Econometric Investigation of DSE＇s Portfolios through Selective Micro and Macro Economic Indicators

This paper examines the impact of the portfolios of Dhaka Stock Exchange （DSE） with corresponding to the selective macro and micro economic indicators of Bangladesh. The microeconomic indicators are Invested Market Capital （1MC） （US$） and the number of Total Enlisted Company （TEC） which have direct and immediate impact on the Stock Turnover Ratio （STR） of DSE and the macroeconomic indicators are Gross Domestic Product （GDP）, Gross National Income （GNI）, Gross Saving （GS）, Gross Inflation （GI）, Deposit Interest Rate （DIR）, and Gross Foreign Investment （GFI） which have indirect and long run impact on DSE portfolios. To investigate the direct impact on DSE＇s turnover ratio, the Cobb-Douglas production function is applied and to investigate the indirect and long run impact, multiple linear regression models are also applied. The estimated results are diagnosed using magnitudes of derivatives, gradient and Wald＇s coefficient restriction with respect to the macro and micro economic indicators.

Vector Autoregressive （VAR） Modeling and Projection of DSE

In this paper, vector autoregressive （VAR） models have been recognized for the selected indicators of Dhaka stock exchange （DSE）. Bangladesh uses the micro economic variables, such as stock trade, invested stock capital, stock volume, current market value, and DSE general indexes which have the direct impact on DSE prices. The data were collected for the period from June 2004 to July 2013 as the basis on daily scale. But to get the maximum explorative information and reduction of volatility, the data have been transformed to the monthly scale. The outliers and extreme values of the study variables are detected through box and whisker plot. To detect the unit root property of the study variables, various unit root tests have been applied. The forecast performance of the different VAR models is compared to have the minimum residual. Moreover, the dynamics of this financial market is analyzed through Granger causality and impulse response analysis.

ARIMA With GARCH Family Modeling and Projection on Share Volume of DSE

A suitable statistical model has been explored for the investors as well as the researchers to resolve the future estimation of share volume by using daily stock volume data from Dhaka Stock Exchange （DSE）. The daily volume data from the June 1, 2004 to April 19, 2010 were retrieved from DSE website as a secondary data source. The Maximum Likelihood---Autoregressive Conditional Heteroskedasticity （ARCH） （Marquardt） method has been applied to construct the models for the stock volume data of DSE by using statistical package software E-Views of verson-5. First of all, an ＂Auto Regressive Integrated Moving Average （ARIMA） model＂ was fitted and observed that heteroscedastic volatilities were still present there. To eliminate this dilemma, ARCH class of volatility models has been used and finally the ARIMA with EGARCH model has been explored. Findings of this study have recognized that ARIMA with EGARCH model implies low mean square error, low mean absolute error, low bias proportion, and low variance proportion for share volume data with comparing to other models. Hence, the modelling concept established in this study would be a decisive study for the investors as well as the researchers.