Intrinsic Valuation

Two playbooks: discounted cash flow and sum-of-the-parts. Both discount or capitalise cash flows to a value and both end at the same equity bridge, given once below. Sum-of-the-parts also draws on references/market-valuation.md for segments valued on peer multiples. The method in SKILL.md applies throughout.

---

Shared mechanics

Match the cash flow to the discount rate. Unlevered free cash flow discounts at WACC and gives enterprise value. Levered free cash flow discounts at the cost of equity and gives equity value directly. Mixing them — discounting FCFF at the cost of equity, or subtracting net debt from a value derived from FCFE — is the most common structural error in a DCF and it does not announce itself. State which you are using in Phase 1 and hold it.

WACC construction. Cost of equity by CAPM: risk-free rate with the instrument and tenor named, equity risk premium with its basis, and a beta re-levered to the target capital structure rather than the current one. Cost of debt at the pre-tax yield the company would issue at today — not the coupon on legacy debt — then after tax. Weights on market values, never book. Report the WACC with the value impact of ±50bps, because that band is usually wider than the reader expects.

Terminal value discipline. Gordon growth requires WACC > g or the formula breaks; if the assumption set produces g close to WACC, the answer is not large, it is undefined. Terminal growth cannot exceed long-run nominal GDP for any sustained period — a company growing faster than the economy forever eventually becomes the economy. Where an exit multiple is used instead, source it from the comps set and back out the growth rate it implies, then check that against the Gordon assumption.

Report the terminal value share of enterprise value. If terminal value exceeds roughly three quarters of total value, the projection period is decorative and the output is a terminal value estimate wearing a DCF's clothing. Say so, and either extend the explicit forecast until the business reaches a steady state or present the result as what it is.

Equity bridge. Enterprise value − net debt − minority interests − unfunded pension and other debt-like items − preferred + associates and non-operating assets = equity value, ÷ diluted shares (treasury method) = value per share. Show every line even where it is zero; a bridge with items silently omitted cannot be checked.

---

Discounted Cash Flow

Intrinsic value from projected cash flows. The output's credibility rests on whether the assumptions are visible and defensible, not on the precision of the arithmetic.

When: "DCF", "intrinsic value", "what's it actually worth", "discounted cash flow", "WACC", "terminal value", "is the market mispricing this".

Phases:

1. Structure. State the free cash flow definition and why; the number of projection stages and why; the WACC methodology; whether you are applying the mid-year convention; and every data gap with the assumption that will close it.
2. Cash flow projection, years one to five. Line by line: revenue, EBITDA, EBIT, NOPAT, add back D&A, less capex, less the movement in net working capital, to unlevered free cash flow. State the growth rate for each year separately rather than a single CAGR, and flag which drivers carry the most assumption risk. Where capex sits persistently below D&A in the terminal year, the business is being modelled as shrinking its asset base forever — check that this is intended.
3. WACC. Per the shared mechanics.
4. Terminal value, both methods. Gordon growth and exit multiple. Report both values and the divergence between them as a percentage. Where divergence exceeds roughly 20%, name the driver in one line and say which method you are anchoring on and why.
5. Bridge. Per the shared mechanics.
6. Sensitivity. At least two tables: WACC against terminal growth, and revenue CAGR against EBITDA margin, each producing enterprise value.
7. Scenarios. Bear, base and bull as narratives with the assumption change that defines each, then enterprise value, equity value and value per share.
8. Challenge. The three assumptions that most move the output, each with the level at which the valuation meets the current market price.

Required tables: the cash flow build; WACC build; terminal value both methods with divergence; equity bridge; two sensitivity tables; scenario summary.

Challenge test: the terminal growth rate and WACC combination implied by today's share price. If the market's implied assumptions are more plausible than yours, the analysis has found something worth saying.

Inputs to gather: company name and ticker; industry and sub-sector; LTM revenue; LTM EBITDA and margin; LTM capex; net debt; diluted shares; tax rate; any published guidance or analyst estimates, kept separate from your own view; closest public comps for benchmarking the exit multiple.

---

Sum-of-the-Parts

Value each segment on its own merits and aggregate. Most useful where a consolidated multiple is masking a mix of businesses the market cannot price together — the job is to quantify that gap and say what would close it.

When: "sum of the parts", "SOTP", "break-up value", "conglomerate discount", "is the market undervaluing the group", "should they spin it off".

Phases:

1. Segment identification. For each segment: is it genuinely distinct in management, customers and cost structure; do listed pure-play comparables exist; could it be sold or separated; and what is its margin after a fair share of overhead. Segments that fail these tests should be combined, not valued separately for the sake of the exercise.
2. Segment valuation. Per segment: revenue, EBITDA, margin, methodology, multiple or discount rate, value, and value per share. Justify each methodology — peer multiples where genuine pure-plays exist, DCF for a segment with no comparables, asset value for property, resources or run-off books, liquidation value where a segment is being wound down. Different methods across segments are correct; an unexplained mix is not.
3. Corporate centre. Overhead not allocated to segments is a real drag and must be valued as one. Capitalise it at a multiple — conventionally below the operating segments — and carry it as a negative. Then the point most SOTP analyses miss: stranded costs. A segment that is sold does not take its share of central cost with it, so the remaining group carries overhead against a smaller base. Quantify that, or say you have not.
4. Bridge to equity. Segment values, less capitalised overhead, to enterprise value; then the standard equity bridge from the shared mechanics.
5. Integrity check. Segment EBITDA plus unallocated corporate cost must reconcile to consolidated EBITDA. Report the residual as a figure. A segment build that does not tie to the group accounts has either double-counted or lost something.
6. Discount and catalysts. Current market value against SOTP value, as a percentage. Compare it to the typical conglomerate discount in this sector rather than treating any discount as an anomaly — most diversified groups trade below the sum of their parts, persistently, and the discount is only an opportunity if something will close it. Then the candidate catalysts — spin-off, segment sale, activist involvement, re-rating — each with a realistic timeline. Note that a disposal realises value net of tax and separation costs, so headline SOTP value is not proceeds.
7. Sensitivity. SOTP value at 25th against 75th percentile multiples for each segment, identifying which segment carries the widest range. That segment is where the analysis should concentrate.

Required tables: segment valuation; corporate centre; equity bridge; EBITDA reconciliation; discount and catalyst summary; per-segment sensitivity.

Challenge test: whether the discount to SOTP has persisted for years without closing. If it has, the burden is on the analysis to explain what is different now rather than to restate the gap.

Inputs to gather: company name, current share price and market enterprise value; segments with revenue and EBITDA for each; corporate overhead if disclosed; net debt and non-operating assets and liabilities; diluted shares; pure-play comparables per segment; any announced strategic review, activist position or restructuring plan.