The Data Collection From State Bank Of Pakistan Financial Analysis

3106 Words Aug 25th, 2014 13 Pages
Methodology
Data:
This study is based on the data collection from state bank of Pakistan financial analysis of non-financial companies listed on Karachi stock exchange (2006-2011).The publication provides the data of 411 companies in around 10 sectors from year 2006 to 2011. In this research we are not including the financial firms because their structure is totally different from non-financial firms and financial firms are services firms.

Methodology
In this study we are using the panel data regression analysis and it considers both cross section and time series features. It helps to create the clear picture of the relationship between different variables because panel data considers multiple variables for different time periods. Panel
…show more content…
We could have run 10 time series regression, one for each sector or we could run 6 cross sectional one for each year. Instead of that we will be using the over all data of all the sectors for six cross sectional one for each year. So in addition the model will help us to find out the problems that cannot only deal with time series or cross section and it will help us to mange the problem of ommited variables. Accordind to (Kremp et al.,1999) the uses of panel data methodology brings in front the two important decision to be made that of balanced panel data utilisation or not and another decision to eliminate some observations that appears to be outliers. The balanced panel is used in this study with instrumental variables and GMM, which reduces or eliminates any bias problems. The data for the companies which was not availabe were eliminated from the sample. Gaud et al., (2005) also utilized the Panel Data model in their work. According to the authors, preference for the model arose due to the possibility of including the effects of time and of controlling the heterogeneity of firms, via the inclusion of specific effects (random and fixed) of firms. The same was done by Martin et al., (2005). Accordingly, the same models already tested by Gaud et al., (2005) and by Martin et al., (2005) are adopted in this study. Pooled regression model

We used a regression model for panel data analysis. This is also called a form factor fixed to Panel data analysis on the
Open Document