syrupUSDC — Retail Risk Report
Moderate-to-low risk · 6.75/10
Live pool backing, peg deviation, and exit-liquidity tiers are on the dashboard.
What’s pinned in this report is structural risk — architecture, the issuer menu, the risk axes, and the scores. Current magnitudes (pool split, per-issuer allocation, collateral ratio, concentration, exit tiers) drift weekly and are live on the dashboard. This report is written to stay correct across that drift.
| Yield | Exit method | Primary redemption | Age | Chains |
|---|---|---|---|---|
| ~4.5–5% live (organic loan interest) | DEX aggregator (sub-minute) or queue | Permissionless (no KYC) | ~2 years | Ethereum, Solana, Arbitrum, Base, Plasma |
Summary
syrupUSDC is Maple Finance’s onchain yield-bearing stablecoin. You deposit USDC, you receive syrupUSDC, and the price slowly accrues yield. The Ethereum pool holds roughly $1B+, with the syrupUSDT sibling materially smaller; live TVL on the dashboard. Maple’s “Syrup” product line has reported zero principal losses across ~3 years. Yield is real (interest paid by real institutional borrowers), not emissions.
Important framing — verified against Maple’s own AUM Details page: the pool actually has two structurally different components, which Maple itself splits as Loans (~75–80% of pool) + Liquidity (~20–25%) — live split on the dashboard:
- Loans = third-party institutional credit (BTC/XRP/cbBTC/HYPE collateral at 125–333% — overcollateralized in the traditional sense). This is the “Maple lending pool” that the marketing describes.
- Liquidity = pool-owned positions in stablecoin and tokenized T-bill yield strategies (PYUSD, USTB, USDC/USDT in DEX pools). At-par with the underlying asset, NOT third-party credit. Risk axis: issuer/RWA/AMM, not borrower default.
The “overcollateralized at all times” framing in Maple’s marketing applies to the Loans bucket. The Liquidity bucket is intentionally at par — it’s not collateralized credit, it’s pool-owned strategies parking capital in yield-generating positions while remaining nominally redeemable. The Pool Delegate rotates this layer across a menu of issuers — Paxos (PYUSD), Superstate (USTB, a NAV-accruing tokenized T-bill), Circle/Tether (USDC/USDT AMM-LP) — and the mix drifts under discretion within days, with no governance gate. Through 2026 the layer rotated substantially out of USTB into PYUSD; any single issuer can come to dominate the layer (recently ~90%+ Paxos PYUSD). Current per-issuer allocation is live on the dashboard; the durable facts are the issuer menu, that the mix is delegate-discretionary, and that single-issuer concentration inside the layer can run high.
What you actually earn
~4.5–5% APY (verified live from Maple’s GraphQL syrupGlobals.apyTimeSeries over the past week). This is the durable yield from institutional borrowers paying interest on their loans, net of Maple’s protocol take and 3.33% delegate fee. It varies by loan-book composition and rate environment; ~3–9% range across individual loans, weighted-avg currently ~4.5–5%.
At ~4.5–5%, syrupUSDC is currently competitive with or above the comparable yield set: 3-month T-bills are around 3.7–4.0% (US Treasury fiscal data, March 2026 average 3.70%), tokenized T-bill products (BUIDL, USTB, USYC) net ~3.5–4.0% after management fees, and onchain USDC lending on Aave V3 / Morpho is in the 3.5–4.5% range. The Maple value proposition is “real institutional credit yield, ~4.5–5%, with the risk profile of overcollateralized loans + at-par Liquidity strategies.” That’s a ~70–100 bp spread above T-bills, which is appropriate compensation for institutional credit risk rather than a yield-chase number.
syrupUSDC also has a structural advantage that pure tokenized-T-bill products don’t: it’s accepted as collateral on Morpho, Euler, and other DeFi lending markets, so a holder can borrow against it and lever the yield if the borrow rate is favorable. That makes it more capital-efficient than locked T-bill exposure, and is a meaningful part of the value proposition for allocators with a use for the borrowed capital.
How exit works
Two paths, both permissionless (no KYC required for either):
1. DEX aggregator (preferred for retail): Use KyberSwap, 1inch, or any DEX aggregator. Empirical exit cost is single-digit-to-low-double-digit bps at retail-to-low-institutional notional; live tiered slippage on the dashboard. Sub-minute settlement. Aggregators route across Uniswap v3/v4, Balancer, and other listed pools — significantly more depth than the strict “Uniswap+Balancer pool TVL” headline implies.
2. Direct redemption: Submit a redemption request to the vault contract; the WithdrawalManager processes it. Maple claims average withdrawal time under 5 minutes during normal markets. This is the path for sizes that exceed aggregator-route depth.
The honest qualifier: aggregator routing is excellent in normal market conditions but less reliable during stress. If many holders try to exit at once (a credit event, a crypto-cycle drawdown that hits institutional borrowers simultaneously), aggregator slippage widens and the redemption queue becomes the binding constraint. Queue speed depends on free USDC in the pool versus outstanding loan principal — if loans are fully deployed, the queue lengthens until borrowers repay or get margin-called.
What the contracts are doing
syrupUSDC is an ERC-4626 vault (the standard “deposit → get share token” pattern). The Ethereum deployment is canonical; the multi-chain versions (Solana, Arbitrum, Base, Plasma) are bridged extensions of the same product.
What sits behind the scenes:
- Borrowers post collateral that’s held off-chain by custodians under Pool Delegate policy. The smart contract itself does not hold or price the collateral — Maple’s GraphQL exposes the asset, amount, and required collateralization level per loan
- A single Pool Delegate (an institutional credit firm vetted by Maple) sets all loan terms, monitors borrower health, and has the right to call any loan with a 24-hour notice + 48-hour grace period before default
- The smart contracts handle loan accounting, payment scheduling, and the time-based default trigger — but the credit-relevant decisions (who to lend to, on what terms, when to call) are human-discretionary
- WithdrawalManager handles the redemption queue; LoanManager tracks per-loan principal and payment state
The Pool Delegate model is the structural difference from purely algorithmic protocols (like Aave). It adds discretionary credit risk — a delegate’s bad loan, a delegate-borrower conflict of interest, or a delegate mistake during a fast-moving market can produce losses that algorithmic protocols wouldn’t face. The delegate is also a single externally-owned address (single private key) and Maple’s first-loss cover requirement for the pool is currently $0 — meaning no on-chain protocol equity absorbs losses before depositors. Mitigated by: Maple’s delegate vetting, public delegate identity, and Maple Labs’ reputational stake.
Backing and how verifiable it is
The credit read, as at the August 2026 check, is reassuring. The loans-only collateral ratio across the Syrup family came in at 175%, above the 145–170% band this report describes as typical. Pool collateral ratio is 100%, unrealized losses are zero, and no loan is impaired, called, or in default. Deployment runs around 97% and the pool is using roughly two-fifths of its liquidity cap. Live figures are on the dashboard; the durable read is that the loan book is comfortably overcollateralized and the loss-recognition signal is clean.
That matters for how you read the rest of this report and the August rubric change: nothing about credit deteriorated. The change was that a question this report had never scored — redemption — got its own dial.
Two things do belong on the debit side, and they are about verification rather than credit.
The collateral is off-chain. Borrowers post collateral to custodians under Pool Delegate policy. The smart contract does not hold it and does not price it. What you can read is what Maple publishes — asset, amount, and required collateralization per loan — so your assurance runs through Maple’s reporting and the delegate’s monitoring, not through an on-chain balance you can check yourself. This is normal for institutional credit and abnormal for DeFi, and it is the main reason the Underlying axis sits at 7.0 rather than higher.
Maple’s public loan data has a known artifact, and it is not small. A subset of loans periodically read below 100% collateralization in Maple’s public GraphQL while every authoritative signal says they are fine — unrealized losses zero, none impaired, none called, none defaulted. At the August 2026 check that subset covered about 39% of the loan book, up from roughly a quarter in July. Read it correctly: the artifact grew, the credit did not deteriorate. The pool collateral ratio is the authoritative figure and it is clean. But a reader trying to verify this book independently should know that the most obvious public data source disagrees with the authoritative one on a material share of it, and that the discrepancy is a reporting issue rather than a solvency one.
One live indicator is worth naming. A single loan of about $25M — under 3% of the book — sat in the 100–120% collateralization range at the August check. That is the “tightest loan approaching par” signal working as designed at small size against a 175% book. It is the indicator firing, not a credit event; what would matter is many loans compressing toward par at once.
Audits & security
Strong by DeFi-stablecoin standards:
- 8+ audits total. The current v2/Syrup contracts specifically reviewed by Spearbit and Trail of Bits (both top-tier), plus Three Sigma and Peckshield.
- $1M+ Immunefi bug bounty active.
- ERC-4626 standard architecture; well-understood pattern.
Caveats:
- Multi-chain bridging uses Chainlink CCIP with the CCT (Cross-Chain Token) standard — burn-and-mint native deployments on each chain (Ethereum canonical + Solana, Arbitrum, Base, Plasma), not a wrapped/lockbox model. This is structurally a different attack class from the April 2026 LayerZero OFT incidents (rsETH, Drift, Volo) — those exploits hit single-DVN OFT configurations and admin-key compromises that don’t map to CCIP’s architecture. CCIP has a clean track record at scale through April 2026 (no public exploits since 2023 launch) with a Risk Management Network as an anti-fraud backstop. What you trade off vs LayerZero: concentrated trust in Chainlink as a single (mature) provider, rather than distributed quorum across DVNs. Per-chain DD still warranted: verify CCT pool addresses on Chainlink’s CCIP directory and check per-chain pool depth before sizing on non-Ethereum venues.
- The Pool Delegate roster is what you’re trusting at the credit-judgment layer. Maple publishes current delegates; cross-check that they’re active and reputable before sizing institutional positions.
Score breakdown
| Dimension | Score | Notes |
|---|---|---|
| Stability | 8.5 | NAV-accruing share, organic yield from loan interest, no rebase, and zero principal losses to date across the Syrup product line. The share price only climbs in normal operation; the path to a drawdown is a credit loss on the loan book, which is scored under Underlying. |
| Contract & Admin | 6.5 | ERC-4626 standard, 8+ audits including Spearbit and Trail of Bits, $1M+ Immunefi bounty. Held back by the Pool Delegate model, which adds human discretion that purely algorithmic protocols do not carry, a non-trivial cross-contract surface (WithdrawalManager + LoanManager + PoolDelegate), a v2 architecture that is only about three years old, and custody consolidated under Maple Labs across the family’s control addresses. |
| Redemption | 7.5 | Newly visible on this page — see the note below. Two permissionless paths, no KYC on either: DEX aggregator routing at single-digit-to-low-double-digit bps in the base case, and a direct WithdrawalManager queue that Maple claims averages under five minutes in normal markets. That combination — permissionless on both paths, no minimum, multi-chain — is materially better than KYC-gated RWA peers. The score sits at 7.5 rather than higher because the queue is the binding path under stress, and queue speed depends on free USDC against outstanding loan principal. |
| Underlying | 7.0 | The collateral is the loan book: roughly 75–80% Loans (institutional credit against BTC/XRP/cbBTC/HYPE at 125–333% funding-time collateralization) plus roughly 20–25% Liquidity layer (pool-owned PYUSD/USTB/AMM positions, at par with the underlying rather than overcollateralized). Loans-only collateral ratio typically runs 145–170% and was above that band at the August 2026 check. Two things hold this below an 8: the collateral is held off-chain by custodians — the contract neither holds nor prices it — and Maple’s own public loan data has a documented reporting artifact on a material share of the book (see Backing and how verifiable it is). Pool Delegate discretion is real but is priced under Contract & Admin and Issuer, not here. |
| Liquidity | 7.5 | $1B+ pool; aggregator route single-digit-to-low-double-digit bps in the base case (live tiers); permissionless mint/redeem at the vault layer (no KYC gating) is a meaningful advantage over RWA peers like KYC-gated thBILL — anyone retail or institutional can enter/exit at NAV without slippage on the primary path. Stress-case binds at pool depth (queue-bound), but the base case is excellent. |
| Issuer | 5.5 | Maple Labs Cayman, doxxed team, 8+ audits, $1M bounty. ~3-year clean record on the Syrup product line. This axis scores the entity, so it is deliberately identical to syrupUSDT — Maple curates both. Per-pool differences belong under Contract & Admin. |
| Overall | 6.75 | Moderate-to-low risk |
A note on the axes, because they changed in August 2026. This report used to score on the stablecoin rubric — peg mechanism and backing. That was the wrong rubric for what syrupUSDC actually is. It is not a pegged dollar; it is a share in a lending vault whose price tracks NAV, and it is now scored on the same six axes as every other vault-share report on this site.
The specific cost of the old rubric was that the stablecoin rubric has no redemption axis. For a vault that runs roughly 97% deployed into loans, “can I get my money out, and how fast?” is not a secondary question — it is the whole question. The material was always in this report (the two exit paths, the sub-five-minute claim, the honest stress caveat about queue behaviour), but it never reached a score, so a reader skimming the table saw a peg-mechanism number where the most important number should have been. Redemption 7.5 is not a new judgement — it is a number that existed internally and was never shown here. Nothing was rescored to fit the new table: every axis that existed on both sides already agreed.
The old backing axis has become Underlying, which is the axis that owns collateral quality. It also now carries the question backing used to answer — is the collateral actually verifiable? — because there is no longer a separate dial for it.
Who it’s for
DeFi-comfortable users who want yield well above stablecoin-savings rates and are comfortable with the trade-offs: real institutional credit risk on the loan book (overcollateralized but discretionary at the underwriting layer), Maple-controlled custody on the Liquidity layer, and no bankruptcy remoteness. Good fit for sizing where low-bps base-case exit slippage is acceptable and a queue-tolerant fallback is OK during stress windows.
Who should avoid
- Anyone needing instant guaranteed exit at NAV regardless of market conditions — DEX-route slippage is bounded in normal markets but can widen during stress
- Position sizing above the low-MM range without explicit queue tolerance — for institutional sizes during correlated outflows, the redemption queue is the binding path
What to watch
- Peg discount on DEX pools. A widening (>50 bps) discount on syrupUSDC vs NAV is the leading indicator that aggregator routing is deteriorating and queue exit is becoming binding.
- Liquidity layer issuer events. ~20–25% of the pool is the Liquidity layer, across the Paxos (PYUSD) / Superstate (USTB) / Circle-Tether (AMM) issuer menu; the delegate rotates the mix under discretion and any one issuer can come to dominate (recently ~90%+ Paxos PYUSD). These positions don’t have a crypto-cycle buffer — the stress is the dominant issuer’s own peg/issuer event. Live per-issuer mix on the dashboard.
- Loan collateral approaching par. The health signal to watch is how close a loan’s current collateral sits to par (100%) — the point where it stops covering the loan — not how far it has drifted below its funding-time level (a loan can be “below init” and still hugely overcollateralized, so that on its own tells you little). Watch for the tightest loan nearing par, and for many crypto-collateralized loans compressing toward par together (a correlated-drawdown warning). The delegate has the right but not the obligation to call these loans; current counts are shown live on the dashboard.
- Multi-chain bridge surface. syrupUSDC bridges via Chainlink CCIP (different attack class from the April 2026 LayerZero OFT incidents). Still verify per-chain CCT pool addresses on Chainlink’s CCIP directory and check per-chain pool depth before sizing on non-Ethereum venues — depth thins fast outside Ethereum.
- Pool Delegate roster changes. Delegate identity is the structural credit-judgment trust assumption.
- Cross-pool concentration if you also hold syrupUSDT. Maple’s materially smaller sibling pool shares the same Pool Delegate firm and several of its largest loan-book borrowers. On a Loans-only basis the top-3 cross-pool borrowers run persistently well above the 10%-per-counterparty institutional norm, with the single largest near ~1/5 of the family loan book. Holding both syrupUSDC and syrupUSDT does not diversify credit risk for those borrowers — it concentrates it. If you allocate across both, compute combined exposure to each borrower rather than treating the pools as independent; live figures on the dashboard.
Live dashboard
A live monitoring view is available at tidresearch.com/dashboards/?asset=syrupusdc — refreshed hourly from on-chain reads. It shows pool backing, separate Loan Book and Liquidity Layer health panels (status flags + distance to par / current collateralization + custody addresses), peg deviation, and exit-liquidity tiers. The signals worth watching listed above all map to specific panels there.
A note on Maple’s history
Maple v1 (2021–2022) lent on an undercollateralized basis and lost LPs ~$50M+ during the 2022 credit cycle (Orthogonal Trading default, M11 Credit / Babel Finance defaults). The Syrup product line is Maple’s structural response — overcollateralized loans on the third-party credit book, vetted Pool Delegates, active margin calls. Same legal entity (Maple Labs, Cayman Islands), same broader team. The v2 Syrup product has run cleanly for ~3 years through May 2026. Some institutional risk frameworks dock entity-track-record regardless of product changes; this report treats it as background context rather than a leading risk factor.
Revision history
- 2026-08-18 — moved onto the correct scoring rubric; no score changed. This report was filed as a stablecoin and therefore rendered the stablecoin axes (peg mechanism, backing). syrupUSDC is a vault share, and the practical cost of the mistake was that the stablecoin rubric has no redemption axis — so for a vault roughly 97% deployed into loans, the binding retail question was discussed in prose but never scored. The page now carries Stability 8.5 / Contract & Admin 6.5 / Redemption 7.5 / Underlying 7.0 / Liquidity 7.5 / Issuer 5.5, Overall unchanged at 6.75. Stability, Contract & Admin and Redemption are newly visible, not newly assigned — they existed internally and every axis published on both sides already agreed. Backing 7.0 became Underlying 7.0, which now also carries the verifiability question backing used to answer. Figures refreshed to the August 2026 check, and the credit read is reassuring: loans-only collateral ratio 175%, above its 145–170% band, with pool collateral ratio 100% and zero unrealized losses. Also recorded: Maple’s public loan-data artifact has grown to about 39% of the book while remaining an artifact, and one loan under 3% of the book sits in the 100–120% collateralization range.
For institutional-grade risk analysis — Pool Delegate identities, specific contract addresses, custody EOA inventory, on-chain monitoring patterns — the institutional version is available.