The Three Allocations — Summary

The taxonomy of arbitrage is also a taxonomy of capital strategy. Once the four fates are clear, the capital response to each is nearly forced, and the entire transition reduces to a small decision structure: four fates, three honest allocations, one trap. This is not a forecast. It is a discipline for pricing. It says nothing about which industries fall when, only that for any position, once its arbitrage layers are classified by fate, the correct capital response is determined, and the most common capital errors of the moment are failures to match the response to the fate. Dissolving arbitrages are not holds. They are trades or avoidances, because the spread is closing and a hold prices a persistence that will not arrive. Intensifying arbitrages are the instrument play. Delaminating arbitrages are layer-by-layer reads. The surviving arbitrage, relationship, is a durable hold.

The first honest move is to own the durable, the arbitrages that survive or compound: scale where volume thresholds confer a structural cost advantage, relationship-dense businesses, track-record trust that cannot be shortcut because it is made of time, and patient-capital time, which survives because it is a property of capital structure rather than information. These are holds, priced for a durability that is not an illusion the audit will dispel. The discipline within the allocation is to isolate the durable layer from the melting layers bonded to it; a relationship-dense business usually also carries a dissolving information layer, and the correct move is to value the relationship layer as the asset and price the information layer at its melt-adjusted runoff, not to pay a blended multiple for the whole. A second discipline is less obvious: the durable layers are mostly unscalable, which caps the return even as it secures it. Relationship does not roll up; a hundred relationship-dense businesses are a hundred separate small durabilities, not one large one. The durable allocation is safer, because the audit cannot reach it, and smaller, because the same quality that protects it prevents it from compounding into scale. Capital that needs the durable to also be large will be disappointed, and the disappointment will tempt it back toward the melting arbitrages, which do scale, right up until they melt.

The second honest move is to own the instrument that performs the audit, the highest-conviction durable allocation in the transition. The instrument compounds as the businesses it audits melt, creates value twice, and its standalone value exceeds the sum of the operational gains. It sits above the others, because it is the one position whose advantage is fed by the very process destroying the rest. The durable arbitrages survive the audit. The instrument is the audit, capturing a slice of every spread it closes. The discipline is to pay instrument multiples only for the instrument and to treat the businesses it runs across as deployment and data. The error here is the inverse of melting ice: mistaking a deployment surface for the durable position, paying for an operator’s persistence on the theory that proximity to the instrument confers the instrument’s durability, which it does not. It is the rarer error and the more expensive one when it occurs.

The third honest move is the one most easily confused with the trap, and the difference is entirely a matter of discipline. A dissolving arbitrage still throws off margin in its final years. The spread is closing, but closing is not instant, and the runoff is real cash that can be harvested by a party who enters with a scheduled exit and treats the position as a depleting trade rather than a going concern. The distinction from the trap is the date. The melting-ice error is holding a dissolving arbitrage as an investment, pricing it for a persistence it does not have. Allocation three trades the same arbitrage as ice, with the melt modeled explicitly and the exit dated before it completes. The same asset, priced two ways: as a going concern it is a liability; as a depleting trade with a disciplined exit it is a legitimate position. The error is not buying the melting arbitrage. The error is forgetting that it is melting. The funds that run this well treat the melting arbitrage the way a resource extractor treats a depleting field, the value the scheduled drawdown, the asset gone at the end by design. The funds that run it badly are running the trap with a rationalization attached, telling themselves they will exit in time and then holding past the date because the cash still looks good. The melting arbitrage is a trap when held as an investment and a legitimate position when traded as ice, and the only difference is whether the exit is dated before the melt completes.

The full structure is a single decision frame. Classify every arbitrage layer in a target by fate. Own the durable layers as holds priced for their real durability. Own the instrument as the highest-conviction position. Trade the dissolving layers, where the runoff justifies it, as depleting positions on dated exits. And never hold melting ice as though it were a going concern, the one move the structure forbids. It does not tell you which industries are melting fastest, which is the work of the next arc. The matrix is the lens. The industries are what the lens is pointed at. Two properties matter before the arc applies it. First, it operates at the layer, not the business; no real target is purely one fate, and a single business often warrants a durable hold on one layer, a dated trade on another, and a zero on a third. A fund that allocates at the level of the business rather than the layer will misprice every target that is not unusually pure, which is nearly all of them. Second, the matrix is indifferent to sentiment. It does not ask whether a destruction is good or bad, only which fate each layer is on, and it returns the same instruction for a liberating dissolution and a painful one. The next arc opens on agriculture precisely because its dissolution is the clearest case of destruction that helps the powerless, and the matrix will treat it exactly as coldly as it treats every other industry. That is the point: the allocation logic does not change with the moral valence of the melt.