Modern Movement of Money, Part 1: With new hands on rails and a shorter stack, endpoints are the prize and trust becomes product

Evan Thorpe, Principal (LinkedIn)

As global money movement shifts from an access business to a controls business —and small businesses are finally brought into the fold— we’re backing the founders building the stack of trusted money movement: from AiPrise (verified endpoints) to Range (controlled and attested flow) to, now, Diameter (licensed settlement). This is the first of two pieces written on the subject. Part 2 takes up what happens to that stack when the endpoints start acting through software agents.

Every payment network in history has been, at bottom, a network of trusted endpoints —which is to say, a club. Correspondent banking, the card networks, ACH: each drew a line between the parties inside, who could transact with the network’s full confidence behind them, and everyone else. Individuals, merchants, suppliers and small businesses the world over have defined their financial lives by which side of that line they stood on.

For the last decade, the defining question in payments was how to get more of them inside: how to widen access to the club. We think the question has changed. The value in the next decade will accrue not to whoever widens access to the rails, but to whoever can establish, govern and prove trust between the endpoints transacting on them. The segment with the most to gain is small businesses and the financial institutions that serve them.

The shift: from access to controls

The rails didn’t change; the hands on them did

Start with what actually happened to cross-border money movement, because it is widely misread. Over the past decade global banks pulled back from correspondent banking — sometimes corridor by corridor, sometimes wholesale. The instinct is to say the rails changed. They didn’t, not fundamentally. Dollars still clear through dollar infrastructure.

What changed is who is taking charge of the rails. As incumbent banks retreated, stablecoin issuers, payment platforms and a new class of licensed players (PSPs, neobanks, vertical banks, etc) stepped into the space they vacated. The pipes are much the same; the roles that operators play are shuffling and the lines previously drawn don’t apply quite the same as they once did. Our view is shaped by several questions and chief among them: how much of what comes next is collaboration with the incumbents, and how much is replacement?

That distinction reframes the opportunity. This was never simply a technology upgrade, old rail out and new rail in. It was a transfer of stewardship and a re-alignment of partnerships. Which players have their hands on the controls? What do controls consist of in a modern, AI-native world? And —a question we hold for Part 2— how many of the hands on the hand-off chain —which in sum, constitute controls— and on the payments themselves, will soon be performed by programs rather than people?

Whoever takes charge of a series of handoffs inherits the job the retrenching banks were actually doing on it. That job was never moving the money; that part is easy. It was deciding who is allowed on, screening what flows across, and standing behind the result. The banks pulled back because that job had become expensive relative to the margin on movement. The new operators inherit the same equation, on rails where movement is now nearly free. The economic weight of the business shifts onto the one thing that isn’t free: trust.

“The rails didn’t change. The hands on them did —and whoever takes charge inherits the job the banks were really doing: not moving the money, but standing behind it.”

Rails vs wallets: triangulating a new trust graph

Here is the structural shift underneath all of it. In the old world, the rail (how value moves) and the wallet (where value sits, and whose it is) were separate businesses run by separate institutions. A bank held the account (singular, though today that is a whole other story), a network moved the money, and a chain of intermediaries sat in between reconciling the two. Stablecoins collapsed that separation in some areas but also extended the transaction chain in others, resulting in a greater number of permutations of hand-offs taking place and, again, bridging many of the same rails as before but with more unique combinations of counterparties. It posited simplicity but did not guarantee it.

What renewed infrastructure makes possible is a trust graph: a network in which verified endpoints and their wallets are the nodes, and the record of how they transact with one another forms the edges. In effect, it is a running map of who can be trusted to move money, with whom, and on what terms. It is possible because, on this infrastructure, wallet + account + rail is the settlement itself: to hold the asset is to be on the rail, and to move it is to settle with finality in the same motion. Every movement of value therefore adds to the graph. Triangulate what is known about each node against the evidence accumulating on its edges, and you arrive at an enriched representation of identity, one that engenders the confidence to meet the standard of trust requisite to transact. The more collapsed the stack, the less the data is drawn from the intermediaries and the more it shifts to the endpoints.

You can see the industry pricing this in. The largest platforms are no longer satisfied with assembling faster rails; they are assembling two halves of a new coin together —issuance and wallet infrastructure on one side, verified endpoints at scale on the other. Traditional networks combining with the means for on-chain transactions marches on —Mastercard and BVNK, Polygon and Coinme, and the rumored Visa – Fireblocks deal— and Stripe’s acquisitions of Bridge and Privy, and its attempted bid for PayPal earlier this year, read as one strategy: rails are plentiful, but hundreds of millions of verified, KYC’d and KYB’d accounts with years of transaction history behind them —and the wallets those accounts transact from— are not; endpoints already acquired and/or grandfathered in are scarcer and they are worth acquiring en masse since they come pre-assembled and primed for deeper triangulation. 

That is the tell. The most sophisticated access companies are repositioning themselves (and perhaps leading to future repricing) toward endpoints and wallets —toward the trust graph— because they can see what we can see. Note what has passed for a trust graph in years prior: simply, years of evidence about how known people and businesses behave. Hold that thought; Part 2 asks what it is worth when the party transacting is an agent acting on someone’s behalf. More data to triangulate greater trust. 

The stack is compressing -bringing within reach the prize: endpoints (though this is not the whole prize)

Put those two observations together —rails and wallets merging, endpoints becoming the prize — and you get the force that will define the next several years of competition: the parties are proliferating, and the trust graph is coalescing, and the stack is compressing.

When ownership of funds and the means of moving them collapse into the same infrastructure (endpoint, wallet, rail), the long chain of intermediaries between payer and payee has less and less to do across fewer steps of a given transaction. Here, the lines become blurred and we have banking infrastructure utilized as payments infrastructure, particularly where USD settlement is concerned. Correspondent bank, sponsor bank, network, processor, gateway: each link existed, in part, because holding value and moving value were separate actions performed by separate parties joined by practice and trust. Hence formed the club of the access-based network model. Compress those actions into one motion and the transaction resolves through fewer hands but perhaps new chains emerge across more endpoints and the burden of trust persists all the fiercer. The old clubs and the access-based paradigm is breaking down; enter the controls-based network paradigm across more endpoints, made possible at speed with flatter infrastructure thanks to a high-definition trust graph to underpin fewer hand-offs. 

This is not gentle news for payment players; it is an escalation. We are at a point today of greater expansion of players prior and greater distribution of transacting parties to future consolidation as trust coalesces around a sleeker infrastructure with fewer intermediaries supporting more endpoints. Fewer intermediary slots means the fight to occupy the remaining ones intensifies. Issuers, processors, platforms, banks and the new licensed non-banks are all racing for a defensible position in a shorter stack, because the positions that survive will capture economics once spread across five parties. You don’t win a seat in a compressed stack by moving money more cheaply; that race is over, and its prize for transacting itself will be zero. The players who win will do so by owning something compression can’t eliminate: the verification of the endpoints, the governance of the flow, or the license under which it settles; this is the turning point towards the controls paradigm. A paradigm where the intermediaries that endure will provide trust, not simply transportation.

So has all the value moved to the endpoints? Not quite. It has moved to whoever can vouch for them: who verifies the endpoints, who governs the flows between them, and who stands behind the settlement underneath.

“Compress the stack and you don’t soften the competition —you raise the ante. The winning intermediaries will be the ones providing trust, not transport.”

The circle widens, to the benefit of small businesses

The two ends of the market (enterprise and SMB) are living this shift in opposite ways. Small businesses are most exposed because they carry enterprise-grade risk borne by consumer-grade tooling. The shorter stack and more open controls-based paradigm is a massive upgrade for the payment platforms upon which small businesses run. The re-tooling of the stack and the paradigm shift described widens the circle to incorporate more of these businesses into an expanded network of trusted transacting parties. 

Enterprises are already buying the controls layer directly: treasury platforms that unify every account into a real-time view, pre-execution screening on outbound flows, AI deployed in compliance functions. For the enterprise, controls are a product category with a budget line and a team to run them.

Small businesses carry the same risks with none of the apparatus. An importer in Lagos, a SaaS company in São Paulo billing US clients, a supplier in Ho Chi Minh City: each faces enterprise-grade cross-border exposure (fees, FX, settlement lag, counterparty fraud) with consumer-grade tooling and minimal compliance function. On rails where settlement is final, speed without controls doesn’t remove their risk; it transfers risk onto the party least equipped to screen for it. The financial institutions serving these businesses are squeezed from both sides, with correspondent relationships contracting beneath them and compliance expectations rising above them.

“Small businesses carry enterprise-grade risk with consumer-grade tooling. That gap is the largest whitespace in financial services.”

Consolidation at the top of the market does nothing for any of this. A larger merchant acquirer, whatever its scale, does not verify a supplier in a de-risked corridor or give a community bank the infrastructure to bank exporters. Left alone, the compression of the stack concentrates capability upmarket and leaves the long tail of global commerce running twenty-first-century volumes on trust infrastructure built for a smaller, more exclusive club of parties protected by a verifiable paper trail; consistent with de-globalization but inconsistent with a more interconnected world, which is the game on the field. 

The gap that has emerged and what closes it

There is an important open question in all of this, and being direct about it strengthens the thesis. Stablecoins bring the rhetorical potential for near-free, instant, global movement. What they do not yet reliably bring is everything that made the old rails safe to use: the rulebooks, the dispute and reversal mechanisms, the interchange and fee logic, the compliance reporting, the netting and reconciliation that banks and networks spent decades encoding into the correspondent system. Old habits are hard to break because substitution for incumbency is a difficult gap to overcome. The vacuum left by old rulebooks, abandoned with closed access-based systems must be filled by new ones: though what will fill that gap will take new form: rules embedded in controls. Rulebooks are the internally-embedded and network-based counterparts of external regulatory requirements and they vary from company to company and network to network based on variable risk assessment and appetite. 

Much of the rulebook will be rebuilt as payment controls natively on stablecoin rails and treasury / trade controls; some may not, or not for years. That gap presents whitespace. Instant settlement is also irreversible settlement. A wire to a fraudulent counterparty might be recalled; a stablecoin transfer will not be. So the screening that used to happen lazily, overnight, after the money cleared, now has to happen before it moves, or it may as well not happen at all. The rules, calculations and controls that lived inside a closed network must be reconstituted as software controls, and that work that can be broken down into three parts:

• Timing. Controls have to run pre-execution, at the edge, in the milliseconds before finality.
• Signal. Disputes, refunds and chargebacks are clues that inform trust, as well as costs.
• Tuning. A shorter chain of handoffs collapses cost and reduces broken payments; but broken payments also provide valuable signals for developing a high-definition view of the parties transacting. The speed vs trust compromise? Sure, that will still exist; but the deeper open question is: how do we take out the cost without losing signal? The network will need to triangulate sufficiently deep and broad data to surface signals that can be tuned to match the controls of the network, and those of the counterparties. 

This turns what was a tradeoff (speed vs trust) into an equilibrium (enhanced trust and enhanced speed reinforcing one another). Whoever can enhance the reach, speed, and affordability of transacting while also enhancing trust —not as a tradeoff but as mutually-reinforcing value propositions— will establish themselves as the next iteration of the global transaction system, not just the movement of messages regarding money at a lower cost but the harvesting of insights based on that rapid movement and multiplicity of endpoints. 

Trust becomes the product

The reframe: a network of trusted transacting endpoints

The way forward starts with redefining what a payment is. For fifty years, the question was “did it arrive?” The questions that now matter are: was each endpoint verified, on what basis was the flow screened before it moved, and can it be attested after?

Attestation deserves a word, because it is the least glamorous and most important of the three. An account is, in the end, the sum of what can be proven about the money that moved through it. Attestation is not paperwork about the account; it is the account’s essence. A business that can prove the provenance and compliance of its flows is a trusted endpoint. One that cannot is back outside the club, whatever rails it can reach.

“An account is, in the end, the sum of what can be proven about the money that moved through it.”

For small businesses, though, upgraded tooling is a relief only if it doesn’t become a second job. Controls that must be micromanaged are a curse wearing the costume of a cure. The answer is controls delivered embedded and invisible, inside the products SMBs and their banks already use, not as a compliance console demanding attention. Verification happens at onboarding, screening happens before execution, attestation accumulates automatically, and the small business simply experiences a payment that works.

Three companies, one controls chain

At SixThirty, this is the through-line of a series of recent investments: three companies, one controls chain for money moving between trusted endpoints.

AiPrise verifies the endpoints. KYB and KYC orchestration across 150+ countries and 80+ registries and data sources, through one API. It turns onboarding from a corridor-by-corridor grind into minutes, and establishes the trusted in “trusted endpoint.”

Range governs and attests the flows. One real-time ledger across bank accounts, custodians, wallets and exchanges —one ledger: yours. Every transaction is screened for sanctions, fraud and counterparty risk before it executes, across fiat and stablecoin rails, and every record is reconciled and audit-ready after. In a world of irreversible settlement, screening before the money moves is the whole game and the crossroads for risk ops. The more that one can triangulate, a high-definition trust graph clarifies, and better the identity behind flows can be understood. 

Diameter settles under license. Embedded, compliance-automated payments that let banks and financial institutions transact in USD in the corridors correspondent banking left behind, with settlement that runs on continuous verification, attestation and control rather than compliance bolted on afterward. Diameter’s own evolution, from remittance rails to a compliance-spined settlement platform, is the access-to-controls migration playing out inside a single company.

Three companies, three links: verified endpoints, governed flows, licensed settlement. Each earns its place in the compressed stack by owning a piece of trust that commoditized movement cannot erase, and each strengthens the others.

The next great payments company won’t be a checkout button

The repricing underway is permanent: movement is a commodity; endpoints, wallets and the trust between them are the assets.

“If software can prove trust, software can also ask for it.”

The next generation-defining payments company will not be a checkout button. It will be the network that lets any two verified businesses, anywhere, transact as safely as two Fortune 500 treasuries did in years past. The Fortune 500 bought that safety with an army of treasurers and controllers. The trust layer we’re investing in works continuously and invisibly underneath, its orchestration shifting to AI, so that the club, at last, is open to everyone who can be trusted —and being trusted is something that technology can now help prove.

Where we’ll pick up for Part 2: If tech properties can prove trust, then those properties can also ask for it. When the party at the endpoint is an agent acting for a business, or for another agent, what exactly is being verified, who is accountable, and where does the human analyst sit in the loop? The bar for trust doesn’t just move to a new set of operators. It has to be recalibrated for a new kind of counterparty.