casemark/analyzing-structured-product-ratings

Analyzing Structured Product Ratings

Evaluates rating agency methodology application with loss model inputs, correlation assumptions, and tranche-level credit assessment across ABS, MBS, CLO, and other securitized products.

When To Use

• Reviewing a new-issue presale report or rating letter to understand tranche-level credit support
• Comparing Moody's, S&P, Fitch, or DBRS/Kroll methodologies on the same transaction
• Assessing whether a rating action (upgrade, downgrade, watch placement) is consistent with stated methodology
• Evaluating sensitivity of ratings to changes in default, recovery, or correlation assumptions
• Preparing investment committee memos that incorporate third-party rating analysis
• Reviewing surveillance reports for deteriorating collateral performance vs. original rating assumptions

Inputs To Gather

• Rating reports: Presale/new-issue reports, rating letters, surveillance updates, and methodology publications from each relevant agency
• Deal documents: Offering memorandum/circular, waterfall structure, priority of payments, trigger definitions, and reserve fund mechanics
• Collateral data: Pool stratification tables (FICO/DSCR/LTV distributions for MBS; industry/rating/recovery for CLOs; obligor concentration for ABS), historical vintage performance data
• Loss model inputs: Base-case default rate, default timing curve, recovery rate, recovery lag, prepayment speed (CPR/CDR for MBS; CPR for auto/student loan ABS)
• Structural features: Credit enhancement levels (subordination, OC, excess spread), tranche thickness, amortization schedule, reinvestment period (CLOs), step-down triggers
• Correlation assumptions: Inter-sector and intra-sector asset correlation matrices, geographic concentration adjustments [VERIFY: agency-specific correlation frameworks differ materially]

Workflow

1. Map the capital structure — Diagram tranche seniority, attachment/detachment points, and credit enhancement (CE) levels. Calculate initial CE as a percentage of the pool balance for each rated tranche.

2. Identify applicable methodology — Determine which agency criteria apply. Key frameworks include:

   • Moody's: Idealized expected loss tables, Cdp (correlated default probability), MILAN CE for RMBS, CLO Monitor/WARF for CLOs
   • S&P: CDO Evaluator (Monte Carlo), LEVELS for RMBS, SROC (scenario-based rating on capital) for CLOs under surveillance
   • Fitch: Portfolio Credit Model (PCM), ResiGlobal for RMBS, asset-specific default models
   • DBRS/Kroll: KBRA CLO methodology, DBRS master trust criteria [VERIFY: confirm current methodology version dates]

3. Analyze loss model inputs — For each agency:

   • Extract base-case and stressed default/loss assumptions
   • Compare assumed severity/recovery rates against historical realized performance for the asset class
   • Evaluate prepayment and default timing vectors and their impact on excess spread capture
   • Identify whether the agency uses a deterministic scenario ladder or stochastic (Monte Carlo) simulation

4. Assess correlation and concentration — Review how each agency models:

   • Asset correlation (flat vs. sector-based matrices)
   • Obligor/geographic/industry concentration penalties
   • Large obligor tests (Moody's Binomial Expansion Technique vs. S&P supplemental tests)
   • Impact of correlation assumptions on tail-risk scenarios at senior vs. mezzanine tranche levels

5. Evaluate structural protections — Map agency-specific treatment of:

   • Cash flow waterfall priorities (pre- vs. post-acceleration)
   • OC/IC trigger mechanics and cure provisions
   • Liquidity facilities, reserve funds, and guaranteed investment contracts
   • Counterparty risk (swap provider, servicer, account bank) and replacement triggers [VERIFY: counterparty criteria vary by agency and jurisdiction]

6. Run sensitivity analysis — Stress key variables independently and in combination:

   • Default rate: +25%, +50%, +100% of base case
   • Recovery rate: –10pp, –20pp from base assumption
   • Correlation: increase by 5–10pp across sectors
   • Prepayment speed: 0.5x and 2.0x base CPR
   • Document which tranches experience notch changes under each scenario

7. Compare cross-agency outcomes — Build a comparison matrix showing:

   • Rating assigned by each agency per tranche
   • Key divergence drivers (e.g., differing recovery assumptions, correlation treatment, or structural credit given for excess spread)
   • Identify split ratings and explain the methodological basis for divergence

Output

Produce a structured rating analysis report containing:

• Executive summary: Transaction overview, agencies involved, rating snapshot, and key findings (1–2 paragraphs)
• Capital structure table: Tranche name, size, rating by each agency, CE level, attachment/detachment points
• Methodology comparison matrix: Side-by-side comparison of loss assumptions, correlation inputs, structural credit, and stress scenarios per agency
• Sensitivity grid: Table showing rating impact of stressed defaults, recoveries, correlations, and prepayments by tranche
• Key risk factors: Concentration risks, structural weaknesses, servicer/counterparty dependencies, and triggers approaching breach
• Split rating commentary: Explanation of any rating divergences between agencies, with identification of the analytical driver
• Surveillance flags: Collateral performance metrics to monitor (delinquency trends, cumulative loss curves vs. original projections, OC/IC test cushion)

Quality Checks

• Confirm all referenced methodology publications are current versions — agencies frequently update criteria [VERIFY: check publication dates against agency websites]
• Verify that CE calculations match offering document definitions (some deals define CE net of defaulted assets, others gross)
• Cross-check that loss model inputs extracted from presale reports are internally consistent (e.g., default rate × severity = expected loss)
• Ensure sensitivity analysis covers the full rated spectrum — senior tranches may be insensitive to moderate stresses but mezzanine tranches may be highly sensitive
• Confirm that waterfall modeling accounts for all payment dates, not just a single snapshot
• Flag any reliance on manager/servicer discretion (e.g., CLO reinvestment flexibility, workout assumptions) as a qualitative risk factor
• Validate that counterparty rating triggers are consistent with agency counterparty criteria for the relevant jurisdiction