Credit Risk Analysis

Which loan segments carry the most credit risk, and how does underwriting quality vary across grades and vintage cohorts? This analysis examines default and loss patterns across loan grades and risk tiers to identify where underwriting standards have the greatest impact on portfolio performance.

Critical Risk Finding: Grade G loans default at approximately 5x the rate of Grade A loans, with cumulative loss rates amplified by significantly lower FICO scores and higher debt-to-income ratios at origination. This concentration of risk in lower-grade segments requires urgent policy review.

Methodology: Default rates are calculated as (default_count / loan_count) and loss rates as (total_loss_amount / total_originated). Cohorts are tracked by origination quarter to measure performance across underwriting cycles and economic conditions.

Grade G Loans Default at 5x the Rate of Grade A

Default Rate Trends Across Cohort Quarters

Cumulative Loss Acceleration by Grade

FICO Score vs. Default Rate Relationship

Risk Tier Breakdown: Default, Loss, and Underwriting Quality

No Results

Loss and Underwriting Metrics by Grade

No Results

So What? Key Business Recommendations

1. Tighten Grade G Origination: With default rates 5x higher than Grade A, Grade G loans are capital-inefficient. Consider either raising underwriting standards for this grade (minimum FICO, maximum DTI) or reducing origination volume to acceptable loss thresholds. Current loss rates do not support the spreads charged.

2. Implement FICO-Based Pricing Floors: The strong inverse correlation between FICO and default rate suggests mispricing in lower-FICO segments. Establish minimum FICO thresholds by grade or implement dynamic pricing that better reflects credit risk, particularly for grades D–G.

3. Monitor Underwriting Cohort Drift: Vintage cohort analysis reveals variation in default performance across quarters. Establish quarterly underwriting quality monitoring to detect shifts in loan origination standards and catch deteriorating underwriting before it manifests in loss rates.

Data Sources: Credit risk metrics aggregated from dbt mart_credit_risk and mart_default_rates tables. Default rates reflect loans aged 24+ months to allow sufficient time for default observation. Loss rates include principal and accrued interest write-offs by grade and cohort.