Credit Risk Loan Portfolio Attribute Remediation Matrix

Role: Senior Risk Data Governance Specialist and Quantitative Credit Risk Analyst specializing in BCBS 239 regulatory data aggregation and Basel model dataset hygiene.

Context

• Credit portfolio class: {{portfolio_class}}
• Loan origination feeds: {{origination_systems}}
• Key dirty fields: {{critical_fields_at_risk}}
• Target accounting & risk models: {{target_credit_models}}
• Benchmark replacement standards: {{benchmark_standards}}
• Capital adequacy framework: {{regulatory_framework}}

Task

Develop a comprehensive credit risk attribute remediation matrix that systematically addresses anomalous loan balances, broken interest rate indices, and missing collateral attributes across {{origination_systems}} to prepare reliable modeling inputs for {{target_credit_models}}.

Method

1. Define numerical clamping and outlier detection boundaries for debt-to-income (DTI), loan-to-value (LTV), and credit scores.
2. Standardize corrupted or non-standard amortization schedules and payment frequency codes across {{portfolio_class}}.
3. Construct replacement logic for legacy floating benchmark rates transition based on {{benchmark_standards}}.
4. Design multivariate imputation rules for missing property valuations, factoring in regional index movement and property vintage.
5. Reconcile contradictory delinquency status codes against actual transactional payment histories and days-past-due counters.
6. Standardize collateral lien prioritization fields, resolving conflicts between multiple originating servicing platforms.
7. Establish strict boundary validation rules ensuring clean datasets satisfy {{regulatory_framework}} risk data aggregation standards.

Constraints

• MUST explicitly categorize whether each remediation is an Imputation, Normalization, Replacement, or Capping action.
• MUST NOT extrapolate values for primary exposure amounts without a deterministic loan ledger reference.
• Imputed variables must be tagged with unambiguous statistical lineage metadata.
• Matrix must specify risk-model impact ratings (High/Medium/Low) for every altered parameter.

Output format

A rigorous Markdown document structured as: 1) Executive Remediation Summary (max 120 words); 2) Credit Portfolio Data Sanitization Matrix (table with columns: Field Name, Input Issue / Anomaly, Detection Boundary, Remediation Method, Model Impact Rating, Fallback Handling, Audit Evidence Code); 3) Model Ingestion Assertions (exactly 4 mathematical assertions).

Self-review

• Verify that all attributes listed in {{critical_fields_at_risk}} have concrete detection boundaries specified.
• Check that legacy benchmark adjustments fully follow {{benchmark_standards}}.
• Confirm that no remediation technique obscures default classification required under {{regulatory_framework}}.