Interactions between Exchange Rate and Financial Performance Indicators in Nigeria Beer Industry : Evidence from Nigeria Breweries Plc

The study aims at evaluating the nature and magnitude of the interactions between foreign exchange rate and selected financial performance indicators in Nigeria beer industry from 2000 to 2013. Granger Causality procedure, 2-step cointegration and error correction model of Engle and Granger and correlation approach was adopted in the analysis. The Augmented Dickey Fuller (ADF), Phillips-Perron (PP) and the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) procedures were applied to test for series stationary. All the series were found non stationary but achieved stationary at second difference. Earnings Per Share has a short term negative and insignificant effect while prices of equity shares, net asset value per share and price-earnings ratio are positively and insignificantly related to foreign exchange rate in the short run. All the variables shared negative and insignificant long run relationship with exchange rate. There is a unidirectional causality running from earnings per share to exchange rate and from exchange fluctuations rate to net asset value per share. The implication of the findings is that foreign exchange rate adversely affects earnings of firms in Nigeria brewery industry. Growth in earnings of firms in Nigeria brewery industry can cumulatively contribute to lowering the rate at which naira exchanges for other foreign currencies.


Introduction
The need for policy makers at both micro and macroeconomic levels to have sufficient information as to the interactions of foreign exchange rate and key financial performance indicators is very crucial especially in the brewery sector.Citing Dornbusch and Fischer (1980), Nath and Samanta (2003) submits that under the goods market approaches, movements in foreign exchange rates influences the competitiveness of a firm, the value of the earnings, it's transaction exposure and cost of its funds as many companies borrow in foreign currencies to fund their operations.In support, Phylaktis and Ravazzolo (2005) argues that aside the economic exposure, which arises from variations in firm's discounted cash flows anytime foreign exchange rates fluctuate, companies also face transaction exposure due to gains or losses emanating from settlement of investment transactions stated in foreign currency terms.
Nigerian naira depreciated drastically against the united states' dollar from 2006 downwards as a result of decrease in foreign exchange inflows due to remarkable decrease in the demand for oil (the country's major source of foreign reserve); reduced inflow from independent sources; increased repatriation of investments by foreign investors and negative accretion to foreign reserve (Zubair, 2013).Citing Zakaree(2012), Inyiama and Ekwe(2014) opine that the growth and development of the economy is largely determined by sound fiscal and monetary policies which by extension influences international trade and investment as well as the growth of public and private sector of the economy.
However, Okwo and Ugwunta (2012) while citing Ogbadu (2000) argues that one of the major problems facing manufacturing companies in Nigeria, including the brewery sector, is the growing trend of high input costs which erodes their bottom line and leads to intermittent interruptions in their operational capacity.This trend was observed to have led to the closing down of production facilities by many of the brewery firms in Nigeria since they could no longer break even.Glauben and Loy (2008) argue that variations of currency values as a consequence of fluctuations in foreign exchange rate, work as cost or profit shifters to expose the pricing behavior of the beer producing firm.
Khan, Akhatar and Rana (2002) emphasized that exchange rate exerts strong influence on the balance of trade, production levels of manufacturing firms, mode of resource allocation to key sectors, as it responds to exogenous factors and shocks triggered by monetary, fiscal and other policies.Citing Dornbusch and Fischer (1980), Phylaktis and Ravazzolo (2005) explained that the flow oriented models of exchange rate determination affirm that currency movements affect international competitiveness and balance of trade position and as a result, the real output of the economy and consequently the current and future cash flows of companies; which could be measured by financial performance indicators.
This study aims at examining the interactions between foreign exchange rate and selected financial performance indicators such as earnings per share, price/earnings ratio and net asset value per share using Nigeria Breweries Plc as a test case.The rest of the paper is organized into four distinct sections: Section 2, reviews related literature; Section 3 establishes the methodology for data analysis; section 4 presents and discusses the empirical results while section 5 summarizes and finally concludes.

Review of Related Literature
Euromonitor (2014) observed that new lifestyle trends are globally encouraging a broad shift away from heavy drinking of beer and associated products as health consciousness increases.Fashion trends has also encouraged alternative drinks, such as red wine, Champaign and bottled water.In Nigeria, the slogan 'if you drink, don't drive and if you drive, don't drink' is actually stepping down the rate of beer intake.However, despite all these negative campaigns, the beer industry in Nigeria, more than ever before is witnessing a boom having contributed about 28 percent of Manufactured Value Added (MVA) and providing direct employment for over 30,000 persons and indirect employment close to 300,000 Nigerians and expatriates as argued by Okwo and Ugwunta (2012) while citing Ola (2001).
The brewery industry in Nigeria is highly competitive, mostly in terms of quality and price of beer brands.This could be the reason why firms in the industry are always reluctant in increasing the price of their products even as supply and demand for their products fluctuate but rather stick to a prevailing price for a long time.Goldberg and Hellerstein (2007) submits that the managerial time to determine a new optimal price, the cost of printing new price tags and of advertising a new price as well as the risk of losing long-term customers when the price increases are reasons why firms stick to their prices even when the bottom line is being eroded.However, before the banning of barley importation by Federal Government of Nigeria in 1988, some firms in Nigeria brewery sector import the product from independent maltsters from the European Union, Australia and Canada.
The relationship between equity returns and currency exposure for a sample of U.S., U.K. and Japanese banks and insurance firms was studied by Gounopoulos, Molyneuk, Staikouras, Wilson and Zhao(2013) using the VAR-BEKK methodology.The result indicates that U.S. (Japanese) banking sector equity returns are negatively related to changes in foreign currency value while equity returns of U.S./U.K. insurers are negatively linked to changes in the value of Japanese Yen.Overall, the negative relationship between the foreign currency value and bank/insurance equity returns supports the "flight to quality" hypothesis from the U.S./U.K. to Japan.Electronic copy available at: http://ssrn.com/abstract=1971540 The level of foreign exchange exposure and its determinants in Indian firms was examined by Kanagaraj and Ekta Sikarwar (2011).It was found that only 16 percent of the firms are exposed to exchange rate exposure at 10 percent level of significance.About 86 percent of the firms are negatively affected by an appreciation of the rupee which confirms that Indian firms are net exporters.On the determinants of exchange rate exposure, the study reveals that export ratio is positively and hedging activity is negatively related to the exchange rate exposure of pure exporter firms.
Using a simultaneous equation model within a fully specified macroeconomic model, and a vector-autoregressive model, Eme and Olugboyega (2012) analysed the impact of exchange rate on macroeconomic aggregates in Nigeria.The research examines the possible direct and indirect relationship between the real exchange rates and GDP growth.The results reveal that there is no evidence of a strong direct relationship between changes in the exchange rate and GDP growth but rather, Nigeria's economic growth has been directly affected by fiscal and monetary policies.Doidge, Griffin and Williamson (2002) re-examines the nature and the economic significance of the exchange rate to firm value relation and found that firms' foreign activity is broadly and significantly related to exchange rate exposure and that after controlling for this activity, large firms are more sensitive to currency movements than small firms.The study further reveals that firms with high international sales outperform those with no international sales during periods of large currency depreciations.However, exchange rate movements have an economically significant impact on firm value in ways that are consistent with theory.
Sensitivity of foreign exchange exposure and its determinants was studied by Agyei-Ampomaha, Mazouzb and Yin (2012) using different estimation methods.The study reveals that the determinants of currency exposure are modeldependent while the cross-sectional results suggest very little or no relationship between firm-specific factors and currency exposure.Zakaria (2013) in a related study examines the impact of exchange rate volatility on trade using regression analysis of standard export demand models while exchange rate volatilities were measured by GARCH(1,1) models.The results from regression analysis show that Malaysian exports to the US and Japan are significantly related with exchange rates volatility.The impact of exchange rate volatility on Malaysia export to US was found negative; while for Japan, it's positive.Malaysia's export to the UK and Singapore was found not significantly related to the volatility in the exchange rates.The findings from this study clearly indicate that the relationship between export performance and exchange rates volatility is ambiguous.
AR(k)-EGARCH(p,q) models was employed by Jamil, Streissler and Kunst(2012) as they explore the impact of exchange rate volatility on industrial production before and after the introduction of common currency for eleven European countries included in European Monetary Union and for four European countries that did not adopt 'Euro' as common currency.They conclude that all the countries enjoyed benefits after the introduction of common currency by reduction in negative impacts of real exchange rate volatility.
Huchet-Bourdon and Korinek ( 2011) examines the impact of exchange rates and their volatility on trade flows in China, the Euro area and the United States in agriculture, manufacturing and mining.It finds that exchange volatility impacts trade flows only slightly.Exchange rate levels, on the other hand, affect trade in both agriculture and manufacturing and mining sectors but do not explain in their entirety the trade imbalances in the three countries examined.

Methodology
The study adopts the Ferson and Harvey (1998) Asset Pricing Model as the theoretical Framework.The Engle-Granger (1987) two-step error correction model procedure discussed in Rao (2005); cited and adopted in Abraham ( 2013) is employed for model estimation.The models are as specified below: Where: represents the first difference computation on the respective variables; a1, b1, c1, and d1 denote the coefficients indicating the short run equilibrium relationship linking the variables; a2, b2, c2, and d2 denote the coefficient indicating the long run relationship linking the variables with a priori expectation of -1; Ut-1 or RES-1 is the residual obtained from the linear regression of variables integrated in order 1(1).The residual is lagged by one to fulfill the requirement of the granger representation theorem.
t, vt, ut and et are the disturbance term for the models.In Granger representation theorem as opine by Engle and Granger (1987) and cited in Abraham(2013), cointegration was referred to as a linear combination of two non-stationary variables which are integrated of the same order.Stationary of time series data has to be achieved to avoid spurious regression results, Time series data is said to be stationary if its mean and variance are constant over time, and the value of the covariance between the two time periods depends only on the distance k (lag) between the two time periods and not the actual time t itself and that the expected value of the time series should, therefore be constant and finite (Baumohl and Lyocsa, 2009).Therefore, the variables under study are first tested to determine their order of integration through unit root test applying the Augmented Dickey Fuller (ADF), Phillips-Perron (PP) and the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) procedures.The view by some researchers that Augmented Dickey Fuller test have a low power in differentiating a situation of stationary and near stationary, gave rise to Phillips-Perron (PP) test while Kwiatkowski-Phillips-Schmidt-Shin (KPSS) was a check for robustness.If the variables are integrated of the same order, there will be likelihood of a cointegration but if the integration is of different orders, the variables may not be cointegrated.If cointegration is established among the variables, then the second step of Engle and Granger (1987) procedure could be specified as an Error Correction Model (ECM) representing each equation in the study.The procedure demands that two variables only will be considered at a time.
However, Johansen (1988) and Johansen and Juselius (1990) developed the maximum likelihood procedure for multivariate cointegration testing.The two basic test statistics in this procedure are the trace test and the maximal eigenvalue test.This procedure is only applied when there are more than two variables and these characteristics make it more robust for such models with above two variables only.The Engle and Granger two -step procedure is adopted for the study.

Unit Root Test
The Augmented Dickey Fuller (ADF), Phillips-Perron (PP) and the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) procedures are applied in testing for existence of unit root and by extension, the order of integration of the variables.This will help to determine whether the variables are likely to be cointegrated.The graphs below were plotted to give the researcher initial idea as to the presence of unit root in the time series data.The graphs in Figure 1, clearly show that the variables have unit root as the line graphs failed to severally cross the zero-line and with remarkable departures from it.Source: Researcher's EView 8.0 Computation Tables 2, 3 and 4 above reveal that exchange rate, earnings per share, net asset value per share and share price time series data under the Augmented Dickey Fuller (ADF), Phillips-Perron (PP) and the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) procedure, achieved stationary at second difference.The Augmented Dickey Fuller (ADF) and Phillips-Perron (PP) tests also confirm that the time series data for price-earnings ratio are also stationary at second difference.When time series data are integrated of the same order, the data tend to cointegrate (Engle and Granger, 1985).They document that when two time series are integrated of the same order and some linear combinations of them are stationary, then the two series are cointegrated.
• Cointegrated series share a stochastic component and a long term equilibrium relationship.
• Deviations from this equilibrium relationship as a result of shocks will be corrected over time.
• We can think of Yt as responding to shocks to X over the short and long term.Therefore the outcome of the unit root tests resulted in the generation of data series, free from unit root as shown in the graphs below:   Where; p is the maximum number of lagged observations included in the equation, the matrix A contains the coefficients of the equation (i.e., the contributions of each lagged observation to the predicted values of X1(t) and X2(t) , X1 is the share price which is constant while X2 takes the form of various financial performance indicators identified above and, E1 and E2 are residuals (prediction errors) for each time series.On causalities as shown in Table 6, there is a unidirectional causality running from earnings per share to exchange rate and another unidirectional causality running from exchange rate to net asset value per share.The implication is that earnings per share granger causes exchange rate at 10% level of significance while exchange rate granger causes net asset value at 5% level of significance.Source: EView 8.0 Computation Table 7 reveals that the variables are co-integrated at 10 percent significance level.According to the Granger Representation Theorem, when the variables under study are integrated in the same order and are found to be cointegrated, an error correction model can then be estimated as in equations 1 to 4 above.The output of the regression analysis is presented below.Source: EView 8.0 Computation Table 9 indicates that foreign exchange rate exerts a positive influence on net asset value per share; though not to a significant extent.The long term coefficient reveals a negative and also insignificant relationship.The error correction mechanism in this case suggests that the deviations from equilibrium and the distortions influencing foreign exchange rate could be corrected in approximately 53% per annum by improving consistently on net asset value per share.This implies that it could take approximately one year and eleven months (approximately 23 months) to readjust to equilibrium and attend to the distortions if net asset value per share is properly managed.The insignificant association places doubt as to the ability of the error correction mechanism to correct the deviations in the long term.10 reveals that foreign exchange rate is positively but not significantly with price-to-earnings ratio in the short run.However, the long term coefficient shows a negative and insignificant relationship between exchange rate and priceearnings ratio.In this case, the distortions influencing the rate of foreign exchange and the error correction mechanism suggest, through the long term coefficient, that approximately 68% of the distortions, all things being equal, could be corrected in a year.Though the long term association is not significant, it implies that in about 18 months (a year and 6 months), the distortions and deviations from equilibrium could be corrected by the error correction mechanism.Source: EView 8.0 Computation Output.
Table 12, reveals a positive correlation between net asset value per share, price-earnings ratio, share price and exchange rate.A negative correlation is found between exchange rate and earnings per share.However, amongst all the variables under study, only price-earnings ratio indicates a significant 23% strength of correlation.

Summary and Conclusion
The study aims at determining the extent of association and causalities amongst the variables under study.The researcher applied the 2-step cointegration and error correction model of Engle and Granger (1987) in a simple regression framework.Earnings Per Share has a short term negative and insignificant effect on exchange rate while the long run coefficient show a negative and significant influence at 10 percent level of significance.It was found that foreign exchange rate exerts a positive influence on net asset value per share; though not to a significant extent while the long term coefficient reveals a negative and also insignificant relationship.In the case of price-earnings ratio, it shares a positive but insignificant relationship with foreign exchange rate in the short run while its long term coefficient shows a negative and insignificant relationship.Prices of equity shares are positively and insignificantly related to foreign exchange rate in the short run while in the long run, share price and exchange rate relate negatively and insignificantly too.
On causalities, there is a unidirectional causality running from earnings per share to exchange rate and another unidirectional causality running from exchange rate to net asset value per share.The implication is that earnings per share granger causes exchange rate at 10% level of significance while exchange rate granger causes net asset value per 351 share at 5% level of significance.The long term coefficient of all the variables are negative and this is in tandem with our earlier (a priori ) expectations that in the long run, EPS, NAVPS, P/E Ratio, and Share Price will relate negatively with foreign exchange rate.Earnings prove to have a stronger long run impact on foreign exchange rate.The implication of the negative short run and significant long term relationship between earnings and exchange rate in Nigeria brewery industry is that foreign exchange rate could be maintained at a low rate if there is a consistent growth in corporate earnings.It can also be argued that since they share an inverse relationship, a rise in the foreign exchange rate may adversely affect earnings of firms in Nigeria brewery industry as a result of high demand of foreign exchange in the sector.

Figure 1 :
Figure 1: Graphical Representation of the Variables Source: Author's EView 8.0 Output.The Augmented Dickey Fuller (ADF), Phillips-Perron (PP) and the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) procedures are therefore applied to determine the order of their integration or stationary status.

Figure 3
below reveals that the line graph of the fitted observations is very close to the graph of the corresponding observed values.

Figure 3 :
Figure 3: Residual graph of the parsimonious model Source: EViews 8.0 Output

Table 1 :
Description of Variables

Table 2 :
Augmented Dickey Fuller (ADF) Unit Root Test Results

Table 5
Inyiama(2013) statistics of the study.It shows the mean values of the data used for the study as well as deviations from the mean.It indicates the maximum and minimum values for the time series data under consideration.The coefficient of skewness for exchange rate, earnings per share and net asset value per share have values far below unity (1) while price-earnings ratio and share prices have skewness above 1.Therefore, a normal frequency distribution is suggested in the case of exchange rate, earnings per share and net asset value per share while that of price-earnings ratio and share price reveal that the frequency is not normally distributed.Kurtosis coefficient is 1.937015, 1.690474 and 1.951168 for exchange rate, earnings per share and net asset value per share respectively while the values are 5.748259 and 3.776167 for price-earnings ratio and share price respectively.Jarque-Bera statistic shows that price-earnings ratio and share price indicate significant p-value of 0.003789 and 0.077196.Both Kurtosis and Jarque-Bera statistic confirm the data series mode of distribution.Exchange rate, price-earnings ratio and share price show volatile standard deviations.Granger-causality test is conducted in the context of linear regression models and specified in bivariate linear autoregressive model of two variables X1 and X2 based on lagged values as applied byPasquale (2006)and cited inInyiama(2013):

Table 7 :
Residual Test for Stationary

Table 8 :
Regression Results

Table 8
reveals that Earnings Per Share has a short term negative and insignificant effects on exchange rate while the long run coefficient show a negative and significant influence at 10 percent level of significance.The error correction mechanism suggests that deviations from equilibrium are corrected at approximately 82% per annum.This implies that the distortions affecting foreign exchange rate in the long run could be corrected in approximately one year and three months ( approximately 15 months) if consistent efforts are made to continually improve on earnings per share.

Table 10 :
Regression Results

Table 11
documents that share prices of equity shares are positively and insignificantly related to foreign exchange rate in the short run.The long run coefficient of the relationship between share price and exchange rate reveals a negative and insignificant relationship.It further reveals that about 74% of the deviations and distortions from equilibrium of exchange rate could be readjusted by the error correction mechanism in an accounting year.This implies that the entire deviations could be corrected in about 16 months (a year and 4 months) except for the insignificant long term coefficient which could impair the return to equilibrium.

Table 12 :
Correlation Results