The seven-stage funnel: (1) industry — TAM, structure, profit-pool shape, competitive intensity, regulatory contour; (2) company — what the business actually does and where it sits in the value chain; (3) financials — quality of revenue, margin walk, working-capital intensity, cash conversion; (4) valuation — DCF, comps, and what is priced in vs what is not; (5) risk register — what would have to be true for the bear case to land; (6) position-sizing — convert the asymmetry into a fraction of capital; (7) exit plan — name the conditions that close the trade.
| Stage | Output | When to backtrack |
|---|---|---|
| Industry scan | One-paragraph framing of structure, profit-pool shape, and the rate-of-change driver you are watching | If the industry frame keeps shifting after stage 3, you do not yet understand the business well enough |
| Company-specific lens | Unit-economics walk: per-customer / per-store / per-rig economics, gross-margin contribution, retention | If the unit math does not foot to consolidated revenue and gross margin within 10%, restart the segment build |
| Financial-statement reading | FCF normalization, working-capital trend, earnings-quality flags, MD&A read | If FCF and net income diverge persistently without a working-capital story, treat earnings as suspect |
| Valuation walk | Bull / base / bear price targets with explicit driver assumptions | If the bear case is not a specific failure of a stage-2 assumption, the bear is not yet falsifiable |
| Risk register | Five-to-eight named risks with severity and likelihood | If you cannot name what would make you wrong, you do not yet have a thesis |
| Position-sizing | Fraction of capital tied to reward / risk asymmetry, NOT to conviction | If sizing is driven by enthusiasm rather than the asymmetry math, restart at stage 5 |
| Exit plan | Thesis-completion exits AND thesis-broken exits, named in advance | If you cannot name a price or a fact that closes the trade, you are improvising, not investing |
Worked example — Brimwood Lumber (fictional regional building-products operator). Industry: housing-starts-driven, fragmented, with a structural lumber-cost pass-through lag of one to two quarters. Company: 14 yards in the southeast, 62% revenue from contractor accounts with stickier pricing, 38% from spot retail with more volatile margin. Financials: GAAP operating margin 7.4%, working-capital-adjusted FCF margin 5.1%, working-capital intensity rising the last three quarters as raw-cost spikes outpace pass-through. Valuation: base $54 (DCF with 6.2% WACC, terminal 2.5%), bull $73 (faster pass-through), bear $32 (housing-starts roll over before pass-through completes). Risk register: pass-through speed, two-customer concentration in contractor segment, single-yard concentration in two metros, integration risk on pending tuck-in deal. Sizing: 1.5x reward / risk asymmetry, 14-month window — starter position 1.2% of capital, full position 2.5% pending pass-through visibility. Exit: scale out into $54-58, full exit if next-quarter receivables-vs-revenue gap exceeds 4 percentage points without a contracting-mix explanation.
Common integration failures: (a) treating the seven stages as a checklist rather than a logic chain — each stage is supposed to feed evidence into the next, not just get ticked off; (b) discovering a stage-6 sizing problem and patching it by inflating the bull case rather than restarting from the industry frame; (c) running the workflow in parallel rather than in sequence, which causes later stages to be built on assumptions that earlier stages have not yet supported.
## See also: deeper references - **Working-capital walk and FCF normalization:** `fin-8` in `financials-101` — for the SBC-adjusted FCF mechanics referenced in stage 3 here. - **Earnings-quality flags read in the MD&A:** `fin-7` and `fin-11` in `financials-101` — for the receivables-vs-revenue divergence and the explanation-by-omission pattern. - **WACC build and re-levering:** `corpval-1` through `corpval-5` in `corpval-wacc-301` — for the cost-of-capital walk used in the DCF here, including the size-premium adjustment for a sub-$500M name like Brimwood. - **Comps cross-check on the DCF:** `comps-201` — for the EV/EBITDA and EV/Sales sanity-check you should run alongside the DCF before committing to the base-case target. - **Bull / base / bear sizing math:** `ptk-4` in `practitioner-toolkit-201` — for the asymmetry framework applied here. - **Industry-structure framing:** `micro-3` and `micro-4` in `microeconomics-101` — for the profit-pool and competitive-intensity lens used at stage 1.
Sit with the ideas.
You have completed the industry scan, the company unit-economics walk, the FCF normalization, the WACC build, the risk register, and the position-sizing math on Brimwood Lumber. The base-case implies $54 against a current price of $41 with a 14-month catalyst window. Three of your six diligence items remain open, including the largest single uncertainty — whether Brimwood can pass through a recent raw-cost spike. The portfolio manager asks for your recommendation. Which response is the practitioner-grade answer?