Stablecoin execution methods are the three dominant ways institutions and trading desks move size in dollar-pegged tokens: atomic onchain swaps, offchain OTC quotes, and multi-dealer RFQ. Each prices and settles differently, and each fits a distinct slice of flow. With total stablecoin supply at $315.3B as of Jun 5, 2026 (DeFiLlama), led by USDT at $187.2B and USDC at $75.6B, the question for treasurers and trading desks is no longer whether to use stablecoins, but which execution surface clears a given order at the best all-in cost.
​

This piece breaks down the mechanics, the trade-offs, and a decision framework drawn from how primary issuers like Circle and Tether, OTC desks like B2C2 and Cumberland DRW, and onchain venues like Uniswap v4 and CoW Protocol actually transact. It closes with where neutral orchestration fits, and what regulators expect under best-execution frameworks including FINRA Rule 5310, MiFID II, and the FSB and IOSCO guidance on global stablecoin arrangements.
​

Stablecoin execution methods divide into three surfaces. Atomic swaps settle onchain in a single transaction at an algorithmic or pool-derived price. OTC desks quote bilaterally offchain and settle by wire or token transfer against credit lines. RFQ venues solicit competing quotes from multiple market makers and route to the best, then settle either onchain or through custodial rails. Sizing, latency, and custody differ across all three.
​

The atomic path runs through AMMs and order-book DEXs. Uniswap v4 introduced hooks that let liquidity providers customize curves for stablecoin pairs, and monthly stablecoin DEX volume across the Uniswap protocol has run in the tens of billions through 2025 (Uniswap Labs, Dune dashboards). Settlement is the swap itself: the same Ethereum transaction that debits one stablecoin credits another, with no separate clearing step.
​

OTC execution is closer to a foreign-exchange voice market. Desks including B2C2, Wintermute, Cumberland DRW, and Galaxy Digital Trading stream indicative quotes and firm prices on request, often through API endpoints or chat. Trades clear bilaterally against pre-established credit, then settle by stablecoin transfer or wire. RFQ formalizes that workflow into a multi-dealer auction: protocols like Hashflow, CoW Protocol, and 0x route a single request to several professional market makers and lift the best response, capturing best-execution evidence in the process.
​

For background on stablecoin design and the policy framing that recognizes these three surfaces as distinct arrangements, see the Federal Reserve's "The Stable in Stablecoins" FEDS Note from December 2023.
​

Stablecoin atomic swaps price using deterministic onchain logic, typically a constant-product or stableswap curve, and settle in the same block as the trade. The taker submits a transaction, the pool's invariant updates, and tokens move in one atomic step. There is no separate clearing party, no T+0 wire, and no counterparty quote. Slippage is the only price uncertainty.
​

The mechanics matter for execution quality. A Curve or Uniswap v4 stableswap pool concentrates liquidity near the peg, so small clips trade at near-zero spread and large clips push price along a known curve. The taker sees the entire cost function before signing. That transparency is why atomic swaps dominate retail and small-institutional flow under roughly $1M per clip, where wire frictions and OTC minimums would dwarf onchain slippage. Uniswap Labs reports monthly stablecoin pair volume in the tens of billions through 2025, and DeFiLlama's stablecoins dashboard ranks the resulting onchain depth across chains.
​

Settlement risk in atomic swaps collapses to inclusion risk. If the transaction lands in a block, the trade is final at the prevailing pool state. Maximal extractable value, or MEV, is the dominant residual cost: searchers reorder or sandwich large clips, leaking value to block builders. Aggregators like 1inch and CoW Protocol route around this with private order flow and batch settlement, but the underlying pricing primitive is still the onchain curve.
​

OTC desks execute stablecoin flow bilaterally offchain. A client requests a price for a specific notional, the desk quotes a two-way market or a one-way firm price, and the trade clears against credit lines or pre-funded accounts. Settlement happens by stablecoin transfer or fiat wire after the trade is booked. The model handles size, custody flexibility, and FX-style relationship pricing that onchain pools cannot.
​

Desks like B2C2, Wintermute, Cumberland DRW, and Galaxy Digital Trading run principal books, meaning they take the other side and warehouse risk. Quotes reflect their inventory, hedging cost, and view on flow. For a $25M USDC-to-USDT trade, a desk can typically print a price tighter than the onchain curve at that size, because it nets the position against existing inventory rather than walking a pool. Custody is whatever the client and desk agree to: Fireblocks, BitGo, Anchorage, or self-custody on either side.
​

The trade-off is bilateral counterparty risk and operational lift. Each desk relationship requires KYB, credit setup, and legal documentation. Settlement is not atomic. The client wires or sends tokens, the desk reciprocates, and the gap between the two creates Herstatt-style risk that institutions manage through net settlement windows and trusted custodial intermediaries. The BIS CPMI and IOSCO addressed exactly this surface in their final guidance on stablecoin arrangements issued in July 2022 and reaffirmed in 2024, applying the Principles for Financial Market Infrastructures to settlement assets used at scale.
​

RFQ, or request-for-quote, is a structured auction where one taker request fans out to multiple market makers and the system lifts the best response. It pairs the size-friendliness of OTC with the competitive discovery of an order book. For stablecoin flow, RFQ venues capture quotes, response times, and fill prices as structured data, which is the raw material for best-execution analytics under MiFID II and FINRA Rule 5310.
​

The protocol surface here is diverse. CoW Protocol runs batch auctions where solvers compete to settle a set of orders at uniform clearing prices, with surplus returned to users. Hashflow uses signed quotes from professional market makers and settles onchain without slippage. 0x and its consumer front end Matcha aggregate dealer liquidity and onchain pools into a single RFQ-style quote. For an institutional desk routing a $10M PYUSD-to-USDC trade, an RFQ run across five makers typically produces a tighter all-in spread than any single OTC desk would quote, while still respecting size.
​

The audit trail is the institutional unlock. FINRA Rule 5310 requires firms to use reasonable diligence to ascertain the best market for a security and to execute customer orders so the price is as favorable as possible under prevailing conditions. RFQ logs every quote received, the spread between best and second-best, and the time stamps. For stablecoins, where the asset is a tokenized liability of an issuer like Circle (USDC) or Tether (USDT), that record is what compliance and the trading committee want to see.
​

Sizing, latency, and custody are the three operational axes that determine which execution method fits a given stablecoin order. Atomic swaps clear small clips in seconds against self-custodied wallets. OTC desks absorb large size with bespoke custody and minute-to-hour settlement. RFQ sits between, executing mid-to-large size with onchain or custodial settlement and capturing dealer competition in a structured workflow.
​

The matrix below summarizes how the three methods compare across the dimensions institutions weigh when routing flow. It reflects how desks at the scale of B2C2, Wintermute, and Galaxy Digital quote against the kind of onchain depth tracked on the DeFiLlama stablecoins dashboard, and how RFQ venues like CoW Protocol, Hashflow, and 0x position themselves in between.

Two patterns fall out of the matrix. First, no single method dominates across all dimensions. Second, the custody and onboarding columns explain why institutional buyers consolidate around platforms that surface all three surfaces under one integration, rather than running KYB with twelve venues independently.
​

Each stablecoin execution method wins a specific flow. Atomic swaps win small, latency-sensitive, self-custodial flow. OTC desks win large, relationship-priced flow with custom settlement terms. RFQ wins mid-sized institutional flow that needs competitive pricing plus a clean best-execution record. Treasurers and trading desks should pick by clip size, custody constraint, and the regulatory record they owe their committee.
​

Treasury rebalancing between operating stablecoins typically lands on RFQ. A payments company moving $5M from USDC to PYUSD weekly cares about spread, atomicity, and an exportable record. CoW Protocol's batch auctions or Hashflow's signed quotes give a clean fill and a structured log. For the same firm's daily $50K sweeps between subsidiary wallets, an atomic swap on Uniswap v4 is cheaper than the operational cost of an RFQ session.
​

Block-size flow is OTC's stronghold. An asset manager reallocating $100M from a tokenized cash product like BlackRock's BUIDL (supply $3.0B as of Jun 5, 2026, DeFiLlama) into USDC for deployment will redeem at primary and execute the residual leg with a desk that can quote on inventory and settle through qualified custody. Onchain depth at that clip would imply meaningful slippage; an RFQ at that size narrows the maker set so far that the auction loses its edge.
​

Arbitrage and market-making flow inverts the picture. A maker hedging a USDT position uses atomic swaps for the leg that needs immediate inclusion and an OTC line for the residual, with RFQ filling in the middle when a counterparty offers a price improvement. The Federal Reserve's FEDS Note on stablecoins describes exactly this stratification of primary, secondary, and bilateral surfaces as the secondary market matures.
​

Hybrid execution splits a single stablecoin order across atomic, OTC, and RFQ legs to optimize the all-in cost. A $40M USDC-to-USDT trade might fill $5M atomically on Curve and Uniswap v4 to capture immediate depth, route $15M through an RFQ across CoW, Hashflow, and 0x for competitive mid-tier pricing, and clear the residual $20M with B2C2 or Wintermute on a bilateral quote. The taker sees one blended fill.
​

The routing logic is the hard part. A solver needs real-time inventory at each venue, expected slippage at each clip size, dealer response-time distributions, and the cost of custody movement between surfaces. CoW Protocol publishes solver competition data, and 0x exposes API-level routing across pools and RFQ. Aggregators including 1inch and the LiFi orchestration layer compose onchain legs across chains. Onchain plus offchain blending, though, is still mostly proprietary to large desks and a small set of platforms.
​

The institutional value of hybrid routing is twofold. First, it shrinks effective spread on flows that don't fit cleanly in one surface, which is most flows above $1M. Second, it produces a unified best-execution record across all three surfaces, satisfying internal trading committees and external supervisors. The FSB's 2023 high-level recommendations on global stablecoin arrangements emphasize comparable expectations across functionally similar activities, which makes a single, exportable record more valuable than three siloed ones.
​

Best-execution frameworks expect firms to seek the most favorable terms reasonably available for client orders across the markets where the asset trades. For stablecoins, that means firms must consider atomic, OTC, and RFQ surfaces, document the comparison, and route accordingly. FINRA Rule 5310, MiFID II RTS 27 and 28, and the IOSCO objectives all impose a process expectation, not a specific venue requirement.
​

FINRA's Rule 5310 sets the US standard: members must use reasonable diligence to ascertain the best market for the subject security and execute orders so the resulting price is as favorable as possible under prevailing conditions. Stablecoins themselves are not securities under US law as currently interpreted, but firms that handle stablecoins inside broker-dealer or alternative trading system structures inherit 5310-style obligations through state money-transmitter rules and bank supervisory expectations. In Europe, MiFID II's best-execution framework requires firms to take "all sufficient steps" and publish execution-quality reports.
​

The supranational layer adds prudential expectations. The BIS CPMI and IOSCO final guidance on stablecoin arrangements from July 2022, reaffirmed through 2024 work, applies the Principles for Financial Market Infrastructures to systemic stablecoin systems. The FSB's 2023 final report sets ten high-level recommendations, including comprehensive oversight of governance, risk management, and redemption mechanisms. Firms executing stablecoin flow at scale should treat these as the framing within which their venue selection and audit-trail design will be examined.
​

Eco is building toward neutral orchestration across atomic, OTC, and RFQ surfaces, so a single institutional integration can clear flow on whichever method fits the order. Eco does not trade its own book, does not act as a market maker, and does not take principal risk. The platform aggregates execution methods, surfaces best-execution comparisons, and routes across primary mint access and secondary liquidity without preferring one venue over another.
​

That neutrality matters because the stratification described above is accelerating. Stablecoin supply at $315.3B as of Jun 5, 2026 (DeFiLlama) is split across dozens of issuers, dozens of chains, and a growing set of onchain and offchain venues. Institutional buyers don't want to run KYB with twelve platforms, integrate twelve APIs, and assemble twelve best-execution records. They want one integration that resolves to whichever combination of atomic, OTC, and RFQ legs clears their order at the best all-in cost, with a single exportable audit trail.
​

The platform thesis follows the market structure, not the other way around. As primary-market access through tokenized cash products like BUIDL ($3.0B, Jun 5, 2026) grows, as RFQ venues like CoW and Hashflow widen their maker sets, and as OTC desks formalize their stablecoin inventory, the orchestration layer that connects them becomes the highest-leverage piece of infrastructure. For a deeper look at how primary and secondary markets fit together under one neutral surface, see Primary vs Secondary Stablecoin Markets.
​

Stablecoin supply figures, including the $315.3B total, USDT at $187.2B, USDC at $75.6B, and BUIDL at $3.0B, are pulled from the DeFiLlama stablecoins dashboard snapshot dated Jun 5, 2026. Onchain volume references draw on Uniswap Labs disclosures and public Dune dashboards for 2025. Regulatory references are sourced directly from BIS CPMI-IOSCO (July 2022, reaffirmed 2024), the FSB final report (July 2023), and FINRA Rule 5310. Where a specific number could not be verified against a primary source, the article uses quarter qualifiers rather than precise figures.
​