Regression Analysis v. Ratios in the Cross-section Analysis of Financial Statements

The assumption that the calculation of a ratio generates a useful number is widespread among analysts. The recent literature on financial statement analysis has highlighted this assumption and raised doubts about its validity on both theoretical and empirical grounds. This paper aims both to summarise and extend this discussion. An examination of the literature is followed by empirical work based around UK data on the brewing industry. A concluding section assesses the relevance of this strand of research for the practice of financial statement analysis.