You are on page 1of 4

Model and Methodology

Sample
To conduct this research there are some financial and non financial items are required for each company. The data of variables of the companies have been collected from the State Bank of Pakistan (SBP) and Karachi Stock Exchange (KSE). The Time period selected for this research is from 2000 to 2010. The fundamental signals are required to calculate: 2003 to 2009 for short term analysis 2003 to 2005 for long term analysis

The samples of 410 companies has been selected which are operating in Pakistan. In the next stage of this research, the contextual factors will be discussed. The contextual factors are the state of economy or industry that may effect on the prediction of future earnings or stock returns. Contextual factors are: prior earning news; industry membership; macroeconomic conditions; and country of incorporation.

Measurement of the Variables


Lev & Thiagarajan (1993) and Abarbanell and Bushee (1997), provide test for this research. The independent variables consist of nine fundamental signals and a current change in EPS. Each of fundamental signals has been designed by Lev & Thiagarajan (1993) as:

Analysis The tax and selling & administrating expense signals have negative relationship with earning per share. All the variables have insignificant impact on the EPS expect tax. R-squared value shows that 44% change in EPS is due to independent variables. Also the Durbin Watson stat which is near to 2 shows that there is no auto correlation.
Table.1 Dependent Variable: J_? Method: Pooled Least Squares Date: 06/01/12 Time: 12:36 Sample (adjusted): 2001 2004 Included observations: 4 after adjustments Cross-sections included: 197 Total pool (unbalanced) observations: 389 Cross sections without valid observations dropped Variable J_?(-1) A_? B_? Coefficient 0.255387 0.002753 0.025987 Std. Error 0.032714 0.003283 0.015761 t-Statistic 7.806562 0.838622 1.648788 Prob. 0.0000 0.4022 0.1000

D_? E_? F_? G_? H_? I_? R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat

0.003032 -0.049566 6.79E-07 7.84E-05 0.000930 -0.002921 0.446218 0.434560 11.57031 50871.38 -1499.858 2.349932

0.012043 0.005936 6.38E-06 0.000155 0.008685 0.003306 Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion F-statistic Prob(F-statistic)

0.251775 -8.349928 0.106439 0.504912 0.107057 -0.883596

0.8014 0.0000 0.9153 0.6139 0.9148 0.3775 2.419794 15.38692 7.757626 7.849328 38.27387 0.000000

Table 1

Statistical Test:
To test hypothesis one, regression is run for future earning changes on the current year earnings change and fundamental signals that is cross sectional regression is run for each year of the sample and mean coefficient calculated for the overall sample. The regression model is as follows

EPS t+,I

change in earning per share this can be either the change in one year ahead EPS or the imputed annual growth rate in earning over five years subsequent to year t current change in EPS between years t-1 and t the fundamental signals defined above Error term

CHGEPS Signals U

= = =

The regression is run for each year of the sample period 2003-2009, for the short term future earning changes with CEPSI as dependent variable. The regression is run again for the sample period 2003-2005 for the long term growth in earning with CEPSI as the dependent variable.

You might also like