Explanation

Hazard models best address the issue of the going-concern assumption in using historical financial statements.

Key points:

• Hazard models (also called survival analysis models) explicitly incorporate the time dimension and can handle the fact that companies may not go bankrupt during the observation period
• They account for censored data - companies that survive beyond the study period
• Accounting-based models (like Altman's Z-score) rely on historical financial ratios but don't explicitly address the going-concern assumption
• Market-based models use market data but may not specifically handle the time dimension and censoring issues

Hazard models are particularly well-suited for bankruptcy prediction because they can model the probability of bankruptcy over time while properly accounting for companies that remain solvent, thus directly addressing the going-concern assumption limitation in traditional financial statement analysis.