Pocket
Structural
Inefficiency

A POSI is not a cheap stock. It is a place where somebody has to trade without a view — a mandate, an index rule, a redemption, a calendar, a language barrier — and the price moves for a reason that has nothing to do with the business. The original talk named fourteen. The investor who coined the term claimed to have isolated seventy-eight. Here are forty-seven more, for listed equities only, each with the forced actor, the tell, the clock and the trap.

47new archetypes

7families

14originals covered

5exploitability tests

61stackable in total

The one question

Every entry below answers the same question in a different way: who is trading this without an opinion, and what makes them stop? If you cannot name the forced actor, you do not have a POSI — you have a stock you like. And if you cannot name what ends the forcing, you have a discount with no expiry, which is a yield, not a trade.

There are pockets in the market where investors are behaving a certain way structurally, and it’s creating inefficiency… The more of them are embedded in there, the more inefficiency there is. Source: the archetypes talk — wC3h6-KWrAU

The filter

Before adding an archetype to the list, it has to survive these. They come out of the Perold conversation, and they are what separate a pocket you can harvest from a market that is merely strange. An inefficient market is not automatically an exploitable one.

The best inefficient market is one where everything is perfect except the price is wrong… The more you can reduce investing to just hard work, it’s awesome. Andre Perold — Capital Allocators, QwFfDzSw1is

Everything perfect but the price

You can see the business completely, handicap it, and the only thing wrong is the quote. The treasury bond at 90 with an identical twin at par.

Fails: frontier markets. The accounting may be wrong, the related-party tunnelling invisible, the governance unknowable. Even if the price is wrong you cannot prove it, and you cannot name a catalyst.

The big money structurally can’t play

Size, liquidity, mandate or benchmark keeps large capital out. The cat is gone and the mice are out to play — and the tourists who do come move prices when they leave.

Fails: anything a $50bn manager can build a full position in. If they can be there, they are there, and you are competing on insight rather than access.

Paid for work, not epiphanies

The return is compensation for labour someone else declined to do: reading the docket, learning the language, visiting the site, building the model no one has built.

Fails: theses that rest on being smarter about a widely-modelled number. Genius is not a repeatable input; hard work is.

Specialists exist, and aren’t swamped

Someone credible already operates here, and the pocket has not been flooded with capital. The ongoing job is to keep re-affirming both halves.

Fails: once the archetype becomes a conference topic and the capacity fills. Merger arb and index-rebalance front-running both did this.

You can name the clock

A dated, mechanical event ends the forcing: a rebalance print, a lockup expiry, a rating action, a subscription deadline, a readout.

Fails: “eventually the market will notice.” Without a clock you are underwriting patience you cannot fund and your LPs cannot sit through.

The original fourteen

Covered in the source talk, listed here so the stack builder can count them — click any of these too. Everything after this section is new.

Forty-seven more

Click a card to add it to a stack. The bar at the bottom counts what you have selected — because the whole point, per the source, is embeddedness: one archetype is a story, three in the same name is an edge.

Mandate mechanics

8 · a rule sells, not a person

The cleanest POSIs in public markets. A charter, an index rule or an eligibility screen forces a trade on a published date, by an actor with no opinion and no discretion. You are not out-thinking anyone here — you are taking the other side of a document.

Dividend-cut exile

Forced actor: Income funds, dividend-aristocrat ETFs and retiree accounts that hold by charter. A cut or suspension makes the stock ineligible overnight, and they sell inside a defined window whether or not the cut was the right capital-allocation call.
The tell: Suspension and cut announcements; the share of the register held by income-mandated funds in the 13F and ETF holder lists.
The clock: The forced sale exhausts in weeks. The re-rating waits for the base to turn over into value and GARP hands — two to four quarters.
The trap: Most cuts are a symptom. You need an offensive cut — redirected to buybacks, deleveraging or reinvestment — not a defensive one.

Index deletion overhang

Forced actor: Passive and closet-indexed money that must sell at the rebalance print, price-insensitively, on a date published in advance.
The tell: Russell reconstitution rank day, S&P and MSCI and FTSE deletion announcements, float-adjusted cap sitting near a band edge.
The clock: The print itself. Lows tend to come after the rebalance, not before it — the crowded trade is the anticipation.
The trap: Deletion is usually earned. Underwrite the business as though the index did not exist, then treat the flow as the entry timing.

Float inflation

Forced actor: Index funds on the buy side. A secondary, a lockup expiry or a share-class collapse raises free float, the float-adjusted index weight rises, and mechanical buying arrives with nothing having changed in the business.
The tell: Registered secondaries, free float as a share of shares outstanding, the calendar of index float reviews.
The clock: The next float review date for each index the name sits in.
The trap: The float rose because somebody chose to sell. Find out who and why before you take the other side of the mechanical bid.

Size-band no-man’s-land

Forced actor: Nobody — and that is the inefficiency. Fallen out of the large-cap benchmarks, still too big and too liquid to interest microcap specialists. It is on no one’s benchmark and in no one’s hunting ground.
The tell: A $300m–$1.5bn cap that was above $5bn five years ago, with institutional ownership falling for no operational reason.
The clock: Growing back into a band, or being acquired. There is no automatic mechanism.
The trap: The band can hold for years. You need self-help — a buyback, insider buying, a sale process — or you are just waiting.

Exclusion-list orphans

Forced actor: ESG, sin and sharia screens remove a permanent slice of the buyer pool by charter: tobacco, defence, coal, gambling, prisons, and any company that trips an Islamic leverage or interest-income threshold.
The tell: Published exclusion policies of the largest domestic pension funds; the leverage ratio of a name sitting just under a sharia screen limit.
The clock: None — that is the point. You are paid a permanently higher discount rate, and you harvest it through buybacks and dividends rather than a re-rating.
The trap: This is a cash-return trade, not a multiple trade. It only works if management returns the capital instead of diversifying out of the excluded business.

Rating trip-wires

Forced actor: Investment-grade bond mandates on a downgrade through to high yield. The equity is then sold in sympathy by people reading the headline rather than the covenant.
The tell: Agency outlook changes, leverage measured against explicitly stated downgrade triggers, index-eligibility rules on the debt itself.
The clock: The downgrade date, then the refinancing that either confirms or kills the concern.
The trap: Sometimes the credit market is right, and what you are buying is an option on a recapitalisation, not a stock. Size it as an option.

The liquidity floor

Forced actor: Again, nobody. Average daily volume is too thin for an institution to build a position that matters without moving the price. The work has been done by several people; none of them can act at size.
The tell: ADV under roughly $1–3m against a business that plainly deserves attention.
The clock: Anything that thickens the tape: an uplisting, an index add, an equity raise, a split, or one large new holder arriving.
The trap: You inherit the same problem on the way out. Size the position to your exit, not your entry.

Dividend / buyback initiation

Forced actor: The mirror of 01. Initiating a dividend, or running the first real buyback, makes a stock eligible for an entire buyer class that was structurally barred from owning it — income funds, yield screens, certain retail channels abroad.
The tell: Net cash, improving free-cash conversion, a new CFO, capital-markets-day language shifting from growth to returns.
The clock: The announcement, then the first index and screen refresh that catches it — usually one to two quarters behind.
The trap: An initiation forced onto a business by activists when that business still needs the capital. Check reinvestment runway first.

Coverage voids

7 · nobody’s job to look

Not mispricing by disagreement — mispricing by absence. Something in the plumbing of research, classification or data means the name never reaches the desk of anyone paid to have a view. These are the most durable POSIs on the list, because the barrier is labour and labour does not arbitrage itself away.

Coverage cliff

Forced actor: The last sell-side analyst drops the name — usually because their broker was acquired or exited the sector, and because research unbundling destroyed the economics of covering small caps. No coverage means no maintenance notes and no reason for anyone to be reminded the company exists.
The tell: Analyst count going 4 to 2 to 0. Check whether the broker exited the whole sector rather than the company deteriorating.
The clock: An initiation, a new house broker, or an index add that forces attention.
The trap: Sometimes they left because there is nothing there. Coverage loss is a filter, never a thesis.

No consensus estimate

Forced actor: Every systematic strategy that requires an estimate field. No consensus EPS means exclusion from factor, quality, revision and momentum universes entirely — a vast pool of capital that cannot own what it cannot screen.
The tell: Query your data vendor for null estimate fields on companies that are actually profitable. The list is shorter than you expect and stranger than you expect.
The clock: The first published estimate. That alone is a re-rating event, independent of what the estimate says.
The trap: You have to build the model nobody has built. That is the work you are being paid for — if you skip it, you are just buying obscurity.

Miscoded GICS

Forced actor: Not a tweener sitting between two sectors — flatly filed in the wrong one. The wrong analysts cover it, the wrong comp set is applied to it, and the wrong sector ETF owns it or fails to.
The tell: Compare the GICS sub-industry against the actual revenue mix. Then listen to who asks questions on the earnings call — the coverage tells you the classification.
The clock: Reclassification, or a segment-reporting change that forces the issue into the open.
The trap: The miscode may be flattering rather than punitive. Check which direction the error runs before assuming it is your friend.

Holdco / SOTP orphan

Forced actor: The parts span three sectors, so no single analyst can own the whole thing and each one discounts the pieces outside their coverage. Nobody is responsible for the total.
The tell: A persistent discount to sum-of-parts; segments with genuinely different natural multiples; a listed subsidiary inside a parent whose own market cap is less than the stake.
The clock: A spin, a stake sale, a collapse of the structure, or a disclosure change that forces segment visibility.
The trap: Holdco discounts persist for decades where the controlling family is comfortable. Without governance pressure the discount is the permanent state, not the anomaly.

Language barrier

Forced actor: Every investor who reads only the English release. The disclosure that settles the question exists only in the local filing — the Japanese tanshin, the Korean DART submission, the Brazilian ITR — while the translated deck is thin and late.
The tell: Compare the page count of the local statutory filing to the English press release. A five-to-one ratio is a standing invitation.
The clock: None. This one is a moat rather than a trade, if you build the pipeline once and keep it.
The trap: Translation without local context produces the wrong answer held with more confidence. You need a reader, not a translator.

Off-cycle reporting

Forced actor: The entire earnings-season workflow. Semiannual reporters and odd fiscal year-ends fall out of the quarterly rhythm, out of revision screens, and off the calendar everyone actually works from. Their news lands in a week when nobody is looking at anything.
The tell: Reporting frequency of two per year; a fiscal year ending in an unusual month; results dated in the dead zone between seasons.
The clock: The report itself — which is precisely when attention is lowest.
The trap: Less frequent disclosure genuinely is more risk, and six months is a long time to be wrong in silence. Price the opacity.

Identity break

Forced actor: The databases. A rename, ticker change, share-class merger or re-domicile snaps the price and fundamental history, so the five-year chart renders as a two-year chart and every screen with a long-term lookback silently drops the name.
The tell: Recent name or ticker changes; charts with an abrupt start date; fundamental history that begins mid-cycle for no operational reason.
The clock: Roughly twelve to thirty-six months, until the new history is long enough to satisfy the screens again.
The trap: Names get changed to bury things. Find out what the old name was and what happened under it before you decide the break is innocent.

III

Event plumbing

9 · beyond the spin-off

The spin-off is the original POSI because a corporate action hands shares to people who did not choose them. That mechanism has nine other forms, and most of them are less crowded than the spin precisely because the spin is the one everybody was taught.

The RemainCo

Forced actor: Everyone studies the spinco — it is new, clean and has a story. The parent is left with changed leverage, changed mix and a shareholder base that owned it for the division that just walked out the door.
The tell: Any spin. Model the parent first, from scratch, as a fresh business — not as “what is left over”.
The clock: Two to four quarters, once the parent reports standalone numbers that can be compared to nothing.
The trap: Sometimes the parent kept the liabilities on purpose, and the spin was the point of the exercise. Read the separation agreement.

Fresh-start equity

Forced actor: Distressed credit funds that wanted the bonds and were handed the stock. They have no equity mandate, no reason to stay, and no view on the multiple — they sell into any bid that appears.
The tell: Recently emerged issuers, often quoted OTC before relisting, carrying a fresh-start balance sheet with no comparable history at all.
The clock: Relisting, index eligibility, and the credit funds finishing their distribution — visible in the 13D and 13G amendments.
The trap: The business that went bankrupt is frequently still the business that went bankrupt. The balance sheet was fixed; the industry was not.

The arb liquidation

Forced actor: A deal breaks and the entire merger-arb book exits on one day, mechanically. Nobody in the departing base ever underwrote the company as a company — they underwrote a spread.
The tell: Terminated or blocked transactions; arbitrage ownership concentration visible in the previous quarter’s 13F filings.
The clock: Days to weeks for the flush, then two to three quarters for a fundamental base to assemble.
The trap: The deal broke for a reason, and often that reason is a regulator who knows something about the industry that you do not.

De-SPAC after redemption

Forced actor: A real operating business left with a shattered cap table: ninety per cent trust redemption, a warrant overhang, PIPE holders unwinding, and a register that was only ever there for the trust value.
The tell: Post-redemption share count against the original; warrants outstanding as a share of the float; PIPE lockup dates in the proxy.
The clock: Warrant expiry or redemption, plus the first two clean reported quarters as an ordinary company.
The trap: Most de-SPACs deserved what happened to them. This is a needle in a haystack, not a screen — and the haystack is enormous.

Rights-issue wreckage

Forced actor: Pre-emption mechanics create holders who cannot or will not participate: funds without cash, mandates that cannot take up, retail that misses the deadline. All of them sell the rights and often the stock, on a fixed calendar.
The tell: Announced rights issues, especially deeply discounted ones; the nil-paid rights price against theoretical value.
The clock: The subscription deadline, then the rump placement that clears whatever was left.
The trap: The dilution is real and permanent. Underwrite the post-money share count, and ask why the money was needed at this price.

The stub

Forced actor: The data, again. After a large asset sale or a special dividend, share counts, per-share history and index weights all break at once, and every screen returns nonsense for several quarters.
The tell: Special dividends worth more than twenty per cent of market cap; large divestitures; per-share history with an unexplained step change.
The clock: The first clean annual report after the event, when the comparatives finally make sense.
The trap: What remains is often the worse half — that is usually why it is what remains. The stub has to be underwritten on its own merits.

Listing migration

Forced actor: Moving a primary listing, or re-domiciling, flips the eligible-buyer set wholesale: one index universe out, another in, and a gap in between during which neither owns it.
The tell: Announced listing moves, ADR to ordinary conversions, uplistings from OTC to a national exchange.
The clock: The effective date, then the next index review of the destination benchmark.
The trap: Companies move listings hoping for a better valuation. Hope is an argument, not a mechanism — check that real index eligibility actually follows.

Dual-class collapse

Forced actor: Several major benchmarks will not hold non-voting or low-vote shares, so those classes carry a structural discount unrelated to the business. Collapsing the classes, or a sunset provision triggering, deletes the discount overnight.
The tell: A persistent A-versus-B spread wider than five per cent; sunset provisions buried in the charter, usually time- or ownership-triggered; the founder’s age.
The clock: The sunset date, a collapse vote, or the controlling holder’s exit or death.
The trap: Control holders rarely surrender control for nothing. Read the conversion terms — the premium often goes to them, not to you.

Going dark

Forced actor: Deregistration — typically a reverse split engineered to cut the holders of record, then a Form 15. The company vanishes from every screen, index and database while the business carries on exactly as before, and liquidity collapses by design.
The tell: Reverse split proposals that state an intent to reduce holders of record; deregistration filings; a controlling holder buying beforehand.
The clock: Usually years. You are buying a private company that happens to have a quote.
The trap: Minority shareholders have very little protection in the dark, and the whole point may be to squeeze you out cheaply. Only where you trust the operator completely.

The ownership base

9 · a specific human has to sell

Here the forcing is not a rule but a person or an institution with a problem of their own. The discipline is the same as with a mandate: name the seller. “Somebody must be selling” is not analysis. A seller you can identify has a size, a deadline and an end.

The wind-down seller

Forced actor: A fund gating, liquidating or facing heavy redemptions sells the entire book regardless of view. Its best position leaves at the same discount as its worst, because the redemption does not distinguish.
The tell: A single holder above five per cent of shares outstanding in the 13F or 13D data, combined with news of that manager’s difficulties.
The clock: Knowable and finite: it ends when the position is gone. Watch the 13D and 13G amendments tick down.
The trap: You must identify the actual seller. Inferring one from a falling price is how you end up on the wrong side of somebody who knows more than you.

The lockup calendar

Forced actor: Post-IPO expiries, plus venture funds distributing shares in kind to limited partners who are not equity managers and sell on receipt. The most predictable, most published, most price-insensitive seller in public markets.
The tell: The lockup schedule printed in the prospectus; the venture holder base; employee vest cliffs disclosed in the proxy.
The clock: Printed in the S-1, to the day.
The trap: The first tranche is usually well priced in, because everyone can read. The edge, if there is one, sits in the second and third tranches that nobody diaries.

The tax calendar

Forced actor: Loss harvesting into the US December, the Japanese March fiscal year-end, and the January reversal afterwards. Selling driven purely by the date, applied hardest to whatever is already down most.
The tell: Down more than thirty per cent year to date by mid-November, small cap, with a retail-heavy register.
The clock: Four to eight weeks, mechanically, every single year.
The trap: Everyone knows this one, which caps the edge. It is a modifier on the timing of a thesis you already had — never a thesis by itself.

The estate sale

Forced actor: A death, a divorce or a family settlement pushes a founder block out at whatever discount clears it. The tape reads the block trade as a business event and marks the whole company down.
The tell: Form 4 and 13D changes by a founder or a founder’s estate; obituaries; the sudden appearance of a large block at a discount to the last print.
The clock: One to three blocks and it is finished. Estate sales are not open-ended.
The trap: Sometimes the family is selling because the founder was the business, and the estate is telling you something about succession.

Cross-shareholding unwind

Forced actor: Japanese and Korean corporates selling down policy shareholdings on a governance timeline — a dated, disclosed, price-insensitive seller, frequently met by the issuer’s own buyback on the other side of the trade.
The tell: Policy-shareholding disclosure in Japanese annual reports; exchange governance pressure; explicitly announced unwind targets and timetables.
The clock: Multi-year, but disclosed. Both sides of the flow are knowable in advance, which is rare.
The trap: The unwind is the easy half. The real thesis is what the freed capital does next — if the answer is nothing, you have bought a flow with no follow-through.

The privatization pipeline

Forced actor: A government selling a residual stake on a schedule. Every rally meets state supply, so the stock is capped until the programme finishes — and then the cap comes off all at once.
The tell: Stated privatisation programmes, residual state stakes above ten per cent, sovereign fund disposal plans in the budget documents.
The clock: The final tranche.
The trap: Political timelines slip badly, and the state is frequently also the regulator, the largest customer and the price-setter. The overhang may be the least of it.

Retail-dominated float

Forced actor: Individuals setting the marginal price — China A-shares, Korea, parts of India, and the meme-adjacent corners of the US tape. Perold’s single exception to the rule that inefficient markets are small ones.
The tell: Retail share of turnover, account-opening data, turnover velocity running far above what the float should support.
The clock: None. This is a standing condition rather than an event, which makes it a strategy rather than a trade.
The trap: It requires local presence and local data. A foreigner’s read on domestic retail sentiment is usually a year stale and always second-hand.

The closed-end discount

Forced actor: Nothing — and the absence is the point. With no create-and-redeem mechanism there is no arbitrage force to close the gap. An ETF’s discount self-corrects; a closed-end fund’s or a holding company’s does not.
The tell: A persistent discount to a directly observable NAV; a listed asset sitting inside a listed wrapper.
The clock: A tender, a wind-up, an open-ending, or an activist with a real stake and real patience.
The trap: Without a catalyst the discount is a permanent feature and your return collapses to the underlying’s, minus fees. The discount alone pays you nothing.

Permanent shunning

Forced actor: Career risk. A company that blew up years ago, healed, and remains uninvestable because a portfolio manager would have to explain the name to an investment committee. That is not analysis, it is the cost of being asked a question.
The tell: An accounting scandal, fraud allegation or near-death event three to seven years back, with a fully replaced board and management since.
The clock: Slow — a full turnover of the institutional base, a coverage return, or an index re-entry.
The trap: Check that the people actually changed. If the auditor, the CFO and the audit committee chair are the same, nothing has healed and the discount is correct.

Optics

6 · the screened number is the wrong number

Accounting conventions and screening conventions collide. The company is fine; the field the screen reads is not. Every one of these is a place where a mechanical universe filter excludes a business for a reason that has no economic content.

Expensed growth

Forced actor: Every profitability screen in existence. A company running customer acquisition, R&D or new-unit openings through the P&L reports no profit, while an identical business capitalising the same spend reports plenty. The first one is excluded from the universe entirely.
The tell: Negative operating profit alongside strong gross margin and improving cohort or new-unit economics. Disclosed acquisition cost and payback is the giveaway.
The clock: The year they stop spending, or the year they start disclosing contribution margin separately — whichever comes first.
The trap: “Investing for growth” is also exactly what a bad business says. Demand cohort data; if it does not exist, assume there is nothing to show.

Purchase-accounting drag

Forced actor: Price-to-earnings screens. Heavy intangible amortisation from historic acquisitions crushes reported EPS while cash generation is untouched, so the stock screens expensive and reads cheap.
The tell: Depreciation and amortisation running far above maintenance capex; large intangible balances; a wide and persistent gap between reported and cash earnings.
The clock: The amortisation schedule runs off — and the date is disclosed in the notes. This is one of the few POSIs with a published expiry.
The trap: Some amortisation is genuinely recurring, because the acquisitions have to keep coming for the business to stand still. Serial acquirers are not exempt from their own spending.

Consolidation distortion

Forced actor: Per-share metrics. A partly-owned subsidiary consolidated at a hundred per cent inflates revenue you do not own; an equity-method stake carried near zero hides value that is plainly there. Both break every ratio a screen computes.
The tell: Large minority-interest lines relative to net income; equity-method investments with a carrying value far below observable market value.
The clock: A stake sale, an IPO of the subsidiary, or a buy-in of the minority.
The trap: The hidden value only counts if it can be extracted. Read the shareholder agreement for the drag, tag and consent rights before you underwrite it.

Historic-cost assets

Forced actor: Price-to-book screens, running in both directions at once. Land, buildings and licences carried at decades-old cost make book value meaningless — sometimes hiding value, sometimes flattering a wreck.
The tell: Property-heavy balance sheets held for a long time: railroads, ports, broadcasters, legacy industrials, old retail estates.
The clock: A revaluation, a sale-leaseback, an appraisal disclosed in a proxy, or a takeover bid that prices the assets for you.
The trap: Assets that cannot be sold without killing the operating business are not worth their appraisal. Separable value is the only value.

The model break

Forced actor: Analysts modelling the reported line. Perpetual to subscription, gross to net revenue recognition, franchising a company-owned estate — reported revenue collapses for two to three years while the underlying business improves.
The tell: A stated transition, deferred revenue growing while revenue falls, billings disclosed separately from revenue and diverging from it.
The clock: The trough of the transition, typically six to ten quarters in, when the comparatives finally lap the change.
The trap: Some transitions are cover for genuine demand loss. Watch billings rather than revenue — if billings are falling too, there is no transition, only decline.

One-times that actually stop

Forced actor: A market that has correctly learned to ignore the adjustment — and then keeps ignoring it in the one year the charges genuinely end. The scepticism is well earned and badly timed.
The tell: A restructuring programme with a stated end date, a disclosed total, and a charge trajectory that is actually declining quarter over quarter.
The clock: The last programme quarter, then the first genuinely clean print.
The trap: Serial adjusters stay serial adjusters. You need a stated, dated, shrinking programme — not a management assurance that this time is the last time.

The clock

4 · right, but not by Friday

Perold’s observation that it always surprises him how short-term markets are: managers say it, everyone shrugs, and it stays true. These four are the archetypes where the analysis is not hard — the holding period is what nobody can fund.

The kitchen sink

Forced actor: Estimate-revision and momentum owners. A new management team resets guidance to the floor, and the revision-driven base is mechanically forced out at exactly the moment the bar is lowest.
The tell: Guidance withdrawn or heavily reset within two quarters of a new chief executive or chief financial officer arriving.
The clock: Two to three quarters, until the first beat against the reset bar re-opens the name to revision buyers.
The trap: Distinguish a reset bar from a falling bar. If a second cut follows, the first was not a sink — it was the beginning of the truth.

The J-curve trough

Forced actor: Anyone who has to explain a position quarterly. Mid-investment-cycle, the spend is fully visible in the P&L and the return is visible nowhere. Nobody gets paid for owning a stock through the trough.
The tell: A capex or opex step-up with a stated payback period — a new plant, format, platform or geography being funded out of current earnings.
The clock: The stated payback date, plus one quarter of proof.
The trap: The return may never arrive. Insist on a smaller, already-completed version of the same investment demonstrably working before you fund the big one.

Cyclical multiple inversion

Forced actor: Every mechanical value screen. Peak price-to-earnings at the bottom of the cycle, trough multiples at the top — so the screens buy cyclicals at precisely the wrong time and sell them at precisely the wrong time, forever.
The tell: Multiple against mid-cycle normalised earnings; price against replacement cost; capacity utilisation and industry-wide capex intentions.
The clock: The capex cycle, not the earnings cycle. Watch supply, because demand is what everyone else is watching.
The trap: “Mid-cycle” is an assumption wearing the clothes of a fact. Anchor to physical capacity and cost curves, not to an average of the last ten years.

Dead money duration

Forced actor: Institutional mandates themselves. The outcome is knowable, the date is not, and almost nobody can hold a flat position for eighteen months and keep their capital. The market does not look past next quarter even when everyone agrees that it should.
The tell: A thesis whose catalyst is structural — a contract renewal, a patent expiry, capacity coming online — with a genuinely wide date range around it.
The clock: Unknown by definition. That is what you are being paid to bear.
The trap: This is the archetype most often used to excuse being wrong. Write down in advance what would falsify it, and the date by which you will admit it.

VII

Work barriers

4 · paid for labour, not insight

Perold’s favourite category, and the most durable: the more you can reduce investing to just hard work, the better. Nobody arbitrages away an inefficiency whose only entry fee is effort somebody else declined to spend. The customer-investor mismatch from the original fourteen is one instance of this; here is the general form.

Scientific adjudication

Forced actor: Generalists trading a technical result they cannot read. Was the dose wrong, is the ore grade real, does the process node yield — all answerable, but only with domain training. Perold’s example: the trial failed on dosing, generalists dumped it, specialists doubled up.
The tell: Binary technical events with published protocols; feasibility studies; qualification milestones. Small caps where the science is the whole company.
The clock: The readout, the assay, the qualification — dated in advance and violent on the day.
The trap: Pseudo-expertise is worse than none at all. If you cannot read the primary literature yourself or hire someone who can, you are the tourist in this trade.

The legal docket

Forced actor: A market that prices litigation as an unbounded liability. The realistic range is usually bounded, and the documents that bound it are public: the docket, comparable settlements, the insurance coverage.
The tell: A market capitalisation discounting a liability far above the historical settlement range for that claim type — or a company whose disclosed coverage nobody has read.
The clock: Motion rulings, class certification, bellwether trials, and settlement.
The trap: Appeal risk and coverage disputes are where amateurs lose. Get real counsel on the read, and assume the tail is fatter than the comparables suggest.

Primary-source only

Forced actor: Everyone working from filings, decks and transcripts. The fact that settles the thesis lives in a regulatory docket, a patent file, a customs manifest, a trial registry, a spectrum or approval database, or a public tender award — and in none of the company’s own materials.
The tell: Any thesis where the decisive variable is a licence, an approval, a shipment or a contract award rather than a reported number.
The clock: Whenever the primary source updates — which is almost always before the company says anything about it.
The trap: Build the pipeline before you need it. Doing this once, by hand, in a hurry, reliably produces a confident wrong number.

Physical diligence

Forced actor: The general form of the customer-investor mismatch. You have to use the product, visit the store, count the trucks or walk the site — and the product is only available in a geography the marginal investor does not live in. Ocado’s short sellers were in New York; the service was in Britain.
The tell: Any consumer or local-services business whose product is not available where its shareholders are. Ask where the register lives, then ask where the customers do.
The clock: None. It is a standing barrier and it renews itself every time the register turns over.
The trap: One store visit is an anecdote with a plane ticket attached. Build a repeatable count, or you have bought a feeling.

Embeddedness

The synthesis from the original talk, and the reason this is a list rather than a screen: situations carry more than one archetype at a time, and the edge scales with how many are embedded in the same name. One archetype is a reason to open a file. Three overlapping ones, each forcing a different holder to act for a different non-economic reason, is the thing you drop everything for.

If you get a new CEO in a business that has a customer-investor mismatch, that has a country arbitrage in it, and that is a high short interest stock — that can be a very explosive piece if you get it right. The archetypes talk — the original worked example

The same construction using the new list. A company that cut its dividend (01, income funds gone), lost its last analyst six months later (09, nobody left to model it), reports semiannually (14, and reported into a dead week), and whose largest holder is a fund in wind-down (25, a seller with a name and an end date) — four separate groups of people, none of whom has an opinion about the business, all leaving at once. Nothing there is a view on the company. That is the whole point.

The counting discipline: they must be independent forcings. If your four archetypes are four descriptions of one selling event, you have a stack of one. And every stack still has to clear the five tests above — embeddedness multiplies an edge, it does not create one.

The alert rack

Most of these archetypes announce themselves in language. Standing alerts and saved screens do the finding; you do the work afterwards. Set the alerts once.

Standing alert or screen	Catches	Where it lives
“suspends its quarterly dividend”	Dividend-cut exile	News alert · 8-K
“strategic alternatives”	Original archetype — and it front-runs half of family III	News alert
“appointed chief executive officer”	New CEO, then the kitchen sink two quarters later	News alert · 8-K
“discontinues coverage” / analyst count → 0	Coverage cliff	Vendor field, monthly diff
consensus EPS is null & net income > 0	No consensus estimate	Saved screen
“terminates the merger agreement”	The arb liquidation	News alert
“emerged from Chapter 11”	Fresh-start equity	News alert · docket
“reverse stock split” + “holders of record”	Going dark	Proxy alert
Form 15 filings	Going dark, confirmed	EDGAR full-text
“nil-paid rights” / rights issue announcements	Rights-issue wreckage	RNS · regulatory feeds
“transfer its listing” / uplisting notices	Listing migration	News alert · exchange notices
“special cash dividend” > 20% of cap	The stub	Corporate-actions feed
13D/G amendments where holder > 5%	The wind-down seller, the estate sale	EDGAR, daily
S-1 lockup schedules, diarised	The lockup calendar — second and third tranches	Your own calendar
Index review dates: recon, quarterly rebalances	Deletion overhang, float inflation	Index provider calendars
“policy shareholdings” reduction targets	Cross-shareholding unwind	Japanese annual reports
A/B class spread > 5%, screened weekly	Dual-class collapse	Saved screen
ADV < $2m & cap > $200m & FCF > 0	The liquidity floor, size-band exile	Saved screen
D&A > 2× capex, sustained	Purchase-accounting drag	Saved screen
Deferred revenue growth > revenue growth	The model break	Saved screen
Reporting frequency = 2/yr	Off-cycle reporting	Saved screen · one-off list
Down > 30% YTD, small cap, mid-November	The tax calendar	Seasonal screen

What this list is not

It is not a screen for cheap stocks, and none of these entries is a reason to own anything on its own. A POSI tells you where to look and why the price might be wrong — the business still has to be worth owning at the price the forcing creates. Half the entries above are also perfectly good descriptions of a company in genuine decline, which is why every card carries a trap. The forced seller and the correct seller often stand in the same queue.