Financial Ratios and Financial Sustainability in EU Non-Financial Companies: A Literature Review and Meta-Analysis

Background and purpose: Financial ratios assess non-financial companies’ financial sustainability, but evidence on their predictive power across EU Member States remains fragmented and heterogeneous. To the authors’ knowledge, this is among the first integrated EU-focused reviews and meta-analyses covering seven ratio groups. Methodology: A structured research identified 29 studies, yielding 230 poolable effect sizes from 22 articles. Effects were standardized as Fisher’s z-transformed partial correlations and pooled using a random-effects DerSimonian–Laird model. Dependence was addressed through robust variance estimation (RVE) and a three-level model. Results: The analysis reveals that the overall pooled effect is small and, once effect-size dependence is accounted for, estimated with far less precision, although still statistically significant. Liquidity and profitability ratios show the largest pooled effects, followed by solvency ratios, whereas cash-flow and activity ratios show only small associations. Under dependence-robust inference, solvency and cash-flow ratios remain significant in both dependence-aware models, liquidity and profitability ratios in only one of them, and activity ratios in neither. At the indicator level, equity-to-assets, EBIT/total assets, and working capital/total assets show the largest and most consistent baseline effects, but each rest on one to six studies. Pooled effects are smaller at longer prediction horizons and larger in large-firm samples, but neither moderator is statistically significant once dependence is accounted for; only the type of outcome measure (continuous versus binary) moderates the effects significantly. Conclusions: Financial sustainability in EU non-financial companies is best captured by a core set of liquidity, profitability and solvency ratios, whereas the larger effects observed at short horizons and in large-firm samples are suggestive rather than conclusive. Associations differ sharply by outcome type, weak for binary bankruptcy and larger for distress scores, so individual ratios are weak and highly context-dependent rather than uniformly predictive.