Treasury diversification across stablecoin issuers, USDC (Circle), USDT (Tether), USDS/DAI (Sky), and USDe (Ethena), is the primary risk-management lever for any treasury holding more than $5M in dollar-denominated tokens. The March 2023 USDC depeg to $0.87 made the case empirically: a treasury holding 100% of reserves in a single issuer is exposed to that issuer's specific reserve composition, banking partners, regulatory posture, and operational continuity. As of the 2026-09-15 pull from DeFiLlama, the four major issuers collectively account for about $268B in stablecoin supply, with very different risk profiles.
​

This guide compares the four dominant treasury stablecoins along reserve composition, regulatory status, depeg history, and DeFi/CeFi acceptance. The goal is a working framework for setting an asset whitelist with defensible per-issuer caps. The reader should walk away able to articulate why a 40/30/20/10 split is more or less appropriate than a 60/30/10/0 split given a particular treasury's constraints.
​

Stablecoin treasury diversification is the practice of allocating treasury reserves across multiple stablecoin issuers to reduce single-issuer risk. The risk being diversified is the issuer being unable to redeem one dollar of its token for one dollar of value, either because reserves are insufficient, frozen by regulators, or temporarily inaccessible.
​

The relevant historical events: USDC broke its peg in March 2023 after $3.3B of Circle's reserves, about 8% of the total, were stuck at Silicon Valley Bank, per Circle's March 13 2023 statement; UST collapsed to near-zero in May 2022 (algorithmic, not asset-backed, and not in scope here); USDT dipped below peg during the same LUNA contagion; BUSD was wound down after Paxos received a stop-issuance order from NYDFS.
​

None of these events caused permanent loss for redemption-eligible holders, but each created mark-to-market loss for treasuries forced to transact during the dislocation. That is the whole case for diversification: the loss falls on whoever has to move money while the peg is broken.
​

Stablecoin supply by issuer (pulled 2026-09-15 from DeFiLlama): USDT $183.3B, USDC $73.4B, USDS $6.5B, DAI $4.8B, USDe $4.7B, RLUSD $2.3B, PYUSD $2.8B, FDUSD $326M, FRAX $148M.
​

The four stablecoins compared here represent four different issuance models. Conflating them is the most common error in treasury policy.
​

Issued by Circle Internet Financial. Reserves disclosed monthly via Circle's transparency reports; ~80% in short-duration US Treasury bills held in the BlackRock Circle Reserve Fund, ~20% in cash at regulated US banks (BNY Mellon, Cross River, Customers Bank). Audited monthly by Deloitte.
​

Regulatory posture: Circle is a federally-chartered Money Services Business with state money-transmitter licenses. EU-licensed under MiCA from June 2024. Filed for IPO in April 2025; completed IPO June 5, 2025 and trades on NYSE under ticker CRCL.
​

Native chain coverage: Ethereum, Solana, Base, Arbitrum, Optimism, Polygon, Avalanche, Stellar, Algorand, Hedera, Noble, Sui, NEAR, Polkadot, Aptos, Linea, Sonic, Unichain, World Chain, ZKsync. Cross-chain via CCTP V2.
​

Depeg history: one event, the March 2023 SVB exposure, closed once US regulators confirmed depositors would be made whole, per Circle's March 11 2023 update. DeFiLlama's stablecoin price feed showed USDC at $0.9997 on 2026-09-15, per DeFiLlama.
​

Issued by Tether Limited (BVI-incorporated). Reserves disclosed quarterly via Tether's transparency reports; the mix spans US Treasury bills, reverse repos, bitcoin, gold, secured loans and other investments. Read the current period's figures off the attestation itself rather than a secondhand summary, because the composition moves quarter to quarter. Attested by BDO; not audited.
​

Regulatory posture: Not licensed in the US. Banned for retail in the EU under MiCA (delisted from major EU exchanges in early 2025). Settled with NYAG in 2021 ($18.5M), CFTC in 2021 ($41M). Cooperates with US law enforcement on address freezes; Tether does not publish a running total of frozen value that can be cited here.
​

Native chain coverage: Tron (~$74B), Ethereum (~$50B), Solana, BSC, Avalanche, Polygon, Arbitrum, Optimism, Aptos, Cosmos, Algorand, Liquid, Omni, EOS, NEAR, Tezos, Telos, Statemine, Hedera. Bridged elsewhere.
​

Depeg history: brief sub-peg trading during the May 2022 LUNA contagion and during the 2017-2019 Bitfinex litigation. DeFiLlama's price feed showed USDT at $0.9994 on 2026-09-15, per DeFiLlama.
​

Issued by Sky Protocol (rebranded from MakerDAO in August 2024). Asset-backed by a basket of tokenized real-world assets, stablecoin collateral and crypto collateral. Reserves are publicly verifiable onchain at any block, and the live collateral breakdown is published on the Sky ecosystem analytics dashboard.
​

Regulatory posture: Sky operates as a decentralized protocol; no single legal issuer. The Sky Foundation is a non-profit Cayman entity that supports protocol development. USDS is not regulated as an issued stablecoin in any jurisdiction.
​

Native chain coverage: Ethereum mainnet (canonical issuance). Bridged to Solana, Base, Arbitrum, Optimism, BSC, Polygon, Avalanche.
​

Depeg history: DAI traded above peg during the March 2023 USDC break, because it was substantially USDC-collateralized at the time, and again during the May 2022 LUNA contagion. DeFiLlama's price feed showed DAI at $1.0001 and USDS at $0.9994 on 2026-09-15, per DeFiLlama, and the current collateral mix is on the Sky ecosystem analytics dashboard.
​

Issued by Ethena Labs. Backed by a delta-neutral basis trade: long staked ETH (collateral) plus short ETH perpetual position (hedge), held by institutional custodians (Copper, Ceffu, Cobo) under off-exchange settlement arrangements. Reserves disclosed via Ethena's transparency dashboard.
​

Regulatory posture: Operates from BVI. Not registered or licensed in major jurisdictions; explicitly not available to US users. The basis-trade model has not been classified by major regulators yet.
​

Native chain coverage: Ethereum (canonical), with extensions to Arbitrum, Base, Optimism, Mantle, BSC.
​

Depeg history: brief sub-peg dips during sharp ETH funding-rate inversions, with no sustained depeg event to date. DeFiLlama's price feed showed USDe at $0.9991 on 2026-09-15, per DeFiLlama. Track record shorter than USDC, USDT, or DAI/USDS.
​

No single allocation is right for every treasury. Three frameworks fit different operating contexts.
​

USDC 60%, USDS 25%, PYUSD 10%, USDT 5%. Heavy weight on US-regulated issuers (Circle, Paxos), partial weight on the asset-backed decentralized option (USDS), minimal exposure to USDT given regulatory posture and lack of US licensing. Caps at 60% per issuer.
​

USDC 40%, USDT 30%, USDS 20%, USDe 10%. Spreads risk across reserve models, fully fiat-backed (USDC), mixed-asset (USDT), crypto-collateralized (USDS), synthetic (USDe). USDT weight is justified by operational need (Tron stablecoin payments, deep CEX liquidity).
​

USDC 35%, USDS 30%, USDe 20%, USDT 10%, PYUSD 5%. Highest weight to decentralized and crypto-native options (USDS, USDe). USDT minimum operational allocation for CEX-bridge transactions. PYUSD allocation for emerging payment-rail integration.
​

The numbers in each framework are starting points, not prescriptions. The right split depends on the treasury's specific operational chains, customer base, regulatory constraints, and risk tolerance.
​

The dominant policy mistake observed across DAO and corporate treasuries in 2023-2024 was the absence of explicit per-issuer caps. A retrospective by Steakhouse Financial on 33 onchain treasury policies found 19 had no cap at all; the median single-issuer exposure was 71%.
​

A defensible cap structure: a 50% per-issuer maximum for larger treasuries, loosening to 60% for smaller ones, where the operational complexity of multi-issuer holdings outweighs the marginal risk reduction. Synthetic-dollar issuers (USDe) capped at 20% regardless of size, because the basis-trade model has not been stress-tested through a full bear cycle.
​

Caps need to be enforced with regular rebalancing. A treasury that drifts to 75% USDC over two quarters because USDC is more operationally convenient on Base hasn't followed its own policy. Quarterly rebalance checkpoints, review balances, rebalance if any issuer exceeds cap by more than 5 percentage points, are the lightest viable enforcement.
​

The diversification analysis extends to yield-bearing variants. sUSDC doesn't exist natively; USDC yield comes through Aave, Morpho, Spark, and similar money markets. sUSDS is the Sky Savings Rate variant, paying the savings rate directly (3.60% on 2026-09-16 per DeFiLlama). sUSDe is the staked Ethena variant, paying the basis-trade yield (5.00% on the same date per the same source).
​

Holding sUSDS and sUSDe directly inherits the issuer risk of USDS and USDe respectively, plus the protocol-specific staking risk. The diversification analysis should treat sUSDS as USDS (with a yield wrapper) and sUSDe as USDe (with a yield wrapper), not as new asset categories. See the yield options guide for the full breakdown.
​

Holding four stablecoins across multiple chains creates operational overhead that single-issuer treasuries don't face. Three categories matter.
​

Swap costs. Customers and counterparties may require payments in a specific stablecoin. A treasury holding USDC may need to swap to USDT to pay a Tron-based vendor, then swap back. Those swap costs are small per trade on deep stable-stable pools, and they add up across high-volume operations.
​

Reporting complexity. Every stablecoin and every chain is a separate reporting line. A four-issuer, four-chain treasury has 16 line items per snapshot. Tools like Den, Steakhouse Financial, and Karpatkey aggregate this; without them, reconciliation is manual and error-prone.
​

Custody coverage. Different custodians support different stablecoins natively. A treasury that requires institutional custody on all four issuers may need to use multiple custodians (Coinbase Custody for USDC, Fireblocks for USDT/USDe, self-custody Safe for USDS).
​

The operational cost of multi-issuer holdings is real and should be priced into the diversification decision. For a small treasury the operational cost likely exceeds the diversification benefit. Past the point where a single issuer failure would be material to the business, the diversification benefit dominates.
​

Treasuries that hold across multiple stablecoin issuers face constant inter-stable swap workflows: receive USDC on Base from a customer, pay a vendor in USDT on Tron, hold a yield position in sUSDS on Ethereum. Eco is the stablecoin execution network that handles those cross-stable, cross-chain movements. A treasury team integrates Eco once and gets unified routing across 15 chains and major stablecoin pairs; the intent, "swap $500K USDC on Base for USDT on Tron, settle in under 60 seconds", goes in, settlement comes out. For the broader treasury management context, see the treasury management pillar; for cross-chain mechanics, see Eco Routes.
​

USDC has stronger US regulatory coverage (Circle is licensed; Tether is not) and more transparent reserve attestations (monthly Deloitte attestation versus quarterly BDO attestation). USDT has a longer track record (since 2014) and deeper CEX liquidity. Most institutional treasuries weight toward USDC for the regulatory profile while keeping operational USDT for liquidity. See the treasury management guide.
​

USDS is the rebranded DAI under Sky's August 2024 protocol upgrade. New treasuries should hold USDS rather than DAI; existing DAI holders can convert 1:1 through the Sky protocol. The economic exposure is identical; USDS has additional features including direct sUSDS savings rate access. The Sky ecosystem analytics dashboard tracks USDS and DAI supply in real time.
​

Most defensible policies cap USDe at 10-20% of stablecoin reserves because the basis-trade backing model has not been stress-tested through a full bear cycle. The yield on sUSDe is variable and was 5.00% on 2026-09-16 per DeFiLlama. It is compensation for funding-rate risk, not a money-market position. See the yield strategy guide for more.
​

USDC traded at $0.87-0.93 for ~48 hours after Silicon Valley Bank failed with $3.3B of Circle's USDC reserves on deposit. The FDIC backstop on March 12 restored the peg. Treasuries holding 100% USDC absorbed the mark-to-market loss for the duration. The event is the empirical case for issuer diversification, see the multi-chain treasury guide for related risk analysis.
​

PYUSD (Paxos) and FDUSD (First Digital) are smaller and newer than USDC, USDT and USDS. On 2026-09-15, PYUSD supply was $2.8B and FDUSD supply was $326M, per DeFiLlama. Both are usable for operational allocations (typically 5-10% caps) but lack the deep DeFi liquidity needed for large-scale yield deployment.
​