agentsclimarketplace

Valuation engine

Skill alzadjaliaafra-hash/murshidi-knowledge-layer/models/valuation-engine

Corporate valuation engine grounded in academic finance — DCF family (FCFF, FCFE, APV), CAPM and WACC, beta levering, growth and terminal value, relative valuation multiples, and distressed-firm valuation including equity-as-call-option and private-company adjustments. Activate with /val, "valuation", "DCF", "WACC", "fair value", or "what is it worth".From its SKILL.md

Install
npx -y skills add alzadjaliaafra-hash/murshidi-knowledge-layer --skill valuation-engine

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.8 KB, 914 tokens by cl100k_base, as published. Nobody here has run it

Corporate Valuation Engine

A complete valuation framework grounded in peer-reviewed academic finance (Damodaran, Fernández equivalence proofs, Koller/McKinsey ROIC-growth framework, Ohlson/Feltham residual income). The consistency principle governs everything: match cash flows to their correct discount rate, use market values for weights, normalize cyclical earnings, and adjust explicitly for survival probability in young or distressed firms.

Discounted Cash Flow Family

FrameworkFormula
Equity valuationΣ CF_to_equity_t / (1 + k_e)^t
Firm valuationΣ CF_to_firm_t / (1 + WACC)^t
APVUnlevered firm value + PV(tax shield) − expected bankruptcy costs
FCFFEBIT(1−t) + Depreciation − CapEx − ΔWorking Capital
FCFENet Income − (CapEx − Depreciation) − ΔWC + (New Debt − Debt Repaid)

Fernández's equivalence proofs show all discounting variants yield identical intrinsic value when assumptions are consistent — divergent answers signal inconsistent assumptions, not model choice.

Cost of Capital

ComponentFormula
CAPME(R_i) = R_f + β_i [E(R_m) − R_f]
WACCk_e·E/(D+E) + k_d(1−t)·D/(D+E)
Beta unleveringβ_u = β_L / [1 + (1−t)(D/E)]
Beta releveringβ_L = β_u [1 + (1−t)(D/E)]

Growth & Terminal Value

EPS growth = Retention Ratio × ROE. EBIT growth = Reinvestment Rate × ROC. Terminal value follows Gordon Growth: TV_n = CF_{n+1} / (r − g_n). Only growth funded at ROIC > WACC creates value (McKinsey diagnostic) — growth at ROIC below WACC destroys it.

Relative Valuation (fundamentals-anchored multiples)

MultipleDriver formula
P/EPayout × (1+g) / (k_e − g)
P/BVROE × Payout × (1+g) / (k_e − g)
P/SNet Margin × Payout × (1+g) / (k_e − g)
EV/SalesAfter-tax Op Margin × (1 − Reinvestment) × (1+g) / (WACC − g)

Forward-looking multiples outperform historical ones; always adjust comparisons for the fundamental drivers above.

Distressed & Private Company Valuation

Equity as a call option (Black-Scholes): Equity = V·N(d1) − D·e^(−rt)·N(d2) — equity in a distressed firm has option value even when firm value sits below the face value of debt.

Distress-adjusted value: Going Concern Value × (1 − π_distress) + Distress Sale Value × π_distress.

Private-company adjustments (Damodaran): Total Beta = sector unlevered beta ÷ correlation with market, relevered at company D/E (owners cannot diversify firm-specific risk); synthetic credit rating from interest-coverage lookup → default spread → pre-tax k_d = R_f + default spread + country risk premium; illiquidity discount 0.25 − 0.04 × ln(Revenue in USD millions), adjusted for profitability and block size, bounded [5%, 40%]; key-person discount 0–25% based on founder dependence and succession readiness.

Output Protocol

Lead with the number that matters (enterprise or equity value, valuation gap, implied ROIC premium). Support with the structural reason. Close with the action it implies. Triangulate at least two methods, name every trade-off, and show the full calculation chain.

Resources

  • knowledge/methodology.md — formula register with worked examples
  • dataset/train.jsonl — instruction-tuning pairs
  • eval/eval.jsonl — held-out evaluation questions

What ships with it: 3 files

12.5 KB alongside SKILL.md

dataset/

eval/

knowledge/

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.