The article Financial Ratios, Discriminant Analysis, and Predicting Business Failure was written in 1968 by Edward I. Altman. The purpose of the article is to address the quality of ratio analysis as an analytical technique. At that time some academics were moving away from proportional analysis and towards statistical analysis. The article attempted to determine whether ratio analysis should be continued, eliminated, and replaced by statistical analysis or serve alongside statistical analysis as cofactors in financial analysis. The example case used in the article was predicting business failure. The ratios traditionally measure major factors such as liquidity, solvency and profitability, as well as other solvency measures. Several studies have found that various ratios are the most efficient indicators of creditworthiness. Ratio analysis studies began in the 1930s, with several studies concluding that companies with the potential to file for bankruptcy all exhibited different ratios than financially sound companies. Among the findings of the study was that the decisive factor of the bankruptcy predictor should be not just a few ratios, since the measure of a company's financial solvency can vary as the company's situations change. The important question is which ratios should be used and, among those chosen, which ratios are given priority. After discussion, multiple discriminant analysis (MDA), a statistical technique, was chosen. MDA has primarily been used to classify and make predictions in problems where the dependent variable was in qualitative form, such as bankrupt or not bankrupt. The main advantage of MDA was its ability to sequentially examine individual elements such as: purpose of the loan, maturity of the pledged collateral, the customer's history with the company and the unique characteristics that the bank's customers might have. The author concluded that financial ratios, when combined with statistical analysis, still remain a valuable tool. The theoretical conclusion was that ratios used within a multivariate framework take on a more influential role than when used in isolation. The discriminant model was very accurate in the initial sample of 66 firms, correctly predicting 94% of the initially failed firms. Suggested potential uses of the model included: corporate credit ratings, investment guidelines, and internal control procedures. The MDA model has also shown the potential to alleviate some problems in portfolio stock selection, but further investigation is recommended.