Aave's Newest Markets Have a 1% Cushion and a 30-Day Exit

A health factor of 1.01 isn't a margin of safety. It's a rounding error. And two of the largest syrupUSDC borrowers on Aave's Arc market were sitting at exactly 1.01 and 1.02 on Oct. 9, per LlamaRisk's risk snapshots from that day.

Those two positions held roughly 97% of all supplied syrupUSDC on Arc. Not 97% of borrows. Ninety-seven percent of the collateral side. That's not a market. That's a pair of whales sharing a pool.

Here's the twist. The asset backing those loans can take up to 30 days to redeem. Liquidation doesn't wait 30 days.

Two markets, one trap

LlamaRisk published risk reviews of two Aave deployments on Oct. 9. The first covers Monad, where borrowers are posting PT-AUSD-17DEC2026 as collateral. That's a Pendle principal token representing a claim on AUSD at its Dec. 17 maturity. The second covers Arc, where the collateral is syrupUSDC, a bridged share in Maple's Ethereum yield vault.

Different chains, different assets, same structural flaw. The oracle sees one price. The exit sees another. And on both markets, the gap between those two numbers is where liquidators either make money or walk away.

Start with Monad. The top PT-AUSD suppliers showed health factors between 1.01 and 1.18, with a median of 1.03. Their debt was mostly USDC, some USDT0. The 30 million PT supply cap was fully used, and LlamaRisk recommended doubling it to 60 million.

Now look at what happens if one of those borrowers tips below 1. A liquidator has to repay the debt in stablecoins, take the PT, and then convert it back to cash. Before December maturity, Pendle's documented route sells PT into SY, its standardized yield wrapper, then redeems SY into a supported output token. That sale runs through a pool that was 47% PT and 53% SY on Oct. 9. A big sale draws entirely from the SY side, which means price impact scales with size. Nobody knows the exit price until they try it.

The numbers tell the story. The liquidation threshold on this reserve is 95%. The stablecoin E-mode bonus is 2.62%. So a liquidator's entire profit margin has to cover slippage on a market sale of an asset with one meaningful venue. That's tight. Frankly, it's tighter than the health factor suggests.

The exit is the real risk

Arc is a different flavor of the same problem. There's one local Uniswap V4 syrupUSDC/USDC venue, and LlamaRisk's September assessment found its proceeds saturated near $500,000 as the USDC side ran dry. Half a million dollars. Against a Maple Spoke add cap heading from 25 million to 50 million shares.

Do you see the mismatch? The cap is expanding. The exit liquidity isn't.

If a liquidator wants an Ethereum redemption instead, the route is a bridge transfer that takes two to five minutes under normal conditions and moves roughly $10 million an hour, followed by a Maple withdrawal queue that runs first-in, first-out. Most withdrawals clear in under 24 hours. Some take up to 30 days. The liquidator has to front the USDC to repay Aave before any of that money comes back.

So the real question isn't whether these positions are close to liquidation. It's who's willing to finance a multi-day unwind for a bonus capped at 4%.

From a risk perspective, the oracle design makes this worse. The December PT uses a linear discount oracle on AUSD/USD, which follows a predictable path toward maturity no matter what the AMM is doing. That's a clean valuation for Aave's accounting. It's also a valuation that can diverge from reality for weeks at a time. A liquidator sells at market. The protocol marks at curve. Those two lines don't have to meet.

Arc has its own version of this. Collateral value comes from Chainlink's syrupUSDC/USDC exchange rate, which reflects the Ethereum vault's exit value. But local trading depth is what determines how much of that value a sale can actually realize. One number is a promise. The other is a receipt. They're not the same thing.

And here's the part that should get lenders' attention. The Maple Spoke on Arc draws from the same Core Hub USDC reserve as the Main Spoke. Arc had 143.45 million USDC added to the Hub, 83.82 million drawn, and 59.63 million available on Oct. 9. That available balance is shared solvency. Spoke-level caps constrain exposure. The Hub is where it all pools. Monad runs on V3 with different boundaries, so the two markets don't cross-contaminate. But within Arc, they absolutely do.

What the street is missing

Everyone's watching health factors. Fewer people are watching the exit route, and the exit route is where the loss actually shows up. A position at 1.03 doesn't liquidate in a vacuum. It liquidates into whatever buyers happen to be standing there, at whatever price they're willing to pay, on whatever timeline the redemption queue allows.

So here's my read. The cap increases being proposed, 30 million to 60 million PT on Monad, 23 million to 46 million USDC draw on Arc, are being pushed while the exit depth underneath the existing positions hasn't been stress-tested at size. That's backwards. Grow the market after you've proven the door works, not before.

Who wins here? Borrowers, as long as nothing moves. Correlated collateral and debt lets them run higher tap into with lower health factors, and Aave itself notes that lower readings can be appropriate when assets are correlated. Fair enough on paper.

Who loses? Lenders and liquidators if a position tips and the unwind turns out to cost more than the bonus pays. That's when you find out whether the oracle was a valuation or a wish.

The takeaway is simple. On these markets, collateral value and collateral liquidity are two separate things, and only one of them shows up in the health factor. If you're lending against syrupUSDC or PT-AUSD, the number you need isn't 1.01. It's the size of the exit, the cost of the conversion, and who's holding the bag while the redemption queue clears.

Check that before you check the cushion. The cushion is smaller than it looks.