alirezarezvani/claude-skills / finance/skills/stock-analysis
stock-analysis
UnverifiedAnalyze, research or value a listed stock or company (US, global, NSE or BSE) with verified data, ratios, scores and valuation. Use whenever asked if a stock is a buy, what it is worth, or how a company looks.
references/13-situations.md
View on GitHub# Situation and Lifecycle Playbooks
Use this when: you are at Stage 2 classifying the company, or at any later stage where the standard metric set is producing an answer that feels mechanically correct and economically absurd.
The sector playbook tells you which metrics exist for this industry. The situation playbook tells you which of them are *currently meaningful*. These are overlays, not substitutes: a steel company is a metals company **and** a deep cyclical, and both files apply. Almost every large valuation error in this skill's domain comes from applying the wrong lens rather than from arithmetic — a trailing P/E screen buys cyclicals at the top, sells them at the bottom, discards every loss-making grower, and rates a holding company on consolidated numbers it does not actually own. Choosing the framework is a bigger decision than the inputs you feed it.
**Every indicative range in this file is indicative only.** Ranges shift with market, cycle stage, interest-rate regime and accounting period. A peer-set comparison and the company's own 5–10 year history override any absolute band printed here, always.
## Contents
- [0. Classify first — the step that prevents the error](#0-classify-first--the-step-that-prevents-the-error)
- [1. Loss-making growth](#1-loss-making-growth)
- [2. Deep cyclicals](#2-deep-cyclicals)
- [2.1 The low-P/E-at-peak trap](#21-the-low-pe-at-peak-trap)
- [2.2 Normalised earnings power](#22-normalised-earnings-power)
- [2.3 The capacity cycle and cost-curve position](#23-the-capacity-cycle-and-cost-curve-position)
- [2.4 Valuation anchors that work at the trough](#24-valuation-anchors-that-work-at-the-trough)
- [2.5 Surviving the trough](#25-surviving-the-trough)
- [3. Turnarounds](#3-turnarounds)
- [4. Distressed and restructuring](#4-distressed-and-restructuring)
- [5. Spin-offs and demergers](#5-spin-offs-and-demergers)
- [6. Merger arbitrage and open offers](#6-merger-arbitrage-and-open-offers)
- [7. Holding companies](#7-holding-companies)
- [8. Recent IPOs](#8-recent-ipos)
- [9. Micro and small caps](#9-micro-and-small-caps)
- [10. Asset-heavy vs asset-light](#10-asset-heavy-vs-asset-light)
- [11. Promoter and family-controlled companies](#11-promoter-and-family-controlled-companies)
- [12. State-owned enterprises (PSUs)](#12-state-owned-enterprises-psus)
- [13. Large treasury and investment books](#13-large-treasury-and-investment-books)
- [14. Serial acquirers and roll-ups](#14-serial-acquirers-and-roll-ups)
- [15. SPACs and de-SPACs](#15-spacs-and-de-spacs)
- [16. Catalyst, horizon and falsification](#16-catalyst-horizon-and-falsification)
- [Checklist](#checklist)
---
## 0. Classify first — the step that prevents the error
Before computing a single multiple, write down three lines in your working notes:
1. **Situation label(s).** Pick from: normal operating company / loss-making growth / deep cyclical / turnaround / distressed / spin-off / merger-arb target / holdco / recent IPO / micro-cap / promoter-controlled / PSU / treasury-heavy / serial acquirer / de-SPAC. **Multiple labels are normal.** A newly demerged PSU cement company is four overlays at once.
2. **Metrics deliberately switched off.** State them: "trailing P/E is switched off — mid-cycle margins are ~40% below current; valuing on EV/tonne and normalised EPS."
3. **Re-classification date.** Situations expire. A turnaround that works becomes a normal company; a compounder that stops reinvesting becomes a treasury-heavy stub; a cyclical migrates from trough to peak. Re-label at least annually and on any transformative event.
### Recognition triggers — run this scan
| If you observe | Suspect | Go to |
|---|---|---|
| Negative EBIT with >25% revenue growth, heavy SBC or ad spend | Loss-making growth | §1 |
| Margin range across 10 years spans 3x or more; commodity or spread-driven revenue | Deep cyclical | §2 |
| Margins well below own history and peers; new management; restructuring charges | Turnaround | §3 |
| Net debt/EBITDA >5x with falling EBITDA, going-concern note, rating downgrade, covenant waiver | Distressed | §4 |
| Listing in the last 12 months without an IPO; scheme of arrangement in filings | Spin-off / demerger | §5 |
| Price pinned just below a fixed offer price; open-offer or scheme announcement | Merger arb | §6 |
| Principal assets are stakes in other listed companies; revenue mostly dividend income | Holdco | §7 |
| Listed <24 months; restated financials only; offer-for-sale in the prospectus | Recent IPO | §8 |
| Market cap below roughly ₹5,000 crore / $2bn, thin volume, no analyst coverage | Small/micro-cap | §9 |
| Capex/sales persistently >10% or persistently <2% | Asset-heavy / asset-light | §10 |
| Promoter or family holding >40%; group companies in the RPT note | Promoter-controlled | §11 |
| Government is the controlling shareholder; administered pricing; subsidy receivables | PSU | §12 |
| Cash + investments >25% of market cap | Treasury-heavy | §13 |
| Goodwill + intangibles >30% of assets; growth bridge dominated by acquisitions | Serial acquirer | §14 |
| Listed via SPAC merger; warrants outstanding; projections in an investor deck | De-SPAC | §15 |
If two labels conflict on the same metric, the **more conservative overlay wins**. A cyclical holdco is valued on trough-adjusted NAV, not peak look-through earnings.
---
## 1. Loss-making growth
**Recognise it.** Negative operating profit, revenue growth typically >20%, gross margin positive, opex dominated by sales/marketing and R&D, heavy share-based compensation, funding from equity rather than operating cash. Common in SaaS, consumer internet, D2C, quick-commerce, biotech, EV.
**What changes.** P/E, ROE and EV/EBITDA are undefined or meaningless. You are underwriting two separate questions that must be answered in order: (a) *are the unit economics positive?* and (b) *is there enough money to reach scale before the funding window closes?* A company losing money because it is spending ahead of a profitable unit is investable. One losing money on every unit sold is a subsidy, and growth makes it worse. Answer (a) first — if it fails, (b) is irrelevant.
**Primary metrics.**
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| Contribution margin per unit | Revenue per order/customer − all direct delivery costs (COGS, fulfilment, payment gateway, support, hosting) | Positive and widening year on year | The single go/no-go test. Negative and static = the business scales its losses. |
| CAC payback | Fully-loaded S&M spend ÷ (new customers × monthly gross profit) | <12 months (SaaS <18) | Longer payback means growth must be funded externally for years; ties directly to dilution. |
| LTV/CAC | (ARPU × gross margin ÷ monthly churn) ÷ CAC | >3x on *gross* margin and realistic churn | Compute it yourself; management LTV usually uses revenue margin and optimistic churn. |
| Net revenue retention | Revenue from a cohort this year ÷ same cohort last year | >100% (SaaS >110%) | Above 100% the installed base grows without new sales spend — the strongest signal available. |
| Quarterly cash burn / runway | (OCF − capex − lease payments); runway = (cash + undrawn committed lines) ÷ burn | >8 quarters, or clear path to breakeven inside runway | Under ~4 quarters, the next financing sets the price, not the fundamentals. |
| SBC as % of revenue | Share-based comp ÷ revenue | <10%; >20% is a material transfer to employees | SBC is a real cost paid in your ownership. |
| Fully-diluted future share count | Today's diluted count + shares implied by the capital still needed to reach breakeven | — | Value the company **per future share**, never per today's share. |
**Signals that invert.**
- **Revenue growth is a negative** if contribution margin is negative — every incremental sale enlarges the loss.
- **"Adjusted EBITDA" positive** is not profitability. Rebuild it to GAAP/Ind-AS operating profit and list every add-back. SBC, annually recurring "one-off" restructuring, capitalised software and marketing reclassified as "investment" are the four usual culprits.
- **Falling CAC** can mean brand strength — or that the company has stopped growing and is harvesting easy demand. Check volume, not just cost.
- **Rising capitalised R&D / content / development cost as a share of spend** flatters the P&L while cash burn is unchanged. Watch the ratio drift, not the level.
**The trap.** Modelling the business correctly and ignoring dilution. You can be right about the company and still lose most of your money because your claim on it shrank through successive down-rounds and rescue raises. The dominant risk for a pre-profit company is not competition — it is a closed funding window. Ask explicitly: at today's gross margin and a normalised opex base, at what revenue level does this break even, and does the cash on hand plus committed lines get there?
---
## 2. Deep cyclicals
**Recognise it.** Selling price is set by an external market, not by the company: steel, aluminium, copper, cement, sugar, paper, shipping rates, refining and petrochemical spreads, memory chips, PVC/soda ash, dry-bulk freight, hotels, airlines, real-estate developers, capital goods with long order cycles. Diagnostic test: pull 10–15 years of EBITDA margin — if the range spans a factor of three or more, and peaks/troughs line up across all peers simultaneously, it is a deep cyclical.
**What changes.** Nothing in the trailing P&L describes the business. Profits are set by the industry supply–demand balance, not company quality. Your entire job is to locate the company in the cycle, estimate mid-cycle earnings power, and determine whether the balance sheet survives the next trough.
### 2.1 The low-P/E-at-peak trap
**This is the single most expensive valuation error in equity analysis. Treat any cyclical trading at a low trailing P/E as a red flag until proven otherwise.**
The mechanics: at the cycle peak, realisations and margins are at decade highs, so **E** is inflated far above sustainable levels. The market, which can see this, correctly assigns a **low multiple** to an earnings number it knows is temporary. The resulting 4–6x P/E is not a mispricing — **it is the market's forecast of an earnings collapse.** The stock then de-rates on falling earnings *and* often a falling multiple, producing losses far larger than the apparent cheapness implied.
The inverse is equally important and equally counter-intuitive: at the trough, earnings are near zero or negative, so the P/E is enormous, meaningless or undefined. **A high or infinite P/E on trough earnings frequently marks the best entry point in the entire cycle.** Cyclical returns are made by buying at high/no P/E on depressed earnings and selling at low P/E on peak earnings — the exact opposite of what a screen instructs.
Confirm which end of the cycle you are at, using at least three independent checks:
| Check | Peak signature | Trough signature |
|---|---|---|
| EBITDA margin vs own 10-year band | At or above 90th percentile | Bottom decile, or negative |
| Trailing P/E | Low (4–8x) and falling | Very high, negative or undefined |
| P/B | Above 10-year median, often 2–4x | Near or below 1x |
| Capacity utilisation / occupancy / freight rate | >90%, spot above contract | Below cash-cost for marginal producers |
| Industry capex and new-project announcements | Booming; greenfield everywhere | Projects cancelled; consolidation, closures |
| Dividend yield and payout | High yield on peak EPS (looks attractive) | Cut or suspended |
| Balance sheets across the sector | Deleveraging fast, net cash appearing | Rights issues, covenant waivers, bankruptcies |
If margins are at a decade high and the P/E is 5x, **treat the stock as expensive**, and write that sentence explicitly in the report so the conclusion is not quietly reversed later by a screen output.
**Cross-check against the commodity itself.** Back out the realisation or spread implied by current earnings, and compare it to the long-run marginal cost of production (typically the 90th-percentile producer's all-in cost plus a return on capital). Prices persistently far above marginal cost are unsustainable by construction — high prices call forth the supply that destroys them.
### 2.2 Normalised earnings power
Build mid-cycle earnings before you value anything:
1. Pull 10–15 years (at minimum one full peak-to-peak cycle) of volume, realisation, EBITDA margin, EPS and ROCE.
2. Take the **median** (not mean — means are dragged by peak outliers) EBITDA margin across the cycle. Adjust upward only for structural, evidenced improvements: a new low-cost plant commissioned, captive power, backward integration, a permanent shift in the cost curve. Do not adjust for "management's efficiency programme".
3. Apply the mid-cycle margin to *current* capacity and a realistic utilisation, not to peak revenue. Volume growth is real and permanent; price is not.
4. Deduct maintenance capex-adjusted depreciation and a normalised tax rate to get mid-cycle EPS and mid-cycle ROCE.
5. Value on 10–14x mid-cycle EPS (indicative; varies widely by industry and rate regime), and cross-check with §2.4 anchors.
State clearly that mid-cycle EPS is an estimate, show the margin assumption, and show what the value is at ±200bps of margin.
### 2.3 The capacity cycle and cost-curve position
Cyclical profits are set by supply. Map it directly:
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| Industry capacity pipeline | Announced capacity additions over next 3–5 years ÷ current industry capacity, vs expected demand CAGR | Additions below demand growth | Peer capex booms are the leading indicator of the next downturn, visible years before the P&L. |
| Cost-curve quartile | Company cash cost per tonne/unit vs industry cost curve | First quartile | A first-quartile producer survives the trough and buys distressed assets; a fourth-quartile one is a call option that can expire worthless. |
| Capacity utilisation | Output ÷ rated capacity | Compare to the 10-year band, not an absolute | Utilisation above ~90% precedes price spikes; below ~70% precedes closures. |
| Contract vs spot mix | Share of volume on long-term contracts | Higher = smoother, lower peak | Determines how violently earnings move; changes the correct multiple. |
| Channel inventory | Distributor/port/warehouse stocks in weeks of demand | Below historical average | Restocking and destocking amplify the apparent cycle by months. |
| Vertical integration | Share of key input self-supplied (ore, power, feedstock) | Higher in input-cost-driven cycles | Integration moves the company down the cost curve and stabilises the spread. |
Also track imports/exports, anti-dumping and safeguard duties (**India:** DGTR investigations and Finance Ministry notifications; **US:** Section 232/301 tariffs and ITC determinations). Trade protection is often the difference between a domestic producer's peak and trough, and it expires on a known date.
### 2.4 Valuation anchors that work at the trough
At the trough, earnings-based multiples break entirely. Switch to:
| Anchor | How to compute | Indicative use | Why it matters |
|---|---|---|---|
| P/B | Market cap ÷ tangible book | Compare to the company's own trough (often 0.5–0.8x) and peak (2–4x) | Book is a stable measuring stick when earnings are near zero. |
| EV per unit of capacity | EV ÷ tonnes, barrels/day, MW, rigs, rooms, TEU, wafer starts | Versus greenfield build cost per unit | Directly comparable across peers and against new supply economics. |
| EV / replacement cost (Tobin's Q) | EV ÷ estimated cost to rebuild the asset base today | <1 at trough; >1 invites new supply | This is *both* the valuation floor and the mechanism of recovery: below replacement cost, nobody builds, supply stops growing, and prices eventually recover. |
| Price / normalised earnings power | Market cap ÷ mid-cycle net profit (§2.2) | 10–14x indicative | Keeps you honest about what the business earns through a cycle. |
| EV / mid-cycle EBITDA | Use mid-cycle, never trailing | 5–8x indicative for heavy industry | The only EV/EBITDA formulation that is not cycle-contaminated. |
Sanity check every asset-based anchor: what would it cost, and how many years would it take, to build this asset base today — including land, environmental clearances and grid/port connectivity? An asset that cannot be replicated for regulatory reasons is worth more than its accounting value; an obsolete asset is worth less than book regardless of what the balance sheet says.
### 2.5 Surviving the trough
The cycle can be right and the equity still be wiped out if leverage was set against peak cash flow. Survivorship *is* the thesis: the surviving low-leverage operator captures the exiting competitor's volume and earns super-normal returns for years.
- Compute **net debt ÷ mid-cycle EBITDA**, never trailing peak EBITDA. Indicative comfort: below 2.5x for a deep cyclical; above 4x on mid-cycle numbers is a solvency question, not a leverage ratio.
- Stress-test at the **worst realisation of the last downcycle**: does EBITDA cover interest plus maintenance capex? If not, count the quarters of liquidity available.
- Line up the **debt maturity schedule against the expected trough years**. A refinancing due in the trough is the single most common route to permanent equity loss.
- Check covenant headroom (usually net debt/EBITDA and interest cover), fixed-vs-floating mix, and contractually committed capex that cannot be cancelled.
- Do the same for the two or three weakest peers. Their distress is your company's opportunity — or the trigger for a sector-wide price collapse as they dump inventory.
**The trap, restated.** Buying a cyclical because it screens cheap on trailing earnings, high ROE and a fat dividend yield — all three of which peak together, at exactly the wrong moment.
---
## 3. Turnarounds
**Recognise it.** Margins materially below both peer median and the company's own history, a new CEO/CFO, restructuring provisions, asset sales announced, and a narrative of "transformation". Price is well off its highs.
**What changes.** Before any valuation, diagnose *which of three* problems this is — the diagnosis determines whether there is a thesis at all:
| Type | Signature | Fixable? |
|---|---|---|
| **Operational** | Revenue and volumes intact; margin gap traceable to identifiable causes — cost bloat, a bad plant, loss-making contracts, over-distribution | Yes, and usually within 2–4 years |
| **Financial** | Business economics fine at the EBIT line; capital structure broken — too much debt, wrong maturity, wrong currency | Yes, via refinancing, asset sale or equity raise — but ownership may transfer (see §4) |
| **Secular** | End-market shrinking, technology substitution, regulatory ban, permanent share loss to a structurally better model | **No.** This is not a turnaround, it is a melting ice cube |
Distinguishing test: look at **volumes/units, not revenue** (price rises can mask volume collapse for years), market share trend, and whether peers are also struggling (industry problem, possibly cyclical) or thriving (company-specific, possibly fixable). Most "value trap" losses are secular decline misdiagnosed as an operational turnaround.
**Primary metrics.** Gross margin trend quarter over quarter (the first place a real fix shows); fixed-cost base in absolute currency terms year on year; working-capital days; net debt reduction in absolute terms; capacity utilisation; share of revenue from products launched in the last 3 years.
**Demand evidence, not intention.** Turnarounds re-rate on hard data points, and the market almost always gives you time to buy *after* the first verified inflection at a far better risk-adjusted price than at the announcement. Waiting filters out the large majority of plans that fail and costs you only a small part of the return. Hard markers to require:
- A new CEO/CFO with a *relevant* prior track record, and their actual incentive plan (are targets tied to margin and ROCE, or to revenue and TSR over one year?).
- Divisions or assets **actually sold with cash received** — not "strategic review initiated".
- Headcount and fixed cost reductions visible **in the reported P&L**, not in a slide.
- Debt actually repaid on the balance sheet.
- Two consecutive quarters of sequential gross-margin or working-capital improvement.
**Signals that invert.**
- **Restructuring charges every year for five years** means restructuring is a permanent operating cost, and "adjusted" earnings are fiction.
- **Promised savings** that never appear in reported opex have been "reinvested" — i.e. they did not exist.
- **Revenue stabilisation** is not evidence if it came from price increases while volumes fell; that is harvesting, and it accelerates the decline.
- A **low P/B on a shrinking asset base** is not a floor if the assets are industry-specific and the industry is disappearing.
**The trap.** Underwriting the announcement instead of the evidence, and confusing a cheap price with a changed business.
---
## 4. Distressed and restructuring
**Recognise it.** Going-concern qualification, covenant breach or waiver, rating downgrade to junk, interest cover below 1.5x, debt trading well below par, auditor resignation, delayed filings, insolvency petition admitted.
**What changes.** You are no longer valuing a business; you are valuing a **residual claim that ranks last**. Equity value = enterprise value minus everything senior to it, and that number is frequently zero or negative even when the business survives and prospers. Build the liability stack before anything else:
Secured debt (by security and seniority) → unsecured debt → subordinated debt and converts → capitalised leases → pension/gratuity deficits → statutory and tax dues (often super-priority) → trade creditors → guarantees given to group entities and other off-balance-sheet exposures → preference shares → equity.
**Primary metrics.** Enterprise value under a realistic (not hopeful) recovery scenario; the resulting equity stub after the stack; nearest maturity wall in months; liquidity runway; the "fulcrum" security — the layer at which value runs out, because that layer typically ends up owning the reorganised company.
**Regime matters — check which applies.**
- **India:** Insolvency and Bankruptcy Code 2016. Once the NCLT admits a case, the CIRP timeline runs to 330 days including litigation, an insolvency professional displaces the board, and resolution plans routinely write existing equity down to a nominal value or extinguish it entirely. Section 29A bars defaulting promoters from bidding for their own company. RBI's prudential framework governs pre-IBC restructuring by lenders. Watch for the "delisting after resolution plan" outcome — minority shareholders can be cashed out at a token price.
- **US:** Chapter 11 (reorganisation, debtor-in-possession financing, absolute priority rule) versus Chapter 7 (liquidation). Equity receives nothing until creditors are made whole, though small "gifted" recoveries occur in negotiated plans. Watch for a shareholders' committee being *denied* — that is the court signalling the equity is out of the money.
**Signals that invert.**
- **A very low price and a tiny market cap** make the equity look like a cheap option. It is an option, but the strike is the entire liability stack, and it can expire worthless while the business continues under new owners.
- **"Promoter/management is infusing capital"** dilutes you unless you can participate on the same terms. Rescue rights issues and debt-to-equity conversions routinely dilute existing holders by 90%+.
- **Asset value exceeding debt on the balance sheet** is not protection: distressed asset sales clear at a fraction of book, and the write-down happens on the way to the sale.
- **Rising share price on restructuring news** frequently reflects retail speculation, not a recovery calculation.
**The trap.** Buying "cheap" distressed equity without doing the capital-structure arithmetic. In distress, value transfers to creditors or to the new-money provider; the operating recovery you correctly predicted accrues to someone else.
---
## 5. Spin-offs and demergers
**Recognise it.** A scheme of arrangement in the filings, a share entitlement ratio, a record date, and a new line item appearing in the shareholder's account. **India:** demerger under Sections 230–232 of the Companies Act 2013, approved by NCLT and SEBI, with a listing lag between record date and the new entity's first trade — often weeks to months, then a special pre-open price-discovery session. **US:** Form 10 registration statement (the primary diligence document) or Form S-1, usually tax-free under IRC §355.
**What changes.** There are two independent things to analyse: the *mechanical mispricing* and the *standalone economics*.
Mechanical: recipients did not choose the new stock, there is no analyst coverage, and index funds plus mandate-restricted holders must sell an unwanted small-cap stub regardless of value. This forced selling is indiscriminate and time-bounded — it is the structural reason spin-offs are a documented source of mispricing. Watch price behaviour in the first several weeks and note when index inclusion decisions land.
Standalone: **the pro-forma segment disclosure is not the standalone P&L.** Add back the corporate costs the unit never carried (its own CFO, board, listing, audit, insurance, treasury, IT) and remove arbitrary parent allocations. Then read the scheme for how debt, pension/gratuity, tax liabilities, contingent liabilities and shared costs were split, what transitional supply or service agreements bind the two entities and on what terms, and any retained stake or cross-holding.
**Primary metrics.** Standalone EBIT after real corporate costs; net debt allocated to each entity vs its EBITDA; the share of each entity's revenue that is a captive sale to the other; the parent's retained stake and its likely disposal path; the implied stub value of the parent after deducting the market value of the spun entity.
**Signals that invert.**
- **Heavy selling with no news** in the first weeks is the opportunity, not a warning — but only if the standalone economics stand up.
- **The parent looking "cleaner"** is often because it kept the cash and pushed the debt and legacy liabilities into the spun entity. Read the allocation, do not assume symmetry.
- **A long transitional services agreement** flatters the spun entity's near-term costs; model the step-up when it lapses.
- **India-specific:** the cost basis of your original holding is apportioned between the two entities per the scheme — relevant to any after-tax return calculation. Verify the ratio in the scheme document, not from a news report.
**The trap.** Valuing the new entity off the parent's old segment margins. Also: assuming the "good" business was spun out. Check where management's incentives, the better assets and the growth capex budget ended up — occasionally the spin is a disposal of a problem dressed as a value-unlock.
---
## 6. Merger arbitrage and open offers
**Recognise it.** The stock trades in a narrow band just below a stated offer price, volatility collapses, and volume spikes around approval news.
**What changes.** Business quality becomes almost irrelevant. You are pricing a probability-weighted event, and the payoff is asymmetric in the wrong direction: **you win a few percent many times and lose 20–40% when a deal breaks.** Position sizing and break probability, not the headline spread, determine the outcome.
**Primary metrics.**
| Metric | Definition / how to compute | Indicative range | Why it matters |
|---|---|---|---|
| Gross spread | (Offer price − market price) ÷ market price | 1–4% for clean deals | The raw compensation. |
| Annualised return | Gross spread × (365 ÷ expected days to close), net of costs and adjusted for any dividends received in the interim | Compare to the risk-free rate plus a break-risk premium | A 3% spread over 3 months is very different from 3% over 18. |
| Downside to undisturbed price | Pre-announcement price (adjusted for market/sector moves since) vs current price | Typically −20% to −40% | This is the actual loss if the deal breaks — the number that governs sizing. |
| Break-even probability | Downside ÷ (spread + downside) | Compare to your own estimate of deal risk | Makes the implied market probability explicit. |
**Enumerate every condition precedent** and give each a probability: antitrust/competition approval in each jurisdiction (**India:** CCI; **US:** HSR waiting period, DOJ/FTC second request), foreign-investment screening (**India:** Press Note 3 for land-border countries; **US:** CFIUS), sector regulator consent (RBI, IRDAI, TRAI, FERC, FCC), shareholder vote thresholds, financing certainty (committed facilities vs "best efforts"), MAC/MAE clause wording, break fees on each side, and litigation.
**India-specific structures.** A SEBI SAST-mandated **open offer** is triggered at 25% acquisition or on change of control, and is for a minimum 26% of shares — so tendering holders typically receive *proportionate acceptance*, not a full exit; model the residual stub you will still own at the post-offer price. **Delisting** via reverse book-building has a different payoff shape entirely, driven by the discovered price and the 90% threshold. **Schemes of arrangement** require NCLT approval and majority-of-minority voting, which adds months and a real rejection risk.
**Signals that invert.**
- **A wide spread is a warning, not a bargain.** It is almost always the market pricing a real regulatory, financing or political risk it understands better than you do.
- **Stock-for-stock deals** carry the acquirer's own risk and require a short hedge; an unhedged position is a bet on the acquirer, not an arbitrage.
- **Repeated extensions of the long-stop date** signal a deal in trouble even while the spread looks stable.
**The trap.** Sizing on the spread rather than the downside, and treating a wide spread as free money.
---
## 7. Holding companies
**Recognise it.** The principal assets are equity stakes in other companies (listed or unlisted); standalone revenue is largely dividend and interest income; the market cap sits far below the market value of the stakes. Common in Indian promoter group structures and in European/Asian family conglomerates. See also `references/sectors/holdco-assetmgr.md`.
**What changes.** **Consolidated financials are largely meaningless to a minority holder.** Consolidated revenue and EPS include subsidiaries whose cash you do not control and, in the case of associates, businesses you own a slice of but do not consolidate. You own a claim on cash that must travel *up* through dividend policy, tax and holdco operating costs. Valuation is sum-of-the-parts, full stop.
**Build the NAV, in this order.**
1. Market value of each listed stake (price × shares held, as of a stated date).
2. A defensible valuation for unlisted stakes — an earnings or book multiple against listed comparables, disclosed clearly as an estimate, with a haircut for illiquidity.
3. Plus net cash / minus net debt at the **holdco standalone** level (not consolidated).
4. Minus holdco operating costs capitalised (annual holdco opex ÷ discount rate) — these are a perpetual leakage.
5. Minus estimated tax leakage on an eventual sale of the stakes (capital-gains tax on the embedded gain; **India:** verify the current LTCG rate and surcharge applicable to listed and unlisted shares separately).
Then: **discount = 1 − (market cap ÷ NAV)**, plotted over 5–10 years to establish the company's *own* normal range.
**Primary metrics.** Discount to NAV vs its own 5–10 year percentile; look-through earnings (your share of each investee's net profit); **cash dividends actually received** at the holdco (the only money that can ever reach you); holdco opex as % of NAV; share count trend.
**Signals that invert.**
- **A low consolidated P/E** on a holdco is meaningless and frequently the reason a screen surfaces it. Switch it off explicitly.
- **A very wide discount is not automatically an opportunity.** Indian holdcos commonly sustain 50–70% discounts for a decade or more. The money is made only when the discount is wide *against its own history* **and** there is a mechanism to close it: buyback, demerger, listing of a subsidiary, dividend policy change, holdco merger, activist pressure. Without a mechanism, a 60% discount can simply stay a 60% discount forever, and your return is only the underlying's return.
- **A holdco continually issuing shares to buy more stakes** widens the discount rather than narrowing it, and dilutes look-through earnings per share.
**The trap.** Treating NAV as a target price. Also: ignoring voting-vs-economic rights and cross-holdings — a circular structure where A owns B owns A inflates apparent NAV by double-counting; net it out.
---
## 8. Recent IPOs
**Recognise it.** Listed within roughly 24 months. No public reporting track record; only restated financials in the prospectus; analyst coverage originating mostly from the bookrunners.
**What changes.** You lack the one thing the rest of this skill depends on — the company's own history under public scrutiny. **The information asymmetry is at its maximum**, and it runs entirely against you: IPOs are sold, not bought, and are timed by informed sellers into favourable markets.
**Read the prospectus for:**
- **Fresh issue vs offer-for-sale split.** Fresh issue money goes into the company; OFS money goes to exiting shareholders. A 100% OFS means no capital is being raised for the business at all.
- **Selling shareholders' cost basis** and the valuation of the last pre-IPO round. A PE holder exiting at 8x their entry price two years later tells you what they think the business is worth.
- **Use-of-proceeds specificity.** "General corporate purposes" and "repayment of promoter loans" are far weaker than a named plant with a named capacity.
- **Lock-up expiry calendar.** **India (SEBI ICDR):** anchor investors — 50% of allotted shares locked for 30 days, remainder for 90 days; minimum promoter contribution locked for 18 months; other pre-issue capital typically 6 months. **US:** typically a 180-day underwriter lock-up, with earlier release triggers. Each expiry is a scheduled supply event.
- **3–5 years of restated financials**, looking specifically for a suspicious margin ramp into the IPO year, a working-capital squeeze (channel stuffing, stretched payables), related-party clean-ups executed just before filing, and one-off revenue booked in the final year.
- **Free float and index inclusion timeline** — small float plus future index addition creates mechanical demand that has nothing to do with value.
**Signals that invert.**
- **A beautiful three-year margin ramp is a warning, not a strength.** Pre-IPO financial dressing is common and reverts within 4–8 quarters of listing. Check whether the trend continues in the first post-listing results.
- **Grey-market premium and listing-day pop** carry zero information about business value.
- **Heavy anchor/institutional subscription** reflects allocation dynamics and momentum, not diligence you can rely on.
**The trap.** Valuing on prospectus-year margins. Wait for two to four quarters of *public* reporting and, where possible, past the first major lock-up expiry, before treating any margin as the base rate. In India, note that **SME-platform IPOs** (NSE Emerge, BSE SME) have materially lighter disclosure and post-listing scrutiny than main-board issues — apply §9 in full.
---
## 9. Micro and small caps
**Recognise it.** Market cap below roughly ₹5,000 crore or $2bn (indicative; adjust to the market), thin volume, few or no analysts, promoter/insider concentration.
**What changes.** Two things simultaneously: the inefficiency is real (institutions structurally cannot participate), and the safeguards are absent (no analyst scrutiny, weaker audit, weaker disclosure). Fraud and governance-failure base rates are far higher here than in large caps, so **verification effort must go up, not down**, exactly where information is scarcest.
**Primary metrics.**
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| Median daily traded value | Median of (volume × price) over 6 months | Enough to exit in <10 trading days at ≤25% of daily volume | A "cheap" stock you cannot sell without moving the price 15% is not cheap. |
| Days to exit | Intended position size ÷ (25% × median daily value) | <10 days | Determines position size before any valuation work. |
| Bid–ask spread | (Ask − bid) ÷ mid | <1% | Wide spreads are a permanent tax on entry and exit. |
| Free float | 1 − (promoter + strategic + locked holdings) | >25% | Low float exaggerates both rallies and collapses. |
| Auditor quality and tenure | Identity of the statutory auditor; any resignation or qualified opinion | No resignations; recognised firm | **Auditor resignation is one of the highest-signal red flags available in this segment.** |
**Governance and existence checks — do these before valuation, not after.** Auditor/CFO turnover; qualified opinions and emphasis-of-matter paragraphs; board independence in substance; related-party transactions; contingent liabilities; regulatory action history. Then verify the business physically exists as described: plant visits or satellite imagery, customer and distributor references, employee headcount vs revenue, GST/tax filings, channel checks. **India:** CARO 2020 clauses (especially on statutory dues, undisclosed income, and diversion of funds), SEBI's ASM/GSM surveillance lists, trade-to-trade segment placement, and circuit filters — a stock locked at upper circuit cannot be sold, which converts a paper gain into a trapped position. **US:** SEC comment letters on EDGAR, Form 8-K Item 4.01 (auditor change) and Item 4.02 (non-reliance on prior financials), and PCAOB inspection status of the auditor.
**Signals that invert.**
- **A strong price move on low volume** is not confirmation of anything; it may be one buyer or a coordinated operation.
- **Very high reported ROE with poor cash conversion** in an unaudited-by-a-major-firm small cap is a fraud pattern, not a quality signal.
- **Concentrated promoter holding**, normally alignment, becomes control risk when combined with thin float and heavy related-party flow (§11).
**The trap.** Sizing a position by conviction rather than by liquidity. Do the days-to-exit calculation first; it caps the position regardless of how good the analysis is.
---
## 10. Asset-heavy vs asset-light
**Recognise it.** Asset-heavy: capex/sales persistently above ~10%, fixed assets several times revenue, high depreciation, long build cycles — utilities, cement, telecom, hotels, shipping, refining. Asset-light: capex/sales below ~2–3%, negative working capital, returns on a tiny capital base — software, franchising, asset-managers, brand licensors, platform businesses.
**What changes.** Capital intensity determines whether growth *creates or destroys* value. Two companies with identical earnings growth can have opposite owner outcomes if one must reinvest 120% of its profit to achieve it.
**Primary metrics.**
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| Capex / depreciation | Full-cycle average | ~1.0–1.3x for a steady business; >2x = heavy expansion | Persistently below 1.0x means the asset base is being run down and reported profit is overstated. |
| Maintenance capex | Floor estimate = depreciation grossed up for asset-price inflation; better = management disclosure or a per-unit engineering estimate | — | Growth capex is optional; maintenance capex is a cost of staying in business. Owner earnings = CFO − maintenance capex. |
| Incremental ROIC | Δ NOPAT over 5 years ÷ Δ invested capital over 5 years | Above WACC, ideally >15% | The only number that says whether *new* money earns a return. Averages hide it. |
| Fixed-cost share | Fixed costs ÷ total costs; test EBIT at −20% volume | — | Quantifies operating leverage in both directions. |
| FCF conversion | FCF ÷ net profit, full cycle | >70% | Asset-heavy companies can report rising profits while consuming every rupee in capex. |
**Signals that invert.**
- **Asset-heavy:** strong EPS growth with FCF persistently negative is value destruction, not compounding. Ask where the cash went and what return it earns.
- **Asset-light:** a spectacular ROIC of 60%+ can signal a **reinvestment ceiling** — high returns on a base you cannot grow — which caps the compounding rate no matter how good the margin. Test whether the moat is real (brand, network effect, IP, switching costs, regulation) or merely an absence of assets, which is not a barrier to entry.
- **Asset-light with off-balance-sheet dependency** (outsourced manufacturing, leased everything, a single cloud provider) has real capital intensity sitting on someone else's balance sheet, with the associated fragility. Capitalise the leases and recompute ROIC before comparing to an asset-heavy peer.
**The trap.** Comparing ROIC across the two models without adjusting for leases and off-balance-sheet capital, and concluding the asset-light business is a better *compounder* when it may simply be a better *cash cow*.
---
## 11. Promoter and family-controlled companies
**Recognise it.** **India:** promoter and promoter-group holding disclosed in the quarterly shareholding pattern, typically >40%. **Global:** founding family holding, dual-class share structures, board seats held by family members. See `references/08-governance.md` for the full treatment; this section covers the situation overlay.
**What changes.** The primary risk shifts from business failure to **value extraction around minority shareholders through legal but one-sided transactions**. Diligence moves from the P&L to the notes.
**Primary metrics.**
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| Related-party transactions | Sum of RPT sales, purchases, loans, guarantees ÷ revenue (and ÷ PAT) | Minimal, disclosed, at arm's length, stable | Every rupee routed through a promoter entity is a rupee that may not reach you. |
| Royalty / brand fee | Payment to promoter-owned entity ÷ revenue and ÷ PAT | Low single-digit % of revenue at most; scrutinise any increase | **India:** SEBI LODR requires shareholder approval where royalty to related parties exceeds 5% of consolidated turnover. |
| Promoter remuneration | Total managerial remuneration ÷ PAT; compare to peers and to the dividend payout | Well below the dividend paid to all shareholders | If the family takes more out as salary than all shareholders receive as dividend, the alignment argument fails. |
| Pledged shares | Pledged ÷ promoter holding (**India:** disclosed quarterly under SAST Reg 31) | 0%; anything above ~20% is a live risk | Pledging converts a price fall into a forced-sale spiral and possible loss of control. |
| Promoter holding trend | 5-year trajectory, and the reason for each change | Stable or rising via open-market purchases | Steady creeping increases signal confidence; a quiet sell-down signals the opposite. |
**Governance-in-substance checks.** Are "independent" directors long-tenured associates, relatives or former employees? Is there a succession plan and which generation is operating? How many other group companies exist, and is cash routinely moved between them? Are there dual-class or differential voting rights? **India:** RPTs above ₹1,000 crore or 10% of consolidated turnover require audit committee and majority-of-minority shareholder approval under LODR Reg 23 — check whether large transactions have been sliced below the threshold.
**Signals that invert.**
- **High promoter holding**, normally a positive alignment signal, becomes a risk when paired with heavy related-party flow, high pledging, or a history of minority-unfriendly schemes.
- **Promoter buying shares** is usually bullish, but ahead of a delisting attempt it is a squeeze on minorities, not an endorsement.
- **A low payout ratio** in a family firm may reflect long-horizon reinvestment (good) or cash being warehoused for the family's other ventures (bad). Distinguish using the ROIC on incremental capital.
**The trap.** Treating "skin in the game" as a governance conclusion rather than a hypothesis. A genuinely aligned family owner with a long horizon is among the best structures to own; an extractive one is among the worst; **the diagnosis rests entirely on the related-party and remuneration record**, not on the size of the stake.
---
## 12. State-owned enterprises (PSUs)
**Recognise it.** Government (central or state) is the controlling shareholder. **India:** CPSEs, state utilities, public-sector banks, often with Maharatna/Navratna/Miniratna status. **Global:** national oil companies, state-owned banks and utilities in many emerging markets.
**What changes.** The controlling shareholder has objectives other than the share price. First determine whether the company is run **for profit or for policy**, because that determines whether any quality metric can be extrapolated at all.
**What to check.** Administered or subsidised pricing and who sets it; obligations to serve unprofitable customers or regions; mandated purchases from other state entities; subsidy receivables from government and the historical collection lag (this is where working capital goes to die); dividends and buybacks demanded by the state to plug its fiscal deficit; forced cross-holding purchases (one PSU made to buy a stake in another); capex mandated for strategic rather than return reasons; the appointment process and tenure of senior management (a chairman with 14 months left will not start a 5-year programme). **India:** DIPAM's disinvestment pipeline, the CPSE dividend policy floor, and the OFS route for government stake sales — announced government selling is a known overhang.
**Primary metrics.** ROCE excluding subsidy receivables and excluding regulatory/policy assets; cash conversion (subsidy-driven receivables make accrual profit unreal); dividend as % of PAT and whether it is discretionary or state-mandated; capex approved on commercial vs strategic grounds; government stake and any announced sell-down; the persistent valuation discount to private peers, plotted over 10 years.
**Signals that invert.**
- **A high dividend yield** is often the state extracting cash rather than a signal of shareholder-friendliness — and it can coexist with underinvestment.
- **Very low P/E and P/B** are usually a rational, permanent discount for policy risk, not a mispricing. The discount has existed for decades in most PSU universes.
- **Large capex programmes** may be strategically mandated and value-destroying; check the approved return assumption, not the size of the number.
- **Strong reported profit at a subsidised-price utility** can reverse the moment the subsidy formula changes — earnings are a policy variable.
**The trap.** Buying statistical cheapness with no catalyst. The upside case in a PSU almost always requires a **specific event**: privatisation or strategic sale, pricing deregulation, a large one-off dividend or buyback, or a formal change in the dividend policy. Without one, the discount is a rational permanent feature and your return is dividends only.
---
## 13. Large treasury and investment books
**Recognise it.** Cash, bonds, listed equity stakes, real estate or non-consolidated subsidiaries exceed roughly 25% of market cap; a meaningful share of reported "other income" is interest, dividends or mark-to-market gains.
**What changes.** Headline P/E and ROE describe a blend of two unrelated businesses. Split the company in two and value each separately:
**(a) Core operating business.** Recompute core operating margin, core ROCE and core EPS **excluding** all investment income and mark-to-market. Value it on an operating multiple. **(b) Investment book.** Value at market (with a haircut for illiquid or strategic holdings) and net off any tax payable on realisation.
Then: implied core P/E = (market cap − investment book value) ÷ core net profit. This is the number that matters, and it is often dramatically different from the headline.
**Primary metrics.** Investments + surplus cash as % of market cap; core ROCE excluding the book; return earned on the investment book vs the cost of capital; share of PAT from non-operating income; capital-allocation history — what management has *actually* done with excess cash over 10 years.
**The critical question: is the cash claimable?** Excess capital may be permanently trapped — held as regulatory capital, sitting in an overseas subsidiary subject to repatriation tax, locked inside a partly-owned entity, or simply reserved by the controlling family with no intention of distribution. **Trapped cash should be haircut heavily or excluded from the sum-of-the-parts**; treating it at face value can overstate value by a large factor.
**Signals that invert.**
- **A low headline P/E** may exist only because a third of the market cap is cash earning treasury yields; the operating business can be far *more* expensive than it looks. It can also work the other way — do the arithmetic before concluding either.
- **A depressed ROE** may be entirely cash drag, and the operating business's return on its own capital may be excellent. Never judge such a company on consolidated ROE.
- **A rising cash pile with no distribution** is frequently the raw material for a value-destroying acquisition. Excess capital is a capital-allocation risk, not automatically a margin of safety.
**The trap.** Double-counting: valuing the company on an EV/EBITDA multiple (which already nets off cash) and then *adding back* the cash. Pick one treatment and state it.
---
## 14. Serial acquirers and roll-ups
**Recognise it.** Several acquisitions a year, goodwill and intangibles above ~30% of total assets (often above 100% of equity), "adjusted" EPS given prominence, and a growth story told in deals.
**What changes.** Reported growth is not a performance measure until you decompose it. Serial acquisition is either one of the best compounding models in existence or a machine for hiding organic decay and manufacturing EPS with cheap debt — **and the two look identical on a revenue chart.**
**Primary metrics.**
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| Growth bridge | Split annual growth into organic / acquired / currency, every year | Organic positive and steady | Decelerating organic growth masked by a larger deal is the first visible symptom of a roll-up breaking. |
| Return on acquisition capital | Cumulative cash spent on acquisitions over 5–10 years ÷ the incremental EBIT or FCF actually generated over the same period | >15% implied return | The single most honest test. It cannot be dressed by accounting choices. |
| ROIC **including** goodwill | NOPAT ÷ (invested capital + accumulated goodwill, including goodwill previously written off) | Stable or rising, above WACC | Excluding goodwill lets a serial overpayer report excellent returns forever. |
| Cash conversion | FCF ÷ net income | >80% sustained | Roll-ups with poor conversion are usually capitalising costs or under-investing in acquired units. |
| Purchase price allocation | Share of consideration allocated to goodwill vs amortising intangibles | More to identifiable intangibles = more conservative | Dumping consideration into non-amortised goodwill flatters reported earnings indefinitely. |
| Earn-outs / contingent consideration | Liability balance and payments due | Disclosed, modest | An off-radar claim on future cash that reduces equity value today. |
**Accounting notes.** Under IFRS/Ind-AS 103 and US GAAP, goodwill is **not amortised** — it is impairment-tested, and impairment is discretionary in timing, so it usually arrives late and all at once. Acquisition costs, restructuring of acquired businesses and amortisation of acquired intangibles are the standard "adjusted EPS" add-backs; if a company acquires every year, these are recurring operating costs and the adjustment is not legitimate.
**Signals that invert.**
- **Strong headline EPS growth** funded by debt at a cost below the target's earnings yield is arithmetic, not value creation. Ask whether ROIC including goodwill improved.
- **Rising goodwill with no impairment ever** is not conservatism; it is a deferred admission.
- **A large acquisition announced immediately after a weak organic quarter** is a pattern worth naming explicitly.
- **A decentralised "we never integrate" model** can be a genuine edge (low overhead, entrepreneurial units) or an excuse for no oversight — distinguish using cash conversion and same-unit organic growth.
**The trap.** Roll-ups break when acquisition multiples rise or funding closes — both external conditions, not company decisions. Model what happens to growth if the company makes **zero** acquisitions next year, and check the funding mix (debt vs equity issuance) against the maturity schedule.
---
## 15. SPACs and de-SPACs
**Recognise it.** The company listed by merging with a special purpose acquisition company rather than through an IPO. Tell-tales: warrants trading alongside the shares, a sponsor holding "founder shares", a PIPE, and an investor presentation containing multi-year revenue projections. Predominantly a US phenomenon (**India:** domestic SPAC listings are not permitted on NSE/BSE; SPAC frameworks exist at IFSCA/GIFT City, and Indian companies have de-SPACed onto US exchanges).
**What changes.** A de-SPAC is a **pre-IPO-quality company with public-market pricing and no IPO-grade diligence**. There is no underwriter due-diligence process of the traditional kind, no restated multi-year track record in many cases, and the merger was marketed on *projections* that a conventional IPO prospectus would not contain. Treat the projections as marketing, not guidance, and rebuild the model from historical financials only.
**What to check.**
- **Redemption rate.** Before the merger closed, shareholders could redeem at the trust value (typically ~$10). Redemption rates above 90% were common, which means the announced cash-to-balance-sheet figure often did not arrive. Check what cash actually landed, in the 8-K/Super 8-K.
- **Dilution stack.** Sponsor promote (classically ~20% of the pre-merger share count, acquired for a nominal sum), public and private warrants, PIPE shares (often issued below the effective deal price), earn-out/vesting shares, and convertible notes taken to bridge redemptions. Compute a **fully-diluted, post-all-instruments** share count and value on that. The headline "enterprise value" of a de-SPAC deal routinely understates the true diluted cost per share.
- **Trust and timeline mechanics** (for a live SPAC, pre-merger): trust value per share, deadline to complete a deal, extension terms. A sponsor facing a deadline has a powerful incentive to complete *any* deal.
- **Post-close reporting.** Look for restatements, material weaknesses in internal control (very common), auditor changes, and delayed filings in the first 18 months.
**Signals that invert.**
- **A projection deck showing revenue growing 10x in four years** is a negative signal about the promoter's seriousness, not evidence of opportunity. Compare the first two years of actual results against the deck — the gap is the most informative number available.
- **Trading near the $10 trust value** does not imply a floor once the merger has closed; the trust is gone and the price can go anywhere.
- **A famous sponsor** is a distribution advantage, not a diligence substitute; the sponsor's economics are earned on *completing* a deal, not on its performance.
**The trap.** Valuing off the announced deal enterprise value and the projection deck, on a pre-dilution share count. Rebuild from actual historicals and a fully-diluted count, then apply §1 (loss-making growth) and §8 (recent IPO) in full — most de-SPACs are both.
---
## 16. Catalyst, horizon and falsification
Every situation in this file is a claim that a **value gap exists** *and* that a **mechanism will close it**. Without the mechanism, cheapness compounds at zero and the IRR decays with time even if your valuation is right. Before finishing, write down four things:
1. **The catalyst.** The specific event: spin listing and index inclusion, deal close, refinancing completed, disinvestment or strategic sale, cycle turn evidenced by capacity closures, first profitable quarter, buyback or dividend policy change, lock-up expiry passing, resolution plan approval.
2. **The timeframe, and the IRR if it takes twice as long.** A 40% gap closing in 12 months is a 40% IRR; the same gap over 4 years is under 9% — often below the index alternative. Run both.
3. **The falsification test.** Two or three *observable* facts that would prove the thesis wrong: contribution margin still negative after two more quarters; mid-cycle margin assumption breached on the downside; the discount widening past its historical extreme with no buyback; organic growth negative ex-acquisitions; the auditor resigning.
4. **The base rate.** How often do situations of this type work? Turnarounds in structurally declining industries: rarely. Announced deals with clean regulatory paths: usually. Roll-ups after acquisition multiples rise: poorly. Anchor your probability to the reference class before adjusting for the specifics.
Set the review trigger on the **catalyst date, not the price**. The characteristic failure mode of situational investing is endlessly re-underwriting a broken thesis because the stock keeps getting cheaper.
---
## Checklist
- [ ] Label the situation(s) explicitly before computing any multiple; multiple overlays are normal and the most conservative one wins conflicts.
- [ ] Write down which standard metrics you are switching off and why, in one sentence, in the report.
- [ ] Set a re-classification date; situations expire and companies migrate between them.
- [ ] **Cyclicals: never conclude "cheap" from a low trailing P/E.** Check margin percentile, P/B, utilisation and industry capex first — a low P/E at peak margins is the market forecasting an earnings collapse.
- [ ] Compute mid-cycle EPS from a median (not mean) full-cycle margin, and value on that; state the assumption and show ±200bps.
- [ ] At the trough, switch to P/B, EV per unit of capacity and EV/replacement cost; earnings multiples are undefined there.
- [ ] Measure cyclical leverage as net debt ÷ **mid-cycle** EBITDA and line maturities up against the expected trough years.
- [ ] Loss-makers: prove contribution margin per unit is positive and improving before anything else; value per **fully-diluted future** share count.
- [ ] Rebuild every "adjusted EBITDA" back to reported operating profit and list each add-back.
- [ ] Turnarounds: classify operational / financial / secular using volumes and peer performance; secular decline is not a turnaround.
- [ ] Require at least one hard, realised evidence marker before underwriting a turnaround — not an announcement.
- [ ] Distressed: build the full liability stack and compute the equity residual before looking at the share price; check IBC (India) or Chapter 11 (US) status and whether equity is typically extinguished.
- [ ] Spin-offs: model each entity standalone with real corporate costs added and parent allocations removed; read the scheme for how debt and liabilities were split.
- [ ] Merger arb: size on the downside to the undisturbed price, not on the spread; enumerate every condition precedent with a probability; check open-offer proportionate acceptance (India).
- [ ] Holdcos: value by SOTP net of holdco costs and tax leakage; plot the discount against its own history; require a named mechanism to close it.
- [ ] Recent IPOs: check fresh-issue vs OFS, seller cost basis, lock-up calendar, and treat a pre-IPO margin ramp as a warning.
- [ ] Micro-caps: compute days-to-exit and cap position size on liquidity before valuing anything; check auditor identity, resignations and qualified opinions.
- [ ] Split maintenance from growth capex and compute incremental ROIC before calling any asset-heavy business a compounder.
- [ ] Test whether an asset-light business's high ROIC comes with a reinvestment ceiling; capitalise leases before cross-model ROIC comparisons.
- [ ] Promoter-controlled: read every RPT, royalty and remuneration line, and check pledged-share percentage and trend.
- [ ] PSUs: identify profit-vs-policy objectives, subsidy receivable collection lags, and require a named catalyst — statistical cheapness alone is a permanent discount.
- [ ] Treasury-heavy: separate core operations from the investment book, compute implied core P/E, and haircut trapped cash.
- [ ] Serial acquirers: build the organic/acquired/currency growth bridge and compute cumulative cash spent on deals vs incremental EBIT generated.
- [ ] De-SPACs: rebuild from historicals, ignore the projection deck, and value on a fully-diluted count including sponsor promote, warrants, PIPE and earn-outs.
- [ ] Name the catalyst, the timeframe, the IRR if it takes twice as long, and two or three facts that would falsify the thesis.
- [ ] Set the review trigger on the catalyst date, not on the price.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |