본문으로 건너뛰기
cccskills
무료GitHub에 공개

analyzing-unfunded-commitment-exposure

Evaluates unfunded commitment obligations with call probability, pacing models, and portfolio-level exposure management. Use when analyzing unfunded exposure, managing commitment pacing, or stress testing capital calls.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

설치 방법 보기

포함된 파일(1)

  • SKILL.md6.9 KB

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Analyzing Unfunded Commitment Exposure

When To Use

  • Evaluating an LP's total unfunded obligation across active fund commitments before making new allocations
  • Stress-testing capital call scenarios for secondary portfolio acquisitions or GP-led continuation vehicles
  • Assessing pacing risk when an LP's liquid reserves may be insufficient to meet projected drawdowns
  • Preparing exposure reports for investment committees ahead of re-up or new commitment decisions
  • Modeling unfunded exposure impact during denominator-effect periods (portfolio NAV declines increasing PE allocation weight)

Inputs To Gather

  • Commitment schedule: Fund name, vintage year, total commitment, called capital to date, unfunded balance, and fund term/extension status
  • GP call history: Historical call cadence per fund (quarterly call amounts as % of commitment) and any known upcoming large calls (e.g., deal closings)
  • Fund lifecycle stage: Investment period vs. harvest period designation for each fund; expected remaining investment period duration
  • Distribution data: Recent distribution rates and whether distributions are being recalled or recycled
  • LP liquidity profile: Available cash, credit facility capacity, expected near-term distributions, and any over-commitment policy limits
  • Market context: Current fundraising environment, deal flow pace, and whether GPs are accelerating or slowing deployment
  • Secondary-specific inputs (if applicable): Unfunded assumptions embedded in secondary pricing, seller's remaining pacing obligations, and any escrow or deferred payment structures

Workflow

  1. Build the commitment inventory — Aggregate all active fund commitments into a single ledger. Categorize each by vintage, strategy (buyout, growth, venture, real assets, credit), fund lifecycle stage, and GP. Calculate unfunded balance as total commitment minus cumulative called capital. Flag any funds with extension provisions or zombie-fund risk.

  2. Assign call probability profiles — For each fund, estimate the likelihood and timing of remaining capital calls:

    • Investment-period funds: Assume 80–100% of unfunded will be called; model call pacing based on GP's historical deployment speed and stated pipeline
    • Post-investment-period funds: Assume 5–20% residual calls (follow-ons, fees, recycling); assign lower call probability
    • Continuation/GP-led vehicles: Assess whether new unfunded tranches exist and model per the vehicle's stated drawdown schedule
    • [VERIFY] Confirm whether each fund's partnership agreement permits capital recycling, which inflates effective callable capital beyond original commitment
  3. Model pacing scenarios — Project quarterly or annual capital calls across the portfolio using at least three scenarios:

    • Base case: Calls follow historical GP pacing patterns; distributions partially offset outflows
    • Accelerated deployment: GPs call capital 30–50% faster (e.g., competitive deal environment); distributions slow
    • Stress case: Simultaneous acceleration across multiple funds, distribution freeze, credit facility unavailable
    • For each scenario, compute net cash flow (calls minus distributions) and cumulative unfunded drawdown over 1-, 2-, and 3-year horizons
  4. Calculate exposure metrics — Derive key ratios:

    • Unfunded-to-NAV ratio: Total unfunded ÷ current portfolio NAV (typical LP range: 0.3x–0.6x; above 0.7x signals elevated risk)
    • Unfunded-to-liquid-assets ratio: Total unfunded ÷ (cash + undrawn credit facility + projected near-term distributions)
    • Over-commitment ratio: Total commitments (funded + unfunded) ÷ target allocation, measuring reliance on distributions to fund future calls
    • Weighted-average call probability: Portfolio-level expected callable amount using fund-by-fund probability weights
    • [VERIFY] Confirm the LP's internal policy limits for each ratio and flag any breaches
  5. Identify concentration risks — Assess unfunded exposure by:

    • Single GP (>15% of total unfunded to one manager is a common concentration threshold)
    • Vintage year clustering (multiple funds from the same vintage calling simultaneously)
    • Strategy overlap (e.g., heavy unfunded in venture during a slow-exit environment)
    • Currency exposure on non-USD commitments with unfunded balances
  6. Synthesize findings and recommend actions — Summarize exposure posture and propose mitigants:

    • Adjust new commitment pacing if stress-case liquidity gap exceeds tolerance
    • Consider secondary sales of fund interests with high unfunded and low call probability (zombie positions)
    • Evaluate credit facility sizing relative to stress-case peak call quarter
    • Flag any GP-led transactions where rolling into a continuation vehicle resets unfunded exposure

Output

  • Unfunded Exposure Summary Table: Fund-by-fund breakdown showing commitment, called, unfunded, % remaining, lifecycle stage, call probability tier, and expected call horizon
  • Portfolio-Level Metrics Dashboard: Unfunded-to-NAV, unfunded-to-liquidity, over-commitment ratio, and weighted-average callable amount
  • Pacing Scenario Analysis: Quarterly projected net cash flows under base, accelerated, and stress cases with cumulative drawdown curves
  • Concentration Heat Map: Unfunded exposure by GP, vintage, strategy, and currency
  • Risk Flags and Recommendations: Bullet-point list of breaches, concentration warnings, and proposed actions (pacing adjustment, secondary sale candidates, credit facility review)

Quality Checks

  • Confirm every active fund commitment is included; cross-check against capital account statements and the LP's commitment ledger
  • Validate that called-capital figures reconcile with GP capital call notices (not just internal estimates)
  • Ensure call probability assumptions are documented with rationale — do not assign blanket percentages without fund-specific justification
  • Verify that distribution assumptions in pacing models are conservative (use net distributions, not gross, and exclude any one-time outlier events)
  • Check that over-commitment ratio calculations use the correct denominator (target allocation vs. total portfolio value) per the LP's policy definition
  • Confirm currency conversion rates are current for non-USD unfunded obligations
  • Mark any data gaps or stale GP reports with [VERIFY] and note the as-of date for all NAV and call figures

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

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.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

CaseMark/skills452026년 10월 17일 업데이트

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.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

CaseMark/skills452026년 10월 17일 업데이트

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.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

CaseMark/skills452026년 10월 17일 업데이트

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.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

CaseMark/skills452026년 10월 17일 업데이트

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.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

CaseMark/skills452026년 10월 17일 업데이트

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.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

CaseMark/skills452026년 10월 17일 업데이트

CaseMark의 스킬 모두 보기

이 스킬의 문제 신고