Explanation

Option B best describes financial models used in predicting default because:

• Statistical Approach: Financial models for default prediction are primarily statistical models that use historical data on loan outcomes (defaults vs. non-defaults)
• Relationship Building: They establish statistical relationships between default probability and various input variables (financial ratios, macroeconomic indicators, etc.)
• Empirical Foundation: These models are empirical in nature, relying on observed historical patterns rather than theoretical constructs

Why other options are incorrect:

• Option A: Describes expert judgment approaches, not quantitative financial models
• Option C: Describes structural models (like Merton's model) which are more theoretical and based on economic principles, but these are not the primary type of financial models for default prediction
• Option D: Describes market-based approaches using bond spreads and CDS prices, which are different from traditional default prediction models

Financial models for default prediction are typically statistical models (like logistic regression, discriminant analysis) that use historical default data to predict future defaults based on observable characteristics.