本文へ移動
cccskills
無料GitHub で公開

modeling-special-dividend-scenarios

Analyzes special dividend catalysts with balance sheet capacity, tax efficiency, and shareholder return comparison analysis. Use when modeling special dividends, evaluating capital return catalysts, or assessing dividend capacity.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md6.9 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。

Modeling Special Dividend Scenarios

When To Use

  • An activist campaign is evaluating whether a target company has excess cash or under-leveraged balance sheet capacity to fund a special dividend
  • Assessing a special dividend as an alternative or complement to share buybacks, debt paydown, or M&A
  • Modeling the shareholder return impact of a one-time cash distribution triggered by an asset sale, litigation recovery, or business unit divestiture
  • Evaluating tax efficiency of a special dividend versus other capital return mechanisms for the shareholder base
  • Stress-testing balance sheet capacity to determine maximum supportable dividend without breaching covenants or impairing operations

Inputs To Gather

  • Balance sheet data: Cash and equivalents, short-term investments, restricted cash, total debt, net debt, and working capital (most recent quarter and trailing four quarters)
  • Debt covenants: Leverage ratio caps (Net Debt / EBITDA), minimum liquidity thresholds, restricted payment baskets, and any incurrence-based vs. maintenance-based distinctions [VERIFY against credit agreement]
  • Cash flow profile: LTM and projected FCF, capex requirements (maintenance vs. growth), mandatory debt amortization, and seasonal working capital swings
  • Share count: Basic and diluted shares outstanding, treasury shares, and any pending conversions or option exercises that affect per-share math
  • Tax parameters: Qualified vs. non-qualified dividend treatment, corporate-level tax on repatriated foreign cash (if applicable), shareholder-level tax rates for institutional vs. retail holders [VERIFY jurisdiction-specific rates]
  • Catalyst details: Source of excess capital (asset sale proceeds, litigation settlement, spin-off, accumulated cash), timing, and any contractual restrictions on use of proceeds
  • Comparables: Precedent special dividends in the sector — size as % of market cap, stock price reaction, and payout ratios

Workflow

  1. Determine excess capital capacity

    • Calculate unrestricted cash, net of minimum operating cash buffer (typically 3–6 months of operating expenses or management-stated floor)
    • Assess incremental debt capacity: headroom under covenant limits at current EBITDA, and pro forma leverage post-dividend
    • Identify any trapped cash (foreign subsidiaries, regulated entity capital requirements, restricted payment baskets) and estimate repatriation cost
  2. Size the special dividend

    • Model three scenarios: conservative (cash-only, no incremental debt), base (partial leverage), and aggressive (full covenant headroom utilization)
    • Express each as total dollar amount, per-share amount, and yield on current stock price
    • Verify that each scenario leaves adequate liquidity for operations, near-term maturities, and a reasonable cushion (typically 1.0–1.5x minimum covenant thresholds)
  3. Compare capital return alternatives

    • Build a side-by-side matrix: special dividend vs. accelerated buyback vs. tender offer vs. debt paydown
    • Key metrics for comparison: after-tax value per share to holders, EPS accretion/dilution, impact on leverage ratios, and signaling effect
    • For buybacks, model at various price assumptions (current, +10%, +20%) to capture execution risk
  4. Assess tax efficiency

    • Model after-tax proceeds to shareholders under qualified dividend rates vs. return-of-capital treatment (if tax basis supports it) [VERIFY current tax rates and holding period requirements]
    • Compare to buyback tax treatment (capital gains rate, timing of realization, and impact of the 1% excise tax on corporate repurchases post-IRA) [VERIFY current excise tax applicability]
    • Note differential impact on tax-exempt holders (pensions, endowments) vs. taxable holders
  5. Run sensitivity and scenario analysis

    • Sensitize on: EBITDA (±10–20%), cash balance assumptions, interest rates on incremental debt, share price (affects yield and buyback comparison), and tax rate changes
    • Model a downside case: what happens if EBITDA declines 20% post-dividend — does the company breach covenants or face a liquidity shortfall?
    • Calculate breakeven: at what share price does a buyback deliver superior per-share value vs. the special dividend?
  6. Build the catalyst timeline

    • Map board approval requirements, record date / ex-date mechanics, and any required lender consent or covenant waiver timeline
    • For activist-driven scenarios, align the dividend proposal with proxy season calendar, 13D filing dates, or settlement negotiation windows

Output

  • Executive summary: Recommended dividend amount per share across scenarios with key supporting rationale
  • Capacity analysis table: Unrestricted cash, debt headroom, total available capital, minimum liquidity floor, and maximum distributable amount
  • Scenario matrix: Conservative / base / aggressive columns showing total payout, per-share amount, dividend yield, pro forma leverage, and pro forma liquidity
  • Capital return comparison: Side-by-side table of special dividend vs. buyback vs. tender offer with after-tax shareholder value, EPS impact, and leverage impact
  • Tax efficiency analysis: After-tax value per share by holder type (taxable individual, tax-exempt, corporate)
  • Sensitivity tables: Tornado charts or data tables on EBITDA, cash, interest rate, and share price assumptions
  • Risk flags: Covenant proximity warnings, liquidity stress results, and any [VERIFY] items requiring confirmation

Quality Checks

  • Per-share amounts reconcile to total payout divided by correct diluted share count
  • Pro forma leverage ratios are calculated on the same EBITDA basis used in the covenant definitions (not adjusted EBITDA unless the covenant permits it) [VERIFY covenant EBITDA definition]
  • Minimum cash / liquidity buffers account for seasonal troughs, not just spot balances
  • Tax rate assumptions match current law and are flagged for any pending legislative changes
  • Buyback comparison uses realistic execution assumptions (not 100% of volume; typically 10–25% of ADTV)
  • All covenant calculations tie to the actual credit agreement terms, not approximations — mark [VERIFY] if working from summary descriptions rather than the executed agreement
  • Downside scenario confirms no covenant breach or, if a breach occurs, the analysis explicitly flags the required waiver or amendment

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