Summary
This proposal seeks to launch a new, dedicated isolated lending market on Neverland for AUSD and Pendle Principal Tokens redeemable in AUSD (PT-AUSD) on Monad. The market is a second, fully self-contained lending pool deployment with its own provider, pool, oracle, and risk surface, entirely separate from Neverland’s canonical pool.
In this market, AUSD is the only borrowable asset, and each approved PT-AUSD maturity is listed as collateral only, with borrowing and flash loans disabled. PT-AUSD is priced through a Neverland-owned deterministic discount oracle that converges to par at maturity and never depends on a Pendle AMM, TWAP, or PT liquidity. The design lets holders of fixed-yield PT-AUSD borrow AUSD against their position (the classic Pendle “PT looping” strategy), which drives AUSD borrow demand and feeds the veDUST flywheel.
This RFC is forward-looking and gated: it activates only when Pendle deploys AUSD Principal Tokens on Monad, and after the oracle and risk parameters have been deployed, reviewed, and signed off. Because the pool is isolated and holds only AUSD and its PT maturities, it runs at high-efficiency parameters directly and needs no separate E-Mode: the pool itself plays the role a dedicated E-Mode would in a mixed pool, and a user participates simply by supplying to it.
Opportunity Statement
Pendle splits a yield-bearing asset into a yield component and a Principal Token (PT) that can be redeemed 1:1 for its accounting asset at maturity. Because the yield has been stripped out, PT trades at a discount before maturity and appreciates toward par. That appreciation is a fixed yield locked in at purchase (Pendle docs).
When Pendle launches AUSD Principal Tokens on Monad, Neverland can be a venue where that fixed-yield, dollar-denominated position is put to work: a holder supplies PT-AUSD as collateral, borrows AUSD against it, and can recycle that AUSD to scale the position. This is the looping strategy that has made Pendle PT collateral a large fixed-yield category in DeFi lending. For Neverland it delivers:
-
New blue-chip borrow demand for AUSD, one of Monad’s ecosystem primary stablecoin, generating borrow-side revenue.
-
A new asset class, fixed-yield stable collateral, that is dollar-correlated and therefore eligible for high capital efficiency.
-
A positive flywheel: better-utilized AUSD liquidity → more revenue → more value to veDUST and LP incentives.
The value to Neverland is structurally simple: 1 PT-AUSD is worth a growing amount of AUSD that converges to exactly 1 AUSD at maturity, and AUSD is worth about $1. It is a slow-appreciating, deterministically-priced dollar instrument rather than a volatile one.
We propose a dedicated isolated market rather than adding PT-AUSD to Neverland’s canonical pool. Pairing AUSD and PT-AUSD at a high loan-to-value would require a dedicated E-Mode, but AUSD already sits in the canonical stablecoin E-Mode (alongside assets such as USDC and USDT0), and the protocol’s reserve model lets each reserve belong to only one E-Mode category. A separate market resolves this cleanly: because it holds only AUSD and its PT maturities, it functions as a purpose-built high-efficiency pool in its own right and needs no E-Mode configured inside it (see Alternatives Considered).
Low Level Details
Market Architecture
The market is a second, independent lending market with its own provider, registered in the shared provider registry alongside the canonical pool. It launches AUSD-only and adds each PT-AUSD maturity as a separate, individually-approved listing.
It is deployed as fully independent infrastructure: its own provider, pool, configurator, access-control manager, oracle, and data provider, reserve tokens, rate strategies, data and UI helpers, descriptor, and subgraph. A canonical-pool upgrade or oracle change does not affect this market, and vice versa. The two pools share no provider, oracle, reserve token, or E-Mode category. The only intentionally shared infrastructure is the immutable shared core contract code, the provider registry, and the incentives controller.
Assets
AUSD (the borrowable and accounting asset). AUSD is Agora’s fully-collateralized U.S. digital dollar. Its reserves (cash, overnight repo and reverse-repo, and short-term U.S. Treasuries) are managed by VanEck ($100B+ AUM) and custodied by State Street ($4T+ AUM), held in a bankruptcy-remote reserve fund legally separated from Agora’s corporate assets (Agora, overview). AUSD is one of the Monad’s primary stablecoins and is already listed and trusted in Neverland’s canonical market.
PT-AUSD (the collateral). A Pendle Principal Token whose accounting (redemption) asset is AUSD. Before maturity it trades at a discount to AUSD; at maturity it is redeemable 1:1 for AUSD (Pendle docs). Each maturity (for example a June-2026 expiry) is a distinct token listed as its own reserve. The token, decimals, maturity timestamp, and Pendle market metadata are fixed inputs supplied at listing time, once Pendle is live on Monad.
Collateral Policy and Risk Controls
PT-AUSD’s value to Neverland is as productive collateral, not as a borrowable. Every PT-AUSD reserve is listed with borrowing disabled, flash loans disabled, borrow cap zero, and is never assignable to a borrow path. AUSD is configured as borrow-side liquidity, so the only collateral in the pool is PT-AUSD.
Because the market is fully isolated, the worst-case blast radius of any PT-specific issue is contained to this market’s AUSD liquidity. Standard protocol controls (freeze, pause, supply-cap-to-zero, and collateral-disable) are wired and routed through the market’s admin, the Governance Timelock (24H delay) for upgradability and Risk Timelock (1H delay) for configurations.
Oracle and Pricing
Pricing is the central risk control, and the design deliberately removes market-liquidity and manipulation risk from the collateral valuation.
Neverland prices PT-AUSD with a purpose-built, Neverland-owned discount oracle adapter, not a raw Pendle feed. It implements the deterministic linear-discount model Pendle itself recommends for lending integrations, the approach used for PT collateral across established lending integrations, because it converges to par at maturity and avoids AMM-TWAP manipulation and liquidity-liveness dependencies (Pendle: Linear Discount Oracle):
PT-AUSD / USD = AUSD/USD × (1 − linear discount to maturity)
Key properties:
-
No Pendle AMM, TWAP, or PY-index dependency. The adapter reads only the AUSD/USD feed and the PT’s on-chain maturity. It works on day one with zero PT liquidity, and liquidations never depend on Pendle market liveness.
-
Deterministic and conservative. The discount decays linearly to zero at maturity (PT → par), and the discount line is calibrated to stay below expected market price, under-pricing collateral, which is the safe direction for a lending market.
-
Standards-compatible and guarded. It exposes a standard Chainlink-style USD price, with staleness and validity checks, a governed annual discount rate bounded by an immutable ceiling, and timelock owners. If the AUSD feed is stale or invalid, it under-prices rather than serving a bad value.
-
Per-maturity, reviewed. A separate oracle adapter is deployed, verified, and risk-reviewed for each maturity before that maturity can be listed.
The oracle adapter intentionally treats AUSD-depeg and PY-index behavior as out of scope of the price path. Risk must explicitly accept that simplification and calibrate caps, LTV, and liquidation threshold to the chosen discount line.
Proposed Market Parameters
Because the pool holds only AUSD and PT-AUSD maturities, all of them dollar-correlated, it runs at high-efficiency parameters directly. There is no separate E-Mode: the values below are the pool’s only parameters, and a user opts in simply by supplying. PT-AUSD is collateral-only, and AUSD is the sole borrowable, configured as borrow-side liquidity rather than collateral.
| Parameter | AUSD (borrowable) | PT-AUSD (collateral, per maturity) |
|---|---|---|
| Borrowable | Yes | No |
| Usable as collateral | No | Yes |
| LTV | — | 91.5% |
| Liquidation Threshold | — | 93.5% |
| Liquidation Bonus | — | 2.5% |
| Liquidation Protocol Fee | 10% | 10% |
| Reserve Factor | 50% | 0% |
| Supply Cap | Dynamic | Conservative, Dynamic |
| Borrow Cap | Dynamic | 0 |
| Oracle | Chainlink AUSD/USD | PT discount adapter (USD) |
These collateral values are in line with how Pendle PT collateral is parameterized across established lending markets (roughly 87% to 91% LTV, 89% to 93.4% liquidation threshold, and a 2.3% to 2.8% liquidation bonus).
Two points of calibration:
-
A high liquidation threshold is justified because the deterministic discount oracle under-prices PT relative to its 1:1 redemption value, giving the protocol a built-in cushion.
-
The liquidation bonus is held at 2.5%, not lower, because PT secondary liquidity on Monad will be shallow at launch. A liquidator must be able to offload PT or hold it to maturity profitably, and the discount cushion shrinks toward zero as maturity approaches. As depth builds, and if a maturity-aware or dynamic discount is later adopted, governance can tighten the bonus and lift LTV further.
Keeping AUSD as borrow-side liquidity (collateral disabled) means the pool’s only collateral is PT-AUSD, which holds the risk surface to exactly the intended pairing.
Launch Gating and Sequencing
This market is treated as not launchable until each gate has direct evidence. The sequence is deliberately staged so nothing is exposed prematurely:
-
Deploy the isolated market, AUSD-only. Separate provider, pool, oracle, access-control manager, data provider, and AUSD reserve, with the AUSD feed reviewed (contract, decimals, a positive and fresh price reading, and the fallback decision). The canonical pool is provably unchanged.
-
When Pendle lists AUSD PTs on Monad, lock the per-maturity inputs: PT token address, decimals, maturity timestamp, and Pendle market metadata.
-
Deploy and review the PT/USD discount oracle for that maturity, verify it on-chain, hand ownership to the risk admin, and obtain risk sign-off on the discount rate, ceiling, and caps.
-
List the PT maturity as collateral-only at the pool’s parameters, and bind the descriptor to the exact reviewed oracle. Tooling rejects the listing until the maturity’s metadata and reviewed oracle adapter are marked ready.
-
Hand off a checksummed market descriptor to frontend, indexer, and monitoring. Nothing is published to users until that handoff and its preflight pass against live state.
Each PT maturity repeats steps 2 through 5 individually. Borrowing of AUSD funds the strategy, and PT borrowing is never enabled.
Risk Considerations
-
Maturity risk. PT pricing and behavior are defined at and after maturity (PT → par). Listings enforce a minimum maturity buffer, and a documented lifecycle (freeze, cap-to-zero, or collateral-disable) governs maturities approaching expiry.
-
Discount-rate risk. The oracle’s discount line must remain conservative versus current and expected lower-bound market prices. It is a governed, ceiling-bounded risk parameter reviewed per maturity.
-
Depeg and PY-index simplification. The price path intentionally ignores AUSD-depeg and PY-index effects for determinism. Risk explicitly accepts this and calibrates caps and LTV accordingly. AUSD’s institutional backing and bankruptcy-remote reserves bound this risk.
-
Liquidation under stress. PT collateral is liquidated against the deterministic discount price (no Pendle liquidity dependency), and the liquidation bonus and protocol fee are set to keep liquidations economic.
-
Containment. Full market isolation, conservative caps, non-borrowable PT, disabled stable borrowing, and independent freeze, pause, and cap controls bound the worst case to this market.
Alternatives Considered
1. Add PT-AUSD to the canonical Neverland pool. Rejected. Reaching a high loan-to-value pairing there would require a dedicated AUSD/PT-AUSD E-Mode, but AUSD already occupies the canonical stablecoin E-Mode and a reserve can belong to only one E-Mode category. It would also mix Pendle maturity and discount risk into the canonical pool’s bad-debt surface. A separate market is the only way to give AUSD a fresh high-efficiency pairing without touching its canonical reserve.
2. A single shared “Neverland Pendle” market hosting all base-and-PT pairs. This is closer to how Pendle PTs are commonly listed inside a single shared pool on Aave and other protocols following this architecture, and is leaner operationally (one provider, oracle, subgraph, and monitoring target). We rejected it in favor of one isolated market per underlying for two reasons specific to how Neverland’s lending protocol handles E-Mode:
-
User experience. A shared pool would need a separate E-Mode per pair to reach high loan-to-value, and the protocol lets a wallet select only one E-Mode at a time, so a user could run only one PT loop per wallet. With one isolated pool per underlying, each pool runs at high parameters directly with no E-Mode to select, so a single wallet can participate in several Pendle pools at once at full efficiency.
-
Risk isolation. Separate pools provide hard bad-debt isolation between underlyings. A problem in one underlying cannot socialize losses onto another’s suppliers, which a shared pool cannot offer.
The trade-off is more per-underlying infrastructure and no shared borrow liquidity across markets.
3. Use Pendle’s oracle directly, or a TWAP oracle. Rejected as the default. Pendle’s linear-discount oracle returns a price in the accounting asset, not USD, and wiring it directly as the price source would misprice collateral. A TWAP or market oracle introduces AMM-liquidity, manipulation-cost, and market-liveness dependencies into liquidations, which is unacceptable for a nascent Monad PT market. We instead use a Neverland-owned adapter that returns a guarded, Chainlink-compatible USD answer derived deterministically from the AUSD feed and the PT maturity. The TWAP path would only be revisited if a separate risk package proved liquidity depth, observation cardinality, and manipulation cost.
4. More aggressive risk parameters (for example a sub-1% liquidation bonus). Considered and not proposed for launch. Although the conservative discount oracle supports a high liquidation threshold, an unusually low liquidation bonus would leave PT positions un-liquidatable on shallow day-one Monad liquidity (a liquidator must offload PT or hold to maturity), and the discount cushion vanishes as maturity approaches. We propose a bonus aligned with established PT listings (about 2.5%), to be tightened by governance as depth builds or once a maturity-aware parameter schedule is adopted.
Open Questions
We are explicitly seeking community and risk-partner input on the following design choices before this proposal advances to a Governance Vote. Where we state a preference, it is a proposal, not a settled decision.
1. Market topology: one isolated market per Pendle underlying (proposed) versus a single shared Pendle market. The reasoning is summarized in Alternatives Considered (2): separate markets give each wallet concurrent multi-pair loops and hard bad-debt isolation, at the cost of more per-underlying infrastructure. Do we confirm per-underlying isolated markets, or prefer the shared-pool model?
2. Oracle simplification (depeg and PY-index). The adapter prices PT as AUSD/USD × (1 − linear discount), intentionally ignoring AUSD-depeg and PY-index effects. Does risk accept this simplification for an institutionally-backed, bankruptcy-remote stablecoin, or should the price path add a PY-index or depeg term, or a TWAP cross-check?
3. Revenue routing. The market is currently designed and pre-configured to the shared Neverland revenue wallet. The collected revenue will feed the current shared revenue-backed incentives; otherwise an isolated collector is used for separate accounting and revenue management. Shared or isolated collector for this market?
4. Parameters to finalize with in-house and external risk partners (per maturity). The discount line and its immutable ceiling (how conservative versus observed market price); the initial supply cap and ramp schedule; the minimum maturity buffer and the unwind policy (freeze, cap-to-zero, or collateral-disable) as a maturity approaches; the liquidation tooling and routes required before listing, given thin PT secondary liquidity.
Closing Statement
This proposal brings Neverland a fixed-yield, dollar-denominated collateral, PT-AUSD, in a fully isolated market that leaves the canonical pool untouched. Holders keep their locked-in Pendle yield while borrowing AUSD against it and earning DUST, and the AUSD borrow demand this unlocks feeds the veDUST flywheel. The collateral is priced by a deterministic, manipulation-resistant, Neverland oracle that converges to par at maturity, the pool runs at conservative, dollar-correlated parameters, and PT borrowing is structurally impossible.
-
Pendle Documentation, Principal Tokens (PT): https://docs.pendle.finance/pendle-v2/ProtocolMechanics/YieldTokenization/PT
-
Pendle Documentation, Oracle Overview: https://docs.pendle.finance/pendle-v2-dev/Oracles/OracleOverview
-
Pendle Documentation, Linear Discount Oracle (recommended for lending): https://docs.pendle.finance/pendle-v2-dev/Oracles/DeterministicOracles/LinearDiscountOracle
-
Pendle Documentation, Choosing Linear Discount Parameters: https://docs.pendle.finance/pendle-v2-dev/Oracles/DeterministicOracles/ChoosingLinearDiscountParams
-
Agora, AUSD, The Digital Dollar: https://www.agora.finance/product/ausd
-
Neverland Documentation, Lending Model, Incentives, and the Neverland Flywheel: https://docs.neverland.money/