The Source Premium, Part 2: The Concept That Explains Every Major Capital Move of This Decade
Part one showed the receipts. This one defines the framework — so you can price it, underwrite with it, and recognize it before the market names it.
The definition
Finance has always priced the distance between a claim and the thing it claims. The gap between a liquid asset and an illiquid twin is the liquidity premium. The extra value of holding a physical commodity over its futures contract — the ability to actually use the barrel, the bushel, the bar — is the convenience yield. The price paid to own enough of a company to direct it, rather than merely ride it, is the control premium. Each of these named its era because each priced the constraint that defined it.
The constraint that defines this decade is physical supply. So we propose the term for what the smartest capital is now demonstrably paying:
The source premium, n. — the difference in value between a direct, secured, verifiable claim on a productive physical asset and a market exposure to the same asset. It is what capital pays to stop trusting the market’s paper and start owning the queue.
Part one of this series documented it empirically: roughly $22 billion in deposits and commitments paid by customers to reserve unbuilt memory under non-cancellable contracts; a conglomerate financing a silver mine’s restart against first-ranking security over its assets and offtake rights over 100% of its concentrate output; automakers holding equity in the ground. This article does the second half of the work: explaining why the premium exists, breaks down what it is made of, and shows the conditions under which it can be opened beyond giants.
Why it appears now
Scarcity converts markets into queues — and a queue position is property.
A spot market performs one service brilliantly: it guarantees that at some price, you can buy. That guarantee quietly assumes supply elasticity — that higher prices call forth more supply fast enough to matter. The defining feature of the 2025–2035 window is that, for the inputs technology needs most, this assumption has failed on the relevant time horizon. Data-center electricity demand is headed from roughly 415 TWh today toward around 945 TWh by 2030 and 700 — 1,700 TWh by 2035, while grid connections queue for years. Advanced memory is sold out years forward. Critical-minerals projects take over a decade from discovery to production.
When supply cannot respond within the horizon that matters, the spot price stops being the scarce thing. The scarce thing becomes the position in line — the connection slot, the offtake right, the reserved production, the permitted site. And positions in line, unlike prices, can be owned, secured, and enforced. That is the moment a market grows a queue, and the queue grows a premium.
Once you see it, you stop reading this decade’s headlines as separate stories. Take-or-pay memory contracts, prepay mine financings, equity stakes in deposits, multi-year GPU leases signed by former Bitcoin miners, grid-connection permits trading like assets — they are one story: capital migrating from claims on markets to claims on sources.
What the premium is made of
If it can be decomposed, it can be priced. If it can be priced, it can be underwritten.
The source premium is not one number; it is three distinct components, each monetizable, each currently captured by whoever sits between capital and the asset.
Component — Economics
What it is: The cumulative fee and spread stack between the asset’s cash flow and the end investor.
Where it is: Management and performance fees, brokerage, structuring margins, the bid-ask of every intermediated hand.
Component — Information
What it is: The difference between marking to the meter and marking to the quarter.
Where it is: Operators read telemetry live; end investors read aggregated NAV letters, quarterly, on trust.
Component — Access
What it is: The right to participate in the direct trade at all.
Where it is: Take-or-pay, prepay offtakes and secured production claims are simply not offered downstream of giant scale.
An illustration, deliberately simplified. Take a grid-scale battery producing a given annual cash yield on invested capital. Route one — the conventional wrapper: the same asset held through an infrastructure fund typically sheds one to two percentage points annually to the fee stack, reports quarterly, and was allocated to institutions first. Route two — a direct tokenized claim: platform fees are a fraction of the fund stack and disclosed as a single line; the asset’s meter, exchange settlements and bank receipts reconcile under a published hash each quarter; and the claim was available to any verified investor from the first euro. The cash flows of the battery are identical in both routes. The difference between what reaches the investor — in yield, in verifiability, in the right to be there at all — is the source premium, made visible.
The premium was never paid to the intermediaries for creating value. It was paid to them for standing where verification used to be impossible.
Who captures it today
So far, the premium has an entry ticket of one billion dollars.
The current holders of the source premium are exactly who you would expect: balance sheets large enough to finance a mine restart for the offtake, to deposit billions against future chips, to buy the refinery rather than the refined product. The first attempts to open the trade downstream are appearing at the edges — regulated products aggregating data-center capacity, pilots tokenizing enterprise GPUs together with their revenue streams — and they are instructive precisely because they cluster in compute, the vertical where scarcity bit first. But energy, the input underneath all of it, remains almost entirely closed: institutions buy operating battery portfolios at hundreds of thousands of pounds per megawatt while everyone below that scale owns, at best, a utility ETF.
This is not a permanent state. It is a technology gap wearing the costume of a natural order.
The conditions for opening it
The premium survives democratization only if verification survives without trust.
Why did the direct trade stay gated? Not regulation alone, and not minimum ticket sizes alone. The binding constraint was that a direct claim is only worth its security if the holder can verify the asset behind it — and verification at retail scale used to be economically impossible. You cannot fly ten thousand small investors to Durango to count concentrate. So the market solved it with layers of delegated trust, and the layers took the premium as payment.
That constraint has expired, and the article-one checklist describes its replacement in full. Compressed to six principles: eligibility and transfer restrictions enforced by the token contract itself, not by a transfer agent’s discretion; supply mapped 1:1 to project capital so the denomination audits in your head; the physical record produced by a regulator-certified meter and committed on-chain through an independently co-signed path; quarterly reconciliation of meter, exchange and bank under a published hash; record-date snapshots and Merkle proofs each holder verifies alone; and return of invested capital ranking first at exit, in code, in that order. When those six hold, the economics, information and access components of the premium stop accruing to the middle layer — because the middle layer no longer performs a function the contract and the meter do not.
One discipline matters more than any feature: open the premium on the most verifiable asset first. A two-hour battery — metered by legal requirement, settled on a public exchange, with no geology narrative and no warehouse to trust — is where direct claims are proven. That is why our first issuance is a battery in Romania, the fourth in Europe on ENTSO-E’s storage revenue benchmark, and why the exotic assets wait their turn behind the boring one.
What this means for an allocator, in one paragraph
You do not need to believe in tokens. You need to believe in queues.
Strip every buzzword and the investment logic stands on two observations you can verify in public filings: first, the most sophisticated buyers of physical supply on Earth are paying, in cash and in advance, for direct claims on sources — which fixes the sign and the rough size of the premium; second, the verification technology that kept that trade gated has been replaced by contract-level enforcement and holder-executable proof — which removes the structural reason the premium must stay wholesale. An allocator does not have to predict the future to act on that. It is enough to notice that the gap between those two facts is a market — and that whoever standardizes it first will set its terms.
The giants priced the source premium. Our work is to open it — one metered, verifiable asset at a time.
The gap TerranOS fills — explained twice
If a framework is true, you can explain it to a ten-year-old and to a CFO — and it is the same explanation wearing different clothes.
For a ten-year-old. Imagine a town where bread is getting scarce. The clever, rich families stop going to the market — they walk to the bakery and pay the baker in advance, for a whole year, straight from the oven. If the baker breaks the promise, they keep his oven. Everyone else still stands at the market, buying bread coupons from a man who bought them from another man who bought them from the baker — and every man in the chain keeps a bite of every loaf, and none of them will let you see the oven. TerranOS is the door that lets anyone walk to the bakery: you pay the baker directly, your name goes in the book at the oven, a counter on the oven itself shows you every loaf it bakes, and if the bakery is ever sold, the people in the book get their money back first. No men in the middle. No bites taken. And you can check the counter yourself, any day, without asking anyone’s permission.
For a CFO. The market has a structural void: there is no rail through which sub-institutional capital can hold what giants now hold — direct, secured, production-linked claims on physical sources. TerranOS fills that void component by component, and the framework in this article is literally its product specification. The access component: a permissioned ERC-3643 token is a queue position made fractional and transferable — eligibility, lock-ups and limits enforced by the contract, not by an intermediary’s discretion, from a EUR 1 denomination that maps 1:1 to project capital. The information component: the position marks to the meter, not to the quarter — a regulator-certified meter produces the record, an independently co-signed path commits it on-chain, a quarterly audit reconciles production, exchange settlements and bank receipts under a published hash, and each holder verifies their own Merkle proof and claims their own distribution. The economics component: the intermediated fee stack collapses into one disclosed platform line, and the waterfall — return of invested capital first, then the published profit split — executes in code, in that order. In other words: the source premium stops being a rent collected by whoever stands between capital and the asset, and becomes a property of the claim itself — underwritable, auditable, and priced. That is the entire company, in one sentence: TerranOS converts the source premium from a privilege of scale into a feature of the instrument — proven first on the most verifiable asset class in existence, a metered, exchange-settled battery, and extended from there to minerals, carbon and the compute those megawatts feed.
Same story, both times: go to the oven, get in the book, watch the counter, get paid first. The rest is engineering.
Figures referenced in article: IEA, Energy and AI (data-centre consumption ~415 TWh 2024, ~945–1,000 TWh 2030, 1,200 TWh base / up to ~1,700 TWh 2035); IEA storage outlook (1,500 GW by 2030, ~1,200 GW batteries); Micron FQ3 2026 disclosures; Silver Storm / Samsung C&T announcements 2025–2026; UK BESS portfolio transactions (PwC, JLL). The fee-stack comparison is illustrative, not a representation of any specific fund. This article is qualitative commentary for professional readers; it is not an offer of securities, and any offering referenced is available only to verified, eligible investors outside restricted jurisdictions through official TerranOS channels.
About TerranOS
TerranOS is the operating system for real-world energy assets: grid-scale battery storage, documented mineral reserves, AI computing power and carbon, brought on-chain for a global community of verified members. Built on real infrastructure in Europe. Learn more at www.terranos.com.
Follow our evolution: X (Twitter) | LinkedIn | Facebook | Medium
Article written by Razvan Laichici and edited by Rouă Denis for TerranOS.
