Primary stablecoin markets are where regulated issuers like Circle, Tether, and Paxos mint new supply against fiat collateral and redeem it back to dollars. Secondary stablecoin markets are everywhere those tokens trade after issuance: OTC desks, centralized exchange order books, and onchain liquidity pools. The total stablecoin market capitalization sits at $315.3B as of 2026-06-05 (DeFiLlama), and almost every institutional flow above $10M touches both tiers before it settles. Understanding which tier you are using, and why, is the difference between a clean basis point and a thirty basis point slippage event.
​

Stablecoins clear through a two-tier system. The primary market is the issuer relationship where supply is created and destroyed at par. The secondary market is the venue ecosystem where that supply trades, gets borrowed, and finds price. Both tiers exist for every major stablecoin, and the choice between them drives execution quality on every flow above retail size.
​

The structure is older than crypto. Treasury markets, money market funds, and ETF shares all clear through a primary-secondary split. An authorized participant creates or redeems ETF shares at NAV with the issuer. Everyone else trades the existing float on exchange. The arbitrage between those two prices is what keeps the secondary price tethered to the underlying.
​

Stablecoins inherited this design. USDT supply is $187.2B and USDC supply is $75.6B as of 2026-06-05 (DeFiLlama). That float trades across hundreds of secondary venues every second. But the float itself only grows or shrinks when an institution with a primary account wires fiat to Circle, Tether, or Paxos and receives newly minted tokens in return, or wires tokens back for fiat.
​

The Bank for International Settlements has documented this tiering in its 2024 working paper on stablecoin market structure, which treats primary issuance as the anchor and secondary trading as the price-discovery layer. The Federal Reserve's Regulation II framework and the BIS working paper 1281 both treat the primary leg as the systemic risk surface and the secondary leg as the liquidity surface. Anyone designing a stablecoin execution stack needs to internalize that division.
​

A primary stablecoin market is the direct relationship between an institution and a regulated issuer. The institution wires fiat to the issuer's bank account and receives newly minted tokens at par, or returns tokens for fiat redemption. Primary access is gated by KYB, minimums, and bank rails, but it delivers exact-par execution with no slippage.
​

Circle's primary market for USDC runs through Circle Mint. An approved institution wires USD via Fedwire or SWIFT, and Circle credits newly minted USDC to a designated wallet, typically within the same business day. Redemption runs in reverse: send USDC to Circle's burn address, receive USD wire back. Tether operates a similar primary venue documented in its transparency disclosures, and Paxos runs primary issuance for PYUSD and USDG against segregated reserves held in trust.
​

Primary access is not retail. Onboarding requires entity-level KYB, source-of-funds attestation, and in most cases a relationship with a qualified custodian like Anchorage Digital to hold the wallet that receives mint proceeds. Circle Mint has institutional tiers with specific minimums set per the issuer's onboarding process; the issuer does not publicly disclose tier-specific thresholds. The benefit is execution quality. A primary mint is a par-priced creation of new supply. There is no order book, no maker fee, no slippage curve. One million USD wires in, one million USDC is created.
​

The drawback is latency and cost on the rails side. Wires settle on bank hours, not block time. Cutoff windows matter. A Friday afternoon mint request may not credit until Monday. And the fiat leg carries wire fees, FX spread if the source currency is not the issuer's reserve currency, and counterparty banking risk. Primary access is a wholesale channel, not a real-time liquidity solution.
​

Secondary stablecoin markets are every venue where issued stablecoin float trades after mint. The three sub-tiers are over-the-counter desks (B2C2, Wintermute, Cumberland DRW, GSR), centralized exchange order books, and onchain liquidity pools on DEXs and bridges. Each sub-tier prices differently and serves a different flow profile.
​

OTC desks are the institutional secondary venue. A trading desk like B2C2, Wintermute, Cumberland DRW, or GSR will quote a two-sided market on $5M to $100M+ blocks with no public order book impact. The desk warehouses the risk on its own balance sheet and works the offsetting flow across its venues. Execution is bilateral, settlement is T+0 or T+1, and the spread compensates the desk for inventory and information risk. For large flows where market impact matters more than the last basis point of spread, OTC is the default secondary channel.
​

Centralized exchange order books are the price-discovery layer most people see. Binance, Coinbase, Kraken, and a long tail of regional venues host continuous double auctions for USDT, USDC, and other stablecoin pairs. Order book depth at the top of book is thin compared to spot crypto pairs because stablecoin trading is mostly used as a settlement leg rather than a directional bet. A $5M USDT-USDC market order on most CEX venues will walk the book.
​

Onchain liquidity is the third secondary tier. Decentralized exchanges like Curve, Uniswap, and Balancer host concentrated liquidity pools where stablecoin pairs trade against pricing curves. Onchain venues are composable, atomically settled, and queryable in real time. They also leak information through the mempool and incur gas costs. LayerZero V2 alone moved $7.5B in bridge TVL as of 2026-06-05 (DeFiLlama), and a meaningful share of that is stablecoin liquidity routed across chains. The onchain tier is the only secondary venue where settlement and trading happen in the same block.
​

Price discovery for stablecoins happens primarily on centralized exchange order books and secondarily on onchain DEX pools. The primary mint price is fixed at par by design and contributes no price information. OTC desks reference exchange and onchain prices to construct their quotes. The peg holds because primary mint and redemption create an arbitrage corridor around the secondary price.
​

This is the part most narratives get wrong. Price is not set at the mint. Price is set on the secondary venues, and the mint window is what enforces the peg. When USDC trades at $0.998 on Binance, an authorized institution can buy at $0.998, redeem at $1.000 with Circle, and pocket twenty basis points minus rails costs. That arbitrage is what pulls the secondary price back to par.
​

The implication for execution: if you only have secondary access, you are a price taker on whatever the order book and the OTC desks are showing. If you have primary access, you have an outside option that caps your downside on the redemption side and an upside option that caps your fill cost on the creation side. The BIS working paper documents how this dual-tier structure compresses peg volatility versus a single-tier design, and it is the same mechanism that keeps ETF shares trading near NAV.
​

Anchorage Digital and other qualified custodians sit in this picture as the wallet layer that lets an institution hold both the primary mint relationship and the secondary venue connections under one custody footprint. The custodian does not set price. It enables the institution to move between tiers without breaking custody.
​

Institutional best execution requires both primary and secondary access because each tier solves a different problem. Primary delivers par pricing and unlimited size, but slow rails and high minimums. Secondary delivers real-time fills and small clip sizes, but pays spread to the desk or the curve. The optimal stack routes each order to the cheaper tier given the flow's size, urgency, and direction.
​

Consider a $25M USDC inflow that needs to deploy onchain by end of day. The naive path is to wire fiat to Circle and wait for the mint. That is par execution, but it takes hours and the deployment window may close. The alternative is to hit an OTC desk for the full clip at a quoted spread of two to five basis points, settle T+0, and deploy immediately. The spread is the cost of speed. A serious treasury operation runs the math both ways on every flow.
​

Now consider a $5M USDT redemption back to USD. Hitting the Tether primary market is par, but the wire may take a day. Selling $5M USDT on a CEX order book will move the price two to four basis points and surface the flow publicly. An OTC quote splits the difference: T+0 settlement, a known spread, no public footprint. Each route has a profile, and treasury teams that only run one route are leaving execution quality on the table.
​

This is where the analogue to TradFi best-execution analytics holds. Equity execution desks route the same order across dark pools, lit venues, and capital commitment from a bank desk, and they measure realized cost against arrival price. Stablecoin execution is heading to the same discipline. BlackRock's BUIDL fund, with $3.0B in supply as of 2026-06-05 (DeFiLlama), is one signal that the institutional discipline is arriving. The flows that move BUIDL touch primary mint, OTC, and onchain rails across a single business day.
​

The orchestration gap is the operational distance between holding a primary mint relationship, a stack of OTC desk integrations, and a set of onchain liquidity routes. Most institutions today maintain these as three separate workstreams with three different ops teams. A unified orchestration layer that prices and routes across all three is the missing piece in the institutional stablecoin stack.
​

The gap shows up in small ways. A treasury team has a Circle Mint account, a Wintermute API, and a wallet on Coinbase Custody that talks to Curve. The primary mint sits in a banking ops queue. The OTC quote comes through a chat or an API. The onchain route is a separate engineering project. Each lane has its own reconciliation, its own counterparty review, its own settlement footprint.
​

The orchestration layer is the piece that asks: given a $25M flow with a four-hour deadline, what is the cost-minimizing route across primary mint, the three OTC desks I have on file, and the four onchain pools I can reach without bridging? That question has a precise answer at every moment. Computing it requires a real-time feed of primary mint windows, current OTC quotes via RFQ, and onchain depth across the venues that matter. DeFiLlama's stablecoin dashboard gives a starting view of the float, but live execution requires live pricing.
​

The clearing layer underneath orchestration is where the three lanes converge into a single executable. Primary mint clears against issuer reserves. OTC clears against desk inventory. Onchain clears atomically against pool state. A clean orchestration design treats all three as clearing venues, prices them against each other in real time, and selects the best route on a per-order basis.
​

Eco operates as a neutral orchestration platform. It does not issue stablecoins, does not run an OTC desk, and does not take principal risk on flows. It connects to primary mint windows at major issuers, to OTC desk RFQ feeds, and to onchain liquidity across chains, and it routes each institutional flow to the venue with the best realized cost. The neutrality is structural, not a marketing claim.
​

Neutrality matters because every other layer of the stablecoin stack has a directional interest. Issuers want flow to mint and stay minted. OTC desks want to capture spread on as many clips as possible. Onchain venues want TVL. An orchestration layer that owns any of those positions cannot honestly route against them. Eco's role is to sit above the clearing layer and price the alternatives without owning any of them.
​

The composability point matters too. Because Eco connects to onchain liquidity natively, flows that originate in a primary mint or an OTC fill can settle into a wallet that is immediately usable in onchain venues. The institution does not have to bridge, rewrap, or re-custody to move between tiers. That continuity is what makes the orchestration layer worth more than the sum of its parts.
​

The benefits read in institutional language: efficiency from routing every order to its lowest-cost venue, neutrality from a platform with no principal book, and composability from a unified settlement footprint across onchain and offchain rails. None of those benefits require the institution to take a view on which issuer wins or which chain dominates. The design is venue-agnostic by construction.
​

A complete stablecoin execution stack covers primary mint access at the major issuers, RFQ connectivity to at least three OTC desks, onchain routing across the chains where the institution settles, qualified custody for the wallets, and an orchestration layer that prices across all of the above. Missing any one of these forces a fallback to a suboptimal route on every flow that needs the missing piece.
​

Use the following table when auditing or building the stack. Each row is a capability that an institutional desk should be able to check off, with the venue type that delivers it and the failure mode if the row is empty.

Two diligence questions sit on top of this checklist. First, does any single counterparty in the stack require the institution to route flow through that counterparty's affiliated venue? If yes, the neutrality of the orchestration layer is compromised. Second, can the orchestration layer produce a post-trade report that compares realized cost against the alternative routes at the moment of execution? Without that report, best-execution claims are not auditable.
​

The Bank for International Settlements working paper on stablecoin market structure makes a related point at the systemic level. The institutions that survive the next regulatory cycle will be the ones that can demonstrate, with logs and timestamps, that every flow was routed to the best available venue at the time. That capability is built at the orchestration layer or it is not built at all.
​

Stablecoin market capitalization, issuer-level supply figures, and bridge TVL in this article are sourced from DeFiLlama's stablecoin dashboard snapshot dated 2026-06-05. Total stablecoin market capitalization of $315.3B, USDT supply of $187.2B, USDC supply of $75.6B, BlackRock BUIDL supply of $3.0B, and LayerZero V2 bridge TVL of $7.5B all reference that snapshot. Regulatory framing draws on the Federal Reserve's Regulation II overview and the Bank for International Settlements working paper 1281 on stablecoin market structure. Issuer transparency and mint mechanics reference public disclosures from Circle, Tether, and Paxos. OTC desk capabilities reference public service descriptions from B2C2, Wintermute, Cumberland DRW, and GSR. Qualified custody references public disclosures from Anchorage Digital. All numbers are point-in-time and will drift; readers running this analysis on a different date should pull a fresh DeFiLlama snapshot.
​