International Journal of Empirical Finance Vol. 2, No. 2, 2014, 75-82

Stock Market Return Volatility and Macroeconomic Variables in Nigeria Emenike Kalu O.1, Odili Okwuchukwu2 Abstract This paper examines the impact of macroeconomic variables on stock market return volatility in Nigeria using GARCH-X model. Five macroeconomic variables: broad money supply, consumer price index, credit to the private sector, US dollar/ Naira exchange rate, and the net foreign assets, were included in the conditional variance model of the Nigerian Stock Exchange (NSE) All-share Index from January 1996 to March 2013. Results of the GARCH-X model suggest that the NSE return volatility is positively influenced by changes US dollar/ Naira exchange rates and credit to private sector but negatively influenced by changes broad money supply and inflation. On the other hand, changes in net foreign assets shows negative but not significant influence on changes in stock market return volatility. The key implication is for investors to adjust their portfolio to changes in these macroeconomic variables. Key words: Stock return, macroeconomic variables, volatility, GARCH-X model, Nigeria JEL Classification: G11, E44, C22 1. Introduction The relationship between stock market return volatility and macroeconomic variables has engaged the attention of academics, stock market professionals and stock market regulators for so many years. This is because of the strong link between the macroeconomy and the stock market. Engle and Paton (2001), for instance, posit that financial assets prices do not evolve independently of the market around them. Chowdhury, Mollik and Akhter (2006) observe that any shock in macroeconomic variables will present a source of systematic risk which will affect any market portfolio, irrespective of how well diversified the portfolio is. Many empirical studies have being conducted to examine this relationship in both developed and emerging stock markets. In a seminal paper, Schwert (1989) ascribes the changes of the returns volatility to the macroeconomic variables and posits that bond returns, short term interest rate, producer prices or industrial production growth rate have incremental information for monthly market volatility. Whitelaw (1994) finds a strong positive relationship between the commercial paper-Treasury yield spread and stock market volatility. Other studies have also found evidence to show that macroeconomic variables relate with stock market returns volatility (see for example, Perez and Timmermann, 2000; Al-Raimony and El-Nader, 2012; Okoli, 2012). These studies thus, imply that macroeconomic variables contain relevant information about stock market return volatility. Since the stock market is a veritable source of long-term capital, identifying the macroeconomic variables that influence stock market return volatility and the nature of influence will help in correctly pricing stocks and managing associated risks. The effectiveness of macroeconomic policies should therefore be anchored on the potency of their influence on reducing risk in the stock market as engine of long-term capital formation. Evidence of the relationship between macroeconomic variables and stock market return volatility therefore, is important for investors, academics, stock market professionals and stock market regulators. It will not only help to develop a better understanding of potential macroeconomic determinants 1 2

Department of Banking and Finance Rhema University Aba, Abia State, Nigeria Department of Banking and Finance Michael Okpara University of Agriculture Umudike, Abia State, Nigeria

© 2014 Research Academy of Social Sciences http://www.rassweb.com

75

K. O. Emenike & O. Okwuchukwu of systematic financial sector risk but may also help to refine theories of stock pricing and to forecast stock market volatility thereby efficiently monitor and manage financial risks. Although several studies have examined the relationship between stock market return volatility and macroeconomic variables in both developed and emerging stock markets, there is a very scant evidence for Nigeria. The few authors in the case of Nigeria concentrate their studies on one or at most two macroeconomic variable(s) to explain this relationship. While Yaya and Shittu (2010), for instance, examined the impact of exchange rate and inflation on stock market return volatility, Aliyu (2012) studied the effect of inflation on volatility of stock market returns. Other studies that investigate the impact of macroeconomic variables on stock market volatility include, amongst others, Oseni and Nwosa (2011). In order to close this lacuna, this paper contributes to existing literature on the impact of macroeconomic variables on stock market return volatility by applying GARCH-X model to recent monthly macroeconomic and stock market data. The major objective is to identify the macroeconomic variables that influence stock market return volatility and the nature of the influence on the stock market. The paper also adds variables, such as credit to private sector, broad money supply and net foreign assets, whose impact on volatility have not been previously examined in Nigeria. The remainder of this paper is organised as follows. Section 2 presents brief review of empirical literature. Section 3 provides methodology and data. Section 4 presents results and discussions. Section 5 provides the conclusion. 2. Brief Review of Empirical Literature One of the earliest attempts at examining the impact of macroeconomic variable on stock return volatility is Schwert (1989) who obtains three important empirical results in an attempt to find why stock market volatility change overtime. First, he finds evidence of a positive linkage between macroeconomic volatility and stock market volatility, with the direction of causality being stronger from the stock market to the macroeconomic variables. Secondly, he shows evidence of stock market uncertainty being higher during recessions than expansions. This result may be explained through an operating leverage effect, i.e. profits would tend to fall more rapidly than revenues during recessions if fixed costs are large. Thirdly, he finds that the level of macroeconomic volatility explains less than half of the volatility of stock returns. Similarly, Whitelaw (1994) find a strong positive relationship between the commercial paper-Treasury yield spread and stock market volatility, with the former leading and affecting negatively output as well. Since the commercial paper-Treasury yield spread is a proxy for the stance of monetary policy, the negative linkage between the business cycle and stock market volatility, appears to be caused by monetary policy, with monetary tightening leading to both a recession and an increase in stock market volatility. Morelli (2002) finds that conditional macroeconomic variables volatility does not explain the conditional stock market volatility. In emerging stock market studies, Chowdhury, Mollik and Akhter (2006) employed both the GARCH and VAR models, and showed that a significant unidirectional causality exists, namely from industrial production volatility to market return volatility and from market return volatility to inflation volatility. Chinzara (2010) investigates, among others, how systematic risk emanating from the macro-economy is transmitted into stock market volatility using augmented autoregressive GARCH (AR-GARCH) and Vector Autoregression models for the 1997-98 Asian and the 2007-2008 sub-prime financial crises, The findings show, among others, that macroeconomic uncertainty significantly influences stock market volatility and that financial crises increase volatility in the stock market and in most macroeconomic variables and, by so doing, strengthen the effects of changes in macroeconomic variables on the stock market. Kadir, Selamat, Masuga, and Taudi (2011) examine the impact of interest rate volatility and exchange rate volatility on Kuala Lumpur Composite Index (KLCI) returns from 1997 to 2009 using GARCH (1,1) Model. Their results show, among others, that the volatility of KLCI is negatively related to interest rates and positively related to exchange rates however both of these relationships are not significant. They opined therefore, that exchange rates and interest rates cannot be used to predict the volatility of the market. Al-Raimony and El-Nader (2012) examine the impact of macroeconomic variables on the volatility of Amman Stock Exchange (ASE) returns in Jordan by applying GARCH models to monthly data between 1991 and 2010. Their results show that 76

International Journal of Empirical Finance growth in real money supply, growth in consumer price index, real exchange rates, weighted average interest on loans have an adverse impact on the ASE returns volatility, while real gross domestic products played a positive effect. Zakaria and Shamsuddin (2012) examine the relationship between stock market returns volatility in Malaysia with five selected macroeconomic volatilities; GDP, inflation, exchange rate, interest rates, and money supply from January 2000 to June 2012 using GARCH(1,1) and multivariate VAR Granger causality models. They find little support on the existence of the relationship between stock market volatility and macroeconomic volatilities. However, they report that interest rate volatility Granger-cause stock market volatility and that only money supply volatility is significantly related to stock market volatility. In studies using Nigeria data, Yaya and Shittu (2010) provide evidence to show that exchange rate and inflation has significant influence on the volatility of stock returns in Nigeria. Oseni and Nwosa (2011) employ AR(k)-EGARCH (p, q) model to examine the volatility in stock market and macroeconomic variables in Nigeria for the periods 1986 to 2010 using time-series data. Their results reveal existence of bicausal relationship between stock market volatility and real GDP volatility and no causal relationship between stock market volatility and the volatility in interest rate and inflation rate. Aliyu (2012) assesses the impact of inflation on stock market returns and volatility by applying quadratic GARCH model to monthly time series data from Nigeria and Ghana. He finds, amongst other results, that inflation rate and its 3-month average have significant effect on stock market volatility in the two countries. Okoli (2012) evaluates the relationship between stock market and macroeconomic variables and their volatilities using GARCH and VAR methodologies. The study finds, amongst others, that the macroeconomic variables are significant in explaining movements in the stock prices and its volatility. 3. Methodology and Data Methodology To investigate the influence of macroeconomic variables on stock market return volatility in Nigeria, GARCH-X model specification is employed. GARCH-X3 model is an extension of ARCH model developed by Engle (1982) and generalised by Bollerslev (1986). GARCH-X model adds explanatory variable in the conditional variance equation in order to examine directly the impact of the variable on the stock market return volatility. The GARCH-X(1,1) model is specified in the following form: p

Rt    1Rt  i  t i 1

t  t2 Zt ,

(3)

Zt ~ N (0,  t2 )

 t2     1 t21   1 t21   1Mt

(4)

Where Rt is the monthly stock market rate of return,  is the AR (p) term in the mean equation in order to account for the time dependence in returns,  t is the residual term in the mean equation, Zt is the standardized residual sequence of IID random variables with mean zero and variance one, N represents distribution of the stock market returns. In the conditional variance equation (4), ω is the constant variance that correspond to the long run average, α1 refers to a first order ARCH term, β1 is the first order GARCH term, and Mt is represents the macroeconomic variables for month t. The GARCH-X specification in equation (4) involves a direct test of the impact of macroeconomic variables on stock market return volatility.

3

GARCH-X model was first used by Lee (1994) to examine how the short-run disequilibrium affects uncertainty in predicting cointegrated series. Brenner, Harjes and Kroner (1996) used GARCH-X model to model short term interest rates by including the lagged interest rate raised to some power as an explanatory variable in the GARCH conditional variance equation. Hwang and Satchell (2005) also applied GARCHX to model aggregate stock market return volatility by including a measure of the lagged cross-sectional return variation as an explanatory variable in the GARCH conditional variance equation. In all variations of GARCH-X, explanatory variable is added to the conditional variance equation. 77

K. O. Emenike & O. Okwuchukwu The impact is assessed by examining the statistical significance of the coefficient of individual macroeconomic variable (  1 ). Data Monthly data ranging from January 2000 to March 2013 were collected from the Central Bank of Nigeria (CBN) statistics databank. This time period was chosen to capture the effect of fundamental changes made to implement advanced information and communication technology in the operation of the Nigeria stock market. The monthly NSE ASI were used to calculate stock market return (Rt) and the set of macroeconomic variables includes the broad money supply (M2), inflation (CPI), credit to the private sector (CPS), US dollar/ Naira exchange rate (EXr), and the net foreign assets (NFA). The ASI tracks the general market movement of all listed equities on the NSE, including those listed on the Alternative Securities Market (ASeM), regardless of capitalisation. The ASI was converted into monthly returns as follows: Rt = Ln (Pt/Pt-1) X 100 Where: Rt is the daily returns, Pt is closing ASI for Month t, Pt-1 is the previous month closing ASI, and Ln is natural logarithm. 4. Empirical Results and Discussions Descriptive Statistics Figure 1, shows time series plots of log-level and monthly returns of the NSE All-Share Index (ASI) from January 2000 to March 2013. The level of the ASI shows upward trend movement till first quarter of 2008 when it fell below average and started pointing upward from 2012. From figure 1, the very high volatility of stock market return during the global financial crises is obvious. Figure 1 also displays the mean reversion tendency of stock market returns. Figure 24 shows the estimated volatility of the Nigeria stock market return as and the standardized residual series. FIGURE 1: Time Plot of Nigeria Stock Market Index and Returns January 2000 to March 2013

DASI LASI 11.5 40 30 11.0 20 10.5 10 10.0 0 -10 9.5 -20 9.0 -30 -40 8.5 2000 2000

2001 2001

2002 2002

2003 2003

2004 2004

2005 2005

2006 2006

2007 2007

2008 2008

2009 2009

2010 2010

2011 2011

2012

2013

Table 1 summarizes the basic statistical features of the data under consideration including the mean, standard deviation, kurtosis, skewness, and the Jarque-Bera test for the data in their first differences. As Table 1 show, estimates of the standard deviation for the variables under study indicate that stock market return; net foreign assets and broad money supply are more volatile compared to the exchange rate, the inflation and credit to private sector. Furthermore, standard deviation estimates indicate that the US dollar/Naira exchange rate (Ex) and inflation (CPI) are less volatile compared to the rest of the macroeconomic variables during the study period, which perhaps may be due to the Central Bank of Nigeria (CBN) tireless efforts in controlling exchange rate and curtailing inflation in Nigeria. P-values associated with the Jarque-Bera statistics, a test for departure from normality, show that all the variables under study deviates normal distribution. The non-normality of the variables is supported by their skewness and kurtosis 4

Figure 2 is shown in Appendix 1. 78

International Journal of Empirical Finance coefficients. The kurtosis and skewness of a normally distributed series are three and zero respectively. However, all the variables exhibit excess kurtosis and skewed, except the inflation series.

Rt Exr M2 CPS NFA INF

Table 1: Descriptive Statistics of NSE Returns Macroeconomic Variables Mean Std. Dev. Kurtosis Skewness Jarque-Bera 1.115 7.186 6.536 -0.657 292.6 0.294 1.603 38.184 5.023 10263.6 2.012 4.362 8.951 1.791 612.03 2.243 2.419 2.903 1.700 131.62 1.653 6.472 1.927 -0.361 27.912 1.027 1.712 1.627 0.338* 17.481

Note: Marginal Significance Levels displayed as (.). Skewness and Kurtosis are tests for zero skewness and excess kurtosis. Jarque-Bera is for normality.

Unit Root Test A crucial issue in time series modeling is to determine whether or not the variables in question are stationary or not. The augmented Dickey-Fuller (ADF) unit root test was used to examine the log-level of the variables and their first difference. The null hypothesis of the ADF test is that a time series contains a unit root. As shown in Table 2, the calculated values of the ADF test statistics indicate that the level series contain a unit root at the 1% significance level, implying that the level series are non-stationary. However, in the case of the return series, the statistics reject the null hypothesis of a unit root at the 1% significance level, implying that the returns are stationary for the two series.

Level

Table 2: Stationary Tests for Level and First Difference of the Variables Rt Exr M2 CPS NFA INF Test -2.136 -2.473 -2.165 -1.553 -1.016 -3.10 ADF

Ist Diff.

-5.739**

-8.033**

-12.94**

-4.236**

-6.010**

-10.01**

Note: The 5% critical value for ADF test is -3.4 for all the variables.

Results of the Benchmark GARCH (1,1) Model The maximum likelihood estimates for the benchmark GARCH (1,1) model for the NSE return series are presented in Table 3. The coefficients of all the three parameters in the conditional variance equation (ω, α1 and β1) are significant at 95% confidence levels and they all satisfy the non-negativity restrictions of the GARCH model. The significance of α1 indicates that previous period volatility has explanatory power on current volatility. Similarly significance of β1 suggests volatility clustering in the monthly returns of the NSE. Notice also the high persistence in volatility with α1 + β1 = 0.88. High volatility persistence implies that average variance will remain high since increases in conditional variance due to shocks will decay slowly (Rachev et al., 2007: 296).

Rt

Table 3: Results of the Benchmark GARCH (1,1) Model Rt-1 ω α1 β1 Diagnostic Tests 1.640 5.835 0.172 0.708 LM (5) McL(25) WC-Q(25) { 3.485 } { 1.969} { 2.444} { 7.991} 8.043 31.56 33.99 [0.000] [0.048] [0.014] [0.000] [0. 153] [0.082] [0.108]

Notes: Marginal significance level displayed as [.] and t-statistics displayed as {.}.

79

K. O. Emenike & O. Okwuchukwu Results of the GARCH-X Model The results of the GARCH-X (1,1) model are shown in Table 4. Notice that the ω and α1 coefficients are not significant. This suggests that accounting for macroeconomic information reduces the explanatory power of the previous period volatility on current volatility. This is supported by reduction in volatility persistence from 0.88 in the benchmark to 0.71 in the GARCH-X model. Notice also that the β1 coefficient is significant at 99% confidence level and that all the three parameters in the conditional variance equation (ω, α1 and β1) satisfy the non-negativity restrictions of the GARCH model. The GARCH-X results also suggest that changes in US dollar/Naira exchange rates and credit to private sector have positive impact on stock market return volatility whereas changes broad money supply and inflation impact negatively on stock market return volatility. Change in net foreign assets, on the other hand, has negative but statistically insignificant impact on stock market return volatility. These finding supports previous studies that examined the role of macroeconomic information as a prominent driving force of volatility. For instance, Yaya and Shittu (2010) show that exchange rate has positive and significant influence on the volatility of stock returns in Nigeria. The positive coefficient of change in exchange rate indicates that depreciation of exchange rates results in the contemporaneous increase in stock market return volatility. Similarly, Aliyu (2012) reports that inflation rate and its 3-month average have significant effect on stock market volatility in Nigeria and Ghana. The positive relation between stock market volatility and credit to private sector is not surprising given that in a monetised economy, deposit money banks (DMBs) create money through granting of loans and advances. According to Alshogeathri (2011), when DMBs create an excess supply of money, the prices of assets, goods, and services tend to rise. Conversely, when not enough money is created, the prices of assets, goods, and services decrease.

Rt-1

Table 4: Results of GARCH-X Models Rt-1 Exr M2 CPS NFA α0 α1 β1 0.304 7.641 0.077 0.634 6.472 -1.52 3.086 -0.532 {3.592} {1.345} {1.403} {4.04} {2.828} {-2.35} {2.81} {-1.11} [0.000] [0.178] [0.160] [0.000] [0.004] [0.018] [0.004] [0.264]

Heteroscedasticity Serial Correlation Normality

Panel B: Diagnostic Tests Results Test Applied Statistic Lagrange Multiplier (5) 4.341 McLoed-Li(25) 42.44 West-Cho Q(25) 32.71 Ljung-Box Q(8) 7.48 Jarque-Bera 4.0364

CPI -4.914 {-5.25} [0.000]

Signif Lvl. 0.501 0.016 0.138 0.485 0.132

Notes: Marginal significance level displayed as (.) and t-statistics displayed as {.}.

Diagnostic Tests The diagnostic tests calculated to assess robustness of the model estimates are displayed in Table 3 for the benchmark model and panel B of Table 4 for the GARCH-X. As Table 3 shows, West-Cho modified Qstatistic for residuals of the stock market return are not significant, suggesting that there is no serial correlation in the residuals. The ARCH-LM and McLeod-Li results indicate that the null hypotheses of no ARCH effect and no serial correlation in squared residuals are accepted at 1% significance level. Similarly, diagnostic tests results for GARCH-X model show evidence of no serial correlation in both residuals and squared residual as well as no remaining ARCH effect. The benchmark GARCH (1,1) and GARXCH-X models are robust as there appear to no specification error.

80

International Journal of Empirical Finance 5. Conclusions In this paper, we have used the GARCH-X (1.1) approach to analyse the influence of macroeconomic variables on stock market return volatility. Monthly macroeconomic variables including broad money supply, inflation, credit to the private sector, exchange rate, and the net foreign assets were investigated for impact on monthly ASI from January 2000 to March 2013. Descriptive analyses of the NSE log-return series show evidence of a non-normal distribution with an average monthly return of 1.11% and a standard deviation of 7.8%. The results of benchmark GARCH (1,1) model shows evidence of volatility clustering. Results of the GARCH-X model indicates that stock market return volatility is positively influenced by changes US dollar/Naira exchange rates and credit to private sector but negatively influenced by changes broad money supply and inflation. On the other hand, changes in net foreign assets shows negative but not significant influence on changes in stock market return volatility. The key implication of these finding is that investors should adjust their portfolio to changes in these macroeconomic variables so as to reduce stock market volatility and improve stock market returns. References Aliyu, S. U. R. (2012), “Does Inflation have an Impact on Stock Returns and Volatility? Evidence from Nigeria and Ghana”, Applied Financial Economics, 22 (6), 427-435 Al-Raimony, A. D. and El-Nader, H. M. (2012), “The Sources of Stock Market Volatility in Jordan”, International Journal of Economics and Finance; Vol. 4 (11), pp 108-121, doi:10.5539/ijef.v4n11p108 Alshogeathri, M.A. M. (2011). Macroeconomic Determinants of the Stock Market Movements: Empirical Evidence from the Saudi Stock Market. An Abstract of a Doctoral Dissertation Submitted to Department of Economic, Kansas State University, Manhattan Kansas. Bollerslev, T. (1986), “A Generalized Autoregressive Conditional Heteroscedasticity”, Journal of Econometrics, 31, 307-327 Brenner, R. J.; Harjes, R. H. and Kroner, K. F. (1996), “Another Look at Models of the Short-Term Interest Rate”, Journal of Financial and Quantitative Analysis, 31, 85-107 Chinzara, Z. (2010), “Macroeconomic uncertainty and emerging market stock market volatility: The case for South Africa”, Uinversity of Rhodes Working Paper No. 187 Chowdhury, S.; Mollik, A. & Akhter, M. (2006). Does Predicted Macroeconomic Volatility Influence Stock Market Volatility? Evidence from the Bangladesh Capital Market”, Department of Finance and Banking, University of Rajshahi Working Paper, pp.1-15 Engle, R.F. (1982), “Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of the United Kingdom Inflation”, Econometrica, 50,987-1008 Engle, R.F., & Paton A. J. (2001). What Good is a Volatility Model? [On-line] Available: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1296430, (Assessed on 13 October2012). Hasan, A. and Nasir, Z. M. (2008), “Macroeconomic Factors and Equity Prices: An Empirical Investigation by Using ARDL Approach”, The Pakistan Development Review, 47 (4 Part II), pp 501-51 Hwang, S. and Satchell, S. E. (2005), “GARCH Model with Cross-Sectional Volatility: GARCHX”, Applied Financial Economics, 15 (3), 203-216 Kadir, H. B. A., Selamat, Z., Masuga, T. and Taudi, R. (2011), “Predictability Power of Interest Rate and Exchange Rate Volatility on Stock Market Return and Volatility: Evidence from Bursa Malaysia”, International Conference on Economics and Finance Research IPEDR, Vol.4, Singapore. Lee, T. H. (1994), “Spread and Volatility in Spot and Forward Exchange Rates”, Journal of International Money and Finance, 13, 375-383. 81

K. O. Emenike & O. Okwuchukwu Morelli, D. (2002). The Relationship between Conditional Stock Market Volatility and Conditional Macroeconomic Volatility: Empirical Evidence based on UK data. International Review of Financial Analysis, 11, pp. 101-110. Okoli, M. N. (2012), “X-Raying the Impact of Domestic and Global Factors on Stock Return Volatility in the Nigerian Stock Market”, European Scientific Journal, Vol. 8 (12), Pp. 235-250 Oseni , I. O. and Nwosa, P. I. (2011), “Stock Market Volatility and Macroeconomic Variables Volatility in Nigeria: An Exponential GARCH Approach”, Journal of Economics and Sustainable Development, Vol.2 (10), Pp. 28-42. Perez-Quiros, G. and Timmermann, A. (2000), ‘’Firm Size and Cyclical Variations in Stock Returns’’, Journal of Finance, 55, 1229–1262. Schwert, G. W., (1989), “Why Does Stock Market Volatility Change over Time?”, Journal of Finance, 44, 1115-1153. Whitelaw, R. (1994), “Time Variations and Covariations in the Expectation and Volatility of Stock Returns”, Journal of Finance 49, 515–541 Rachev, S. T., Mittnik, S., Fabozzi, F. J., Focardi, S. M. and Jasic, T. (2007), Financial Econometrics: from Basics to Advanced Modeling Techniques, New Jersey: John Wiley & sons Inc. Yaya, O. S. and Shittu, O. I. (2010), “On the Impact of Inflation and Exchange Rate on Conditional Stock Market Volatility: A Reassessment”, American Journal of Scientific and Industrial Research, 1 (2), 115-117. Zakaria, Z. and Shamsuddin, S. (2012), “Empirical Evidence on the Relationship between Stock Market Volatility and Macroeconomics Volatility in Malaysia”, Journal of Business Studies Quarterly, Vol. 4 (2), pp. 61-71

82