本文へ移動
cccskills
無料GitHub で公開

building-leveraged-buyout-models

Constructs LBO models with sources/uses, debt schedules, operating projections, and returns analysis across entry/exit scenarios. Use when modeling leveraged buyouts, calculating sponsor returns, or analyzing leverage capacity.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md5.5 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Building Leveraged Buyout Models

When To Use

  • Modeling a sponsor-backed acquisition to evaluate feasibility and returns
  • Sizing debt capacity and structuring capital across tranches (senior, sub, mezzanine)
  • Stress-testing leverage and coverage ratios under downside scenarios
  • Comparing entry/exit multiple assumptions to derive IRR and MOIC targets
  • Evaluating management rollover, co-invest, or earnout structures alongside sponsor equity

Inputs To Gather

  • Target financials: Last 3 years of revenue, EBITDA, capex, working capital; current-year budget or LTM figures
  • Transaction terms: Purchase price or EV/EBITDA entry multiple, transaction fees (advisory, financing, legal), minimum cash on balance sheet
  • Debt terms: Tranche sizes (revolver, Term Loan A/B, senior notes, mezzanine), interest rates (fixed vs. floating + spread), amortization schedules, mandatory prepayment sweep percentages, commitment fees
  • Operating projections: Revenue growth rates, margin assumptions, capex as % of revenue, working capital days (DSO, DIO, DPO), tax rate [VERIFY jurisdiction-specific rates]
  • Exit assumptions: Holding period (typically 3–7 years), exit multiple range, expected dividend recaps if applicable
  • Equity structure: Sponsor equity contribution, management rollover %, option pool or incentive equity allocation

Workflow

  1. Build sources & uses table

    • Uses: enterprise value, refinanced debt, transaction fees (banker, legal, financing), cash to balance sheet
    • Sources: each debt tranche sized to leverage multiple (e.g., Senior at 4.0x EBITDA, Sub at 1.5x), sponsor equity as residual plug
    • Cross-check total sources = total uses
  2. Construct the debt schedule

    • For each tranche: opening balance, mandatory amortization, optional prepayments (from excess cash flow sweep), closing balance
    • Calculate interest expense per tranche per period (handle PIK vs. cash-pay separately)
    • Model revolver draws/paydowns based on minimum cash balance constraint
    • Track total leverage (Total Debt / EBITDA) and senior leverage through the hold
  3. Build operating model projections

    • Project revenue, EBITDA, and EBIT from base-case growth and margin assumptions
    • Derive unlevered free cash flow: EBITDA − taxes on EBIT − capex − change in net working capital
    • Apply mandatory debt service and sweep mechanics to determine cash available for optional prepayment
    • Confirm debt service coverage ratio (DSCR) exceeds covenant minimums each period [VERIFY lender covenant thresholds]
  4. Calculate returns at exit

    • Apply exit multiple to projected EBITDA at end of hold period
    • Subtract net debt at exit to derive equity value to sponsors
    • Compute gross IRR and MOIC on sponsor equity (include any interim dividends or recap proceeds)
    • Back into implied entry vs. exit multiple arbitrage contribution to returns
  5. Run sensitivity and scenario analysis

    • Two-way sensitivity tables: entry multiple vs. exit multiple, revenue growth vs. margin, leverage vs. exit multiple
    • Downside case: revenue miss of 10–20%, margin compression of 100–200 bps, no multiple expansion — confirm no covenant breach and positive equity value
    • Upside case: outperformance + earlier exit or dividend recap — show max returns
    • Flag any scenario where leverage ratio exceeds typical market thresholds (>6.0x total for middle-market, >7.0x for large-cap) [VERIFY current market leverage norms]
  6. Document the model package

    • Assumption summary page: entry multiple, leverage, growth, margins, exit multiple
    • Key outputs page: IRR/MOIC at base/down/up, credit statistics through the hold, FCF conversion
    • Clearly label cells as input (blue font), formula (black), or linked (green) per modeling convention

Output

  • Sources & Uses table with complete reconciliation
  • Debt schedule showing each tranche's balance, interest, amortization, and covenants through the hold
  • Operating projections (income statement through unlevered FCF) for the hold period
  • Returns summary: IRR, MOIC, and cash-on-cash at base, upside, and downside
  • Sensitivity tables: two-way grids on key return drivers
  • Credit statistics: leverage ratios, coverage ratios, and FCF yield by year

Quality Checks

  • Sources exactly equal uses — no unresolved plugs
  • Debt balances never go negative; revolver does not exceed commitment size
  • Balance sheet balances in every period (if full three-statement model)
  • IRR and MOIC are internally consistent (cross-check IRR via XIRR against discrete cash flows)
  • DSCR and leverage ratios are within stated covenant limits in the base case
  • Circular references from cash sweep → interest → net income → cash sweep are resolved with an iteration toggle or copy-paste macro
  • All hard-coded assumptions are isolated on a single inputs tab, not buried in formulas
  • Tax rate, amortization of financing fees, and D&A treatment are consistent with target jurisdiction [VERIFY]

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Evaluates activist investor campaigns with thesis assessment, proposed changes, and likely outcome analysis. Use when analyzing activist situations, evaluating campaign theses, or assessing activist track records.

日本語の概要は準備中です。原文の説明を表示しています。

CaseMark/skills442026年9月9日 更新

Evaluates corporate vulnerability to shareholder activism with governance assessment, valuation gaps, and operational improvement opportunities. Use when assessing activist risk, identifying vulnerabilities, or preparing defensive analyses.

日本語の概要は準備中です。原文の説明を表示しています。

CaseMark/skills442026年9月9日 更新

Evaluates bolt-on acquisitions for existing platforms with strategic fit, synergy quantification, and return contribution analysis. Use when screening add-on targets, building platform acquisition strategies, or modeling tuck-in economics.

日本語の概要は準備中です。原文の説明を表示しています。

CaseMark/skills442026年9月9日 更新

Evaluates signal half-life, turnover implications, and capacity constraints for systematic alpha factors. Use when analyzing signal persistence, evaluating factor decay, or estimating strategy capacity.

日本語の概要は準備中です。原文の説明を表示しています。

CaseMark/skills442026年9月9日 更新

Evaluates alternative data sources including satellite, NLP sentiment, web scraping, and geolocation for alpha signal generation. Use when analyzing alt data, evaluating new data sources, or integrating non-traditional signals.

日本語の概要は準備中です。原文の説明を表示しています。

CaseMark/skills442026年9月9日 更新

Evaluates competition law exposure with market definition, HHI analysis, and remedy estimation for proposed transactions. Use when assessing merger clearance risk, analyzing market concentration, or preparing HSR filings.

日本語の概要は準備中です。原文の説明を表示しています。

CaseMark/skills442026年9月9日 更新

CaseMark のスキルをすべて見る

このスキルの問題を報告する