Why Hold

A holder who wants verified service in a bad state can hold a Work Credit, hold Bitcoin and buy on the day, or hold the base asset. The case for the third was first set against Bitcoin (Chapter 7: Why Not Something Already Here?). This chapter follows the channels through which service value reaches the asset, and asks how much of that value is money.

Economic Linkage: How Triad Demand Becomes Asset Value

One question separates a systems manifesto from a monetary thesis: why does demand for Privacy, Proofs, and Compute raise the value of holding the asset, rather than merely rewarding consuming a service?

The familiar answer is a chain that runs from fee demand through burns to scarcity to price. That chain fails (Chapter 13: Why Hold the Base Asset?) and is not used here in any form. The answer below is in four parts, of which the first is a cash-flow claim the thesis does not describe as monetary, and the remaining three are the monetary content.

1. What exactly is the asset?

The base asset is the protocol-internal unit in which fees may be required, collateral posted, and settlement denominated. It is a conditional monetary candidate and nothing in this section upgrades it. Work Credits are typed capacity or service claims; FCR, FER and PIDL artifacts are evidence; project notes are duration-bearing credit; LP and staking positions are derivatives or operating claims (Chapter 9: The Triad and the Monetary Candidate). Only the base asset is evaluated as money.

The strong phrasing, that all uses of the triad must flow through the asset, overstates the design. Core protocol operations are denominated in the asset by rule. Equivalent computational content is obtainable elsewhere, always, at a price; bypass is a magnitude rather than a switch (Chapter 13: Why Hold the Base Asset?), and the design constrains what that magnitude is worth rather than abolishing the alternative.

2. Why is it scarce in a monetary sense?

Scarce capacity does not imply a scarce asset. The design forges the link through scheduled issuance, fee retirement, collateral lockup, and Work Credits sold against committed, verifiable capacity rather than by decree.

The link needs two precisions. First, base-asset issuance follows the constitutional schedule and is independent of Work Credit dynamics; it is Work Credits that are bounded by delivered verified capacity. Second, whether burns exceed issuance at steady state (whether net supply actually falls as workloads grow) is an empirical outcome the fee-coverage and issuance boards report, not a property the design confers. Issuance discipline is real. Net deflation is a reading.

3. Why does demand for triad capacity raise the value of holding?

Four channels, stated in ascending order of monetary content, with the first two explicitly conceded not to be monetary at all.

1. A fee stream (not monetary). Every proof, private settlement and verified inference pays a fee in the native unit, and routing rules direct it to retirement, staking, and operator share. What the holder owns through this channel is a competitively priced claim on a cash-flow stream bounded above by differential value $\Delta$ and deflating with unit price (Chapter 13: Why Hold the Base Asset?). A discounted-cash-flow valuation reproduces it exactly. A burn is a buyback and confers no moneyness.

2. A collateral lockup (not monetary). Required lockup in the native unit immobilizes float while operators remain solvent and choose to keep operating. Complete Edition, Chapter 13: Why Hold the Base Asset?^† bounds how large a fact this can be (large-cap-equity magnitudes, not monetary aggregates) and shows it is a slope effect on price impact rather than a level effect on price, because a lockup places no bid.

3. Pledgeability without due diligence (monetary). This is the first channel through which growing triad demand raises the value of holding rather than of receiving. As verified throughput grows and receipts accumulate, the cost to any counterparty of establishing what a pledged position actually is falls toward zero, and a position nobody must investigate can be accepted no questions asked (Chapter 15: The Three Monetary Mechanisms). The holder is not being paid; the holder is holding something that has become easier for everyone else to accept. Nothing is distributed through this channel, so a valuation of distributions alone omits it. Whether a holding service of this kind is monetary is what Chapter 15: The Three Monetary Mechanisms and Red Line 17 test.

4. A unit with something to denominate (monetary). A growing endogenous service economy gives counterparties a reason to reckon in the unit rather than merely settle in it, and denomination is the stickiest monetary function because it is a coordination device rather than a portfolio choice (Chapter 15: The Three Monetary Mechanisms). This channel is available to a service protocol and is not available to a bearer asset with no native goods.

A fifth channel, the state-contingent holder-side service flow, is argued separately at Chapter 15: The Three Monetary Mechanisms, because it is a property of the regime rather than of demand growth.

Channels 1 and 2 pay the holder. Channels 3 and 4 change what holding is. Only the second kind is monetary, and only the second kind survives a correctly specified discounted-cash-flow valuation.

A holder of channels 1 and 2 alone is buying “a share of the fee stream from indispensable workloads,” a position “closer to equity in a utility than to a collectible.” That description is accurate for those channels, and it is the reason they cannot carry the monetary reading: a utility equity earns its cost of capital and carries no monetary premium. Channels 3 and 4 are the reply, and they are a reply about the asset’s acceptability and denomination rather than about its cash flows. Is either sized? Not in this document.

4. What prevents capacity providers from capturing all value?

The classic “utility token trap”: if operators earn the fees and governance can inflate supply, holders are exit liquidity. The design’s answers are routing rules, and they are stated as such: fees split between retirement, stakers and operators so no party takes the whole; base-asset issuance constitutionally scheduled; Work Credit sales bounded by reserved capacity, with overselling rejected and visible; core operations denominated natively.

These govern where the fee goes, and none of them governs how large the fee can be, which is the question Chapter 13: Why Hold the Base Asset? takes up and answers with a bound rather than a promise. Read this paragraph as a description of plumbing. The monetary argument is in channels 3 and 4 above and in Chapter 15: The Three Monetary Mechanisms, and it does not depend on any number in this paragraph.

Summary: the Value Capture Lemma in plain language

Demand for triad capacity becomes value accrual to the native asset only if five conditions hold at once: all core usage requires the asset, usage retires supply, capacity provision requires staking, issuance is constrained, and capacity cannot be cheaply bypassed. These conditions are necessary and not sufficient, and what they establish is accrual — a cash-flow claim and a collateral lockup — not moneyness. If any one fails, the system may be indispensable infrastructure whose economics accrue to operators or customers.

A related but separate claim. The five conditions are properties of the monetary design and say nothing about the durability of the demand being routed. That workload demand is structural rather than speculative is a distinct empirical claim whose failure the thesis treats as its own falsifier: a workload mix that becomes and remains predominantly speculative retires the store-of-value claim through Red Line 5, not through this lemma.

Value Capture Lemma

In the reference design, demand for Privacy, Proofs, and Compute accrues economically to the native asset — as a claim on the fee stream and a collateral lockup — through five channels, and the design relies on all five holding together:

1. Required fee medium: the asset is required for core fees, and users cannot pay in fiat, stablecoins, or other tokens at equivalent service quality.

2. Supply reduction: a meaningful share of fees is burned or permanently retired.

3. Collateral lockup: operators must stake the asset as collateral to provide services.

4. Issuance discipline: gross issuance follows a schedule fixed ex ante, not governance fiat.

5. Non-bypassability: users cannot obtain equivalent triad capacity through bypass channels (cloud contracts, stablecoin-denominated services, direct fiat payment to operators) that avoid touching the asset.

Each channel closes a named leak, and in this design nothing else stands in for a channel that fails. The five are this design’s route to accrual, not the only route economics allows: a protocol can take fees in another unit and use them to buy and retire the asset, distribute revenue instead of burning it, or take native fees without native collateral, and each of those captures value by a different route with different risks. The lemma is the bridge between “these capacities are indispensable” and “therefore this asset accrues the value of that indispensability.” Accrual is a precondition the monetary claim builds on, not the monetary claim itself (Chapter 13: Why Hold the Base Asset?, Chapter 15: The Three Monetary Mechanisms).

Which conditions the monetary channels actually use.

Of the five conditions, four govern accrual, which is a discounted-cash-flow quantity. Issuance discipline is the exception: a discretionary supply is information-sensitive on supply, for the reason Chapter 15: The Three Monetary Mechanisms gives. Condition 4 is therefore doing monetary work that Conditions 1, 2, 3 and 5 are not. That is the same conclusion Chapter 15: The Three Monetary Mechanisms reaches from the other direction, and the two agree for a stated reason rather than by coincidence.

What the five conditions do not address.

Every one of them governs where fees go. None governs how large they can be, and a routing rule applied to a fee of nothing captures nothing. Chapter 13: Why Hold the Base Asset? takes up the level question, finds that the standard objection to taxing a contestable market is wrong for a reason worth stating, and then finds that the correct answer supports a narrower conclusion than the claim that fees, burns, and collateral earn the asset a monetary premium (Chapter 13: Why Hold the Base Asset?). The conditions themselves are unchanged.

The lemma has a frequently-missed converse. It states when native demand accrues economically to the asset. It does not state that movements in the asset’s observed market price are caused by that demand. Those are different machines, and conflating them is the most common way a monetary thesis fools its own author.

Corollary: Market-Price Non-Equivalence

External financial wrappers can create price demand without exercising privacy settlement, purchasing proofs, consuming verified compute, paying native fees, locking collateral, or producing burns. Conversely, wrapper redemptions, leverage unwinds, dealer hedging, or broad risk-off flows can depress the asset’s price while native monetary usage improves. Therefore

$$\text{price appreciation} \;\nRightarrow\; \text{monetary adoption}$$

and

$$\text{monetary adoption} \;\nRightarrow\; \text{immediate price appreciation}.$$

Price is an output to explain, not a protocol KPI to worship. The monetary thesis must be evaluated through native fees, burns, collateral, settlement, receipt volume, verification health, reachability, and non-custodial use. Market-price interpretation requires a separate financial-flow telemetry layer (VerifyFlow, Complete Edition, Chapter 26: Extended Telemetry^†).

Value Capture vs. Price Capture

It is worth separating the two phenomena explicitly, because the thesis is routinely “confirmed” by the wrong one.

Value capture is the existing loop: triad usage $\rightarrow$ native fees $\rightarrow$ burns, collateral lockup, and operator demand $\rightarrow$ scarcity and cash-flow accrual to the asset. It is governed by Chapter 13: Why Hold the Base Asset?. What it produces is a claim on a fee stream plus a collateral lockup; whether the asset also carries monetary premium is a separate question, answered separately in Chapter 15: The Three Monetary Mechanisms.

Price capture is demand for exposure to the asset’s price, arriving through spot ETFs, exchange custody, treasury companies, margin products, options, futures, swaps, leveraged and inverse ETPs, passive indices, and systematic trading rules. Price capture can occur with no protocol use whatsoever.

Native value capture and observed price appreciation are neither equivalent nor mutually necessary. No cell is monetary evidence: that is a separate axis, read on Red Lines 16–18.

The diagonal is what makes the thesis falsifiable. A thesis that treats every price rise as confirmation and every price fall as noise cannot be wrong, and therefore cannot be right either.

Fee Incidence and the Level of the Fee

Every condition of the Value Capture Lemma answers one question: where does the fee go? Required medium, supply reduction, collateral lockup, issuance discipline, and non-bypassability are all routing rules. None of them answers the question a valuation actually needs answered, which is how large the fee can be. A protocol that routes the whole of a fee of nothing into a burn address has satisfied five conditions and captured zero.

The gap has a sharp form, best stated as the objection it invites:

“You have deliberately engineered the most contestable market you could. Workloads are canonical and interchangeable, verification is cheap enough that no trust is required, admission is open, and there are no vendor chokepoints. A contestable market competes economic profit to zero. Zero profit means no surplus, and no surplus means nothing to burn.”

The objection is wrong. It is worth showing exactly why, because the correct answer rescues considerably less than it first appears to, and what survives is not the claim that fees, burns, and collateral earn the asset a monetary premium.

Incidence: A Fee Is a Wedge on Turnover, Not a Claim on Profit

The objection conflates two different bases. A protocol fee is levied on transactions, not on residual profit. It is a tax wedge, and tax wedges do not require anybody to be earning rents.

Take the market the thesis is trying to build: many providers, free entry, roughly constant returns to scale in the relevant range, a canonical workload so that units are fungible. Long-run supply is then approximately horizontal at marginal cost $c$ (equivalently the protocol’s all-in unit cost $c_p$ of Chapter 13: Why Hold the Base Asset?; write $c \equiv c_p$). Impose a per-unit protocol fee $f$ on that market and the supply curve the buyer faces shifts to $c+f$. The new equilibrium has buyers paying $c+f$, providers receiving $c$, and providers earning exactly the zero economic profit they earned before. The protocol collects $f \cdot Q$. Nothing about that revenue came out of operator margin, because there was no operator margin to take it from; in the long-run competitive case the entire statutory burden lands on the buyer.

Operator margin compression is irrelevant to fee revenue. The fee is a wedge on turnover. Turnover is what a contestable market maximizes.

This is standard incidence analysis, and it means the objection proves too much: if contestability destroyed fee bases, no commoditized market could ever be taxed, and every one of them is.

Incidence explains who bears a fee. It does not explain why a positive fee is sustainable, and the two should not be run together. Exchanges, clearing houses, and toll roads are lucrative on top of viciously competitive flows for a further reason: the flow is competitive but the fee layer is not. There is one place to clear the contract, one road across the water, one book where the trade is legally final, so the competition among users of the layer never becomes competition to be the layer. That inescapability is a property those venues possess and a protocol must earn; it is the subject of Chapter 13: Why Hold the Base Asset? and Complete Edition, Chapter 13: Why Hold the Base Asset?^†, and the fork channel is precisely the case where it is not earned.

The short run is different, and the difference matters.

With installed capacity fixed, supply is not horizontal. Deployed hardware has already sunk its capital, so in the short run the supply curve slopes upward and part of the fee is absorbed as a reduction in quasi-rent to existing machines rather than passed to buyers. That is not a problem for this quarter’s fee revenue; it is a problem for the next round of capacity. Quasi-rent is what funds replacement, and an operator whose realized return on deployed hardware is being clipped by a fee re-prices the entry decision accordingly. The long-run pass-through result therefore describes the destination, not the transition, and the transition is where entry and capacity growth are decided. A fee schedule set as though incidence were instantaneously long-run will suppress the capacity it is trying to tax.

The Base, and Why It Deflates

Fee revenue is not $f$. It is

$$\mathrm{Rev} = f \cdot Q(c_p + f),$$

and both terms are endogenous. Two consequences follow, and the second is the more serious one.

Quantity responds to the wedge.

The tollbooth analogy holds only where there is no parallel free road. Raising $f$ raises the delivered price, and at the margin some volume declines, defers, batches, or leaves. What the protocol can extract is bounded not by what buyers would pay for the service but by what they would pay through this channel rather than the next-best one, which is the subject of Chapter 13: Why Hold the Base Asset?.

An ad valorem fee on a deflating unit price is a shrinking real toll.

For this paragraph the protocol’s fee is modeled as ad valorem: a share of unit price rather than the per-unit wedge $f$ used in the incidence analysis above. Under a per-unit fee only the volume term below moves, and the deflation mechanism this paragraph isolates does not arise. Complete Edition, Chapter 11: Compute Through the “AI Money” Lens^† argues that verified compute stays scarce even as raw compute deflates. Grant that argument entirely; it does not touch this one. Even if the verification premium is fully durable, a fee expressed as a percentage of a unit price that falls secularly collects fewer real resources per unit of service delivered every year. Real fee revenue then grows only if

$$\underbrace{\text{growth in verified units delivered}}_{\text{volume}} \;>\; \underbrace{\text{rate of decline in real unit price}}_{\text{deflation}}.$$

That inequality is an empirical bet, not a theorem, and the thesis is making it. It may well be a good bet (the historical pattern in compute is that unit-cost decline is met by more than proportional volume expansion), and a bet must say what would settle it against the thesis. Condition B of Red Line 14 (Chapter 17: Risk Analysis & Failure Modes) does: deflation-adjusted gross native fee turnover that fails to grow while physical throughput grows is this bet losing, in public.

The bet has a macro frame.

The volume-over-deflation bet is usually argued from technology curves: unit costs fall, demand expands more than proportionally, revenue grows. That is the technological channel, and it is the one Complete Edition, Chapter 11: Compute Through the “AI Money” Lens^† treats. There is a second channel with the same signature but a different driver: a financial clearing event that transfers installed capacity from weakly financed owners to strongly financed users at lower capital cost, collapsing the effective price of capacity without any change in the technology [Perez 2002]. The two channels are observationally distinct. The technological channel shows up as gradual unit-cost decline with stable utilization, the financial one as a step-change in the price of installed capacity with utilization resetting. They have the same effect on the inequality above: both push the deflation term up. What the frame adds is timing. The technological channel is smooth; the financial channel is not, and a fee base built on the expectation of smooth deflation being outrun by volume will experience the financial channel as a shock to exactly the term it is betting on. The instrument is unchanged, since Condition B reads on the same series either way. But a reader tracking it should know that the two channels are live simultaneously, and that the financial one can move a quarter’s reading by more than a year of Moore’s law.

The Bound on the Sustainable Fee

Incidence tells us the fee has a base. It does not tell us how high the fee can be set. That ceiling is not a function of operator margin either; it is a function of what the protocol channel is differentially worth.

Differential Value $\Delta$ and the Sustainable Fee Bound

Let $c_p$ be the protocol’s all-in cost of delivering canonical workload $W$ at tier $T$, and $c_b$ the all-in cost of the best bypass channel delivering the same computational content without protocol-grade verification. Let $\Delta$ be the marginal buyer’s willingness to pay for the protocol’s differential properties: cryptographic checkability, neutrality, credible non-discretion, and non-custodial exit.

The buyer routes through the protocol only if $c_p + f - \Delta \le c_b$, so the sustainable fee satisfies

$$\boxed{\; f \;\le\; \Delta - (c_p - c_b) \;}$$

Here $\Delta$ is a scalar wedge in price units, not the difference operator used elsewhere in this Part.

The first term is the one a protocol designer wants to think about. The second is the one that decides the outcome, and it is structurally positive: verified compute costs more to produce than unverified compute, because proving overhead and redundancy are real resources spent on top of the underlying work. So the protocol begins every negotiation with a cost handicap and must recover it out of $\Delta$ before it collects a single unit of fee.

The document’s own tier multipliers are cost pass-through, not evidence of willingness to pay.

Complete Edition, Chapter 22: Layer 4: Truth & Work^† prices assurance on its own axis: Replicated at $1.0\times$, Replicated-plus-audit at $1.4\times$, Probabilistic (M2) at $1.6\times$, Succinct (M1) at $2.5\times$, with a latency multiplier of $1.0$/$1.3$/$1.6$ multiplied in. Read those against what each tier costs to supply. Replicated-plus-audit is three independent re-executions and a randomized audit, so its multiplier is set against roughly three times the production cost; the M1 multiplier is set against a proving overhead that Complete Edition, Appendix E: Hardware Profiles^† publishes as a quarterly series and does not yet know. The multipliers are rationale-set to cover redundancy or proving overhead plus a margin, and the document nowhere establishes that they do. That is a question about $c_p$ and the margin above it, not about demand, and nothing in the document establishes that any buyer’s $\Delta$ for succinct assurance is $2.5\times$ its $\Delta$ for a committee’s word. If it is not, headroom narrows as assurance rises, and at the top of the axis $\Delta - (c_p - c_b)$ can be negative: a tier that is technically superior, honestly priced, and unsellable. That is a testable prediction against the thesis, and a red line measures it (Chapter 17: Risk Analysis & Failure Modes).

Bypass Is a Magnitude, Not a Binary

Condition 5 of the Value Capture Lemma is written as an existence condition (users cannot obtain equivalent capacity through a bypass channel), and Red Line 6 is written as a direction condition: usage grows while native fees do not. Neither is wrong, and neither is changed here. But both are read, in practice, as though bypass were a switch.

It is not. Bypass is never impossible. It is priced. There is always a channel that delivers the same computational content without protocol-grade verification, and the only question is what the buyer gives up by using it, which is exactly $\Delta$. Condition 5 should therefore be read and measured as a magnitude claim: not “no bypass exists,” which is false everywhere, but “$\Delta$ is large enough, for enough of the volume, to support a fee worth burning.” The condition stands as stated. Its interpretation is what this section tightens, and its measurement is what Red Line 14 supplies.

This is why the chain-strength rating for non-bypassability is Medium rather than Strong (Chapter 9: The Triad and the Monetary Candidate). The condition is well specified as a binary and the binary is well enforced. The economics requires a magnitude, and the magnitude is nowhere estimated. A rating of Strong would rate the enforcement machinery, not the quantity the argument depends on.

What Determines $\Delta$, and Why It Is Probably Thin

To first order, the differential value of a proof is the loss it prevents, times the probability that loss survives the alternative:

$$\Delta \;\approx\; \varrho\,\mathcal{L},$$

where $\mathcal{L}$ is the loss from undetected misexecution or denial of service and $\varrho$ is the residual probability of that loss under the best bypass channel. This operationalization lower-bounds $\Delta$. It prices only loss-prevention, while the full definition in Chapter 13: Why Hold the Base Asset? covers four properties: cryptographic checkability, neutrality, credible non-discretion, and non-custodial exit. Neutrality and non-discretion therefore enter as additional terms the $\varrho\,\mathcal{L}$ form omits. ($\varrho$ and $\mathcal{L}$ are used rather than the conventional $\pi$ and $L$, which denote the inflation rate in Chapter 3: The World Forces New Monetary Primitives and the leverage multiple throughout VerifyFlow and Complete Edition, Appendix H: Formal Model of Market Realization, Wrapper Flows, and Price Capture^†.)

$\Delta$ is a distribution, not a number.

Before evaluating either term it is worth fixing what object $\Delta$ is. Every buyer has their own $\varrho$ and their own $\mathcal{L}$, so $\Delta$ is a distribution across the buyer population. The fee bound of Chapter 13: Why Hold the Base Asset? binds at the marginal buyer (the last one the protocol wants to retain at the posted fee), not at the median one. These come apart, and the difference is the whole commercial question. A thin median is entirely consistent with a lucrative fee, provided the upper tail is deep enough to be worth serving. That tail is made of counterparties nobody will indemnify because they are sanctioned or pseudonymous, flows crossing jurisdictions where no single forum will hear the claim, regulated model-risk and audit functions that must evidence correctness to a supervisor rather than merely believe it. For those buyers $\varrho$ is not small, because the indemnitor substitute is unavailable rather than merely inferior. The protocol’s commercial problem is therefore not “is the median workload willing to pay,” which it plainly is not, but “is the high-$\Delta$ tail large enough, and reachable enough, to carry a fee.” That is a sizing question the thesis has not answered, and Red Line 14 measures it capacity-weighted across the served mix rather than assuming it.

And at the median the answer is unflattering.

Written as $\varrho\,\mathcal{L}$, the critical term is $\varrho$, and here the thesis has been flattering itself. The bypass channel is not “unverified.” It is reputationally and contractually verified. A hyperscaler’s alternative to a cryptographic proof is its balance sheet, its SOC 2 and ISO audit reports, its published incident history, and an indemnity clause enforceable in a court that will actually hear the case.

A creditworthy, suable indemnitor is an excellent substitute for a cryptographic proof.

For an ordinary commercial workload in a benign state, the buyer’s exposure to undetected misexecution is not $\mathcal{L}$; it is $\mathcal{L}$ net of a contractual recovery from a solvent counterparty under a legal system that functions. That makes $\varrho$ small, and therefore $\Delta$ thin over the bulk of the distribution. Thin $\Delta$ against a structurally positive $(c_p - c_b)$ is the strongest form of the objection this section opened with, and it is not answered by incidence analysis. It has to be stated plainly, because a critic will otherwise state it first and more crudely: for the median commercial workload in a working legal order, cryptographic verification is a premium product competing against a cheaper substitute that is good enough. What that sentence does not say, and what a fair reading of it must preserve, is that the median is the marginal buyer. It is not, and the fee is set against the tail.

The tail is one limit on the fee; openness is another. Every property the thesis is proud of, from published workload specifications to permissionless participation, also lowers the cost of copying the protocol (Complete Edition, Chapter 13: Why Hold the Base Asset?^†).

The Fork Is an Equilibrium, Not a Competitor

The cheapest bypass is not a hyperscaler. It is this protocol, copied, with the fee removed.

Complete Edition, Chapter 13: Why Hold the Base Asset?^† concedes that, and concedes what actually stops the copy: liquidity depth, anonymity-set size, the installed collateral base, the registry everyone cites, the receipt history worth citing. Network effects, in other words, and not one of the five lemma conditions. The concession is correct. It was filed under the wrong heading.

Modelled as a bypass channel, the fork looks like the end of the fee. Modelled correctly, it changes what the fee is bounded by.

Two different bounds, and the binding one is not always the same.

The sustainable-fee bound of Chapter 13: Why Hold the Base Asset? is $f \leq \Delta - (c_p - c_b)$, where $c_b$ is the all-in cost of the best bypass channel. Against a hyperscaler that expression carries the structurally positive cost handicap the section emphasizes: verified compute costs more to produce than unverified compute, and the protocol must recover the difference out of $\Delta$ before collecting anything.

Against a fork the arithmetic is different in a way worth stating explicitly. The fork runs the same technology, so $c_b = c_p$ and the handicap vanishes. What remains is

$$f \;\leq\; \Delta_{\mathrm{coord}},$$

the differential value of transacting on this instance rather than on an identical copy, which is not a cryptographic property at all. It is liquidity depth, anonymity-set size, the installed collateral base and the operator relationships around it, whether the workload registry is the one everybody cites, and whether a receipt from this chain is worth citing because of what else is in its history.

The two bounds are separate quantities and the protocol is subject to both:

$$f \;\leq\; \min\bigl(\underbrace{\Delta_{\mathrm{tech}} - (c_p - c_b)}_{\text{versus the unverified alternative}},\; \underbrace{\Delta_{\mathrm{coord}}}_{\text{versus the identical copy}}\bigr).$$

This is a real tightening of Chapter 13: Why Hold the Base Asset?, which stated only the first term, and it is the correct place to put the concession Complete Edition, Chapter 13: Why Hold the Base Asset?^† made informally.

Contestable markets and coordination games give opposite predictions, and only one of them is the right model.

The objection that opened Chapter 13: Why Hold the Base Asset? (a contestable market competes economic profit to zero) is the right model for the hyperscaler comparison and the wrong one for the fork. Contestability assumes entrants and incumbents are substitutes at the margin for each buyer independently. They are not, when the good’s value to each buyer depends on how many other buyers chose the same instance. That is a coordination game, it has multiple equilibria, and equilibria are not competed away by a better price but displaced by a focal point moving, which is a different and much slower mechanism.

A contestable market has one equilibrium and prices race toward it. A coordination game has several, and the incumbent’s advantage is not that it is better but that everyone is already there.

Openness cuts both ways, and the thesis has counted only one direction.

Complete Edition, Chapter 13: Why Hold the Base Asset?^† observed that every property the thesis is proud of (published workload specifications, open verifier implementations, permissionless participation) lowers the cost of producing the copy. True, and incomplete. Those same properties operate on the two bounds in opposite directions:

• Openness drives $\Delta_{\mathrm{tech}}$ against a fork to zero. There is no technical differentiator, by construction and by design commitment.

• Openness raises $\Delta_{\mathrm{coord}}$. A canonical workload registry that is public, stable, and cited by everyone is more focal than a proprietary one, not less. Schelling points are made of common knowledge, and publishing is how common knowledge is manufactured. The same is true of an open receipt history: its citability is a function of how many independent parties already read it.

So the honest statement of the trade-off, in the form Complete Edition, Chapter 13: Why Hold the Base Asset?^† uses for collateral:

Openness must be total for the technology and is the mechanism by which the coordination assets accumulate. You cannot have a defensible technical moat and an open specification. You can have an open specification and a focal point, and the second is the only moat this design is entitled to.

And the moat is buyable, which is the part that keeps it honest.

$\Delta_{\mathrm{coord}}$ is not a law of nature. A challenger can rent it: subsidize liquidity, pay operators to migrate, underwrite the anonymity set, and buy the switching costs down. This is the fork’s actual attack and it is a capital question rather than a technical one. The defensible claim is therefore narrow: the fee is bounded by what it costs a challenger to move the focal point, and that quantity is empirical, contestable by capital, and not guaranteed to exceed the fee.

The one case, in the right reference class and with its limits stated.

Complete Edition, Chapter 13: Why Hold the Base Asset?^† cites the 2020 SushiSwap attack on Uniswap: an identical copy of open contracts, paying liquidity providers to migrate, attacking a liquidity-and-coordination moat directly, and winning a share rather than the market.

Two corrections to how that case was used. First, the relevant reference class is not “forks” in general but coordination games in which an installed focal point is challenged by an identical copy with better short-term incentives. In that class the observation is on point rather than cherry-picked, because it is a direct test of $\Delta_{\mathrm{coord}}$ against paid switching. Second, and stated plainly because the thesis holds itself to base rates elsewhere: it is one observation, it was selected after the fact, the incumbent survived, and a single surviving incumbent is not a base rate. It establishes that coordination assets can hold against a funded identical copy. It does not establish how often they do, or that they would here.

What this changes in the instruments.

Two additions, both small and both closing gaps that the bypass framing left open:

• Red Line 14’s reference panel gains a fork row. The pre-committed panel of Chapter 17: Risk Analysis & Failure Modes lists hyperscalers and GPU marketplaces, venues that measure $\Delta_{\mathrm{tech}}$ only. The lowest-fee protocol-grade fork offering the same canonical workload must be priced into $P^{\mathrm{byp}}$ on the same terms, because it is the channel with no cost handicap and therefore the one that binds first. A wedge measured only against unverified alternatives is measured against the weaker competitor.

• Acceptance breadth becomes a published series. The number of independent venues, corridors, and counterparties that quote, settle, and cite this instance rather than any other is the observable form of $\Delta_{\mathrm{coord}}$, and it belongs beside the denomination series of Chapter 15: The Three Monetary Mechanisms on the Value Capture Board (Complete Edition, Chapter 26: Extended Telemetry^†). Both measure coordination; one measures whether people reckon in the unit and the other whether they converge on this instance of it.

Where this leaves link 7.

Non-bypassability was reduced from Strong to Medium in Chapter 9: The Triad and the Monetary Candidate because the economics requires a magnitude that is nowhere estimated. Nothing here raises it. The magnitude is now specified as two magnitudes rather than one, the second of which the document had been treating as an admission rather than a term, and both remain unestimated. What has changed is that there is now an instrument pointed at each.

Openness has one further effect that neither bound prices: a protocol anyone can implement is a credible entrant against the incumbents it contests, and the credible possibility of entry can discipline conduct without any entry being observed [Baumol 1982], which can serve agency while leaving the asset nothing to capture (Chapter 9: The Triad and the Monetary Candidate).

What the Five Conditions Establish, and What They Do Not

Collecting the above fixes what the lemma’s conditions can deliver.

What they establish.

That a contestable market can be taxed and the proceeds routed to a token. Specifically: (i) a competitively priced claim on a cash-flow stream $f \cdot Q$, bounded above by the per-unit wedge $\Delta$ applied to volume ($\bigl(\Delta - (c_p - c_b)\bigr) \cdot Q$ at the sustainable-fee bound of Chapter 13: Why Hold the Base Asset?) and deflating with unit price; and (ii) a balance-sheet lockup of collateral required in the native unit. Both are real. Both are also, precisely, quantities that a discounted-cash-flow valuation captures: a fee stream and a lockup.

What they do not establish.

Monetary premium. A monetary premium is by construction value in excess of discounted cash flows; it is what gold has and a pipeline does not. A holder who buys a share of the fee stream from indispensable workloads owns something closer to equity in a utility than to a collectible, and a utility equity earns its cost of capital; it does not carry monetary premium. The routing rules of Chapter 13: Why Hold the Base Asset? are design commitments; nothing in this thesis claims that they set a price. The incidence question asked of the fee is asked of each monetary mechanism’s benefit as well (Chapter 15: The Three Monetary Mechanisms).

Three corollaries follow:

• A burn is a buyback. Economically, retiring supply with revenue is a pro-rata return of value to an existing claim. Buybacks change per-share value; they do not confer moneyness. No quantity of buyback has ever made an equity a store of value, and no burn schedule will do it either.

• Collateral is a lockup, not a premium. A required lockup denominated in the asset immobilizes float while operators remain solvent and choose to stay, and it places no bid. It is a balance-sheet fact, and Complete Edition, Chapter 13: Why Hold the Base Asset?^† bounds how large a fact it can be.

• A subsidy is a negative fee. Pass [Pass 2026] models proof-of-useful-work economies as three regimes: one in which the work is worthless and the token is pure seigniorage, one in which useful work is sold at cost, and one in which token issuance subsidizes the work below its compute cost. The fee-incidence bound places the second regime at $f \leq \Delta - (c_p - c_b)$ and the third at $f < 0$. A network in the third regime is buying volume with issuance; its receipts are real and its fee coverage is negative, and Red Line 5 and Red Line 14(B) are written to see it. Bitcoin’s work function, which has no external buyer, sits in the first regime by construction, and that is the content of the concession in Chapter 7: Why Not Something Already Here?.

The Test That This Reasoning Does Not Prove Too Much

If durable fee extraction from a commoditized flow made an asset money, the world would be full of monetary assets. CME clears standardized futures; DTCC settles equities; Visa takes a spread on card turnover; MSCI licenses indices that everyone benchmarks to; pipelines and port authorities meter physical throughput. All of them sit on flows more commoditized than canonical MatMul, all of them extract fees durably, several are structurally harder to bypass than any protocol will be, and not one of their equities is money. Visa’s take rate does not make Visa shares a store of value; it makes them a good business.

The comparison is the discipline. If the argument for the asset’s monetary premium is “recurring fees on indispensable throughput,” the argument is an argument for a high-quality utility equity, and it should be labelled as one. Something else has to be doing the monetary work.

From Utility Demand to Monetary Premium

Many indispensable services (electricity, bandwidth, compute instances, legal services, cloud storage) have recurring demand without being stores of value. The monetary argument requires explaining why this utility becomes monetized savings demand, not just operating expense.

The bridge has eight conditions. The middle five are the Value Capture Lemma restated informally; they are named here in one line each and stated formally at Chapter 13: Why Hold the Base Asset?, not argued twice:

1. Utility demand is not enough. Recurring demand for a service does not automatically create a scarce, holdable asset. Bandwidth is indispensable, but claims on bandwidth do not become money.

2. The asset must be required for fees or settlement: required fee medium, Condition 1.

3. Fees must produce burns, retirement, or staking yield: supply reduction, Condition 2.

4. Operators must post the asset as collateral: Condition 3, whose worth Complete Edition, Chapter 13: Why Hold the Base Asset?^† bounds as a slope effect on price impact, not a valuation floor and not a source of premium.

5. Issuance must follow a schedule fixed ex ante: issuance discipline, Condition 4.

6. Users must not be able to bypass the asset at equal service quality: non-bypassability, Condition 5.

7. Telemetry must prove the loop is working. Fee coverage, burn rates, collateral lockups, and workload demand must be publicly verifiable, not asserted.

8. If any value-capture condition fails, the asset does not even capture its own service value. It becomes infrastructure exposure whose economics accrue to operators or customers rather than to holders. Meeting those five conditions makes service value accrue to the asset; it does not make the asset money, establish stress-deliverable service, or create a loss-bearing holder constituency.

This section is the “utility-token trap” defense. The thesis does not claim that useful things automatically become money; it claims that under specific, testable value-capture conditions, a useful asset accrues the value of its usefulness to its holders: a cash-flow claim and a collateral lockup. Whether anything further, a monetary premium in excess of those, is available is argued separately and on different grounds in Chapter 15: The Three Monetary Mechanisms. The Value Capture Lemma formalizes the accrual conditions; the telemetry regime (Complete Edition, Chapter 25: Layer 6: Governance & Telemetry^†) makes them falsifiable.