The Escapist Doctrine
Most of the market is trapped in a game it does not even realize it is playing. Picture two companies staring at each other across a table, each holding the same lever: cut the price, win the quarter. The moment one pulls it, the other has to match, and then both pull again, and again, until the margin that used to fund their future is sitting on a clearance rack. This is the corporate version of the Prisoner's Dilemma, and it is the single most reliable wealth-destroyer in public equities. We do not try to pick the winner of that fight. We try to find the companies that refused to sit at the table.
That is the whole doctrine in one sentence: escape the price war or profit from everyone else's. The market spends enormous energy handicapping who will win the discount race, when the more durable question is who is not racing at all. A business that competes on price is renting its revenue from customers who will leave the instant a cheaper option appears. A business that competes on something a rival cannot copy overnight is, in a real sense, playing a different game with a different scoreboard. The price warrior is measured by share this quarter; the escapist is measured by profit that compounds for years.
It helps to be honest about why the dilemma is so hard to escape. Cutting price always works the first time. It is the most visible, most immediately rewarded move a management team can make, and it shows up in the next earnings call as growth. The trap is that it is equally available to your competitor, who will reach for the same lever the moment it costs them share. What feels like strategy is really just the opening move in a sequence that ends with both sides poorer. The companies we want are the ones who saw that sequence coming and built a business where the lever does not exist.
We do not pick the winner of the price war. We buy the companies that walked away from the table.
So we sort the entire investable universe into three postures. Not sectors, not styles, not market-cap buckets in the usual sense, but postures toward this one inescapable game. Two of those postures are traps we deliberately avoid. The third is the only place we are willing to put money, and even then only when the math and the price both agree. The rest of this piece is how we draw those lines, the lenses we use to turn a posture into a number, and the discipline that keeps us out on the nights when nothing is worth owning.
The Three Postures
The first posture is the Defector — the company that pulls the price lever first and loudest. On the surface it can look like a winner: share is climbing, the top line is growing, management sounds aggressive on the call. But it is winning a contest whose prize is a thinner margin, and the moment a competitor matches, all that share comes with less profit attached to each unit of it. Defectors manufacture the very race to the bottom that poisons their own returns. We avoid them, no matter how exciting the growth headline looks, because growth bought with margin is not growth — it is a slow liquidation dressed up as momentum.
The second posture is the Victim — the company that did not start the fight but is bleeding out in it anyway. You can read it straight off the income statement: operating margins compressing quarter after quarter, the business defending a position it can no longer afford to hold. Victims are sometimes mistaken for value plays because the stock is cheap and getting cheaper, and a falling price tempts people who confuse low with safe. It is cheap for a reason. A collapsing margin is not a discount; it is a leak, and you cannot buy your way to a good outcome by paying less for a business that is structurally giving away its profit.
A collapsing margin is not a discount. It is a leak.
The third posture is the Escapist — and this is the only group we target. Escapists come in two flavors. The first has built a wall the price war cannot reach: a real moat, a brand customers ask for by name, or switching costs so high that leaving for a cheaper rival is more painful than just staying. When a customer has wired your product into their workflow, their billing, their habits, a competitor's lower price is not an offer — it is just noise. The second flavor is more elegant still.
That second flavor is the anti-fragile supplier, the pick-and-shovel business that sells the tools every combatant needs. When the rivals burn cash fighting each other, the supplier collects on every shot fired. They do not pick a side and they do not need to win the downstream war, because their customer is the war itself. They do not just survive the fight; they are paid by it. Whichever Defector ends up bankrupt and whichever survivor ends up with thin margins, the supplier has already been paid for the ammunition both of them bought.
The Hunting Ground
Posture tells us what to look for. Size tells us where to look. We hunt almost exclusively in the roughly two-to-fifteen-billion-dollar mid-cap band, and the boundaries are not arbitrary. Below about two billion you are wading into penny noise: thin liquidity, erratic disclosure, and a signal-to-noise ratio that punishes process. The numbers are too easy to manipulate and too hard to trust, and a good thesis drowns in volatility that has nothing to do with the business. Above about fifteen billion the convexity is mostly gone — a great thesis on a ninety-billion-dollar company might be right and still barely move the stock, because the size is already in the price and a thousand analysts already wrote the same note.
The mid-cap band is the sweet spot for a reason that is almost mechanical. It is large enough that the company is a real business with audited statements, durable customers, and a story you can actually verify, but small enough that the market has not yet fully agreed on what it is worth. That gap between a real business and a settled price is exactly where money is made. A mid-cap with a genuine moat can re-rate hard when the market finally understands it, and that re-rating is the part of the return you cannot manufacture by being merely correct on a giant.
Inside that band, two tailwinds shape almost everything we find interesting right now, and they are physically linked. The first is the build-out of AI compute — the chips, racks, and data centers that the entire industry is racing to stand up. The second is the energy required to feed that compute: SMR and conventional nuclear, the uranium that fuels it, and the grid infrastructure that moves the power to where the servers live. You cannot run the models without the electrons. That dependency is exactly the kind of unavoidable bottleneck where pick-and-shovel Escapists thrive.
Notice how naturally these tailwinds produce Escapists rather than Defectors. The companies fighting to sell the cheapest finished AI product are Defectors in the making — the model layer is already a price war, with each release undercutting the last. The companies selling the irreplaceable input — the cooling, the power, the fuel, the specialized component everyone needs regardless of who wins — are structurally positioned to profit from the war instead of in it. That is the doctrine and the macro pointing in the same direction, which is the kind of alignment we are willing to act on.
The Hardest Bottleneck: Energy
The reason we keep returning to energy is that it is the constraint nobody can engineer around with software. You can write a cleverer model, compress the weights, optimize the data pipeline — but the data center still has to draw real power off a real grid, and that grid was not built for this kind of demand. The supply of electrons is slow, capital-intensive, and regulated, which means the companies that already own generation, fuel, or transmission are sitting on an asset the AI buildout cannot route around. That is the definition of an anti-fragile position: the more frantic the competition upstream, the more valuable your bottleneck becomes.
This is why nuclear, SMR, uranium, and grid infrastructure keep surfacing in our screens. They are not exciting consumer stories and they will never win a marketing war, but they sell something every combatant in the AI race is forced to buy, at terms that improve as demand outruns supply. A uranium producer does not care which AI lab wins; it cares that all of them need baseload power, and that the fuel for that power has a supply chain measured in years, not quarters. Scarcity on the supply side is the moat that the customer cannot dig around.
You cannot run the models without the electrons — and the electrons are the one thing software cannot fake.
None of this means we buy the theme. The macro tells us where to point the telescope; it never tells us what to own. A tailwind is only useful when it lands on a specific company that also passes the posture test and the valuation gate, and plenty of energy names are Victims dressed in a hot narrative — overbuilt, over-promoted, and trading on a story rather than a margin. The tailwind earns a name a closer look. It does not earn the name a place in the portfolio.
Three Lenses That Make Posture a Number
A posture is a story, and stories are easy to fool yourself with. So we force each one through three quantitative lenses that turn a narrative into a measurement. The first is margin stability: the standard deviation of operating margins across five years. A company with low margin volatility and a high margin level is, almost by definition, exercising pricing power — it sets prices and they hold, war or no war. A company whose margins lurch around is telling you the market sets its prices for it. Low volatility plus a high base is the statistical fingerprint of an Escapist, and it is very hard to fake across half a decade of statements.
The second lens is the R&D-to-Capex ratio. This one separates the companies building a new pie from the ones spending to grab a bigger slice of the old one. Heavy capex with thin R&D usually means a business buying more capacity to wage the same commodity fight — a Defector or a Victim funding its own race to the bottom, adding supply to a market that already competes on price. A high R&D-to-capex reading means the company is inventing the thing that does not exist yet, which is precisely how an Escapist stays out of reach: you cannot price-war a product your competitor has not figured out how to build.
High volatility says the market sets your prices. Pricing power says you set theirs.
The third lens is LTV/CAC — lifetime value over acquisition cost. When that ratio sits comfortably above three, it means customers are worth far more than they cost to win, which only happens when they stay, renew, and rarely shop around. That stickiness is lock-in the price war literally cannot reach, because the customer is not in the market to be poached. No single lens is proof on its own — a number can always be flattered for a quarter — but together they let a posture survive contact with the spreadsheet, not just the pitch. When all three agree, the story and the statements are telling you the same thing, and that agreement is what we trade on.
The Reflexivity Check and the Valuation Backstop
Finding a real Escapist is necessary but not sufficient, because a great company is not the same thing as a great investment. The bridge between them is price, and here we apply a reflexivity check. If a name has already sprinted to its fifty-two-week high and the bullish thesis is on every screen and in every newsletter, then the good news is already in the price. We do not pretend that is fresh upside. We mark down the expected return accordingly, and an Escapist that has already been discovered often fails to clear our bar precisely because everyone else found it first. The best businesses are frequently the worst trades, simply because the crowd got there before you.
The backstop under all of it is a DCF frame for expected return: dividend yield plus earnings growth plus multiple expansion. We want that sum north of twelve percent, deliberately above a ten-percent floor, so there is a genuine margin of safety baked in. The extra two points are not greed; they are the buffer that protects us when our assumptions turn out to be too optimistic, which they sometimes will. Only a still-undervalued Escapist carrying a double-digit expected return clears that gate. A wonderful business at a punishing price is a pass, full stop.
Mechanically, the engine does this every night without ego. It pulls real, live prices, grades each name on a fast cloud model against the postures, the three lenses, and the valuation gate, and then publishes only the survivors. There is no narrative override, no falling in love with a logo, no rounding up because a name feels right. The model does not care what was exciting yesterday; it only cares what clears the bar tonight at tonight's price, and that price changes everything from one evening to the next.
Some nights nothing survives — and that is not a bug, it is the entire point. A screen that always has something to say is a screen with no standards, because there is no rule that the market must offer a great business at a fair price on any given Tuesday. The discipline to publish an empty list is what keeps the published list worth reading. Cash is a position, and on the nights when nothing clears, it is the only honest one.
Some nights nothing clears the bar. That silence is the strategy working.
So the doctrine, end to end: refuse the Prisoner's Dilemma, avoid the Defectors who start the price war and the Victims bleeding out in it, target the Escapists who either walked away from the table or get paid for the fighting, hunt them in the mid-cap band where a real business can still re-rate, ride the linked AI-compute and energy tailwinds where pick-and-shovel businesses naturally cluster, prove the posture with margin stability, R&D-to-capex, and LTV/CAC, then demand a double-digit expected return with the reflexivity discount already applied. Get all of that, or hold cash. There is no prize for forcing a trade, and there is no penalty for waiting that comes anywhere close to the cost of buying the wrong thing.
This article is opinion and editorial content, not investment advice; it is not a recommendation to buy or sell any security, and the author accept no liability for decisions made based on it.