Understanding Test Suites & Thresholds¶

A test suite is a collection of compliance tests that run together to evaluate a CLO portfolio against its indenture requirements. CalcBridge provides pre-built test suites for common CLO structures and allows full customization for deal-specific requirements.

What is a Test Suite?¶

A test suite groups related compliance tests into a single executable unit. Each suite contains:

• Test definitions - The specific compliance tests to run
• Threshold configurations - Pass/fail boundaries for each test
• Warning levels - Early warning thresholds (typically 90% of limit)
• Execution order - Dependencies between tests
• Reporting template - How results are formatted

flowchart TB
    subgraph Suite["Test Suite: Standard CLO"]
        direction TB
        A[Concentration Tests]
        B[Rating Tests]
        C[Coverage Tests]
        D[Quality Tests]
        E[Diversity Tests]
    end

    A --> F[Results]
    B --> F
    C --> F
    D --> F
    E --> F

    style Suite fill:#EFF6FF,stroke:#3B82F6
    style F fill:#DCFCE7,stroke:#22C55E

Test Categories¶

CalcBridge organizes compliance tests into five categories, each addressing different aspects of portfolio risk:

1. Concentration Tests¶

Concentration tests ensure the portfolio is not overly exposed to any single obligor, industry, or geographic region.

2. Rating Tests¶

Rating tests monitor the credit quality distribution of the portfolio.

Rating Agency Mappings:

CalcBridge supports all major rating agencies with automatic cross-mapping:

Split Rating Handling¶

A split rating occurs when two or more rating agencies assign different credit ratings to the same loan. For example, a loan might be rated B1 by Moody's but BB+ by S&P -- a meaningful divergence that affects which compliance bucket the loan falls into.

Indentures specify how to resolve split ratings. CalcBridge supports all common methods via the split_rating_method parameter in the test suite configuration.

Resolution Methods¶

Configuration¶

Set the split rating resolution method in the test suite threshold configuration:

{
    "rating": {
        "split_rating_method": "lower_of_two",
        "ccc_bucket_limit_pct": 7.5,
        "maximum_warf": 2850,
        "rating_agencies": ["moodys", "sp"],
        "warf_rating_source": "moodys"
    }
}

Implementation¶

import numpy as np
import pandas as pd


# Numeric scale: lower number = better credit quality
NUMERIC_SCALE = {
    "Aaa": 1, "AAA": 1,
    "Aa1": 2, "AA+": 2, "Aa2": 3, "AA": 3, "Aa3": 4, "AA-": 4,
    "A1": 5, "A+": 5, "A2": 6, "A": 6, "A3": 7, "A-": 7,
    "Baa1": 8, "BBB+": 8, "Baa2": 9, "BBB": 9, "Baa3": 10, "BBB-": 10,
    "Ba1": 11, "BB+": 11, "Ba2": 12, "BB": 12, "Ba3": 13, "BB-": 13,
    "B1": 14, "B+": 14, "B2": 15, "B": 15, "B3": 16, "B-": 16,
    "Caa1": 17, "CCC+": 17, "Caa2": 18, "CCC": 18, "Caa3": 19, "CCC-": 19,
    "Ca": 20, "CC": 20, "C": 21, "D": 21,
}


def resolve_split_rating(
    df: pd.DataFrame,
    method: str = "lower_of_two",
    specified_agency: str | None = None,
) -> pd.Series:
    """
    Resolve split ratings to a single composite rating per loan.

    Args:
        df: DataFrame with 'rating_moodys' and 'rating_sp' columns.
        method: One of 'lower_of_two', 'higher_of_two',
                'worst_of_all', or 'specified_agency'.
        specified_agency: Required when method is 'specified_agency'.
                          One of 'moodys' or 'sp'.

    Returns:
        Series of resolved ratings (using Moody's notation as the
        canonical output scale).
    """
    moodys_numeric = df["rating_moodys"].map(NUMERIC_SCALE)
    sp_numeric = df["rating_sp"].map(NUMERIC_SCALE)

    # Build reverse lookup: numeric -> Moody's rating
    numeric_to_moodys = {v: k for k, v in NUMERIC_SCALE.items() if len(k) <= 4}

    if method == "lower_of_two":
        # Higher numeric = worse credit quality = lower rating
        composite_numeric = np.maximum(
            moodys_numeric.fillna(sp_numeric),
            sp_numeric.fillna(moodys_numeric),
        )
    elif method == "higher_of_two":
        composite_numeric = np.minimum(
            moodys_numeric.fillna(sp_numeric),
            sp_numeric.fillna(moodys_numeric),
        )
    elif method == "worst_of_all":
        # Same as lower_of_two for two agencies; extensible to Fitch
        composite_numeric = np.maximum(
            moodys_numeric.fillna(sp_numeric),
            sp_numeric.fillna(moodys_numeric),
        )
    elif method == "specified_agency":
        if specified_agency == "moodys":
            return df["rating_moodys"]
        elif specified_agency == "sp":
            return df["rating_sp"]
        else:
            raise ValueError(f"Unknown agency: {specified_agency}")
    else:
        raise ValueError(f"Unknown split rating method: {method}")

    return composite_numeric.map(numeric_to_moodys)

Impact on Other Tests¶

The resolved composite rating feeds into multiple downstream tests:

flowchart LR
    A[Moody's<br/>Rating] --> C[Split Rating<br/>Resolution]
    B[S&P<br/>Rating] --> C
    C --> D[Composite<br/>Rating]
    D --> E[WARF<br/>Calculation]
    D --> F[CCC Bucket<br/>Test]
    D --> G[Minimum Rating<br/>Test]
    D --> H[Diversity Score<br/>Calculation]

    style C fill:#FEF3C7,stroke:#F59E0B
    style D fill:#EFF6FF,stroke:#3B82F6

3. Coverage Tests¶

Coverage tests verify that the portfolio provides adequate protection for debt holders.

OC Ratio Calculation:

OC Ratio = (Collateral Principal Balance / Tranche Principal Balance) * 100

Example:
- Collateral Balance: $500,000,000
- Senior Tranche: $400,000,000
- Senior OC Ratio: ($500M / $400M) * 100 = 125%

4. Quality Tests¶

Quality tests assess the overall characteristics of the loan portfolio.

WAS Calculation:

# Weighted Average Spread
WAS = Sum(Loan Par * Loan Spread) / Sum(Loan Par)

# Example
Portfolio:
  - Loan A: $10M par, 3.50% spread
  - Loan B: $15M par, 4.00% spread
  - Loan C: $25M par, 3.75% spread

WAS = ($10M * 3.50% + $15M * 4.00% + $25M * 3.75%) / $50M
WAS = ($350K + $600K + $937.5K) / $50M
WAS = 3.775%

5. Diversity Tests¶

Diversity tests measure how well the portfolio is diversified across obligors and industries.

Diversity Score Calculation:

The Moody's diversity score approximates the effective number of independent obligors using the Herfindahl-Hirschman Index (HHI):

# Simplified diversity score calculation
HHI = Sum(Obligor Weight ^ 2)
Diversity Score = 1 / HHI

# Example
Portfolio with 4 obligors:
  - Obligor A: 40% weight -> 0.16
  - Obligor B: 30% weight -> 0.09
  - Obligor C: 20% weight -> 0.04
  - Obligor D: 10% weight -> 0.01

HHI = 0.16 + 0.09 + 0.04 + 0.01 = 0.30
Diversity Score = 1 / 0.30 = 3.33

Threshold Configuration¶

Setting Threshold Values¶

Thresholds are configured per test suite and can be customized for each deal:

{
  "suite_name": "Standard CLO v2",
  "description": "Standard compliance tests for US BSL CLOs",
  "thresholds": {
    "concentration": {
      "single_obligor_limit_pct": 10.0,
      "top_5_obligor_limit_pct": 40.0,
      "single_industry_limit_pct": 15.0,
      "second_lien_limit_pct": 10.0
    },
    "rating": {
      "ccc_bucket_limit_pct": 7.5,
      "maximum_warf": 2850
    },
    "coverage": {
      "minimum_senior_oc_ratio": 120.0,
      "minimum_mezzanine_oc_ratio": 108.0,
      "minimum_senior_ic_ratio": 150.0
    },
    "quality": {
      "minimum_was": 3.25,
      "maximum_wal": 5.5
    },
    "diversity": {
      "minimum_diversity_score": 45.0,
      "minimum_obligor_count": 120
    }
  },
  "warning_threshold_pct": 90.0
}

Warning Levels¶

CalcBridge uses a three-tier status system:

Test Direction¶

Some tests have minimum thresholds (higher is better), while others have maximum thresholds (lower is better):

Pre-Built Test Suites¶

CalcBridge includes several pre-built test suites:

Standard CLO Suite¶

The most commonly used suite for US Broadly Syndicated Loan (BSL) CLOs:

Suite: Standard CLO
Tests: 24
Categories:
  - Concentration (8 tests)
  - Rating (4 tests)
  - Coverage (6 tests)
  - Quality (3 tests)
  - Diversity (3 tests)

Middle Market CLO Suite¶

Adjusted thresholds for middle market loan portfolios:

Suite: Middle Market CLO
Tests: 20
Key Differences:
  - Higher single obligor limits (15% vs 10%)
  - Lower diversity requirements (30 vs 45)
  - Higher CCC bucket tolerance (10% vs 7.5%)

European CLO Suite¶

Compliance tests for European CLO structures:

Suite: European CLO
Tests: 22
Key Differences:
  - Currency-adjusted calculations
  - EU regulatory requirements
  - Retention rule compliance

Creating Custom Test Suites¶

Via API¶

curl -X POST https://api.calcbridge.io/api/v1/compliance/suites \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Custom Deal Suite",
    "description": "Custom thresholds for Deal ABC 2024-1",
    "base_suite": "standard_clo",
    "threshold_overrides": {
      "concentration": {
        "single_obligor_limit_pct": 8.0
      },
      "rating": {
        "ccc_bucket_limit_pct": 5.0
      }
    },
    "enabled_tests": [
      "single_obligor",
      "top_5_obligor",
      "ccc_bucket",
      "senior_oc_ratio",
      "was",
      "wal",
      "diversity_score"
    ]
  }'

Via UI¶

1. Navigate to Settings > Compliance > Test Suites
2. Click Create New Suite
3. Select a base suite to start from
4. Adjust thresholds as needed
5. Enable/disable individual tests
6. Save and assign to workbooks

Best Practices¶

Example: Typical CLO Compliance Tests¶

Here is a complete example of a typical CLO test suite configuration:

# ComplianceCalculator default thresholds
STANDARD_CLO_THRESHOLDS = {
    # Concentration Limits
    "single_obligor_limit_pct": Decimal("10.0"),
    "top_5_obligor_limit_pct": Decimal("40.0"),
    "single_industry_limit_pct": Decimal("15.0"),

    # Rating Limits
    "ccc_bucket_limit_pct": Decimal("7.5"),

    # Coverage Minimums
    "minimum_oc_ratio": Decimal("120.0"),
    "minimum_ic_ratio": Decimal("150.0"),

    # Quality Metrics
    "minimum_was": Decimal("3.0"),
    "maximum_wal": Decimal("5.0"),

    # Diversity Minimums
    "minimum_diversity_score": Decimal("40.0"),
}

Next: Running Compliance Tests