0142 | |||||
Sum of Squared Estimate | .0002 | MSE | .0001 | F(2,139) | .0201 | |
Sum of Squared residuals | .9440 | MSR | .0067 | P-value (F) | .9841 |
Dependent Variable | Inflation | |||||
Data | 2001m1 to 2012m12 | No of Observation | 143 | |||
Co-efficient | Std. Error | t- value | P-value | 95% confidence Interval | ||
Lag residuals γ_{2} | .0019 | .0032 | .60 | .5480 | -.0044 | .0082 |
Inflation Lag1 (γ_{3}) | .3561 | .0781 | 4.52 | 0.0000 | .2000 | .5101 |
Constant (γ_{1}) | .0031 | .0008 | 4.44 | 0.0000 | .0020 | .0051 |
R^{2} | .1300 | |||||
Adjusted R^{2} | .1201 | |||||
Sum of Squared Estimate | .0015 | MSE | .0007 | F(2,139) | 484.2600 | |
Sum of Squared residuals | .0100 | MSR | .0000 | P-value (F) | 0.0000 |
Dependent Variable | Residuals | |||||
Data | 2001m1 to 2012m12 | No of Observation | 143 | |||
Co-efficient | Std. Error | t- value | P-value | 95% confidence Interval | ||
Lag residuals β_{2} | .9501 | .0300 | 31.09 | .0000 | .8901 | 1.0110 |
Inflation Lag1 (β_{3}) | .1711 | .7500 | .23 | .8202 | -1.3101 | 1.6501 |
Constant (β_{1}) | -.0048 | .0081 | -.60 | .5501 | -.0201 | .0111 |
R^{2} | .8700 | |||||
Adjusted R^{2} | .8701 | |||||
Sum of Squared Estimate | 6.4301 | MSE | 3.2200 | F(2,139) | 484.2601 | |
Sum of Squared residuals | .9303 | MSR | .0067 | P-value (F) | 0.0000 | |
Sum of Squared residuals | .9441 | MSR | .0067 |
Dependent Variable | Inflation | |||||
Data | 2001m1 to 2012m12 | No of Observation | 142 | |||
Co-efficient | Std. Error | Z value | P-value | 95% confidence Interval | ||
Inflation Lag1 (γ_{2}) | .3501 | .0780 | 4.5201 | 0.0000 | .1900 | .5000 |
HP-filtered DGEN lag1 (γ_{3}) | -.0000 | .0000 | .9300 | 0.3550 | .0000 | .0000 |
Constant (γ_{1}) | .0036 | .0008 | 4.5202 | 0.0000 | .0021 | .0054 |
Dependent Variable | First difference of HP-filtered DGEN | |||||
Data | 2001m1 to 2012m12 | No of Observation | 142 | |||
Co-efficient | Std. Error | t- value | P-value | 95% confidence Interval | ||
HP-filtered DGEN Lag1 (β_{2}) | .9300 | .032 | 28.69 | 0.00 | .86 | .99 |
Inflation Lag1 (β_{3}) | 1847 | 3009 | .61 | .54 | -4050 | 7745 |
Constant (β_{1}) | -2217 | 32.42 | -.68 | .49 | -85.71 | 41.37 |
5. Multicollinearity test:
Variable | VIF | 1/VIF |
Lagfdlndse | 1.00 | .99 |
Laginflation | 1.00 | .99 |
Lagfdlndse | Laginflation | |
Lagfdlndse | 1.00 | |
Laginflation | .04 | 1.00 |
Variable | VIF | 1/VIF |
HP-filtered DGEN | 1.01 | .99 |
Laginflation | 1.01 | .99 |
HP-filtered DGEN | Laginflation | |
HP-filtered DGEN | 1.00 | |
Laginflation | -.08 | 1.00 |
6. Auto correlation test:
Model | Lag(p) | Chi^{2} | Df | P-value |
Inflation and lnfddse | 1 | 1.3740 | 1 | .2411 |
Hp-filtered and inflation | 1 | 0.6490 | 1 | .4203 |
Model | Lag(P) | Chi^{2} | df | p-value | Lag(P) | Chi^{2} | df | p-value |
Inflation and lnfddse | 12 | 26.27 | 12 | .01 | 8 | 8.87 | 8 | .35 |
Hp-filtered and inflation | 12 | 30.41 | 12 | .002 |
7. Model specification:
Null hypothesis: the model has no omitted variables:
Model | F-value | p-value |
Inflation and lnfddse | .93 | .43 |
Hp-filtered and inflation | .83 | .48 |
References
Footnotes
^{1}Asset Returns and Inflation, Eugene F Fama and G.William Schwert, 1977, page-1.
^{2}Equity investment as a hedge against inflation: part 1 by Warner Kramer.
^{3}Stock returns and Inflation, a long horizon perspective by Jacob Boudoukh and Matthew Richardson, American Economic Review, December 1993, p-2
^{4}The competitive performance of African stock Markets: Nominal, real or U.S.Dollar return by Alvan E Ikoku and Ahmed Hossaini, International journal of Business (2008).
^{5}Basic Econometrics, 4^{th} Edition, Damodar N Gujrati and Dawn C Porter, page no-696.
^{6}Inflation= (CPI_{t}-CPI_{t-1})/CPI_{t-1}