ISSN: 2319-7285
Rohit Malhotra and Dr. Jimmi Kapadia
The present paper demonstrated that how the impact of “residual correlations” approach helps in selecting the parsimonious regression method. For this the data is tested against heteroskedasicity, Normality, autocorrelation and colinearity before considering for the OLS parameter estimation procedure. For confirming this use of financial ratios, viz. Employee cost per net profit, the operating expenses per net profit, and together several ratios were taken into consideration. The result clearly stated that bi-variate model indeed found better than higher variable regression equations