-

We developed two logit models for prediction the rating of return on stock (ROS) and return on equity (ROE), for accepted firms in Tehran Stock Exchange (TSE).
The first phase of study examines that "do exists any significance statistical differences between the financial statement data of successful and unsuccessful firms?", in a cross-sectional sample. The results proposed that, there are some significant differences between the two groups, from the points of financial activity indexes, liabilities position (financial leverage), average size (in terms of book values), and kind of industries. However, the differences for liquidity not passed the test of significance.
We developed the logit models with a quasi-experimental methodology. They predict the rating of the firms with a reasonable accuracy. However, the ROE based model has higher accuracy than the ROS based model, and its statistics are better, too.