diff --git a/private-credit/.claude-plugin/plugin.json b/private-credit/.claude-plugin/plugin.json new file mode 100644 index 0000000..f864422 --- /dev/null +++ b/private-credit/.claude-plugin/plugin.json @@ -0,0 +1,8 @@ +{ + "name": "private-credit", + "version": "1.0.0", + "description": "Domain-specific skills and commands for private credit analysts and portfolio managers at direct lending funds. Covers covenant compliance, EBITDA adjustments, IC memo generation, quarterly monitoring, deal screening, credit modeling, amendment analysis, and scenario testing.", + "author": { + "name": "Accretive AI" + } +} diff --git a/private-credit/README.md b/private-credit/README.md new file mode 100644 index 0000000..ecfce6c --- /dev/null +++ b/private-credit/README.md @@ -0,0 +1,108 @@ +# Private Credit Plugin + +A Claude plugin for analysts and portfolio managers at direct lending funds. Encodes the domain expertise, workflows, and judgment frameworks used in private credit — from CIM screening through quarterly portfolio monitoring. + +Built for [Claude Cowork](https://claude.com/product/cowork), also compatible with [Claude Code](https://claude.com/product/claude-code). + +> **Disclaimer:** This plugin assists with private credit workflows but does not provide investment, financial, or legal advice. All outputs — including covenant calculations, credit assessments, and investment recommendations — should be reviewed by qualified professionals before use in investment decisions, regulatory filings, or committee presentations. + +## Who This Is For + +- **Credit analysts** doing quarterly monitoring, model updates, and borrower analysis +- **Portfolio managers** reviewing covenant compliance and credit trajectory +- **Investment professionals** screening deals, writing IC memos, and running scenarios +- **Operations teams** processing amendments and tracking covenant schedules + +Works best for middle-market direct lending ($50M–$500M deal size), leveraged finance, ABL, and specialty lending workflows. + +## What's Included + +### Skills (auto-fire when relevant) + +Skills encode domain knowledge Claude draws on automatically — no slash command needed. + +| Skill | What it covers | +|-------|---------------| +| **Covenant Compliance** | Maintenance vs. incurrence testing, compliance cert as source of truth, step-down schedule lookup, headroom analysis (≥20% comfortable, 10–20% elevated, <10% flag), cross-checking financials against cert | +| **EBITDA Add-Back Treatment** | GAAP-to-Adjusted EBITDA waterfall, permitted addbacks per credit agreement, pro forma acquisition adjustments, cap mechanics, push-back framework for management's numbers, addback realization tracking | +| **Credit Memo Standards** | IC memo structure and section order, minimum content bar, advocacy tone calibration, merit/risk framework, base and downside case presentation standards | +| **Quarterly Package Extraction** | Extraction order (P&L → BS → CF), label-matching logic, compliance cert vs. financials reconciliation, handling acquisitions and restatements, common data quality issues | +| **Borrower Monitoring** | Full quarterly workflow (receive → update → narrate → question → deliver), credit trajectory evaluation, watchlist criteria (3-tier: watch closely / watchlist / workout), management call question development | +| **Credit Model Standards** | LBO-based model structure, tab organization, FCF waterfall construction (exact line-item order), case methodology (base / downside / stress), credit statistics, debt schedule roll-forward | + +### Commands (user-triggered) + +| Command | What it produces | +|---------|-----------------| +| `/private-credit:cim-screen` | Pass / Pursue / More Info assessment from a CIM or teaser with supporting credit analysis | +| `/private-credit:ic-memo` | Full investment committee memo in private credit format (requires CIM + model + term sheet) | +| `/private-credit:quarterly-review` | Covenant compliance table, variance summary with MD&A narrative, management call questions | +| `/private-credit:covenant-check` | Comprehensive covenant snapshot: maintenance tests with headroom, incurrence levels, restricted baskets, step-down schedule, cure rights | +| `/private-credit:amendment-summary` | Structured summary of every change in an amendment, categorized as tightened / loosened / new / removed | +| `/private-credit:scenario-analysis` | Base / Downside / Stress case comparison table with covenant breach flagging and liquidity runway | + +## Plugin Structure + +``` +private-credit/ +├── .claude-plugin/ +│ └── plugin.json +├── commands/ +│ ├── amendment-summary.md +│ ├── cim-screen.md +│ ├── covenant-check.md +│ ├── ic-memo.md +│ ├── quarterly-review.md +│ └── scenario-analysis.md +├── skills/ +│ ├── borrower-monitoring/SKILL.md +│ ├── covenant-compliance/SKILL.md +│ ├── credit-memo-standards/SKILL.md +│ ├── credit-model-standards/SKILL.md +│ ├── ebitda-addback-treatment/SKILL.md +│ └── quarterly-package-extraction/SKILL.md +├── LICENSE +└── README.md +``` + +## Installation + +```bash +# From the marketplace +/plugin install private-credit@claude-plugin-directory + +# Or from GitHub +/plugin marketplace add accretive-ai/private-credit-plugin +/plugin install private-credit@accretive-ai +``` + +Once installed, skills activate automatically when Claude detects relevant context (financial data, credit agreements, borrower packages). Commands are available via `/private-credit:command-name`. + +## Optional MCP Connectors + +This plugin works standalone — paste or upload financial data and documents directly. For live data integration, add MCP servers to a `.mcp.json` file in the plugin directory: + +- **PitchBook** — Private equity and deal data for comparable transactions +- **Moody's** — Credit ratings and company financial data +- **FactSet** — Financial data, analytics, and market benchmarks +- **S&P Global / Kensho** — Capital IQ financial data + +See [MCP documentation](https://modelcontextprotocol.io/) for connector setup. + +## Customization + +These skills and commands encode general private credit best practices. They become more powerful when customized for your firm: + +- **Adjust covenant severity rankings** — Edit `covenant-compliance/SKILL.md` to match your firm's headroom thresholds and escalation criteria +- **Add your EBITDA definition** — Modify `ebitda-addback-treatment/SKILL.md` with your standard permitted addbacks and cap structures +- **Match your IC memo format** — Update `credit-memo-standards/SKILL.md` with your firm's section order, required exhibits, and tone preferences +- **Configure watchlist criteria** — Adjust thresholds in `borrower-monitoring/SKILL.md` to match your fund's risk rating framework +- **Set variance callout thresholds** — Modify `quarterly-review.md` to match what your IC considers material + +## About + +Built by [Accretive AI](https://goaccretive.ai). The team behind Accretive AI has billions invested and decades of experience across leveraged finance, ABL, and special situations. + +## License + +Apache-2.0 — see [LICENSE](LICENSE) for details. diff --git a/private-credit/commands/amendment-summary.md b/private-credit/commands/amendment-summary.md new file mode 100644 index 0000000..617b94e --- /dev/null +++ b/private-credit/commands/amendment-summary.md @@ -0,0 +1,160 @@ +--- +description: "Structured summary of every change in a credit amendment" +argument-hint: "" +--- + +# /private-credit:amendment-summary + +Produces a structured, mechanical summary of every change in a credit agreement amendment. Categorizes each change as tightened, loosened, new, or removed. + +--- + +## Trigger + +User invokes `/private-credit:amendment-summary` with an amendment document, amended and restated credit agreement, consent letter, or waiver letter attached or referenced. + +--- + +## Required Inputs + +| Input | Required? | Notes | +|---|---|---| +| Amendment document | **Yes** | The amendment, consent, waiver, or A&R credit agreement | +| Original credit agreement or prior amendment (if available) | Recommended | Enables comparison to prior terms — helpful but not required since amendments typically state what changed | + +--- + +## What This Command Does + +This is a **mechanical extraction and categorization exercise.** It does not assess whether the amendment is borrower-friendly or lender-friendly, does not compare to market terms, and does not provide qualitative judgment on the changes. Those assessments require legal expertise and market data that are outside the scope of this command. + +--- + +## Extraction Framework + +Read the amendment document end to end and extract every material change. Categorize each change into one of four classifications: + +| Classification | Definition | +|---|---| +| **Tightened** | A term that became more restrictive for the borrower (lower leverage threshold, higher coverage requirement, smaller basket, additional restriction) | +| **Loosened** | A term that became less restrictive for the borrower (higher leverage threshold, lower coverage requirement, larger basket, removed restriction) | +| **New** | A term, covenant, basket, or provision that did not exist before and is being added | +| **Removed** | A term, covenant, basket, or provision that existed before and is being eliminated | + +--- + +## What to Capture + +Extract and categorize changes across these categories. Not every amendment will have changes in every category — only include categories that are modified. + +### Pricing Changes +- Spread changes (increase or decrease, by how much) +- OID changes +- SOFR floor changes +- Commitment fee changes +- PIK toggle or PIK rate changes +- Call protection or prepayment premium changes + +### Structural Changes +- Maturity extension or reduction +- Amortization schedule changes (increased, decreased, holiday) +- New tranches added or existing tranches modified +- Revolver commitment changes (increased, decreased) +- Delayed draw commitments added or modified +- Accordion feature changes + +### Financial Covenant Changes +- Maintenance covenant threshold modifications (state old and new levels) +- Step-down schedule modifications (state full revised schedule) +- Test frequency changes +- Covenant holiday or suspension periods +- New financial covenants added +- Existing financial covenants removed + +### Incurrence Test Changes +- Leverage levels for restricted payments modified +- Leverage levels for additional debt incurrence modified +- Leverage levels for investments modified + +### Basket and Definition Changes +- CapEx basket changes (dollar amount, percentage, or elimination) +- Restricted payment basket changes +- Investment basket changes +- Debt incurrence basket changes +- Permitted acquisition thresholds changed +- Builder basket modifications + +### EBITDA / Financial Definition Changes +- Permitted addback modifications (new addbacks, removed addbacks, cap changes) +- Pro forma adjustment methodology changes +- Synergy realization period changes +- Any other defined term modifications that affect financial calculations + +### Waivers +- Waiver of any past breach (note the covenant breached, the period, and that it was waived) +- Temporary waivers with expiration dates + +### Other Material Changes +- Reporting requirement changes +- Collateral or guarantee modifications +- Change of control definition changes +- Assignment or transfer provision changes +- Any other provisions modified by the amendment + +--- + +## Output Format + +### Header + +``` +AMENDMENT SUMMARY +Borrower: [Name] +Amendment: [Amendment No. X / A&R Credit Agreement / Consent / Waiver] +Effective Date: [Date] +``` + +### Priority Items (Top of Output) + +The most important changes go first — these are what the reader needs within 30 seconds: + +1. **Pricing changes** — any change to the cost of borrowing +2. **Structural changes** — maturity, amortization, tranche modifications +3. **Covenant resets** — threshold changes, new covenants, removed covenants + +### Categorized Change List + +For each change, present: + +``` +[TIGHTENED / LOOSENED / NEW / REMOVED] + +Category: [Pricing / Structure / Covenant / Basket / Definition / Waiver / Other] +Item: [Specific term or provision] +Prior: [Previous term, if applicable] +Revised: [New term] +Section: [Credit agreement section reference, e.g., "Section 7.1(a)"] +``` + +Group changes by classification (all tightened together, all loosened together, etc.) or by category (all pricing together, all covenants together) — whichever produces a cleaner, more readable output for the specific amendment. For amendments with many changes, grouping by category is typically clearer. + +### Summary Count + +At the bottom: +``` +Total changes: [X] + Tightened: [X] + Loosened: [X] + New: [X] + Removed: [X] +``` + +--- + +## Calibration Notes + +- **Extract every material change.** Do not skip minor items — let the reader decide what matters. +- **Use precise language from the amendment.** When the amendment states a specific threshold, basket size, or date, reproduce it exactly. Do not paraphrase numbers. +- **State old and new values side by side.** For every modification, the reader must see what it was and what it is now. +- **Section references matter.** Include the credit agreement section reference for each change so the reader can locate it in the document. +- **No qualitative assessment.** Do not characterize the amendment as "favorable" or "unfavorable." Do not compare to market. Summarize the facts. diff --git a/private-credit/commands/cim-screen.md b/private-credit/commands/cim-screen.md new file mode 100644 index 0000000..706f363 --- /dev/null +++ b/private-credit/commands/cim-screen.md @@ -0,0 +1,137 @@ +--- +description: "Screen a CIM or teaser for credit fit: Pass / Pursue / More Info" +argument-hint: "" +--- + +# /private-credit:cim-screen + +Produces a credit fit assessment from a CIM, teaser, or lender presentation. Output: **Pass / Pursue / More Info** with supporting analysis. + +--- + +## Trigger + +User invokes `/private-credit:cim-screen` with a CIM, teaser, or lender presentation attached or pasted. + +--- + +## Required Context + +Before running the screen, Claude must know the firm's credit box. If not already provided, ask for: +1. **Structure mandate:** Senior only? Unitranche? Mezzanine? Second lien? +2. **Size range:** Minimum and maximum hold size (or EBITDA range) +3. **Excluded industries:** Tobacco, firearms, cannabis, etc. +4. **Known IC preferences:** Customer concentration thresholds, geography restrictions, business type preferences or aversions (e.g., "IC does not like software," "IC prefers asset-heavy businesses") +5. **Leverage comfort:** Maximum leverage the firm will consider + +If the credit box has been established in a prior conversation or configuration, use it without re-asking. + +--- + +## Screening Workflow + +### Step 1: Quick Read (Simulate the First 5-10 Minutes) + +Read the document end to end. Identify: +- What is the business? (sector, product/service, scale, geography) +- What is the opportunity? (new financing, refinancing, add-on, dividend recap, acquisition financing) +- What capital is being sought? (amount, structure, position in cap stack) +- Who is the sponsor? (if PE-backed) + +### Step 2: Hard Pass Check + +Check the opportunity against the firm's credit box for hard disqualifiers: +- Structure mismatch (e.g., seeking junior capital, firm only offers senior) +- Excluded industry +- Size outside mandate (EBITDA too small or too large) +- Known IC dealbreaker (customer concentration above threshold, geography exclusion) +- ESG restriction + +If any hard pass criteria are triggered, output **PASS** immediately with the specific reason. Do not continue the analysis. + +### Step 3: Soft Pass Check + +Check for soft disqualifiers — items that are not formal mandate exclusions but pattern-match to known IC aversions: +- Business type that IC has historically rejected +- Characteristics that raise well-known credit concerns for the firm +- Structural features that do not fit the firm's typical deal profile + +If soft pass criteria are triggered, note them but continue the analysis. These influence the final recommendation but do not automatically result in a PASS. + +### Step 4: Credit Assessment + +Analyze the opportunity across these dimensions: + +**Business quality:** +- Market position and competitive dynamics +- Revenue model (recurring vs. project vs. transactional) +- Customer concentration and diversification +- Growth trajectory and drivers +- Margin profile and stability + +**Financial profile (from whatever data is available):** +- Revenue scale and trajectory +- EBITDA and margins +- Free cash flow generation capacity +- Working capital dynamics +- Capex requirements + +**Leverage and structure (if provided, or build back-of-envelope):** +If capital structure and leverage are provided in the CIM, analyze them directly. If not, build a quick back-of-envelope model: +- Project EBITDA forward 3 years with reasonable assumptions for this type of business +- Plug in the proposed debt quantum and estimated market pricing +- Calculate cash interest expense and basic FCF +- Assess whether the business generates positive FCF at the proposed leverage +- If at market leverage the business does not generate positive FCF for the first 3 years, leverage is too high for this business +- Factor in business type: capital-intensive businesses with reinvestment requirements cannot support as much leverage even if near-term FCF looks acceptable + +**Risk identification (at this stage):** +- Top 3-5 risks visible from the CIM +- Potential mitigants if identifiable +- Open questions that would need answers before making a full credit decision + +### Step 5: Determine Output + +| Recommendation | When to Use | +|---|---| +| **PASS** | Hard pass criteria triggered, OR the credit assessment reveals fundamental issues that make the opportunity unattractive regardless of terms (unsustainable leverage, structural mismatch with the firm's lending philosophy, critical business risk with no visible mitigant) | +| **PURSUE** | No hard or soft pass criteria triggered. Business quality, financial profile, and proposed leverage are within acceptable parameters. Risks are identifiable and appear mitigable. The opportunity fits the firm's credit box and appears attractive. | +| **MORE INFO** | Cannot make a definitive call with the information provided. This is the default for teasers (2-5 pages) where you can check hard criteria but cannot assess the full credit picture. For full CIMs, this applies only if 2-3 critical gating questions remain unanswered. | + +**Bias toward PURSUE or MORE INFO.** Analysts should be biased toward learning more and spending time, not toward clearing opportunities quickly. Unless hard-no criteria are hit, lean toward continued evaluation. + +--- + +## Output Format + +Produce a 1-2 page narrative memo structured as follows: + +### 1. Summary (3-5 sentences) +What the business is, what the opportunity is, and the recommendation (Pass / Pursue / More Info) with the primary reason. + +### 2. What's Attractive (3-5 bullet points) +Business and financial merits identified from the CIM. Focus on characteristics that would resonate with a private credit IC: recurring revenue, strong margins, market leadership, low customer concentration, asset coverage, clear FCF generation. + +### 3. Key Risks (3-5 bullet points) +Risks identified at this stage. Pair with mitigants where possible. Acknowledge where mitigants are unknown pending further diligence. + +### 4. Financial Overview +Back-of-envelope summary tying the merits and risks to actual numbers: +- Revenue and EBITDA (current and trajectory) +- Proposed leverage and structure +- FCF assessment (positive/negative, magnitude) +- Covenant cushion estimate (if enough data) +- Sources & Uses if available + +### 5. Open Questions / Next Steps +If PURSUE: What are the next steps? (management call, data room access, model build) +If MORE INFO: What specific information is needed to make a decision? Frame these as specific asks to the broker or sponsor — not vague requests. +If PASS: State the reason clearly. No further analysis needed. + +--- + +## Calibration Notes + +- For a **teaser** (2-5 pages): You can only check hard criteria and form a very preliminary view. Default output is MORE INFO unless a hard pass is triggered. Frame the "More Info" as "this fits our mandate on the surface — requesting the full CIM to evaluate further." +- For a **full CIM** (50-100+ pages): You should be able to make a PURSUE or PASS call. MORE INFO is only appropriate if 2-3 truly critical questions cannot be answered from the materials provided. Even then, the default approach is to PURSUE and flag the gating items rather than wait. +- The output is a **mini investment memo** — it should be clean, concise, and useful enough to share with the IC or deal team to align on whether to proceed. diff --git a/private-credit/commands/covenant-check.md b/private-credit/commands/covenant-check.md new file mode 100644 index 0000000..b2408dd --- /dev/null +++ b/private-credit/commands/covenant-check.md @@ -0,0 +1,120 @@ +--- +description: "Full covenant compliance snapshot with headroom analysis" +argument-hint: "" +--- + +# /private-credit:covenant-check + +Produces a comprehensive covenant compliance snapshot: financial maintenance covenants with headroom, incurrence test levels, restricted covenant baskets, step-down schedule, and cure rights summary. + +--- + +## Trigger + +User invokes `/private-credit:covenant-check` with a borrower's compliance certificate and/or credit model. Optionally, the credit agreement or summary of key terms for incurrence and basket details. + +--- + +## Required Inputs + +| Input | Required? | Notes | +|---|---|---| +| Compliance certificate (current quarter) | **Yes** | Source of truth for actual covenant levels | +| Credit agreement or term sheet summary | **Yes** | For covenant definitions, step-down schedules, incurrence levels, baskets | +| Credit model (for historical quarters) | Recommended | Enables trend analysis across multiple quarters | + +--- + +## Calculation Approach + +### Financial Maintenance Covenants + +**Use the compliance certificate number directly.** The compliance cert is signed by the CFO or company officer and is the authoritative source for covenant calculations. + +Do NOT independently recalculate and substitute a different number. The cert number is the covenant test result. + +**Verification exercise:** Separately attempt to calculate the covenant metrics from the financial statements to verify you can reach the same number. If you cannot reconcile: +- For net debt: investigate OID treatment, lease classification, letter of credit inclusion — this should almost always be bridgeable +- For EBITDA: investigate addback detail, PF adjustments, retroactive adjustments — bridgeability depends on reporting quality +- If the gap cannot be explained, flag it as a question. Do not override the cert. + +--- + +## Output Structure + +### Section 1: Financial Maintenance Covenant Table + +| | Q1 2025 | Q2 2025 | Q3 2025 | Q4 2025 | Q1 2026 | Q2 2026 | +|---|---|---|---|---|---|---| +| **Net Leverage** | | | | | | | +| Covenant Threshold | 5.00x | 5.00x | 5.00x | 4.75x | 4.75x | 4.50x | +| Actual | 4.52x | 4.38x | 4.15x | 4.02x | — | — | +| Cushion | 9.6% | 12.4% | 17.0% | 15.4% | — | — | +| Status | ⚠️ Tight | ✅ | ✅ | ✅ | *step-down* | *step-down* | +| **FCCR** | | | | | | | +| Covenant Threshold | 1.10x | 1.10x | 1.10x | 1.10x | 1.15x | 1.15x | +| Actual | 1.18x | 1.22x | 1.28x | 1.31x | — | — | +| Cushion | 7.3% | 10.9% | 16.4% | 19.1% | — | — | +| Status | ⚠️ Tight | ✅ | ✅ | ✅ | *step-down* | — | + +**Key formatting rules:** +- Columns: quarterly progression left to right (oldest to newest, then 2-4 future quarters) +- Future quarters: show covenant threshold only (actuals blank). Reader sees upcoming step-downs visually. +- Cushion: positive = in compliance. Flag per headroom framework: + - ≥ 20%: no flag + - 10-20%: "⚡ Elevated attention" + - < 10%: "⚠️ Tight — flag to IC" +- Note trajectory in a summary line: "Leverage headroom: improving (9.6% → 17.0% over 3 quarters)" or "FCCR headroom: stable at 10-12%" + +### Section 2: Incurrence Test Levels + +| Incurrence Test | Trigger | Leverage Level | +|---|---|---| +| Additional Senior Debt | Incurrence of new pari passu debt | ≤ 4.00x Total Net Leverage | +| Additional Junior Debt | Incurrence of subordinated debt | ≤ 5.50x Total Net Leverage | +| Restricted Payments / Dividends | Distributions to equity holders | ≤ 3.50x Total Net Leverage | +| Junior Debt Payments | Payments on subordinated obligations | ≤ 4.00x Total Net Leverage | + +**Include current actual leverage for context:** "Current Total Net Leverage: 4.02x — restricted payment incurrence test (3.50x) is not currently available." + +### Section 3: Key Restricted Baskets + +Summarize the most important baskets from the credit agreement: + +| Basket | Limit | Current Utilization | Available | +|---|---|---|---| +| CapEx | $5.0M annually | $3.2M YTD | $1.8M remaining | +| Restricted Payments (general) | $2.0M per year + leverage-based | $0 used | $2.0M available | +| Permitted Investments | $3.0M in aggregate | $1.5M used | $1.5M available | +| Debt Incurrence (incremental) | $10.0M | $0 drawn | $10.0M available | + +Populate with actual figures if available from the compliance cert or credit agreement tracking. If utilization data is not available, show the limit and note "utilization not reported." + +### Section 4: Equity Cure Rights + +| Item | Detail | +|---|---| +| Total cures permitted | [X] over the life of the facility | +| Consecutive quarter limit | No more than [X] in any [Y] consecutive quarters | +| Cures exercised to date | [X] | +| Cures remaining | [X] | + +Present as a factual reference footnote. No analysis unless headroom is tight enough that cure exercise is a realistic near-term possibility. + +--- + +## Summary Statement + +Close with a 2-3 sentence summary: +- Overall compliance status +- Headroom trajectory (improving, stable, deteriorating) +- Any upcoming step-downs that will tighten headroom +- Any incurrence tests currently unavailable due to leverage levels + +Example: "Company X is in compliance with all financial maintenance covenants as of Q4 2025. Leverage headroom improved from 9.6% in Q1 to 15.4% in Q4, driven by EBITDA growth and modest deleveraging. Note: the 4.50x step-down in Q2 2026 will compress headroom to an estimated 10-11% at current run-rate — monitor closely." + +--- + +## Value of This Command + +This command eliminates the need to manually open the credit agreement, look up step-down schedules, cross-check compliance cert numbers, calculate headroom, and compile incurrence levels. One command produces the full covenant picture: current levels, cushion, trend, upcoming changes, incurrence availability, and cure rights — everything needed to assess the covenant position at a glance. diff --git a/private-credit/commands/ic-memo.md b/private-credit/commands/ic-memo.md new file mode 100644 index 0000000..784cae1 --- /dev/null +++ b/private-credit/commands/ic-memo.md @@ -0,0 +1,175 @@ +--- +description: "Generate a full IC memo from deal inputs (CIM + model + term sheet)" +argument-hint: "" +--- + +# /private-credit:ic-memo + +Produces a full investment committee memo in private credit format from deal inputs. The memo is written in advocacy tone — the team is recommending approval to proceed. + +--- + +## Trigger + +User invokes `/private-credit:ic-memo` with deal materials attached or referenced. + +--- + +## Minimum Required Inputs + +| Input | Required? | Notes | +|---|---|---| +| CIM or lender presentation | **Yes** | Foundation for business description and deal context | +| Financial model with base case and downside case | **Yes** | Must be the analyst's own model, not just the sponsor's case | +| Term sheet or summary of key credit terms | **Yes** | Can be 1-2 pages — detailed credit agreement is often post-IC | + +**Without a model and your own case analysis, do not produce an IC memo.** A memo built from the CIM alone is a reorganization of sell-side materials without independent analytical input. Inform the user that a model is required and offer to help build one using `/private-credit:scenario-analysis` first. + +## Inputs That Improve Quality + +| Input | Impact | +|---|---| +| Management call notes | Adds narrative depth, addresses risk mitigants, provides forward-looking context | +| Expert call findings | Industry validation, competitive positioning verification | +| Data cuts (customer, product, vendor, historical) | Deal-specific analysis beyond CIM | +| Comps analysis (public and private) | Valuation context, market positioning | +| Precedent deals (same industry or sponsor) | Structural benchmarking | +| Q&A responses from borrower or sponsor | Answers to diligence questions | + +If these additional inputs are available, incorporate them throughout the memo — not in a separate section. + +--- + +## Output Structure + +The IC memo follows the section order defined in the credit-memo-standards skill. Produce each section with the following specific guidance: + +### Section 1: Executive Summary / Deal Setup (1-2 slides equivalent) + +Write this section **last** — after all other sections are complete. It is a distillation of the entire memo. + +Must answer in the first few sentences: +- What is the business? (one sentence: sector, product/service, revenue scale) +- What is the opportunity? (new deal, refinancing, add-on — amount and structure) +- What are the headline terms? (leverage, spread, maturity) +- Why is this attractive? (2-3 strongest merits from Section 4) +- What are the key risks? (2-3 most important risks from Section 5) + +**Do not** open with company history. Do not spend the first paragraph on industry background. Lead with the deal. + +### Section 2: Sources & Uses + Capital Structure (1 slide equivalent) + +Two tables: + +**Sources & Uses:** +Pull directly from the CIM, lender presentation, or term sheet. If not explicitly provided, construct from available information: +- Sources: each debt tranche (committed amount), equity contribution, rollover equity, seller note, other +- Uses: enterprise value / purchase price, debt refinancing, transaction fees, OID, working capital, cash to balance sheet + +**Capital Structure:** +Each tranche with: facility type, committed amount, drawn amount, spread, OID, maturity, amortization schedule. Calculate and show: +- Senior leverage, total leverage (gross and net) +- LTV at entry (if enterprise value is available) + +### Section 3: Term Sheet Summary (1-2 slides equivalent) + +Summarize from the term sheet or CIM term sheet section: +- Financial maintenance covenants with levels and step-down schedule +- Key incurrence tests (leverage levels for restricted payments, additional debt) +- Important baskets (CapEx, restricted payments, investments, debt incurrence) +- Reporting requirements +- Equity cure rights +- Call protection and prepayment terms +- Change of control + +### Section 4: Investment Merits (multiple slides equivalent) + +4-6 merits. The first 1-3 are the strongest arguments for the deal. + +For each merit: +- Clear heading stating the merit +- 2-4 sentences of narrative explanation with specific evidence from the CIM, data cuts, or model +- Supporting data point, chart description, or table reference where applicable + +Draw from these categories as applicable: +- **Business strengths:** Market position, industry growth, switching costs, customer stickiness, recurring revenue, retention rates +- **Financial strengths:** Revenue CAGR, margin profile, FCF generation, low leverage relative to cash flow, asset coverage +- **Other:** Management quality, sponsor reputation, familiarity with industry or sponsor, attractive structure, clear deleveraging path + +**Always include a merit on FCF and deleveraging.** IC must see the path from entry leverage to a lower level over the projection period. State the base case deleveraging trajectory explicitly: "Base case projects deleveraging from X.Xx at close to Y.Yx by Year 3 through $ZM of cumulative FCF generation." + +### Section 5: Key Risks & Mitigants (multiple slides equivalent) + +3-5 risks specific to this deal. **Every risk must have a mitigant paired with it.** + +For each risk: +- Clear heading stating the risk +- 2-4 sentences explaining why this is a risk for this specific business and this specific structure (not generic platitudes) +- Mitigant: how the risk is addressed — structurally (covenant, basket, collateral), contractually (credit agreement provision), or by the business's characteristics + +**Do not include generic risks.** "An economic downturn could impact revenue" without deal-specific context is not useful. "Revenue concentration: top 3 customers represent 45% of revenue, and loss of the largest (22%) would reduce EBITDA by an estimated $4M, compressing covenant headroom from 18% to 6%" — that is a useful risk statement. + +**This section determines whether the deal survives IC.** If IC identifies a risk the team did not address, it signals insufficient diligence. Identify every risk IC is likely to raise and address it proactively. + +### Section 6: Historical Financials (1 slide equivalent) + +Table: 3 years of historical financials, quarterly if available. +- Revenue by segment (if provided) +- Gross profit and gross margin +- EBITDA and EBITDA margin +- Key balance sheet items (cash, debt, net debt) +- Key credit ratios (leverage, coverage) + +Narrative alongside the table: what drove performance changes, one-time events, trend context. + +### Section 7: EBITDA Adjustments Detail (1 slide equivalent) + +Bridge: Reported EBITDA → each adjustment category → Adjusted EBITDA. +- Show dollar amounts per category +- Show total adjustments as a percentage of Reported EBITDA +- Flag any category that appears recurring (3+ consecutive quarters at similar levels) +- Flag any category exceeding 5% of EBITDA +- Note cash vs. non-cash nature of each adjustment + +### Section 8: Model Output — Base Case and Downside Case (2-4 slides equivalent) + +**Base Case (1-2 slides):** +Key model outputs in table form: +- Revenue, EBITDA, EBITDA margin (3-year projection) +- FCF and cumulative FCF +- Leverage trajectory (entry → Year 1 → Year 2 → Year 3) +- Covenant compliance with headroom at each test date +- Liquidity position (cash + revolver availability) + +Narrative alongside: key assumptions, what drives the case, why the assumptions are reasonable. + +**Downside Case (1-2 slides):** +Same metrics as base case. Plus: +- Specific assumptions and why they are appropriate for this business +- The quarter where covenants breach (if they do) and the leverage at that point +- FCF at trough — positive or negative? Magnitude? +- Liquidity runway at trough +- Recovery trajectory: what does the path back look like? How many quarters to return to base case levels? + +Present the downside factually and descriptively. State the assumptions, state the outcomes, describe how the company is positioned to manage through the trough. IC evaluates whether the structure survives the downside — give them the facts to make that judgment. + +### Section 9: Appendix + +Include as applicable: +- Detailed company and business description +- Industry analysis and market sizing +- Deal-specific analysis: retention metrics, KPI deep-dives, fixed vs. variable cost structure, customer analysis, supplier analysis +- Public and private comparable companies (EV multiples, leverage, growth) +- Precedent transactions +- Management team bios + +--- + +## Tone Calibration + +The memo is **advocacy.** The team has done the diligence and is recommending the deal. The balanced evaluation happened at the CIM screen stage. + +- Executive summary, merits, and financial analysis: confident, clear, advocating for the credit +- Risk section: honest, thorough, balanced — but each risk has a mitigant. The message is "we identified the risks and here is how they are addressed" +- Downside case: factual and descriptive, not editorial. Present the scenario objectively and let IC evaluate +- Overall: professional, credit-focused, no filler language. Every sentence should either advance the argument or address a risk. If a sentence does neither, cut it. diff --git a/private-credit/commands/quarterly-review.md b/private-credit/commands/quarterly-review.md new file mode 100644 index 0000000..8b8d8b2 --- /dev/null +++ b/private-credit/commands/quarterly-review.md @@ -0,0 +1,170 @@ +--- +description: "Produce covenant table, variance summary, and management questions from quarterly package" +argument-hint: "" +--- + +# /private-credit:quarterly-review + +Produces a covenant compliance table, variance summary with MD&A narrative, and management call questions from a borrower's quarterly financial package. + +--- + +## Trigger + +User invokes `/private-credit:quarterly-review` with a borrower's quarterly financial package (Excel and/or PDF) and compliance certificate attached or referenced. The existing credit model for the borrower should also be available. + +--- + +## Required Inputs + +| Input | Required? | Notes | +|---|---|---| +| Quarterly financial package (Excel or PDF) | **Yes** | P&L, balance sheet, cash flow statement for the current quarter | +| Compliance certificate | **Yes** | May be inline with financials or a separate document (PDF) | +| Existing credit model for this borrower | **Yes** | The model being updated with the new quarter's data | +| Prior quarter's review (if available) | Recommended | Enables trend tracking and narrative continuity | +| Annual budget (if available) | Recommended | Enables vs. budget variance analysis | + +--- + +## Output 1: Updated Financial Summary + +### Variance Analysis Framework + +Apply comparisons in this order: + +**1. Quarterly Performance — Year-over-Year** +Compare the current quarter to the same quarter in the prior year. Never do quarter-over-quarter unless it is specifically relevant to the deal (e.g., a seasonal business where sequential comparison is meaningful). + +**2. Quarterly Performance — vs. Budget** +Compare the current quarter to the budgeted quarter (if budget is available). + +**3. Year-to-Date — vs. Prior Year and vs. Budget** +YTD actual vs. prior year YTD. YTD actual vs. budgeted YTD. + +**4. LTM — vs. Annual Budget or Prior LTM** +Rolling LTM vs. the annual budget. Rolling LTM vs. the prior period's LTM. + +### Variance Callout Thresholds + +| Line Item | Callout Threshold | Notes | +|---|---|---| +| **Revenue** | Always call out | Revenue is the most important line item. Note the variance regardless of magnitude, even if the dollar change is <5%. | +| **Gross profit (dollars)** | ≥ 5% dollar change | Below 5% dollar change YoY, not worth calling out unless part of a multi-quarter trend. | +| **Gross margin (percentage)** | Any notable movement | Margin percentage point changes are always significant. A 500bps swing (e.g., 45% → 40%) is massive and must be highlighted. Do not confuse the 5% dollar threshold with margin points — a 2-3 percentage point margin move is material even if the dollar change is small. | +| **EBITDA (dollars)** | ≥ 5% dollar change | Same threshold as gross profit on the dollar amount. | +| **EBITDA margin (percentage)** | Any notable movement | Same as gross margin — percentage point swings are always significant. A 200-300bps compression or expansion warrants a callout and explanation. | +| **Operating expense categories** | ≥ 5% dollar change | Call out which category is driving the variance (SG&A, R&D, sales & marketing, G&A). | +| **Working capital items** | ≥ 5% dollar change or notable trend | AR days, inventory days, AP days — call out if trending in a direction over multiple quarters. | +| **Capex** | ≥ 10% dollar change | Higher threshold — capex is lumpier by nature. | +| **Free cash flow** | Always call out | FCF is a critical credit metric. Note the variance regardless of magnitude. | + +### MD&A Narrative + +Write a narrative summary of the quarter's performance. This is not a table — it is a written analysis that explains the numbers. + +Structure: +1. **Revenue:** What happened this quarter? YoY comparison, vs. budget. What drove the result? Organic growth, acquisition contribution, new customer wins, customer losses, pricing changes. +2. **Margins:** Gross margin and EBITDA margin trends. If margins moved, explain why — COGS changes, OpEx changes, which categories. +3. **Cash flow and liquidity:** FCF this quarter. Working capital dynamics. Cash balance and revolver availability. Any notable cash events (distributions, capex spikes, acquisition payments). +4. **Credit metrics:** Leverage trend. Coverage ratio trend. Any movement toward or away from covenant thresholds. +5. **Notable items:** Acquisitions, restatements, new line items, management commentary themes, anything unusual. + +**For items that are unexpected or material (>20-30% variance), both weave into the narrative AND flag separately** so the analyst does not miss them. + +--- + +## Output 2: Covenant Compliance Table + +### Table Structure + +Columns progress left to right by quarter: Q1, Q2, Q3, Q4, etc. + +Rows for each covenant: + +``` +[Covenant Name] + Covenant Level (threshold) [step-down schedule values per quarter] + Actual Level [calculated from financials / cert] + Cushion (%) [headroom percentage] +``` + +Repeat for each financial maintenance covenant (typically net leverage, FCCR, and any others per the credit agreement). + +### What to Include + +- **Current quarter actuals** from the compliance certificate (source of truth) +- **Prior quarters** (4-8 quarters of history for trend visibility) +- **Covenant thresholds** including any step-downs that have taken effect +- **Upcoming step-downs** — show the threshold for the next 2-4 quarters even though actuals are blank. The reader should see tightening ahead. +- **Headroom percentage** calculated as: (Actual - Threshold) / Threshold for "max" covenants (leverage); (Actual - Threshold) / Threshold for "min" covenants (coverage). Express as a positive percentage when in compliance. + +### Headroom Flags + +Apply the headroom framework from the covenant-compliance skill: +- ≥ 20%: No flag +- 10-20%: Note "elevated attention" +- < 10%: Flag "tight — flag to IC" + +Note the trajectory: improving, stable, or deteriorating. + +### Equity Cure Rights + +Add a footnote: total cure rights under the agreement, number used, number remaining. + +--- + +## Output 3: Management Call Questions + +### Target Length + +Less than one page. More than 1-2 questions, fewer than 20. Typically 6-12 questions depending on the quarter. + +### Categorization + +Group questions by category. Order categories by priority: + +1. **Financial** — questions driven by the numbers (variances, trends, discrepancies) +2. **Strategic** — questions about business direction, market positioning, growth initiatives +3. **Operational** — questions about execution, cost management, integration (for acquisitions) +4. **Acquisition** — questions about tuck-in pipeline, integration progress, synergy realization (for roll-ups) + +Within each category, rank by importance — most critical question first. + +### Question Quality Standards + +Every question must demonstrate that the analyst understands the business and the financials. Apply second-order thinking: + +**Bad:** "EBITDA margin declined this quarter." +This is an observation, not a question. It does not demonstrate analytical depth. + +**Good:** "EBITDA margin compressed 250bps YoY. Gross margin was roughly flat, so the compression is in OpEx. SG&A grew 15% vs. 3% revenue growth — is this the sales team investment discussed in Q2? What is the expected timeline for revenue contribution from the new hires?" +This traces the margin compression to its source, links to prior management commentary, and asks a forward-looking question. + +### Prior Quarter Context + +Every question list must be informed by the prior quarter's: +- Financial performance and trends +- MD&A and management commentary +- Questions asked and answers received +- Commitments or guidance management provided + +If management guided to margin recovery by Q4 and margins are still compressing, ask specifically what changed. If a new initiative was announced in Q2, ask for a status update. + +Track the management team's narrative for consistency. If the story changes quarter over quarter without explanation, that is a question. + +--- + +## "Done" Checklist + +The quarterly review is complete when: + +- [ ] Financial data extracted from the quarterly package and entered into the credit model +- [ ] All model analysis sections updated for the latest quarter +- [ ] Variance analysis complete (YoY, vs. budget, YTD, LTM) +- [ ] MD&A narrative written covering revenue, margins, cash flow, credit metrics, notable items +- [ ] Covenant compliance table updated with current quarter actuals and headroom +- [ ] Material unexpected items flagged both in narrative and as standalone callouts +- [ ] Management call questions compiled, categorized, and ranked +- [ ] Questions reference prior quarter context and management commitments +- [ ] Package is ready for IC or portfolio review meeting diff --git a/private-credit/commands/scenario-analysis.md b/private-credit/commands/scenario-analysis.md new file mode 100644 index 0000000..d348333 --- /dev/null +++ b/private-credit/commands/scenario-analysis.md @@ -0,0 +1,160 @@ +--- +description: "Base / Downside / Stress case comparison with covenant breach flagging" +argument-hint: "" +--- + +# /private-credit:scenario-analysis + +Produces a Base / Downside / Stress case comparison table showing key credit metrics across cases, flags the quarter where covenants breach, and calculates liquidity runway per case. + +--- + +## Trigger + +User invokes `/private-credit:scenario-analysis` with a financial model or sufficient financial data to project forward. The user should specify the business type and key risk factors so the downside and stress cases can be tailored to the specific deal. + +--- + +## Required Inputs + +| Input | Required? | Notes | +|---|---|---| +| Financial model or current financials | **Yes** | Need a starting point for projections (revenue, EBITDA, debt, cash) | +| Debt terms (quantum, pricing, amortization) | **Yes** | Required for interest expense and debt service calculations | +| Covenant schedule (thresholds, step-downs) | **Yes** | Required to flag breach quarters | +| Business description or risk profile | **Yes** | Downside/stress assumptions must be tailored to the specific business | +| Sponsor/vendor base case (if available) | Recommended | Starting point for the lender base case | + +--- + +## Case Construction + +### Base Case + +If the sponsor/vendor case is available, use it as the starting point. The sponsor case is typically sandbagged — it is a conservative version of their realistic expectations, used as the basis for covenant setting. + +Options: +- **Use sponsor case as-is** if it appears reasonable +- **Build a lender case** by adjusting around the edges: a few percentage points less aggressive on revenue growth, slightly less FCF generation, marginally higher operating expenses. Not a fundamentally different view — a more conservative tilt. + +### Downside Case + +**Assumptions must be derived from the specific risk profile of the business.** Do not apply generic percentage haircuts. Identify the most relevant downside scenario for this business: + +| Business Profile | Downside Scenario | +|---|---| +| **Cyclical** (manufacturing, construction, industrials) | Model a full down-cycle. Anchor revenue decline to how this sector performed in 2008-2009. | +| **Customer concentrated** (top customer >15-20% of revenue) | Model loss of the top customer. Revenue impact + margin impact + working capital impact. | +| **Supplier dependent** (favorable contract, single-source) | Model loss of the supply relationship. Margins compress to industry standard. | +| **Acquisition dependent** (roll-up strategy) | Model a pause in acquisitions. Organic growth only. No PF EBITDA additions. Integration costs continue. | +| **Discretionary / consumer-facing** | Model a demand pullback anchored to 2008-2009 recession impact on comparable businesses. | +| **No specific risk driver** | Model a general economic recession. Revenue decline of 10-20% over 4-6 quarters with margin compression. Anchor to historical sector performance. | + +### Stress Case + +More severe than the downside. Combines multiple adverse factors: +- Revenue decline (larger than downside) +- Margin compression (on top of revenue decline) +- Working capital deterioration (cash consumption increases) +- Test: does the business survive? What breaks first? + +### Recovery Assumptions + +Recovery is not always V-shaped: +- **Cyclical downturn:** Typically V-shape or U-shape. Revenue declines then recovers over 2-4 quarters. Model the recovery to pre-downturn levels or close to it. +- **Structural impairment** (technology disruption, permanent market shift): The business may not recover to prior levels. Model a "new normal" at a lower base. +- **Customer loss:** Partial or full replacement over 4-8 quarters depending on the business's ability to acquire new customers. + +Show both the trough and the recovery trajectory in the output. + +--- + +## Output: Comparison Table + +### Metrics to Show (Across All Three Cases) + +Only include metrics relevant to the specific business. Do not show gross profit for a business where it is not a meaningful driver. + +**Standard metrics (always include):** + +| Metric | Year 1 | Year 2 | Year 3 | +|---|---|---|---| +| **Revenue** | | | | +| Base | | | | +| Downside | | | | +| Stress | | | | +| **EBITDA** | | | | +| Base | | | | +| Downside | | | | +| Stress | | | | +| **Total Debt** | | | | +| Base | | | | +| Downside | | | | +| Stress | | | | +| **Net Leverage** | | | | +| Base | | | | +| Downside | | | | +| Stress | | | | +| Covenant Threshold | | | | +| **Free Cash Flow** | | | | +| Base | | | | +| Downside | | | | +| Stress | | | | +| **Liquidity** | | | | +| Base | | | | +| Downside | | | | +| Stress | | | | + +If quarterly granularity is available and relevant (especially for identifying the specific quarter of covenant breach), show quarterly rather than annual. + +**Deal-specific additions** (include when relevant): +- EBITDA margin (if margin compression is a key feature of the downside) +- FCCR (if fixed charge coverage is a tested covenant) +- LTV (if enterprise value is available and relevant) + +### Covenant Breach Flagging + +For each case, identify: +- **Does a covenant breach occur?** Yes or No +- **Which covenant breaches first?** (typically net leverage) +- **In which quarter?** Specific quarter (e.g., "Q3 2027") +- **Leverage at breach:** The actual leverage level when the covenant trips + +Present as a leverage bridge showing the quarterly progression from current to breach: + +``` +Downside — Leverage Bridge: +Q4 2025: 4.02x (current) → Q1 2026: 4.25x → Q2 2026: 4.48x → Q3 2026: 4.62x [BREACH at 4.50x threshold] +``` + +Do not include cure analysis in the breach flagging. Keep it to the factual leverage progression. + +### Liquidity Runway + +Calculate liquidity for each case at each period: + +``` +Beginning Cash Balance + +/- Free Cash Flow Generated + +/- Mandatory Amortization (if not already in FCF) + +/- Revolver Draws or Repayments + +/- Additional Capital Sources (new debt, equity injection) + - Upcoming Maturities (if within the projection period) += Ending Liquidity (Cash + Available Revolver Capacity) +``` + +Present liquidity at each period. Flag the quarter where liquidity reaches its trough in each case and state the dollar amount. + +For the stress case specifically, state: "Liquidity runway: [X] months / quarters from trough before liquidity is exhausted" — if the stress case leads to a liquidity crisis. + +--- + +## Narrative Summary + +Close with a brief (3-5 sentence) comparison across cases: + +- **Base case:** Summarize the deleveraging path and FCF trajectory. "Base case projects deleveraging from X.Xx to Y.Yx by Year 3 with $ZM cumulative FCF." +- **Downside case:** Summarize what happens. "Downside case models [scenario]. Leverage peaks at X.Xx in [quarter], [breaching / approaching] the covenant threshold. The company [generates / consumes] $XM of FCF at trough. Recovery to base case levels by [quarter]." +- **Stress case:** Summarize the extreme outcome. "Stress case combines [factors]. Covenant breach occurs in [quarter] at X.Xx leverage. Liquidity reaches $XM at trough in [quarter], representing [X] months of runway." + +State the assumptions for each case clearly. The value is in well-thought-out, business-specific inputs — not generic percentage haircuts applied to a template. diff --git a/private-credit/hooks/hooks.json b/private-credit/hooks/hooks.json new file mode 100644 index 0000000..fe51488 --- /dev/null +++ b/private-credit/hooks/hooks.json @@ -0,0 +1 @@ +[] diff --git a/private-credit/skills/borrower-monitoring/SKILL.md b/private-credit/skills/borrower-monitoring/SKILL.md new file mode 100644 index 0000000..752532a --- /dev/null +++ b/private-credit/skills/borrower-monitoring/SKILL.md @@ -0,0 +1,209 @@ +# Borrower Monitoring + +This skill fires automatically when performing quarterly portfolio monitoring, updating borrower tracking, evaluating credit trajectory, managing watchlists, or preparing for management calls. + +--- + +## Quarterly Monitoring Workflow — Step by Step + +This is the complete workflow from the moment a borrower's quarterly package arrives to the moment the review is ready for IC or portfolio review. + +### Step 1: Receive and Review the Package + +Open the borrower's quarterly financial package. Read through the first time quickly to get a sense of what is being reported: +- What financial statements are included? +- Is there MD&A or management commentary? +- Is the compliance certificate included or separate? +- What data is available vs. what is missing? + +### Step 2: Extract and Update the Quarterly Model + +Working top-down through the P&L, balance sheet, and cash flow statement (per the quarterly-package-extraction skill): +- Enter the latest quarter's financial data into your quarterly Excel model +- Update all analysis sections within the model based on the new quarterly information +- Follow the existing template mapping from prior quarters + +### Step 3: Note Questions During Processing + +As you enter data and update the model, questions will naturally arise: +- Discrepancies between the financials and the compliance cert +- Unexpected trends or movements you cannot explain +- New line items or reclassifications +- Deviations from budget or from the prior quarter + +Compile these into a quarterly questions list as you work. Do not stop processing to investigate each one — note it and keep going. The questions list will be addressed in Step 6. + +### Step 4: Populate the Quarterly Presentation + +After the model is updated, build the full quarterly review package (typically PowerPoint): +- MD&A narrative discussing the quarter's financial performance +- Revenue and margin discussion with trend context +- Cash flow and liquidity discussion +- Covenant compliance summary +- Narrate each section of the model — do not just paste numbers without commentary +- Add any supplementary analysis performed outside the borrower's reporting: + - Industry analysis or market developments + - Public company comparable valuations (EV multiples, trading levels) + - Estimated enterprise value for the borrower based on current comps + +### Step 5: First Pass Complete + +At this point you have: +- Updated quarterly model +- Draft quarterly presentation with MD&A +- List of questions from processing + +### Step 6: Resolve Questions + +Send the compiled questions list to the management team, the agent (administrative agent for the facility), or whoever manages the relationship. Seek answers to every question on the list. + +Sources for answers: +- Direct management call or email exchange +- Agent or servicer (for facility-level questions) +- Additional research (industry data, public filings, news) +- Sponsor (for PE-backed borrowers) + +### Step 7: Incorporate Answers and Finalize + +Once questions are answered: +- Update the model if answers require adjustments (reclassifications, additional data points) +- Update the quarterly presentation with the additional context +- Incorporate management commentary into the MD&A narrative +- Update comps and supplementary analysis if new information warrants changes + +### Step 8: Done + +The quarterly review is complete when: +- Financial data is extracted and entered into the quarterly model +- All model analysis is updated for the latest quarter +- Questions identified during processing have been answered +- Answers are incorporated into both the model and the presentation +- The quarterly review package is ready for IC or portfolio review meeting + +--- + +## Credit Trajectory Evaluation + +### Evaluate Each Line Item Independently + +Credit trajectory is not a single number — it is an assessment of direction across multiple financial dimensions. Evaluate each independently because they can send conflicting signals: + +- **Revenue trajectory:** Growing, stable, or declining? What is the rate of change? +- **Cost of goods sold / Gross profit trajectory:** Are margins expanding, compressing, or stable? Is the company spending more to generate each dollar of revenue? +- **Operating expense trajectory:** Is OpEx growing faster than revenue? Which categories are driving it? +- **EBITDA trajectory:** Combining the above — is the bottom line improving or deteriorating? +- **Cash flow trajectory:** Is the business converting EBITDA to cash effectively? Are working capital or capex trends changing the cash conversion story? + +### Conflicting Signals + +Revenue can be growing while margins deteriorate. EBITDA can be stable while cash flow declines (due to working capital consumption or increased capex). The credit story is built from the combination of these individual trends, not from any single metric. + +When signals conflict, the analysis must explain why: +- "Revenue up 8% YoY driven by new customer acquisition, but EBITDA margin compressed 200bps due to sales team investment. Management guided to margin recovery in H2 as the sales team matures." +- "EBITDA stable at $15M but free cash flow declined 30% due to a $3M increase in working capital driven by extended receivable terms on two new enterprise contracts." + +### How Many Quarters to Call a Trend + +- **1 quarter:** Not a trend. An observation. +- **2 quarters:** Not a trend. A data point. Worth noting but not actionable as a trajectory. +- **3-4 quarters:** A trend. You can describe the trajectory with confidence and should adjust your forward view accordingly. + +--- + +## Watchlist Criteria + +### Two Categories Drive Watchlist Placement + +**Category 1: EBITDA and Leverage Deterioration** + +| Performance vs. Plan | Classification | Action | +|---|---|---| +| 10-15% below plan | Watch closely | Heightened monitoring. More detailed questions. Track trajectory. | +| 20-25% below plan | Watchlist | Approaching covenant levels. Formal watchlist designation. Active IC flagging. | +| Covenant breach | Watchlist / Workout | 100% on watchlist. May require workout classification depending on severity and cure options. | + +The borrower should be on the watchlist **before** a covenant breach, not after. The purpose of the watchlist is early identification. If a breach is the first time a borrower appears on the watchlist, the monitoring process failed. + +Exception: A sudden, unexpected large decline within a single period that causes an immediate breach without prior warning signs. This happens but should be rare. + +**Category 2: Cash Flow and Liquidity** + +Even if revenue is acceptable and leverage has not deteriorated significantly: +- Insufficient cash flow generation → watchlist +- Thin liquidity with no access to additional sources (fully drawn revolver, no additional capacity) → watchlist +- Upcoming maturity with no clear refinancing path → watchlist + +Cash flow and liquidity can deteriorate independently of EBITDA — working capital consumption, capex spikes, or one-time cash outlays can drain liquidity while the P&L looks fine. + +### Three-Tier System + +| Tier | Criteria | Response | +|---|---|---| +| **Watch closely** | Performance 10-15% off expectations. Early signs of deterioration. | Extra scrutiny on quarterly review. Deeper management questions. Track trajectory quarter by quarter. | +| **Watchlist** | Close to covenant breach (within 5-10% headroom). Performance 20-25% off plan. Liquidity thinning. | Formal IC flagging. Increased monitoring frequency if possible. Proactive engagement with management on remediation. Assessment of cure rights and amendment options. | +| **Workout** | Covenant breached. Significant performance deterioration. Liquidity crisis. | Active remediation plan. Amendment or waiver negotiations. Restructuring considerations. Frequent (monthly or more) updates to IC. | + +### What Gets a Borrower Off the Watchlist + +Sustained improvement over multiple quarters (3+ quarters of trend reversal), covenant headroom rebuilt to comfortable levels (>20%), liquidity restored, and the underlying cause of deterioration addressed — not just temporarily masked. + +--- + +## Management Call Questions + +### What Makes a Good Question + +A good management call question demonstrates that the analyst understands the business and the financials. After receiving the answers, the analyst should be able to answer any question IC might ask about the borrower's performance, trends, and key credit considerations. + +### Two Dimensions of Questions + +**Financial questions — driven by the numbers:** +- Identify discrepancies between periods or between financials and cert +- Identify trends that cannot be explained from the available data +- Dig into specifics: "AR days extended from 45 to 58 over two quarters — is this a change in collection terms, customer mix shift, or concentration in a few large receivables?" +- Unpack margin movements: "Gross margin compressed 300bps QoQ — is this COGS inflation, product mix shift, pricing pressure, or something else?" +- Follow the cash: "Working capital consumed $2M this quarter vs. generating $500K last quarter — what drove the swing?" + +**Narrative questions — driven by the business story:** +- Understand the story behind the numbers: why is revenue up, why is margin down, why is leverage higher or lower +- Probe specific events: "Revenue includes a $1.5M contract win with [Customer] — is this recurring or project-based?" +- Track management's narrative consistency: are they telling the same story quarter over quarter? If guidance from Q2 said margin recovery by Q4 and margins are still compressing, ask specifically what changed. + +### Second-Order Thinking + +The most common problem with junior analyst questions is that they stop at the first-order observation: +- Bad: "EBITDA margin declined this quarter." +- Good: "EBITDA margin declined 250bps QoQ. Gross margin was flat, so the compression is coming from OpEx. SG&A grew 15% while revenue grew 3%. Is this the sales team investment you discussed in Q2, or is there a new expense driver? And when should we expect the revenue contribution from this investment to materialize?" + +The second version shows the analyst traced the margin compression to its source, linked it to prior management commentary, and is asking a forward-looking question about when the investment pays off. + +### Prior Quarter Context Is Mandatory + +Every quarter, the analyst must refresh on the prior quarter's: +- Financial performance and trends +- MD&A and management commentary +- Questions asked and answers received +- Any commitments or guidance management provided + +Compare the current quarter's numbers and narrative to the prior quarter. Track whether management is delivering on what they said. If they guided to margin recovery and margins continued to decline, that is a question. + +When monitoring many portfolio companies, it is easy to forget details from the prior quarter. This is why written question logs and MD&A notes are essential — they provide continuity. + +--- + +## Performing Credit vs. Watchlist Credit + +The workflow is the same. The difference is scrutiny and depth. + +**Performing credit (shelf credit):** +- Standard quarterly review process +- Check-the-box questions if the business is performing well +- Less time investment per quarter — the company is doing fine, acknowledge it, update the model, move on + +**Watchlist credit:** +- Same process, significantly more scrutiny +- Sharper, more detailed questions on every financial line item +- More questions overall — dig into everything, not just the headlines +- Must understand the answers to all components — no surface-level acceptance +- Track management responses with more rigor — are they delivering on commitments? +- Higher priority in portfolio review — present with more detail and urgency to IC diff --git a/private-credit/skills/covenant-compliance/SKILL.md b/private-credit/skills/covenant-compliance/SKILL.md new file mode 100644 index 0000000..0d9f14a --- /dev/null +++ b/private-credit/skills/covenant-compliance/SKILL.md @@ -0,0 +1,136 @@ +# Covenant Compliance + +This skill fires automatically when working with financial covenants, compliance certificates, headroom analysis, or covenant testing for private credit borrowers. + +--- + +## Source of Truth + +The compliance certificate is always the source of truth for covenant calculations. The compliance certificate is a PDF document signed by the CFO or another company officer. Because it is executed by a company officer, it has been reviewed and is authoritative. + +When the borrower provides LTM Compliance EBITDA on the cert, use that number directly for covenant testing. Do not substitute your own independently calculated number. + +However, always attempt to reconcile the cert number by spreading quarterly financials and building up to the LTM figure independently. This reconciliation is a verification exercise, not a replacement. If you cannot bridge to the cert number, that is a question to raise — not an error to correct. + +--- + +## Reconciliation: Why the Numbers Often Don't Tie + +The reconciliation between your independent quarterly build-up and the compliance cert's LTM number is rarely straightforward. The following are the primary sources of discrepancy: + +### EBITDA Discrepancies + +1. **M&A during the LTM period.** When a borrower acquires a company mid-period, the compliance cert pro forma's the acquired company's EBITDA for the pre-acquisition portion of the LTM window. You will not have the acquired company's standalone quarterly financials to spread, so the PF adjustment appears as a plug. This is expected — not an error. + +2. **New adjustment initiatives.** The borrower may introduce new EBITDA addbacks that apply retroactively to the full LTM period but were not present in your prior quarterly spreads. A restructuring program started in Q3 may produce an addback applied to Q1-Q4 on the LTM cert. + +3. **Unclear reported-to-adjusted bridge.** Some borrowers report only Revenue → Gross Profit → Adjusted EBITDA with no detail in between. When the quarterly P&L does not show the full waterfall, you must make assumptions about D&A, pro forma adjustments, and addback allocations across quarters. These assumptions will introduce variance. + +4. **COGS-level adjustments.** Some addback categories include amounts embedded in cost of goods sold. The borrower reports gross profit as a single line (netting these out), then separately lists the full addback category on the compliance cert. A portion of the cert addback is already reflected in gross profit, causing an apparent double-count if you are not careful. + +### Net Debt Discrepancies + +1. **OID treatment.** The balance sheet reports debt net of original issue discount. The compliance certificate reports debt at face value. This difference is mechanical and explainable — not a flag. + +2. **Lease and financing treatment.** The credit agreement defines which obligations count as debt for covenant purposes. Equipment leases, capital leases, sale-leasebacks, and other financing arrangements may be included or excluded depending on the agreement. The balance sheet may classify these differently than the credit agreement. + +3. **Letters of credit.** Drawn and undrawn letters of credit may or may not be included in the covenant debt calculation depending on the credit agreement. + +4. **Revolver balance.** The compliance cert tests as of quarter-end. The revolver balance at quarter-end may differ from average or month-end balances shown elsewhere in the reporting. + +**Rule:** An analyst should almost always be able to bridge net debt between the audited financials, quarterly reporting, and the credit agreement. If you cannot bridge the debt number, something is wrong or missing — escalate. For EBITDA, the ability to bridge depends on the quality and detail of the borrower's reporting. + +--- + +## Headroom Framework + +Calculate covenant headroom as the percentage difference between the actual level and the covenant threshold. + +| Headroom Level | Classification | Action | +|---|---|---| +| ≥ 20% | Comfortable | Standard monitoring. No escalation required. | +| 10% – 20% | Elevated attention | Monitor closely. Note in quarterly review. Track trajectory. | +| < 10% | Tight — flag to IC | High priority. Flag in quarterly review as approaching breach territory. Prioritize management call questions around drivers. | + +### Trajectory Matters + +Headroom assessment must always be read in context of the trend: + +- **Improving** (e.g., 5% → 10% → 15%): Still flag if below 20%, but note the positive trajectory. IC needs to know the direction, not just the level. +- **Deteriorating** (e.g., 18% → 12% → 9%): Escalate early. Do not wait until sub-10% to flag. If the trend line points to sub-10% within one or two quarters, flag now. +- **Stable at tight levels** (e.g., 11% → 12% → 11%): The company is operating near the threshold consistently. Flag and assess whether this is structural or temporary. + +--- + +## Maintenance vs. Incurrence Covenants + +### Financial Maintenance Covenants + +Tested quarterly (or as specified in the credit agreement) based on the most recent LTM period. The compliance certificate reports the results. The borrower must remain in compliance at each test date. + +Standard maintenance covenants in mid-market private credit: +- **Net leverage** (Total Net Debt / LTM Adjusted EBITDA) — most common, usually the binding constraint +- **Fixed charge coverage ratio (FCCR)** (LTM EBITDA / (Cash Interest + Required Amortization)) — second most common +- **Minimum liquidity** (Cash + Revolver Availability ≥ threshold) — unusual; when present, it signals credit stress and is the highest priority covenant + +### Incurrence Covenants + +Not tested periodically. Triggered only when the borrower takes a specific action (incurring new debt, making a restricted payment, paying a dividend, making a junior debt payment). The borrower must demonstrate pro forma compliance with the incurrence level before taking the action. + +**What the analyst must do at close and maintain ongoing:** +- Extract all incurrence test levels from the credit agreement +- Build a reference table: what action is restricted, at what leverage level does the restriction lift, and what baskets are available +- Key incurrence levels to track: + - **Additional debt capacity:** At what leverage level can the borrower incur additional senior, junior, or pari passu debt? + - **Restricted payments / dividends:** At what leverage level can the borrower make distributions, pay dividends, or make junior payments? +- Monitor actual leverage against incurrence levels on a quarterly basis — even though the test is not periodic, you need to know if the borrower is approaching a level where they could take action + +--- + +## Step-Down Schedules + +Covenant step-downs almost always align with quarter-end reporting dates. A step-down effective January 1 applies to the test period ending March 31, with results due approximately 45 days after quarter-end. Mid-quarter step-downs are extremely rare (well under 1% of deals). + +### How to Apply Step-Downs + +Always look up the applicable covenant threshold by test date. Walk the step-down schedule and return the threshold from the most recent effective date that is on or before the test date. Never hardcode a single threshold — always reference the schedule. + +### Revolver Window Dressing + +The most relevant operational consideration for step-downs and covenant testing is **revolver manipulation at quarter-end.** + +Because covenants test as of quarter-end only, borrowers can: +- Draw up the revolver balance significantly during the quarter (increasing net debt) and repay it 1-2 days before quarter-end to reduce net leverage at the test date +- Withhold cash payments (vendor payments, discretionary spending) in the final weeks of the quarter so the cash balance is artificially high at quarter-end, reducing net leverage + +**Watch for this pattern when:** +- Headroom is tight (sub-15%) +- Revolver utilization shows sharp quarter-end drops followed by immediate re-draws +- Cash balances spike at quarter-end and immediately decline in the following month (if monthly reporting is available) + +--- + +## Covenant Severity Hierarchy + +When a borrower is tight on multiple covenants, prioritize in this order: + +1. **Minimum liquidity** — If present, this is the top priority. Minimum liquidity covenants are unusual in mid-market private credit. Their presence signals that lenders had specific concerns about cash flow or liquidity at origination. A liquidity covenant breach can trigger immediate acceleration — there is no cure through operational performance. + +2. **Net leverage** — The binding constraint in the vast majority of mid-market direct lending deals. Whether it is cov-lite (springing leverage on the revolver only) or a full maintenance covenant, leverage is the metric that drives most credit decisions. + +3. **Fixed charge coverage ratio (FCCR)** — Typically tested alongside leverage. Important, but leverage usually trips first because EBITDA declines hit the leverage numerator (debt is relatively fixed in the near term) before they flow through to fixed charge shortfalls. + +4. **Maximum capex** — Second-tier covenant. Relevant for capital-intensive businesses. Rarely the binding constraint but can catch borrowers who are investing through a downturn. + +**General principle:** Lower credit quality borrowers carry higher leverage and tighter covenant packages. The further down the credit spectrum, the more covenants you see and the tighter the thresholds. Many mid-market deals are cov-lite (leverage maintenance only, or springing). When covenants exist, net leverage is almost always the one that matters most. + +--- + +## Equity Cure Rights + +When documenting a covenant check, note the equity cure provision as a factual reference: +- How many cures are permitted under the credit agreement (typically 2-3 over the life of the facility, sometimes limited to 1-2 per 4-quarter period) +- How many have been used to date +- How many remain available + +Present this as a footnote or side note. Do not build cure mechanics into the covenant check output unless headroom is tight enough that a cure is a realistic near-term possibility. diff --git a/private-credit/skills/credit-memo-standards/SKILL.md b/private-credit/skills/credit-memo-standards/SKILL.md new file mode 100644 index 0000000..bed70da --- /dev/null +++ b/private-credit/skills/credit-memo-standards/SKILL.md @@ -0,0 +1,198 @@ +# Credit Memo Standards + +This skill fires automatically when drafting, reviewing, or structuring investment committee memos for private credit transactions. This covers new deal IC memos, amendment memos, and portfolio review memos. + +--- + +## IC Memo Section Order — New Deal + +The reader must understand what the deal is about within the first three slides. Every subsequent section adds depth. This is a PowerPoint-based presentation, not a Word document. + +### 1. Executive Summary / Deal Setup (1-2 slides) + +The first one to two slides must answer: +- What is the business? (one sentence: sector, product/service, scale) +- What is the opportunity? (new financing, refinancing, add-on, amendment) +- How much capital and at what terms? (headline leverage, spread, structure) +- Why is this attractive? (two to three bullet-level reasons) +- What are the key risks? (two to three bullet-level flags) + +This section is a mini-version of the entire memo. A senior IC member who reads nothing else should be able to form an initial view from these slides alone. Do not bury the lead. Do not spend the first slide on company history. + +### 2. Sources & Uses + Capital Structure (1 slide) + +One slide. Two tables. + +**Sources & Uses table:** +- Sources: Term Loan, Revolver (committed/drawn), equity contribution, rollover equity, seller note, any other capital +- Uses: Purchase price / enterprise value, refinancing of existing debt, transaction fees, OID, working capital, cash to balance sheet + +**Capital Structure table:** +- Each tranche: facility type, committed amount, drawn amount, spread, OID, maturity, amortization +- Leverage profile: senior leverage, total leverage, net leverage (gross and net of cash) +- LTV at entry +- Pricing detail: SOFR + spread, floor, OID, call protection + +### 3. Term Sheet Summary (1-2 slides) + +Key credit agreement terms in summary form. Not the full agreement — the items IC needs to evaluate the structure: +- Financial covenants: maintenance levels, step-down schedules, test frequency +- Incurrence covenants: leverage levels for restricted payments, additional debt, investments +- Key baskets: CapEx limits, restricted payment baskets, investment baskets, debt incurrence baskets +- Reporting requirements +- Equity cure rights (number, limitations) +- Change of control provisions +- Call protection / prepayment terms + +### 4. Investment Merits / Credit Highlights (multiple slides) + +4-6 points. The first 1-3 are your strongest arguments. The remaining are supporting. + +**One merit per slide** with narrative text and supporting graphs, tables, or data underneath. Two merits per slide only if both are concise and do not require extensive explanation. + +Categories of merits (not all will apply to every deal): + +**Business strengths:** +- Market leadership or dominant niche position +- Growing industry with secular tailwinds +- High barriers to entry or switching costs +- Sticky customer base with long-term contracts +- Recurring revenue model with high historical retention rates + +**Financial strengths** (business strengths expressed in numbers): +- Strong multi-year revenue CAGR +- New customer growth metrics +- High margin profile (indicative of value-add to customers) +- Low leverage relative to cash flow +- Strong free cash flow generation and conversion +- High asset coverage (for asset-heavy businesses) + +**Other merits:** +- Experienced management team with relevant track record +- Reputable sponsor with sector expertise +- Familiarity — portfolio includes similar businesses or prior deals with the sponsor +- Attractive deal structure and credit documentation +- Clear deleveraging path in base case through FCF generation + +**Always include:** FCF generation and deleveraging trajectory in the base case. IC must see the path from entry leverage to a meaningfully lower leverage level over the hold period. + +### 5. Key Risks & Mitigants (multiple slides) + +Same format as merits slides. Each risk paired with a mitigant. + +**One risk-mitigant pair per slide.** Narrative text explaining the risk, why it matters for this specific credit, and how it is mitigated (structurally, contractually, or by the business's characteristics). + +This is the section that determines whether the deal survives IC. The most common reason deals fail in committee is that the team did not identify the key risks, did not emphasize them sufficiently, or did not have mitigants prepared. If IC identifies a risk you did not address, it signals insufficient diligence. + +**Do not:** List generic risks that apply to every deal. "Economic downturn could impact revenue" without deal-specific context is useless. + +**Do:** Identify the 3-5 risks specific to this business, this structure, and this market that IC will ask about — and address each one directly. + +### 6. Historical Financials (1 slide) + +Summary historical financial performance. Minimum 3 years of history, quarterly basis if available. + +Include narrative descriptions alongside the table: what drove performance changes, any one-time events, trend commentary. + +### 7. EBITDA Adjustments Detail (1 slide) + +Bridge from Reported EBITDA to Adjusted EBITDA. Show every adjustment category, dollar amounts, and the aggregate adjustment as a percentage of Reported EBITDA. + +Flag any adjustment categories that are recurring at similar levels, any categories exceeding 5% of EBITDA individually, and the total adjustment load. + +### 8. Lender Case Model Output (2-4 slides) + +Base case and downside case. Each case gets 1-2 slides. + +Each slide: Excel model output (screenshot or embedded table) with narrative descriptions on the side explaining the key assumptions and takeaways. + +**Base case must show:** +- Revenue and EBITDA trajectory +- FCF generation and cumulative FCF +- Leverage trajectory (deleveraging path) +- Covenant compliance with headroom at each test date +- Liquidity position + +**Downside case must show:** +- The specific assumptions used and why they are appropriate for this business +- Revenue and EBITDA at trough +- FCF impact — does the business generate or consume cash at the trough? +- Leverage at trough and whether covenants are breached (and in which quarter) +- What the recovery looks like and how the company manages through the trough +- Liquidity runway at the trough + +Present the downside case factually and descriptively. The purpose is to show IC what happens if things go wrong so they can evaluate whether the business and structure can withstand it. Do not editorialize — state the assumptions, state the outcomes, and describe how the company is positioned to manage through it. + +### 9. Appendix + +Factual and supplementary materials. These support the memo but are not presented unless IC asks: +- Company and business detail (can draw from sell-side materials with added context) +- Industry analysis and market sizing +- Bespoke deal-specific analysis: retention metrics, KPI deep-dives, fixed vs. variable cost structure, customer concentration analysis, supplier analysis +- Comparable companies: public and private comps with EV multiples, leverage comps, market comps +- Precedent transactions from similar industry or from the same sponsor +- Management team bios + +--- + +## Tone and Voice + +An IC memo is written in an **advocacy tone.** By the time a deal reaches a full IC presentation, the team has already done the screening and decided this is worth pursuing. The balanced discussion happened earlier — at the 1-2 page CIM screen stage. + +The team is presenting to IC with the objective of requesting approval to proceed. The memo should be clear about that intent. This does not mean it is pushy or salesy — risks must be presented honestly and completely — but the overall framing is: "We have done the work. Here is why this is an attractive credit. Here are the risks and how they are mitigated. We recommend proceeding." + +If the team is not recommending approval, they should not be taking the deal to IC. + +--- + +## Senior vs. Junior Analyst Focus + +Understanding the division of labor helps calibrate where to invest the most effort: + +- **Junior analysts** own the descriptive and mechanical sections: financial data, historical performance, EBITDA adjustment detail, appendix materials, model build +- **Senior team members** focus on the executive summary, deal setup, presentation narrative for IC, investment merits framing, key risk identification and mitigation, and the overall "why this deal" argument +- The sections that kill deals in IC are the senior sections — merits and risks framing, not the data tables + +--- + +## Amendment Memos + +The format depends on the scope of the amendment: +- **Major amendment** (covenant reset, structural changes, significant repricing): Can be as comprehensive as a new deal IC memo with full financial analysis, updated model, and risk assessment +- **Minor amendment** (technical amendment, small basket adjustment): 1-2 pages summarizing the change, rationale, and impact +- **Portfolio review integration:** Small amendments may be discussed as part of the quarterly portfolio review rather than warranting a standalone memo + +There is no single template — the depth should match the significance of the change. + +--- + +## Portfolio Review Memos + +These are the quarterly monitoring deliverables: +- PowerPoint format +- Updated quarterly model with latest financial data +- MD&A narrative on financial performance +- Management discussion points and Q&A from the quarter +- Updated comparable company analysis +- Covenant compliance summary + +See the borrower-monitoring skill for the full quarterly monitoring workflow. + +--- + +## Minimum Input Requirements + +To produce a credible IC memo, Claude needs at minimum: +- CIM or lender presentation +- Financial model with base case and downside case (your own model, not just the sponsor's) +- Term sheet or summary of key credit terms + +To produce a comprehensive IC memo, Claude should also have: +- Management call notes +- Expert call findings +- Data cuts on customers, products, vendors, or historical performance +- Comps analysis (public and private) +- Precedent deal references from the same industry or sponsor +- Q&A responses from the borrower or sponsor + +A CIM alone is not sufficient. Without a model and your own case analysis, the memo is just a reorganization of sell-side materials without independent analytical input. diff --git a/private-credit/skills/credit-model-standards/SKILL.md b/private-credit/skills/credit-model-standards/SKILL.md new file mode 100644 index 0000000..8a2dd0f --- /dev/null +++ b/private-credit/skills/credit-model-standards/SKILL.md @@ -0,0 +1,193 @@ +# Credit Model Standards + +This skill fires automatically when building, reviewing, or modifying financial models for private credit transactions — including new deal underwriting models, quarterly monitoring models, and scenario analysis. + +--- + +## Model Structure — Tab Order + +The tab structure for a private credit model follows the same foundational architecture as a leveraged buyout model. Start with the LBO model structure as the base and modify for credit-specific needs. Every firm has its own template, but the fundamental structure is consistent. + +### Standard Tab Order (Left to Right) + +1. **Assumptions / Inputs** — All hardcoded assumptions in one place: revenue growth rates, margin assumptions, working capital assumptions, capex, tax rate, debt terms, pricing. Color-coded as inputs. + +2. **Income Statement** — Revenue by segment → COGS → Gross Profit → Operating Expenses by category → EBIT → D&A → EBITDA → Interest → Taxes → Net Income. Historical actuals and projected periods. + +3. **Balance Sheet** — Standard format. Working capital detail (AR, inventory, AP, accruals). Debt by tranche. Equity. Ensure debt is shown at face value in the model even if balance sheet reports net of OID. + +4. **Cash Flow Statement** — Operating cash flow → Investing → Financing → Net change in cash. Working capital changes flow from the balance sheet. Capex from assumptions. + +5. **Debt Schedule** — Beginning balance → draws → repayments → amortization → PIK additions → ending balance. One section per tranche (Revolver, Term Loan A, Term Loan B, any subordinated debt). Interest calculation per tranche (spread + base rate, floors, PIK if applicable). Mandatory amortization schedule. Revolver draw/repay mechanics. + +6. **Free Cash Flow Waterfall** — See detailed build below. This is a critical tab for private credit — it shows the business's actual ability to service debt and maintain liquidity. + +7. **Credit Statistics** — All key credit ratios calculated from the model outputs. See standard set below. + +8. **Covenant Compliance** — Covenant levels by test date, actual levels from the model, headroom calculation, pass/fail, step-down schedule. + +9. **Summary Output** — One-page view: high-level P&L flowing into FCF, key credit statistics, leverage trajectory, covenant summary. This is the tab the MD reads. + +10. **Cases / Scenarios** — Base, Downside, Stress. Each case has its own assumption set that feeds the model. Output comparison across cases on a summary page. + +### Credit-Specific Additions (vs. Standard LBO Model) + +- **Do not focus on equity returns.** IRR, MOIC, and equity waterfall tabs are PE-specific. A private credit model focuses on debt service capacity, not equity upside. +- **FCF waterfall must be explicit.** Do not bury free cash flow inside the cash flow statement. Build a dedicated section or tab that clearly shows EBITDA → FCF conversion. +- **Summary output tab is mandatory.** IC members and portfolio managers need a single page that tells the story. They will not flip through five tabs of detail. +- **Cases must be well-developed.** Downside and stress cases require specific, business-relevant assumptions — not just a percentage haircut on the base case. + +--- + +## Free Cash Flow Waterfall — Exact Build + +**Start with Reported EBITDA.** This is critical. Do not start with Adjusted EBITDA. + +Adjusted EBITDA includes addbacks that are not cash flow items. If you start with Adjusted EBITDA, you must separately subtract the addback amounts to avoid overstating cash flow. Starting with Reported EBITDA avoids this issue entirely. + +``` +Reported EBITDA + less: Capital Expenditures + (split Maintenance vs. Growth if available) + (split Capitalized Software separately for software companies) + plus/less: Change in Net Working Capital + (Change in AR + Change in Inventory - Change in AP - Change in Accrued Liabilities) + less: Cash Taxes + less: Cash Interest Expense + less: Mandatory Amortization Payments + less: Management Fee (if not already deducted above Reported EBITDA) + less: One-Time Cash Items (if known — highlight as separate line items in red) += Free Cash Flow +``` + +### Common Mistakes + +**Synergies and pro forma adjustments.** This is the single most important item to get right in the FCF waterfall. Leverage is set off of pro forma Adjusted EBITDA, which includes projected synergies. But projected synergies are not generating cash until they are realized and appear in the P&L. + +When building a model with synergies: +- The covenant compliance section uses Adjusted EBITDA (including synergies) for leverage calculation — this is per the credit agreement +- The FCF waterfall must reflect when synergies actually transition from "projected" to "realized in the P&L" +- Track: In which quarter do specific synergies begin contributing to Reported EBITDA? +- Before that quarter, they are addbacks for covenant purposes but do not generate cash — the FCF waterfall should not include them +- After that quarter, they are part of Reported EBITDA and flow naturally into FCF + +**Adjustments double-counting.** If you start from Adjusted EBITDA instead of Reported EBITDA, you must subtract each addback category in the FCF waterfall to get back to actual cash flow. This creates complexity and risk of omission. Starting from Reported EBITDA is cleaner. + +**Operating cash flow as sanity check.** Operating Cash Flow from the CFS less CapEx should be a reasonable proxy for FCF. Compare your FCF waterfall output to this simple calculation. If they diverge significantly, investigate. + +--- + +## Case Build Methodology + +### Starting Point: Sponsor/Vendor Case + +The private equity sponsor provides a base case (sometimes called "management case" or "vendor case"). This is the case used for covenant setting and pricing negotiations. + +**Important context:** The sponsor's base case is typically sandbagged — it is a conservative version of their internal realistic case. They present this to lenders because the more conservative the base case, the more covenant cushion they build against their actual expectations. This means the covenant levels are set relative to a case that the sponsor expects to outperform. + +### Lender Base Case + +Some firms use the sponsor case as-is for their base case. Others build a "lender case" that adjusts the sponsor case around the edges: +- A few percentage points less aggressive on revenue growth +- Slightly less free cash flow generation +- Marginally higher operating expenses + +The lender case is not a fundamentally different view — it is the sponsor case with a more conservative tilt to reflect the lender's perspective. + +### Downside Case — Business-Specific + +**Downside case assumptions must be derived from the specific risk profile of the business.** You cannot reuse a prior deal's downside template. Each downside case must be built from the unique dynamics of the business being analyzed. + +| Business Characteristic | Downside Assumption | +|---|---| +| Cyclical business (manufacturing, construction, industrials) | Model a full down-cycle with revenue decline consistent with historical recession performance for this sector | +| High customer concentration (top customer >15-20% of revenue) | Model loss of the top customer: revenue impact, margin impact, working capital impact | +| Supplier-dependent margins (single-source supplier, favorable contract) | Model loss of the favorable supply arrangement: margins compress to industry standard levels | +| Acquisition-dependent growth (roll-up strategy) | Model a pause in acquisitions: organic growth only, no PF EBITDA additions, integration costs continue | +| Discretionary/consumer-facing revenue | Model a demand pullback: revenue decline anchored to 2008-2009 recession performance for comparable businesses | + +**Generic downturn scenario:** When business-specific risks do not point to a clear downside driver, model a general economic recession. Anchor to how the business or comparable businesses performed during the 2008-2009 financial crisis. Model a similar decline in revenue and margin, followed by recovery. + +### Stress Case + +More severe than the downside. The stress case tests whether the structure survives an extreme scenario: +- Combine multiple adverse factors (revenue decline + margin compression + working capital deterioration) +- Test: In what quarter do covenants breach? What is leverage at trough? Does the business generate positive FCF at trough? What is the liquidity runway? + +### Recovery Assumptions + +Recovery is not always V-shaped: +- **Cyclical downturn:** Typically V-shape or U-shape recovery — revenue declines and then recovers as the cycle turns. Model the recovery period (usually 2-4 quarters from trough to pre-downturn levels). +- **Structural impairment:** If the downside scenario reflects a permanent market shift (e.g., technology disruption, regulatory change), the business may not recover to prior levels. Model a "new normal" at a lower revenue/margin base. +- **Customer loss:** Recovery depends on the business's ability to replace the lost customer. May be partial recovery or full replacement over 4-8 quarters. + +Show both the trough and the recovery trajectory in the output. IC needs to evaluate: (1) can the business survive the trough, and (2) what does the recovery path look like? + +--- + +## Debt Schedule + +Straightforward roll-forward for each tranche: + +``` +Beginning Balance + + Incremental Debt (new draws, add-on terms loans) + - Mandatory Amortization + - Voluntary Prepayments (from excess cash flow sweep or optional paydown) + + PIK Interest (if applicable — interest added to principal balance) += Ending Balance +``` + +Each tranche modeled separately: Revolver, Term Loan A, Term Loan B, mezzanine, subordinated notes, etc. + +Interest calculation per tranche: +- Base rate (SOFR) + spread +- SOFR floor (if applicable) +- PIK component (if applicable — separate cash pay vs. PIK) +- Commitment fees on undrawn revolver + +--- + +## Credit Statistics — Standard Set + +Every private credit model must include: + +| Metric | Calculation | Purpose | +|---|---|---| +| **Senior Net Leverage** | Senior Net Debt / LTM EBITDA | Primary leverage metric for senior lenders | +| **Total Net Leverage** | Total Net Debt / LTM EBITDA | Total leverage including subordinated debt | +| **Gross Leverage** | Total Debt (no cash offset) / LTM EBITDA | Leverage without cash benefit — shows structural exposure | +| **Fixed Charge Coverage Ratio (FCCR)** | LTM EBITDA / (Cash Interest + Mandatory Amortization) | Ability to cover required debt service payments | +| **Interest Coverage** | LTM EBITDA / Cash Interest Expense | Variation of FCCR — some agreements test interest only | +| **EBITDA less CapEx Coverage** | (LTM EBITDA - CapEx) / (Cash Interest + Mandatory Amortization) | Tighter coverage metric — accounts for required reinvestment | +| **Loan-to-Value (LTV)** | Total Debt / Enterprise Value | Risk profile relative to business value. Critical to track over time. | + +### LTV — Why It Matters + +LTV can deteriorate even when leverage is stable. If you originate at 6x leverage on a 10x EV business (60% LTV) and the industry re-rates to 8x EV multiples while leverage stays at 6x, LTV jumps to 75%. The risk profile has changed materially even though the P&L and leverage have not. + +Track LTV quarterly using updated comparable company multiples. Include in the credit statistics section and flag when LTV trends higher. + +### Deal-Specific Additions + +Depending on the business, you may need: +- **Tangible Net Worth** — for asset-heavy or regulated businesses +- **Borrowing Base metrics** — for ABL structures +- **Recurring Revenue coverage** — for software/SaaS businesses +- **Same-store growth** — for multi-location or franchise businesses + +--- + +## Model Hygiene + +Follow standard investment banking and private equity modeling best practices. There are no private credit-specific deviations from standard model hygiene rules: + +- **Input cells** (hardcoded assumptions): Blue font or yellow fill — consistently applied throughout +- **Formula cells:** Black font, never hardcoded values mixed with formulas +- **Linked cells** (references to other sheets): Green font or consistent indicator +- **Error checks:** Build error check rows at the bottom of each sheet (BS balances, CFS reconciles to BS cash, etc.) +- **No circular references.** If the model requires circularity (e.g., interest expense depends on cash balance which depends on interest expense), use an iterative calculation toggle or break the circularity with a prior-period approximation. +- **Assumptions organized in one place.** All key assumptions on the Assumptions/Inputs tab. Do not bury hardcoded numbers in formula cells throughout the model. +- **Sign convention consistent.** Pick a convention (positive = inflow, negative = outflow) and apply it everywhere. Document it on the first tab. +- **Period labels clear.** Every column header shows the period (Q1 2025, FY 2025, LTM Q3 2025, etc.) +- **Units labeled.** Every table header shows units ($000s, $M, %, x) diff --git a/private-credit/skills/ebitda-addback-treatment/SKILL.md b/private-credit/skills/ebitda-addback-treatment/SKILL.md new file mode 100644 index 0000000..c3b2f69 --- /dev/null +++ b/private-credit/skills/ebitda-addback-treatment/SKILL.md @@ -0,0 +1,191 @@ +# EBITDA Add-Back Treatment + +This skill fires automatically when working with EBITDA calculations, addback analysis, compliance certificate reconciliation, or any workflow that requires bridging from reported to adjusted EBITDA. + +--- + +## The EBITDA Waterfall — Build-Up Order + +Always build EBITDA from the bottom up. Start with Net Income and add back each component to verify you can reach Reported EBITDA before layering credit agreement adjustments. + +### Step 1: Build to Reported EBITDA + +``` +Net Income ++ Income Tax Expense (use actual tax, not provision adjustments) ++ Depreciation & Amortization ++ Interest Expense += Reported EBITDA +``` + +This number should tie to whatever the borrower labels as "EBITDA" or "Operating EBITDA" in their financials. If it does not, investigate before proceeding. Common causes of mismatch: the borrower may include or exclude items below the operating line differently than you expect (FX gains/losses, other income/expense, non-cash impairments). + +### Step 2: Layer Credit Agreement Adjustments + +From Reported EBITDA, apply the adjustments permitted under the credit agreement. The compliance certificate will itemize these if the borrower provides a detailed cert. Categories typically include: + +``` +Reported EBITDA ++ Non-cash charges (stock-based comp, non-cash rent, impairments) ++ Permitted restructuring / transformation addbacks (subject to caps) ++ Management fees (if not already above the line) ++ Transaction / M&A costs ++ Pro forma acquisition EBITDA (for tuck-ins during the LTM period) ++ Run-rate cost savings / synergies (subject to credit agreement terms) ++ Other permitted addbacks per credit agreement += Adjusted EBITDA (Covenant Compliance EBITDA) +``` + +### Step 3: Track on an LTM Basis + +Best practice is to track Reported EBITDA on a rolling LTM basis and then layer adjustments separately. This lets you: +- See the trend in the underlying business before adjustments +- Identify when new adjustments are introduced or existing ones change materially +- Verify that LTM Adjusted EBITDA ties to the compliance certificate + +If the LTM build-up does not tie to the cert, investigate. Do not assume an error — it is usually an explainable difference (see reconciliation challenges below). + +### Sanity Check: Cash Flow Statement Cross-Reference + +As a secondary verification, calculate EBITDA from the cash flow statement: + +``` +Net Income (from CFS) ++ Income Taxes (add back) ++ Interest Expense (add back) ++ Depreciation & Amortization (add back from CFS) += EBITDA (CFS-derived) +``` + +Compare this to your P&L-derived Reported EBITDA. They should be close. Differences arise from non-cash items classified differently between the P&L and CFS, but this is a useful gut check on magnitude. + +--- + +## Reconciliation Challenges + +### Worst-Case Reporting + +Some borrowers provide only: Revenue → Gross Profit → Adjusted EBITDA. No detail on operating expenses, no D&A breakout, no interest or tax detail. When you receive this: +- You cannot independently build to Reported EBITDA +- You cannot strip adjustments to see underlying performance +- Flag this as a reporting limitation and note that independent verification of EBITDA adjustments is not possible with the information provided +- Push for more detailed reporting in management calls or through the agent + +### Acquisitions in the LTM Period + +When a borrower completes an acquisition during the LTM period, the compliance cert will include a pro forma adjustment for the acquired company's pre-acquisition EBITDA. This acquired EBITDA figure is usually itself adjusted (the acquired company's adjusted EBITDA, not reported). You will not be able to independently decompose this into the acquired company's revenue, expenses, and addbacks on a historical basis. + +**How to handle:** +- Accept the pro forma acquired EBITDA as presented on the cert — it is CFO-signed +- Note it as a line item in your tracking: "Acquired EBITDA — [Company Name] — $X.XM" +- In your internal analysis, be clear about how much of total Adjusted EBITDA comes from organic operations vs. pro forma acquisition contribution +- Track whether acquired EBITDA contributions are growing, stable, or declining in subsequent quarters as the acquisitions are fully integrated and the PF period rolls off + +### COGS-Level Adjustments + +Some addback categories include amounts that sit within cost of goods sold. The borrower reports gross profit as a single line (net of these), then lists the full addback amount on the compliance cert. A portion of the cert addback is already reflected in gross profit, which can create an apparent mismatch. + +**How to handle:** +- If the P&L shows gross profit only (no COGS detail), accept that you cannot fully decompose the addback between COGS and OpEx +- Note the limitation +- Focus on whether the total addback amount per the cert is reasonable, not whether you can allocate it across P&L sections + +### New Adjustments Applied Retroactively + +Borrowers may introduce a new addback category in Q3 and apply it retroactively to Q1 and Q2 for LTM calculation purposes. This causes your prior quarterly spreads to not tie to the new LTM cert. + +**How to handle:** +- Update your tracking to include the new category +- Note when it was introduced and the retroactive amount +- Flag if the retroactive application is material (>5% of EBITDA) + +--- + +## Cash vs. Non-Cash: Critical Distinction + +When evaluating a borrower's true cash flow generation ability, you must understand which EBITDA adjustments are cash and which are non-cash. + +- **Non-cash addbacks** (stock-based comp, non-cash rent, impairments): These inflate Adjusted EBITDA but do not represent additional cash the business generates. Reported EBITDA, which already includes these non-cash items as expenses, may be a better proxy for the business's true cash-generating capacity. +- **Cash addbacks** (restructuring charges actually paid, transaction costs, management fees paid in cash): These represent real cash outflows that the credit agreement permits to be added back for covenant purposes, but the cash still left the building. + +**Rule:** When assessing the borrower's ability to service debt, generate free cash flow, and maintain liquidity, always distinguish between cash and non-cash adjustments. Adjusted EBITDA per the cert is the right number for covenant compliance. It is not necessarily the right number for assessing debt service capacity. + +--- + +## Addback Caps + +Credit agreements typically impose caps on certain addback categories. Common structures: + +- **Dollar cap:** "Non-recurring charges capped at $2M per annum" +- **Percentage of EBITDA cap:** "Restructuring charges capped at 15% of LTM EBITDA" +- **Basket cap:** "Total permitted addbacks (excluding pro forma acquisition adjustments) capped at 25% of EBITDA" + +### How to Handle + +- Pull caps from the credit agreement at close and record them in your tracking +- The compliance certificate will usually state the applicable cap +- CFOs manage to the caps and are aware of them — this is not an area where analysts need to police +- If adjustments exceed the cap, the borrower will likely request an amendment to increase it +- Note the cap and the current utilization level for reference, but do not flag unless the borrower appears to be structuring around the cap (splitting adjustment categories to avoid triggering) + +--- + +## Push-Back Framework + +You cannot change what the borrower reports on their compliance certificate. Push-back happens in two places: (1) conversations with the management team, and (2) internal credit discussions when you assess the borrower's true free cash flow generation capacity. + +### When to Highlight and Potentially Exclude + +Apply this framework to each addback category: + +| Condition | Action | +|---|---| +| Same addback appearing at similar levels for 3-4 consecutive quarters | Not one-time. Highlight. Consider excluding from internal FCF view. | +| Single category > 5% of EBITDA | Noteworthy. Highlight. Investigate what is included. | +| Single category > ~$500K on a $15-20M EBITDA business | Definitely highlight. May exclude if you believe it is not legitimate. | +| Single category < ~$100K on a $15-20M EBITDA business | Immaterial. Ignore unless pattern across multiple small categories. | +| Multiple categories individually < 5% but aggregating to > 10-15% | Flag the aggregate. Individually small does not mean collectively immaterial. | + +### Most Abused Addback Category + +**"Restructuring / Transformation / Operational Improvement / M&A Expenses"** — This category is the most broadly defined in most credit agreements. Companies can classify a wide range of expenses under these labels. Key watchpoints: +- Is the dollar amount consistent quarter over quarter? If so, it is a run-rate cost, not a one-time charge. +- Are new initiatives continually being added as old ones are supposedly completed? +- Can management specifically identify the projects, their expected cost, and their expected completion date? + +### Tracking Realization + +A good analyst tracks addback realization across quarters: +- What adjustments were flagged as temporary or one-time? +- Have they rolled off, or are they persisting? +- Are new adjustments replacing old ones at similar dollar levels? +- If "restructuring" was $1.5M in Q1 and $1.4M in Q4, with different underlying projects cited each quarter, the total dollar exposure has not actually declined — the company is spending at a run-rate pace and relabeling. + +Flag this trend and notify if new adjustments are coming on or existing ones are trending higher instead of rolling off as expected. + +--- + +## Run-Rate Synergies and Cost Savings + +### Credit Agreement Governs + +The credit agreement specifies the permitted realization window for run-rate synergies and cost savings — typically 12 to 24 months from the date of the action giving rise to the synergy. Do not apply a different timeline than what the agreement specifies unless it has been amended. + +### Tracking Realization vs. Projection + +When the borrower projects the same synergies quarter after quarter without visible P&L improvement: +- Ask specifically: What has been actioned? What is on the come? +- For actioned items: When will the impact appear in reported financials? +- For on-the-come items: What is the timeline and what milestones must be hit? +- Track projected synergies vs. actual realized savings on a quarterly basis +- If the gap between projected and realized is not closing, flag it — the synergies may not materialize + +### Pro Forma Acquisition Synergies (Roll-Ups) + +For acquisition-intensive businesses, pro forma synergies layer on top of pro forma acquired EBITDA. Each tuck-in may come with its own synergy estimate. Track: +- Which acquisitions have associated synergy projections +- What the dollar amount is per acquisition +- Whether realized synergies are appearing in the P&L as the acquisitions integrate +- The total aggregate synergy number vs. what has been achieved + +This is especially important for roll-up businesses where EBITDA can include multiple layers of pro forma adjustments and synergies from acquisitions at different stages of integration. diff --git a/private-credit/skills/quarterly-package-extraction/SKILL.md b/private-credit/skills/quarterly-package-extraction/SKILL.md new file mode 100644 index 0000000..91243c1 --- /dev/null +++ b/private-credit/skills/quarterly-package-extraction/SKILL.md @@ -0,0 +1,184 @@ +# Quarterly Package Extraction + +This skill fires automatically when extracting financial data from borrower quarterly packages, compliance certificates, or any borrower-provided financial reporting for the purpose of updating a credit model. + +--- + +## Extraction Order + +Extract financial data top-down, following the natural order of the financial statements: + +### Step 1: Income Statement / P&L + +Start at the top of the P&L and work down: +- Revenue (by segment or line of business if provided) +- Cost of goods sold / Cost of revenue (if provided — many borrowers report gross profit only) +- Gross profit +- Operating expenses (by category if provided: SG&A, R&D, sales & marketing, G&A) +- Operating income / EBIT +- Depreciation & Amortization (may be embedded above or shown separately) +- EBITDA (reported) +- Adjustments (if shown on the P&L) +- Adjusted EBITDA (if shown) +- Interest expense +- Tax expense +- Net income + +Extract whatever the borrower provides. If they report only Revenue → Gross Profit → Adjusted EBITDA, extract those three lines and note the gaps. + +### Step 2: Balance Sheet + +- Cash and cash equivalents +- Accounts receivable +- Inventory (if applicable) +- Other current assets +- Total current assets +- Property, plant & equipment +- Goodwill and intangibles +- Other long-term assets +- Total assets +- Accounts payable +- Accrued liabilities +- Current portion of long-term debt +- Other current liabilities +- Total current liabilities +- Long-term debt (by tranche if provided) +- Other long-term liabilities +- Total liabilities +- Total equity + +Pay attention to how debt is reported — net of OID on the balance sheet vs. face value on the compliance certificate. Note any lease obligations, sale-leaseback financing, or other off-balance-sheet items that may count as debt for covenant purposes. + +### Step 3: Cash Flow Statement + +- Net income +- D&A add-back +- Stock-based compensation +- Changes in working capital (AR, inventory, AP, accrued liabilities, other) +- Operating cash flow +- Capital expenditures (split maintenance vs. growth if provided) +- Capitalized software (for software companies) +- Other investing activities +- Debt draws / repayments +- Equity contributions / distributions +- Other financing activities +- Net change in cash + +### Step 4: Compliance Certificate + +- Extract all covenant tests: covenant name, actual level, threshold, pass/fail +- Extract the compliance EBITDA calculation if provided (build-up from net income or reported EBITDA with each addback category) +- Extract net debt as calculated for covenant purposes +- Note any items where the cert calculation differs from what you would expect based on the financial statements + +--- + +## Excel vs. PDF — Same Workflow + +The workflow is identical regardless of whether the source is Excel or PDF: +1. Open the source document +2. Review the financials +3. Type the numbers into your own credit model + +Even when the borrower provides an Excel file, it is safer and cleaner to type the numbers manually rather than copy-paste. Copy-paste can introduce hidden formatting issues, linked references to the source file, or misaligned rows. Manually entering the numbers forces the analyst to look at each line item and consider how it maps to the model. + +--- + +## Label Matching and Template Customization + +Borrower financials almost never use the same labels as your credit model template. The correct approach is to customize the template to match the borrower's reporting, not to force the borrower's data into rigid template labels. + +### First-Time Setup (New Borrower) + +When building a quarterly model for a new borrower for the first time: +- Start from your firm's base template (general structure, standard sections) +- Modify line item labels to match the borrower's actual reporting +- Add line items as needed — if the borrower has 10 revenue segments and your template has 2, add the additional segments and bucket the remainder into "Other" +- Apply the same customization to every section: revenue, cost structure, margins, operating expenses, EBITDA adjustments, balance sheet, cash flow +- This first-time setup is the heaviest lift in the monitoring workflow + +### Subsequent Quarters + +After the first quarter is set up: +- The template already matches the borrower's reporting structure +- Each subsequent quarter, you repeat the same mapping and input process +- Much faster and lighter — you already know where every line item goes +- Only adjust if the borrower changes their reporting format (new segments, reclassifications, restatements) + +### Unmapped Line Items + +When a borrower introduces a new line item or label you have not seen before: +- Do not guess where it maps +- Flag it: "New line item — [label] — $X.X — requires mapping decision" +- Determine whether it is a reclassification of an existing item, a genuinely new item, or an acquisition-related addition +- Add to the template once the mapping is confirmed + +--- + +## Handling Discrepancies Between Financials and Compliance Certificate + +### Rule: The Compliance Certificate Is the Source of Truth + +The compliance certificate is a signed PDF document, executed by the CFO or another company officer. It is the authoritative source for covenant calculations. The Excel financial package is less formal. + +### Discrepancies Are Not Errors + +When numbers in the financial package differ from the compliance certificate, this almost always reflects a reporting difference, not an error: +- Debt reported net of OID on the balance sheet vs. face value on the cert +- Lease or financing obligations included in covenant debt per the credit agreement but classified differently on the balance sheet +- Letters of credit included or excluded based on credit agreement definition +- EBITDA adjustments visible in the cert but not separately identifiable in the P&L + +**The analyst's job:** Identify the discrepancy, attempt to bridge it using the credit agreement definitions, and if the bridge is not possible with available information, raise the question with the management team or agent. Do not assume the cert is wrong. Do not assume the financials are wrong. Understand why the numbers differ. + +### Debt Should Always Be Bridgeable + +An analyst should almost always be able to reconcile the debt number between the balance sheet, the compliance certificate, and the credit agreement terms. The ingredients are: balance sheet debt, OID adjustment, lease/financing treatment per the agreement, and letter of credit inclusion. If you cannot bridge the debt, something is missing — escalate. + +### EBITDA May Not Be Fully Bridgeable + +The ability to reconcile EBITDA depends entirely on the quality of the borrower's reporting. With detailed reporting (full P&L down to net income, itemized adjustments), you can bridge. With minimal reporting (Revenue → Gross Profit → Adjusted EBITDA), you cannot — flag the limitation and work with what is available. + +--- + +## Common Data Quality Issues + +### Minimal Reporting + +The most common issue is not data errors but **insufficient data.** Borrowers report the bare minimum required by the credit agreement: +- No historical comparisons unless required +- No budget comparison unless required +- No segment detail unless required +- Summary-level financials with limited line item granularity + +**Analyst response:** Assess what you can build with the data provided. Identify what additional information would be needed for a complete credit assessment. Frame questions to management or the agent to fill the gaps. + +### Acquisitions Embedded in Reporting + +For acquisition-active borrowers, a common trap is that acquired company financials are folded into the reporting without restatement of prior periods. This means: +- Revenue and EBITDA jump significantly from one quarter to the next +- The increase is not labeled as acquisition-driven — it just appears as growth +- QoQ and YoY comparisons become misleading because prior periods do not include the acquired business + +**Analyst response:** +- Track all acquisitions with close dates +- When financials show a material jump, check whether an acquisition closed during the period +- Ask for organic vs. inorganic breakdowns if not provided +- Adjust your QoQ and YoY analysis to separate organic growth from acquisition contribution where possible +- Note in your quarterly review when comparisons are distorted by acquisitions + +### Other Issues (Rare but Worth Noting) + +- **Mixed units:** Thousands vs. actuals vs. millions. Should be obvious from table headers and labels. Almost never an issue in practice. +- **Missing quarters:** If the borrower simply does not report a quarter, there is nothing to extract. Flag the gap and note it in the quarterly review. +- **Restated numbers:** Rare. If prior period numbers change without explanation, flag and investigate. This could indicate a restatement, a reclassification, or an error in the prior period reporting. + +--- + +## Monthly to Quarterly / LTM Aggregation + +When borrowers report monthly: +- Aggregate monthly figures into quarterly totals (3 months per quarter) +- Aggregate quarterly totals into LTM figures (4 quarters) +- No special adjustments or gotchas in the aggregation itself +- Consolidate, aggregate, and analyze — straightforward mechanical exercise