Impact of Foreign Exchange Rate on Stock Returns - Dissertation Example

? Dissertation, Finance and Accounting Impact of foreign exchange rate on stock returns and Chapter 5: findings and analysis 5.1 Introduction The ADF test was applied in this study to test the null hypotheses of unit root in each of the series. Basically, if the ADF statistic exceeds the critical value in each of the tests, then the null hypothesis of the unit root cannot be accepted. On the other hand, if the test statistic does not exceed the critical value, then the null hypothesis is accepted and the series is then said to be non-stationary (Dickey and Fuller, 1981). Table 1 reports the results of Augmented Dickey Fuller test of unit root for the stock indices and exchange rates of both the UK and Pakistan. The test has further been applied at different levels as well as in the first difference form. The findings will be reported in terms of each country starting with the UK first then Pakistan. Both the ADF and regression results will be presented for the country indices and then for each of the companies. This chapter is divided into several parts with the main parts comprising study findings for the UK (both ADF test and regression), study findings from Pakistan (ADF and regression), and finally the analysis section where these findings are discussed and analyzed. 5.2 Findings (UK) Augmented Dickey Fuller (ADF) Test Table 1 below presents ADF results for UK companies and GBP/USD rate. As is evident in the table, the results include the FTSE 100 index and ten major companies operating in the UK. Table 2 provides the findings of the ADF test for the national exchange rate (EX) of the GBP against the GBP. Table 1: Augmented Dickey Fuller test (UK companies and FTSE 100 Index) Augmented Dickey fuller (ADF) test (UK companies and FTSE 100 Index) Index + companies S= Index value/ stock price Critical values of 1%, 5%, 10% at Level ?S= 1st difference Critical values at 1%, 5%, 10% for ?S FTSE 100 -2.250374 -3.486064 -2.885863 -2.579818 -10.25769 -2.584707 -1.943563 -1.614927 Antofagasta -3.066955 -4.036983 -3.448021 -3.149135 -12.97182 -2.584707 -1.943563 -1.614927 Babdock -1.747700 -4.036983 -3.448021 -3.149135 -10.71888 -3.486551 -2.886074 -2.579931 HSBC Holdings -2.867695 -3.486064 -2.885863 -2.579818 -10.42942 -2.584707 -1.943563 -1.614927 Imperial tobacco -2.160689 -4.036983 -3.448021 -3.149135 -11.17921 -2.584707 -1.943563 -1.614927 Johnson matthey -2.543982 -4.036983 -3.448021 -3.149135 -10.41562 -2.584707 -1.943563 -1.614927 Meggit 0.828451 -2.584539 -1.943540 -1.614941 -11.56089 -2.584707 -1.943563 -1.614927 Morrison -2.901854 -4.036983 -3.448021 -3.149135 -11.50647 -2.584707 -1.943563 -1.614927 Pearson plc -1.841245 -4.036983 -3.448021 -3.149135 -12.55195 -3.486551 -2.886074 -2.579931 Standard chartered bank -1.868623 -3.486064 -2.885863 -2.579818 -10.62130 -2.584707 -1.943563 -1.614927 Weir group -1.782673 -4.036983 -3.448021 -3.149135 -9.851895 -2.584707 -1.943563 -1.614927 Table 2: Augmented Dickey fuller (ADF) test (GBP/USD) Augmented Dickey fuller (ADF) test (GBP/USD) Currency X= Exchange rate Critical values of 1%, 5%, 10% at Level ?X= 1st difference Critical values at 1%, 5%, 10% for ?X GBP/USD -2.046301 -4.036983 -3.448021 -3.149135 -9.755815 -2.584707 -1.943563 -1.614927 According to the findings of the study, the ADF test results for the FTSE 100 index show that FTSE statistic value does not exceed the critical values at the 1%, 5%, and 10% critical levels. Therefore, the null hypothesis of the unit root for the index is accepted and thus it contains a unit root all the levels. This stock index is, therefore, non-stationary. However, at the first difference, the stock index becomes non-stationary as results indicate that the index statistic exceeds the critical values at 1%, 5% and 10% levels. For Antofagasta, it is evident from the results that the statistic value at level does not exceed the critical values at 1%, 5%, and 10%. Therefore, the null hypothesis of the unit root for the index cannot be rejected and thus it contains a unit root. However, the same is not true for tests at the first difference because the statistic value to the company exceeds the critical values at all the three levels. In this case, the Antofagasta stock price is deemed to be stationary. In the case of Babdock, the statistic value of their stock at level is lower than the critical values at the 1%, 5%, and 10% levels meaning that the null hypothesis of the unit root for the company is accepted. The index value for the company is, therefore, non-stationery. It is, however, observed that at the first difference, the statistic value of the index for the company is lower than the critical values at 1%, 5%, and 10%. This also indicates that the null hypothesis is rejected and that the index does not contain a root. Consequently, the stock for the company is deemed to be stationary. The same observation is true for HSBC, Imperial Tobacco, John Matthey, Meggit, Morrison, Pearson, Standard Chartered, and Weir Group It can be observed that for all the companies in the UK, at the first difference (?S), the statistic value of the index for the company is lower than the critical values at 1%, 5%, and 10%. This, therefore, indicates that the null hypotheses for all the cases should be rejected and that their indexes do not contain a root at the first difference. Consequently, the stocks for the companies are deemed to be stationary. Regression Analysis Table 3 below shows the regression analysis results of the FTSE 100 index and the 10 companies in the UK included in the study. The regression analysis was carried out with Foreign Exchange Rate (EX) as the independent variable with the dependent variable being Stock return. Table 3: Summary of regression analysis results (UK) Regression Statistics Antof. Babdock HSBC Holdings Imperial Tobacco John Matthey Meggit Morrisson Pearson Plc Std. Chart Weir Grp FTSE 100 Index Multiple R 0.03854 0.06168 0.10825 0.01143 0.01407 0.05136 0.14615 0.11793 0.02821 0.05550 0.02384 R Square 0.00148 0.00380 0.01172 0.00013 0.00019 0.00263 0.02136 0.01390 0.00079 0.00308 0.00056 Adjusted R Square -0.00705 -0.00471 0.00327 -0.00842 -0.00835 -0.00589 0.01299 0.00548 -0.00774 -0.00544 -0.00797 Standard Error 88.0448 29.268 32.8717 82.8903 111.726 18.6362 12.3243 35.0234 90.0953 91.4569 64.8731 Observations 119 119 119 119 119 119 119 119 119 119 119 Significance (95% Conf.) 0.6773 0.5052 0.2412 0.9017 0.8793 0.5790 0.1127 0.2015 0.7607 0.5488 0.7967 The regression analysis results of the FTSE index in the table reveal that the regression model used in this study explains 0.2% of effects of foreign exchange rate (Independent Variable) on Stock Return (Dependent Variable) in the UK. This means that the remaining 99.2% of effects on the index can be explained by other variables (In this case extraneous variables). The level of significance is 0.6773 at the 95% significant level, which is greater than 0.05 indicating that the results are, therefore, not significant. In the case of Antofagasta, R squared is 0.00148; this means that the model can explain 0.15% of the effects of foreign exchange on Stock return. Extraneous variables, therefore, affect more than 98% of the observable trends in stock return. The result is not significant at the 95% level of confidence (Significance=0.6773). Babdock registered an R squared value of 0.0038, which means that the model accounts for about 0.4% of the effects of foreign exchange on stock return for the company. The results are not significant at the 95% confidence level (Significance=0.5052) For HSBC, the model accounts for about 1.2% of the total variability meaning that over 98.8% of the changes in stock return are not affected by foreign exchange, but are attributed to extraneous factors. The findings are non-significant at 95% confidence level (s= 0.2412). On the other hand, for the case of Imperial Tobacco, the model accounts for about 0.13% of variability in foreign exchange rate affecting returns in stock for the company. The rest is affected by other variables outside the model. The results are not significant at the 95% confidence level (Significance=0.9017). For John Matthey, the model accounts for only 0.2% of the effects of exchange rates on stock return. The finding is not significant at 95% confidence level(s=0.8793). Similarly, for Meggit, the model accounts for only 0.26% exchange rate effect on stock returns and the finding is insignificant (s=0.5790). The same trend can be seen for all the other companies with the correlation model accounting for about 2.1% of the effects of exchange rate on stock rates in Morrison, a result which is insignificant at the 95% level of confidence (s=0.1127). For Pearson, the correlation model accounts for about 1.4% of the effects of exchange rate on stock rate, a result which is insignificant at the 95% level of confidence (s=0.2015). For Standard Chartered Bank, our model accounts for about 0.1% of the effects of foreign exchange on Stock return. Extraneous variables, therefore, affect more than 99.9% of the observable trends in stock return. The result is not significant at the 95% level of confidence (Significance=0.7607). Finally, in the case of Weir Group, the model accounts for about 0.31% of the effects of foreign exchange on Stock return - the result is not significant at the 95% level of confidence (Significance=0.5488). 5.3 Findings (Pakistan) Augmented Dickey Fuller (ADF) Test In the case of Pakistan, the ADF findings are presented in exactly the same way as the results for the UK, but these appear in two tables because they were obtained at two different times, therefore, necessitating the difference. The analysis also involves concurrent presentation of the ADF findings and regression results for each company. Table 4: Augmented Dickey fuller (ADF) test (PAKISTANI COMPANIES) Augmented Dickey fuller (ADF) test (PAKISTANI COMPANIES) Index + companies S= Index value/ stock price Critical values of 1%, 5%, 10% at Level ?S= 1st difference Critical values at 1%, 5%, 10% for ?S KSE 100 Index 1.333862 -2.584539 -1.943540 -1.614941 -9.310450 -2.584707 -1.943563 -1.614927 Arif Habib corporation limited -3.531164 -4.036983 -3.448021 -3.149135 -11.60382 -2.584707 1.943563 1.614927 Bank Alfalah limited -2.254333 -4.051450 -3.454919 -3.153171 -9.834523 -2.588292 -1.944072 -1.614616 Bata Pakistan limited -2.415953 -4.036983 -3.448021 -3.149135 -11.24823 -2.584707 -1.943563 -1.614927 Bank of Khyber -3.416011 -3.514426 -2.898145 -2.586351 -9.290606 -2.594563 -1.944969 -1.614082 Engro corporation limited -0.158973 -2.585050 -1.943612 -1.614897 -10.86238 -2.585405 -1.943662 -1.614866 Habib metropolitan bank -1.742879 -2.597025 -1.945324 -1.613876 -9.821114 -2.597476 -1.945389 -1.613838 Indus motor company limited -1.970113 -3.486064 -2.885863 -2.579818 -10.93459 -2.584707 -1.943563 -1.614927 Mari petroleum company limited -2.603732 -3.487550 -2.886509 -2.580163 -13.69172 -2.585226 -1.943637 -1.614882 Nishat power limited 0.918612 -2.627238 -1.949856 -1.611469 -5.312531 -2.628961 -1.950117 -1.611339 Oil and gas development company -2.804505 -4.044415 -3.451568 -3.151211 -9.206209 -3.492523 -2.888669 -2.581313 Table 4 shows the first results of the ADF test on the Pakistani stock exchange and ten companies. According to the findings of the study, the ADF test results for the KSE 100 index show that the KSE statistic value exceeds the critical values at the 1%, 5%, and 10%. Therefore, the null hypothesis of the unit root for the index is rejected, the value does not contain a unit root and the stock index is, therefore, stationary. This is also the case at the 1st difference level. Just like in the case of the UK companies, it can be observed that for all the companies in Pakistan, at the fist difference (?S), the statistic value of the index for the company is lower than the critical values at 1%, 5%, and 10%. This, therefore, indicates that the null hypotheses for all the cases should be rejected and that their indexes do not contain a root at the first difference. Consequently, the stocks for the companies are deemed to be stationary. Table 5: Augmented Dickey fuller (ADF) test (PKR/USD) Augmented Dickey fuller (ADF) test (PKR/USD) PKR/ USD exchange rate X= Exchange rate Critical values of 1%, 5%, 10% at Level ?X= 1st difference Critical values at 1%, 5%, 10% for ?X KSE 100 Index -1.906112 -4.036983 -3.448021 -3.149135 -9.118322 -3.486551 -2.886074 -2.579931 Arif Habib corporation limited -2.215052 -4.039075 -3.449020 -3.149720 -3.514613 -2.585050 -1.943612 -1.614897 Bank Alfalah limited -2.024460 -4.052411 -3.455376 -3.153438 -7.841948 -3.497029 -2.890623 -2.582353 Bata Pakistan limited -1.905737 -4.036983 -3.448021 -3.149135 -5.643942 -3.487046 -2.886290 -2.580046 Bank of Khyber -2.088773 -2.594946 -1.945024 -1.614050 -3.233059 -2.594946 -1.945024 -1.614050 Engro corporation limited -2.229998 -4.039075 -3.449020 -3.149720 -3.473091 -2.585050 -1.943612 -1.614897 Habib metropolitan bank -2.388208 -2.597476 -1.945389 -1.613838 -5.400791 -3.524233 -2.902358 -2.588587 Indus motor company limited -2.218464 -4.039075 -3.449020 -3.149720 -8.563734 -3.486551 -2.886074 -2.579931 Mari petroleum company limited -2.127371 -4.041280 -3.450073 -3.150336 -3.324642 -2.585587 -1.943688 -1.614850 Nishat power limited -1.119396 -4.219126 -3.533083 -3.198312 -5.605217 -3.621023 -2.943427 -2.610263 Oil and gas development company -2.880007 -2.586350 -1.943796 -1.614784 -9.814658 -3.491928 -2.888411 -2.581176 Table 5 shows the second results of the ADF test on the Pakistani stock exchange and ten companies. According to the findings of the study, the ADF test results for the KSE 100 index show that the KSE statistic value is less than the critical values at the 1%, 5%, and 10%. Therefore, the null hypothesis of the unit root for the index is accepted, the value contains a unit root and the KSE index is, therefore, non-stationary. At the 1st difference level the index is stationary at 1%, 5%, and 10%. The results for Arif Habib Corporation Limited, Bank of Alfalah Limited, Bata Pakistan Limited, Engro Corporation, Indus Motor Limited, Mari Petroleum Company, and Nishat Power Limited are all similar to those of the KSE 100 index, at the fist difference (?S), the statistic value of the index for the company is lower than the critical values at 1%, 5%, and 10%. This, therefore, indicates that the null hypotheses for all the cases should be rejected and that their indexes do not contain a root at the first difference. Consequently, the stocks for the companies are deemed to be stationary. Regression Analysis Table 6 below shows the regression analysis results of the KSE 100 index and the 10 companies in Pakistan included in this study. Just like in the case with the UK, the regression analysis was carried out with Foreign Exchange Rate (EX) as the independent variable with the dependent variable being Stock return. Table 6: Summary of regression analysis results (UK) Regression Statistics Bank Alfalah Bata Pakistan Bank of Khyber Engro Corp. Habib Met. Bank Indus Motor Mari Petrol Nishat Power Oil And Gas DC KSE 100 Multiple R 0.1156 0.0221 0.05307 0.204215 0.0076 0.22239 0.298145 0.055866 0.21605 0.37128 R Square 0.0134 0.0005 0.00282 0.041704 5.7805 0.04946 0.08889 0.003121 0.046678 0.13785 Adjusted R Square 0.0034 -0.0080 -0.00997 0.033513 -0.01403 0.04133 0.080898 -0.02457 0.037768 0.13048 Standard Error 5.5364 95.114 1.25158 18.31187 6.744487 23.9611 26.21945 1.178748 11.3596 676.274 Observations 101 119 80 119 73 119 116 38 109 119 Significance L. (95%) 0.2498 0.8113 0.6401 0.0259 0.9491 0.0151 0.0012 0.7390 0.0240 3.2234 The regression analysis results of the KSE index in the table reveal that the regression model used in this study explains 13.8% of effects of foreign exchange rate (Independent Variable) on Stock Return (Dependent Variable) in Pakistan. This means that the remaining 86.2% of effects on stock return in the country can be explained by other variables (In this case extraneous variables). The level of significance is 3.2234 at the 95% confidence level, which is greater than 0.05 indicating that the results are, therefore, not significant in explaining the relationship between foreign exchange and the Stock return for the KSE index. In the case of Bank of Alfalah, R squared is 0.0134. This means that the model can explains 1.34% of the effects of foreign exchange on Stock return. Extraneous variables, therefore, affect more than 98% of the observable trends in stock return. The result is not significant at the 95% level of confidence (Significance=0.2498). Bata Pakistan registered an R squared value of 0.0005, this means that the model accounts for about 0.1% of the effects of foreign exchange on stock return for the company. The results are not significant at the 95% confidence level (Significance=0.8113). For Bank of Khyber, R squared is 0.0028. This means that the model can explain 0.3% of the effects of foreign exchange on Stock return. Extraneous variables, therefore, affect more than 99% of the observable trends in stock return. The result is not significant at the 95% level of confidence (Significance=0.6401). On the other hand, Engro Corporation registered an R squared value of 0.0417, which means that the model accounts for about 4.17% of the effects of foreign exchange on stock return for the company. The results are significant at the 95% confidence level (Significance=0.026). For Habib Metropolitan Bank, our model accounts for about 57.9% of the effects of foreign exchange on Stock return. Extraneous variables, therefore, affect 42.1% of the observable trends in stock return. The result is not significant at the 95% level of confidence (Significance=0.9491). Indus Motors registered an R squared value of 0.04946. This means that the model accounts for about 4.9% of the effects of foreign exchange on stock return for the company. The results are significant at the 95% confidence level (Significance=0.0151). For Mari Petrol, our model accounts for about 8.9% of the effects of foreign exchange on Stock return. Extraneous variables, therefore, affect 91.1% of the observable trends in stock return. The result is significant at the 95% level of confidence (Significance=0.0012). In the case of Nishat Power, our model accounts for about 0.31% of the effects of foreign exchange on Stock return. Extraneous variables therefore affect 99% of the observable trends in stock return. The result is not significant at the 95% level of confidence (Significance=0.7390). Finally, in the case of Oil and Gas Development Company, the model accounts for about 4.7% of the effects of foreign exchange on Stock return, the result is significant at the 95% level of confidence (Significance=0.0240). 5.4 Analysis The findings of this study were aimed at achieving three major objectives. The analysis of results is important for the achievement of all three objectives, but conclusively covers objective one and two. The findings outlined in this chapter are of great importance towards establishing the impact of foreign exchange rate on the returns of the FTSE 100 index of the UK and the KSE 100 index of Pakistan. The findings have also helped the study to know the impact of foreign exchange rate on big companies in the UK and Pakistan in terms of the values of their stock. While regression helps to establish whether or not there is a relationship between foreign exchange and stock values both for the country indexes and individual companies, the Augmented Dickey Fuller Test helps to establish the usefulness of the regression results (Amare and Mohsin, 2000). The result of regression may be spurious if it is assumed that the time series data being tested is stationary when it is not. It has been established that spurious correlation has great likelihood of existing in developing markets like Pakistan because each of the nominal variables adjusted for inflation may have large inflationary components in them. Spurious correlation between independent variables is known to inflate values of R2 and t-statistics, thus giving wrong regression results. Study of the effects of exchange rate between the USD and the GBP on the stock market in the UK provides a number of important findings. First the FTSE 100 index was found to be non-stationary meaning that the regression results are not spurious and thus useful. However, the regression results reveal that the exchange rate with the USD accounts for only 0.2% of the variability in the FTSE 100 index, this leaves over 99.8% to factors outside the model. The regression results are not significant and, therefore, it means that even though they are not spurious the relationship does not explain anything important. This trend can be observed in a number of other results involving the companies. While the ADF tests at level for all the companies indicate that some of their stock return values are stationary while some are not, the overall observation for all the companies from the UK included in the study is that at 1st difference level all the company stock values are stationary. This means that testing this data for relationship between exchange rate and stock values is valid as there is no likelihood of getting spurious regression results. The assumption, therefore, is that all the regression results for the companies are correct and that the conclusions made on their basis are valid. Consequently, it has been established in the findings from regression that the relationship between exchange rate (USD/GBP) and the return on stock values of the companies participating in the stock exchange is quite small. In all the ten companies included in the study, the regression model explains very small proportions of variability in the stock values; Antofagasta (0.15%), Babdock (0.4%), HSBC (1.2%), Imperial Tobacco (0.13%), John Matthey (0.2%), Meggit (0.26%), Morrison (2.1%), Pearson (1.4%), Standard Chartered (0.1%), and Weir Group (0.1%). This implies that the fluctuations in stock prices of the company stocks in the market are not significantly affected by the exchange rate. All the regression findings were found to be insignificant at the 95% confidence level. Just like in the case with the UK market, while the ADF tests at level for all the companies indicate that some of their stock return values are stationary while some are not, the overall observation for all the companies from Pakistan included in the study is that at 1st difference level all the company stock values are stationary meaning that the regression results for the market are valid and there was no likelihood of getting spurious regression results. The regression model reveals that the exchange rate accounts for 13.8% of the variability in the KSE index of Pakistan, however, the level of significance (3.2234) indicates that this finding is not significant in explaining the relationship between foreign exchange and the Stock return for the KSE index. As for the companies examined in Pakistan, results indicate that the rate of exchange has very little effect on the returns on their stock. All of the findings were found not to be significant at the 95% confidence level used apart from those of four companies, which were significant. These included Engro, Indus Motors, Mari Petrol, and Gas Development Company. The model results indicated that for Engro Company, 4.17% of stock value variability was affected by the exchange rate with the USD. For Indus Motors, the exchange rate accounted for 4.9% of the variability in stock values while for Mari Petrol it accounted for 8.9% of the variability. Finally, for Gas Development Company about 4.7% of stock value variability was affected by the exchange rate with the USD. The main implication of these results is that the impact of foreign exchange rate on the returns of the FTSE 100 index of the UK is very small because overall there is no significant relationship between the two. The relationships established through regression are almost negligible. On the other hand, although regression shows that the exchange rate with the USD accounts for about 13.8% of variability in the KSE 100 index of Pakistan, the result is statistically insignificant. The findings also indicate that the impact of foreign exchange rate on big companies in the UK is very small as their stock is not significantly affected by the exchange rate. For the case of Pakistan, it is evident that although most of the companies examined were not affected significantly by the exchange rate, it significantly affected the values of the stock of four companies. This accounts for 40% of all the companies studied in Pakistan and it therefore indicates that the exchange rate may have a significant impact overall on stocks of Pakistani companies in the stock market. The study basically aimed to compare the effects of exchange rate on the stock index and company stocks of developed market and developing market, in this case the UK and Pakistan respectively. This study reveals that although significant relationship has not been established between exchange rate and the KSE index in Pakistan, company stocks are affected by exchange rates. However, companies do not suffer any such effect in the UK. References Amare, M. and Mohsin, A., 2000. Stock Prices and Exchange Rates in Leading Asian Economies: Short Versus Long Run Dynamics. Singapore Economic Review, 45, pp.165-181 Dickey, D. and Fuller W. 1981. Likelihood ratios statistics for autoregressive time series with a unit root’, Econometrica, 49, pp.1057-1072 Read More