On Oct. 8, $67.4 million of collateral sitting on Aave V3's Monad market stops earning yield. Whether it stays on-chain is the only question that matters this week.
Those are Pendle principal tokens โ PT-AUSD-8OCT2026. LlamaRisk tallied 67.4 million of them supplied as collateral as of Oct. 2. When they mature, each redeems for exactly one AUSD. The fixed-yield appreciation ends there. The debt stacked on top of them doesn't.
This is the first real stress test of a loop that quietly became DeFi's most elegant capital machine. Lock a return through Pendle. Use the position as Aave collateral. Borrow stablecoins against it. Roll into the next maturity. As DeFi researcher Andree put it: "Fixed yield becomes collateral. Collateral creates credit. Then the next maturity keeps the cycle moving."
Speed isn't the pulse of the market. Structure is.
Context: how a $20M market became a $67M one
The October market didn't start big.
Aave launched the PT-AUSD collateral market with a 20 million-token supply cap. Users filled it by late August. LlamaRisk recommended a bump to 40 million. That filled within days. Another recommendation pushed the cap to 80 million. By Oct. 2, 67.4 million PT was live as collateral.
That cap trajectory is the story. It shows a market that keeps outrunning its own risk limits โ and risk stewards willing to chase it. TokenLogic has already proposed listing a Dec. 17 AUSD principal token, with a fresh 20 million initial cap. LlamaRisk wants to start at 30 million. Both are less than half the expiring position, but the October precedent says that ceiling is a starting line, not a wall. LlamaRisk explicitly named the December PT the rollover destination for the October position, estimating as much as 67.4 million of Aave collateral could migrate into it.
The demand underneath is real and accelerating. On Oct. 3, TokenLogic reported active AUSD loans on Aave jumped 113% โ from $4.1 million to $8.7 million in 15 days. User deposits more than doubled to $11.2 million. That's separate from the stablecoin debt raised specifically against PT-AUSD, but both point the same direction: AUSD credit on Monad is heating up fast, and it's pulling fresh capital onto the chain rather than recycling old positions. This isn't a slow migration. It's a 15-day spike, and it landed right before a 67-million-token maturity.
Core: the mechanics that actually carry risk
Here's where it gets interesting โ and where most coverage stops short.
In an Aug. 31 assessment, LlamaRisk found that the 18 largest PT suppliers all carried debt. Primarily USDC. Some GHO, USDe, USDT0. Their median health factor was 1.02.
Read that again. 1.02. That is a whisker from liquidation.
It's survivable because both sides of the book are dollar-denominated. Collateral and debt move together. Borrowers can run high loan-to-value positions with far less directional risk than crypto-backed leverage. That's the whole trick โ and it's a good one. But it only works while the collateral keeps redeeming near par and the borrow rate stays below the fixed yield. The moment either leg slips, a 1.02 health factor has no cushion left.
Maturity changes the geometry. Expiry itself doesn't trigger liquidation. Borrowers can redeem PT for AUSD after Oct. 8, repay loans, or post other collateral. The catch: a user with debt against the October PT can't necessarily withdraw the collateral until the position stays adequately covered. So the exit isn't free. It's a refinancing decision, and it lands all at once.
Based on my years tracking these markets, this is the moment where spread math beats narrative. And the spread is already thinning.
LlamaRisk put the December PT's implied yield at 5.64% on Oct. 2. A temporary one-percentage-point campaign incentive lifted the effective rate to 6.64%. Against borrowing costs โ mUSD at 4.28%, GHO at 4.64%, USDT0 at 5.10%, USDC at 6.09% โ there's positive carry before transaction costs and price impact. Against USDe's 6.82% borrow rate, there isn't.
Those spreads move fast. Aave borrow rates track utilization. PT yields swing as traders buy and sell. The incentive propping up December returns is temporary and expires on its own schedule. A carry trade this tight doesn't need a shock to break. It needs a shrug.
Contrarian: the rollover story has a liquidity hole
Everyone is framing December as the rollover destination. Almost nobody is pricing the gap.
As of Oct. 2, the December Pendle pool held just $1.61 million of liquidity, 904,717 PT outstanding, and $44,000 of trading volume since deployment. That's not tens of millions. That's a rounding error next to the October position.
The bull case says it doesn't matter: Pendle users can mint new PT by splitting yield-bearing AUSD into principal and yield tokens, so pool liquidity isn't a hard cap on how much collateral can exist. True. But large-scale migration still moves execution prices and compresses the fixed return available to buyers. The deeper the migration, the worse the entry โ which is exactly the wrong incentive for a market trying to absorb 67 million tokens.
So the "capital never leaves DeFi" framing is hype. What actually keeps capital in place is a positive spread, and a positive spread is a rates trade, not a structural moat. Regulation doesn't create that spread. Temporary incentives that expire don't defend it either.
We didn't get a clean test of cross-maturity rollover until now. And the first one arrives with the replacement market at roughly 2% of the expiring market's size.
Takeaway
Watch the first days after Oct. 8. If December PT fills its initial cap as fast as October did, Aave's risk stewards face pressure to raise the limit again โ and their answer will hinge on Pendle liquidity, borrower health, and whether the new market builds real depth.
Exchange leads see the wave before it breaks. This one is already curling.
For borrowers, the clock is simpler: repay at maturity, replace the collateral, or grab space in December while the spread still pays. Three doors. One closes Friday.