The stablecoin custody execution settlement stack is unbundling along the same fault lines that separated TradFi market infrastructure half a century ago. Stablecoin custody execution settlement refers to four distinct functions: holding client assets (custody), matching buyers and sellers (execution), netting and risk management between trade and finality (clearing), and the final irrevocable transfer of value (settlement). As of Q2 2026 the stablecoin float sits at roughly $315B (DeFiLlama, June 2026), and institutional buyers are now sourcing each function from a different provider for the same reasons NYSE, the DTCC, and BNY Mellon were never one company.
​

This piece walks through how TradFi arrived at that separation, where the onchain analogs now sit, and how regulators in Brussels and Washington are codifying the split through MiCA Article 75 and the GENIUS Act framework.
​

Traditional capital markets run on a four-function model. Execution venues match orders and discover price. Clearing houses novate trades, net exposures, and manage counterparty risk. Custodians hold securities and cash on behalf of beneficial owners. Settlement systems move final title and funds. Each function sits at a different firm, under different regulation, for different reasons.
​

NYSE matches trades. The National Securities Clearing Corporation, an arm of the DTCC, novates and nets them. State Street, BNY Mellon, and Northern Trust custody the underlying securities. The DTC, also under DTCC, books final settlement through book-entry transfer. Fedwire moves the cash leg for institutional dollar payments, with the Federal Reserve operating a real-time gross settlement system that processes trillions per business day (see Federal Reserve, Fedwire Funds Service).
​

The Bank for International Settlements has codified this layered model through its Principles for Financial Market Infrastructures, the global rulebook for how clearing, settlement, and custody interlock without re-introducing the single-point-of-failure risk that vertical integration creates (BIS CPMI-IOSCO PFMI).
​

Vertical integration broke in US equities because conflicts of interest, operational fragility, and concentration risk became untenable. The 1968 paperwork crisis forced the creation of a central depository. The 1975 Securities Acts Amendments mandated a national market system and required separation between execution venues and the clearing and settlement layer. The result is the architecture securities markets still run on today.
​

Before 1968, brokers held physical certificates and re-delivered them after every trade. Volume growth broke the model. Trading floors were closed on Wednesdays so back offices could catch up. Congress responded by chartering what became the DTCC, immobilizing certificates in a single depository, and forcing the operational functions out of the executing brokers.
​

The 1975 amendments went further. They directed the SEC to facilitate a national market system in which price discovery happened at competing execution venues while clearing and settlement consolidated at neutral utilities. The IOSCO implementation reports trace how this pattern then propagated to almost every major jurisdiction (IOSCO PFMI Implementation Monitoring). The lesson institutional buyers absorbed is durable: the venue that prices the trade should not also be the entity that holds the asset, novates the obligation, and books finality. Concentration of those functions creates conflicts that no disclosure regime can neutralize.
​

In stablecoin infrastructure, separation of functions means that the entity routing the trade, the entity holding the underlying tokens, the entity orchestrating multi-venue clearing, and the chain that records final settlement are deliberately distinct. The buyer integrates each layer through standard interfaces rather than depending on a vertically integrated provider whose incentives may diverge from best execution.
​

Operationally, that looks like an asset manager whose treasury team holds USDC at Anchorage Digital, sources liquidity through an orchestration layer that polls Circle's primary mint, several OTC desks, and onchain venues, and receives final settlement on Ethereum or a Layer 2. No single counterparty sees the entire flow. The custodian does not see the orders. The orchestrator does not hold the assets. The chain does not match or net.
​

The European Central Bank has flagged this layering explicitly in its macroprudential work on stablecoins, arguing that the functional separation seen in TradFi is a prerequisite for stablecoins to scale into mainstream payment and settlement use (ECB Macroprudential Bulletin).
​

The onchain stack now mirrors the four TradFi functions with named providers in each layer. Centralized exchanges and decentralized exchanges run execution. Orchestration layers handle multi-venue routing, netting, and best-execution analytics analogous to clearing. Qualified custodians hold the assets. Public chains, occasionally bridged through messaging protocols, provide final settlement.
​

Execution sits across Binance, Coinbase, OKX, and Bybit on the centralized side, with Binance alone reporting roughly $137.8B in custodied client balances as of Q2 2026 (DeFiLlama, June 2026). On the decentralized side, Uniswap, Curve, and CoW Protocol match flow against onchain liquidity. Custody is concentrated at Anchorage Digital, BitGo, Fireblocks, and Coinbase Custody. Fireblocks alone has secured more than $7T in cumulative digital-asset transfers (Fireblocks corporate disclosure, 2024). Settlement happens on Ethereum, where roughly $37.1B in DeFi TVL is currently anchored (DeFiLlama, June 2026), plus Tron, Solana, Base, and the broader L2 set.
​

The table below shows the four-function map across TradFi and stablecoin providers.

The Dune community dashboards covering stablecoin settlement volume are the most-cited primary source for the chain layer's growth curve (Dune stablecoin settlement queries).
​

Institutional buyers demand separation because integrated providers create conflicts in best execution, concentrate operational risk, and complicate audit. A treasury that custodies, executes, and settles inside one venue cannot prove independently that it received the best price, that its assets are segregated from the venue's own book, or that its records reconcile against an external source of truth.
​

Three institutional outcomes drive the demand. The first is neutrality. A custodian that also runs an execution desk has an incentive to internalize flow rather than route to the venue offering the tightest spread. Splitting the functions removes that incentive entirely. The second is audit trail. When custody, execution, and settlement sit at different counterparties, every leg generates an independent record that auditors and regulators can reconcile against. The third is capital efficiency through orchestration. The DTCC processed approximately $3.0 quadrillion in securities transactions in 2023 by netting obligations across thousands of execution venues into a single end-of-day book-entry transfer (DTCC 2023 Annual Report). The same logic applies onchain: an orchestrator that nets mint, burn, and secondary flow across Circle, Tether, and OTC desks reduces the gross settlement load institutions push to the chain layer.
​

BIS Triennial 2022 puts daily FX turnover at $7.5T. CLS settles roughly $6.5T of that.5T daily across 18 currencies, neutralizing principal risk for the largest pairs (BIS, October 2022). Stablecoin markets are now reaching the scale where the same neutral-utility logic begins to dominate buyer procurement decisions.
​

The lines remain blurry where custodial exchanges hold client balances on the same balance sheet that runs their matching engine, where wallets bundle custody with execution routing, and where re-hypothecation lets a venue lend out client tokens without explicit consent. Each pattern recreates the conflicts that TradFi separation was designed to remove.
​

Custodial exchanges are the most consequential blur. When a buyer holds USDT or USDC at a venue that also runs the order book, the same legal entity owns the custody obligation and the trading book. The collapse cycle of 2022 illustrated what happens when those functions fuse without segregation: client assets and house assets become a single bankruptcy pool. Anchorage Digital, BitGo, and Fireblocks have built around that risk by offering off-exchange settlement, where collateral stays at the custodian and only mirrored balances trade at the venue.
​

Vertically integrated wallets that include built-in swap routing introduce a softer version of the same conflict. Re-hypothecation, the practice of lending out customer-deposited tokens to fund venue operations, sits at the most aggressive end. Public disclosures here are thin, which is itself a signal: TradFi prime brokerage agreements specify re-hypothecation limits and disclosure terms in detail, and onchain equivalents are only beginning to be standardized.
​

Regulators are codifying the custody-execution-settlement split through three parallel tracks. The EU's Markets in Crypto-Assets Regulation requires segregation of client assets. The US GENIUS Act framework draws a bright line between issuer reserves and trading venues. The Office of the Comptroller of the Currency has confirmed that national banks may custody crypto-assets, opening the door to bank-grade custody at scale.
​

MiCA Article 75 requires Crypto-Asset Service Providers to hold client crypto separate from proprietary assets, with the rule taking effect for CASPs on December 30, 2024 (European Commission, MiCA implementing measures). The practical effect: an EU-licensed exchange that also wants to custody client tokens must structurally segregate them, with the operational and legal independence that implies. Bundled custody-execution business models do not survive contact with the rule unless the custody function is structurally walled off.
​

In the United States, the OCC's 2020 interpretive letter on crypto custody and subsequent clarifications established that nationally chartered banks may provide crypto custody services, which has driven the channel of institutional buyers toward Anchorage Digital (the first federally chartered crypto bank) and bank-affiliated custodians. Stablecoin-specific legislation under the GENIUS framework continues the same separation logic at the issuer layer, restricting permitted issuers from running trading venues against their own reserves.
​

The separation thesis holds that institutional stablecoin buyers should procure custody, execution, clearing, and settlement from distinct providers, integrate them through standard interfaces, and resist any single counterparty offering the full stack. The framework lets buyers benchmark spread, measure best execution, audit segregation, and avoid the concentration risk that TradFi spent five decades engineering out.
​

A practical reference procurement looks like this. Custody at a qualified custodian with explicit segregation language and an SOC 2 Type II report. Execution sourced from a panel that includes at least one CEX, one onchain venue, and one OTC desk, with RFQ workflows for tickets above a threshold. Orchestration through a neutral layer that exposes best-execution analytics, mint access at primary issuers, and netting across venues. Settlement on a chain selected for finality and audit characteristics rather than venue affiliation.
​

Raw onchain stablecoin transfer volume reached roughly $27T in 2024 (Visa Onchain Analytics), with adjusted estimates filtered for bots and MEV closer to $5-15T (Artemis, January 2025); current Q2 2026 float is $315.3B (DeFiLlama, June 2026). At that scale, the cost of conflicted execution and unsegregated custody is no longer hypothetical. The TradFi pattern is the reference, and the onchain stack is converging on it faster than most procurement teams have updated their vendor lists.
​

Eco operates as a neutral orchestration layer in this stack. It sits between execution venues, primary issuers, OTC desks, and custodians, providing routing, mint access, and best-execution analytics without taking principal risk or running its own custody or trading book. Institutional buyers integrate once and reach primary and secondary stablecoin markets across the providers named throughout this piece.
​

Stablecoin supply, market cap, and chain TVL figures are pulled from the Eco live data snapshot dated June 2026, which sources DeFiLlama and CoinGecko. Settlement-volume figures attributed to 2024 are drawn from Artemis Onchain Analytics aggregations cross-referenced against Visa's published onchain dashboards. DTCC processing volume is from the DTCC 2023 Annual Report. Fireblocks cumulative transfer volume is from Fireblocks corporate disclosures (2024). BIS turnover and CLS settlement coverage are from the BIS Triennial Central Bank Survey (October 2022). MiCA Article 75 effective date reflects the European Commission's published implementing timeline. Quarter qualifiers are used wherever a figure post-dates the last verified snapshot.
​