Study Of Impact Of Pakistan's Macroeconomic Variables On Volatility Of Stock Returns Of Karachi Stock Exchange Based On Improved GARCH-MIDAS Model

Presently,in the classical econometric research methods,there are two main approaches to study the relationship of macroeconomic variables with the stock market price volatility.The first one employs the same frequency data for carrying out the research process,while the second one employs the low frequency stock market data,which renders the stock market data and macroeconomic variables data to have the same frequency.Usually,handling of time series data having different frequencies requires conversion of high frequency data into low frequency or same-frequency data,which causes loss of valuable information contained in the high-frequency variables.Therefore,the current study employs mixed data sampling(MIDAS)approach involving mixed-frequency data i.e.both low frequency(monthly)and high frequency(daily)data to study the relationship between the macroeconomic explanatory variables and stock market price volatility.Due to problem of data frequency limitation the traditional co-frequency model cannot be used to study the causal relationship between macroeconomic explanatory variables and stock market volatility.Therefore,the classical GARCH-MIDAS(Generalized autoregressive conditional heteroskedasticity-Mixed data sampling)model,proposed by Engle,is integrated into the traditional GARCH model to decompose the stock price volatility into short-term and long-term components.Furthermore,to ensure efficient and effective utilization of data;the long-term volatility is explained through macroeconomic variables.Moreover,the current study differs in this respect,since it expands the classical GARCH-MIDAS model through integration of high frequency data of USD-PKR exchange rate and uses the realized volatility to explain the long-term volatility of the Karachi stock exchange(KSE).Additionally,in the empirical section,several single-factor and multi-factor GARCH-MIDAS models had been established to analyze and estimate stock market volatility of KSE from perspectives of horizontal effect;volatility effect.Furthermore,the current study chooses the money supply and consumer price index as monthly(low frequency)indicators and USD-PKR exchange rate as daily(high frequency)indicator to study KSE stock price volatility.Moreover,based on the improved GARCH-MIDAS model,the empirical results indicate that both the horizontal and volatility values of money supply have significant positive correlation with the KSE(Karachi Stock Exchange)stock price volatility.Additionally,horizontal value of CPI has significant negative correlation,while the volatility value has no significant relationship with KSE stock price volatility.Moreover,both horizontal and volatility values of USD-PKR exchange rate have significant negative correlation with KSE stock price volatility.The empirical results revealed that the estimation results of multi-factor models were roughly identical to those of the single-factor models.However,the multi-factor models involve estimation of a large number of parameters,which could give rise to problems like over-parameterization,and hence render some coefficients values no longer significant.Simultaneously,the forecasting capability analysis reveals that both single-factor and multi-factor models possess strong forecasting capability.Furthermore,forecasting capability of the multi-factor horizontal effect model is superior then that of single-factor horizontal effect model.Moreover,cross comparison between the single-factor mixed effect model and multi-factor mixed effect model revealed that the multi-factor mixed effect model more proficient in explaining the long-term components of the Pakistan's stock market price volatility.Keeping in view the actual economic conditions of Pakistan,several policy recommendations were put forward e.g.effective enactment of regulatory policies(monetary policy aspect,inflation aspect and exchange rate aspect);improvement in stock market efficiency;guiding rational investment;and improvement in disclosure mechanism.Finally,this study states the research deficiencies and difficulties faced during the course of this research along with the future prospects.