Developing a Model for the Assessment and Evaluation of Financial Health in Iran

Financial health is defined as capability of making profit and the continuation of industrial entity activity. This research aims at the determination and recognition of financial components and features affecting health and presentation of a model based on Multinomial logistic Regression (MLR) approach to assess and evaluate financial health in Iran.
companies in this research have been divided into three groups as to their financial health state called; healthy, intermediate and distressed.In order to test the accuracy of prediction and trath-value of the extracted model , the selected companies at every level of health were divided as symmetrical pairs in to control and experimental groups.Using Excel software 15 financial ratios were calculated in 5 states of liquidity, leverage, activity , market value and Value creation .
SPSS (17) was used to test hypotheses , (ANOVA) and Kruskal-Wallis statistical variance analysis to compare the average and SD of ratios at 5% level of error. Research findings show that there is significant statistical differences among these companies according to different levels of financial health based on leverage, activity and market value ratios. But the liquidity and Value creation state difference of these three levels is significant and the components affecting financial health are as quick ,current , debt , the proportion of net working capital to total assets ratios, EVA and MVA. Based on Dunnett and Tukey tests the distressed companies are the factors yielding the difference. i.e. distressed companies are statistically on one side and intermediate and healthy ones on the other side of distribution. As a resalt two models were presented, one for distressed companies and another one for intermediate and healthy ones.
The test of model prediction accuracy in the experimental group at three levels of distressed , intermediate , healthy and total are subsequently 70% , 92% , 100%, 88.88% and shows 66% , 79.16% , 77.7% and 76.92% of prediction accuracy and classification in the control group.