Why Pooled Lending Markets Matter: A Case Study in Building on Aave

Aave V3's pooled market design is under attack. The rsETH incident has revived a familiar argument: that pooled, multi-collateral lending markets are fundamentally unsafe, and that the future of on-chain lending belongs entirely to isolated markets. We disagree — not because Aave is beyond criticism, but because the critique is aimed at the wrong target.

The rsETH episode was a collateral approval failure, not a design failure. Aave was, in effect, serving a customer who wanted to max-lever a liquid staking token whose economic value was questionable from the start. Once approved, that aggressive looping demand became the dominant use case for rsETH on Aave — and when the asset broke, the leverage multiplied the damage across the entire pool, including for users who had never heard of rsETH and were simply lending USDC or ETH.

We have no problem with rsETH looping. We just don't see why it should have been happening in a pooled market in the first place. That's exactly what isolated markets are for.

But the inverse is also true: there are use cases that only a pooled market can serve well, and the way Bitcoin Dollar's USD Vault is built is one of them.

What Pooled Markets Do That Isolated Markets Can't

A pooled market like Aave V3 lets a single deposit do two jobs at once. When you supply wBTC to Aave, that wBTC is collateral for whatever you want to borrow against it — and it's also lendable supply that earns interest from anyone in the broader pool who wants to borrow wBTC. Your capital is never sitting idle. You get leverage on your own position and yield on the same coins, simultaneously.

This is impossible in an isolated market by construction. An isolated USDC/wBTC market could let someone borrow wBTC against USDC, sure. But the wBTC supplying that market would have to be parked there exclusively — it couldn't simultaneously serve as collateral for that supplier's own borrowing somewhere else. The capital fragments. Every isolated pair you participate in costs you the optionality of using that asset as collateral elsewhere.

The result is a structural asymmetry that benefits everyone using the pooled market correctly. wBTC on Aave has many natural suppliers — Bitcoin holders who want USD liquidity against their BTC — and very few natural borrowers. So wBTC borrow rates stay low, wBTC suppliers earn a small yield on collateral they were going to deposit anyway, and protocols that genuinely need to borrow wBTC at scale can do so cheaply and reliably. None of those properties survive in a fragmented isolated-market world.

This is the part of Aave worth defending. Not every collateral approval decision they've made — but the architecture itself, which is irreplaceable infrastructure for any protocol that needs deep blue-chip liquidity.

How the USD Vault Uses Aave the Way It's Meant to Be Used

Bitcoin Dollar's USD Vault gives depositors USD-denominated yield by depositing sBTCD as collateral on Morpho and borrowing wBTC against it. That structure needs wBTC supply on Morpho. While sBTCD is still earning credibility in the market, organic wBTC suppliers won't show up at scale on day one — so the protocol manufactures the supply itself, in three different ways, depending on which part of the system has the spare capacity.

Mechanism 1: Direct supply from the BTCD Portfolio's BTC sleeve. The simplest version. The portfolio holds wBTC as part of its 50% BTC exposure; some of that wBTC is supplied directly to the Morpho market and earns lending yield while the vault borrows against it. No Aave involved. The portfolio's BTC sleeve becomes mildly more productive.

Mechanism 2: Carry trade from the BTCD Portfolio's USD sleeve. Here the portfolio uses USDC from its 50% USD sleeve to run an Aave carry. It deposits USDC on Aave, borrows wBTC against it at Aave's structurally low wBTC rate, and supplies that wBTC to the Morpho market. The wBTC borrowed from Aave is matched by wBTC supplied to Morpho, so the position is BTC-flat — the portfolio's omega is unchanged. But the same dollars are now earning USDC supply yield on Aave and the spread between Aave's wBTC borrow rate and Morpho's wBTC lend rate. It's pure USD exposure with stacked yield.

It's worth being explicit about what this means for Aave's existing wBTC suppliers: they have no exposure to sBTCD. None. We are not pledging sBTCD to Aave — we are pledging USDC. We borrow wBTC against that USDC at conservative LTV. If we mismanage the position, Aave liquidates our USDC collateral to make wBTC lenders whole, exactly as the protocol is designed to do. The sBTCD risk lives entirely inside the isolated Morpho market, where every participant has explicitly opted into it. This is the architectural point: blue-chip collateral on Aave, novel collateral on Morpho, no risk bleed across the boundary.

Mechanism 3: The vault's self-balancing carry. This is the one that makes the architecture click.

The USD Vault wants every dollar of depositor capital to be in the sBTCD loop — that's where the highest yield is. But the loop only works if there's wBTC available to borrow on Morpho. If the loop saturates available wBTC supply, the vault has a choice: turn away deposits, or create more capacity itself.

It creates more. With the marginal dollar of depositor capital, the vault deposits USDC on Aave, borrows wBTC against it, and supplies that wBTC into its own Morpho market — where the loop then borrows it. The vault has effectively become its own counterparty: one leg of the vault demands wBTC, another leg manufactures it.

Depositors in this scenario earn a weighted average across the two strategies. Capital running in the sBTCD loop earns the leveraged sBTCD yield. Capital running the Aave carry earns USDC supply yield plus the wBTC borrow/lend spread. As loop demand grows, more capital flows into the carry. As organic wBTC suppliers arrive on Morpho — which they will, as sBTCD matures — the carry naturally unwinds and capital migrates back into the loop. The vault breathes.

The elegance is that none of this requires governance intervention or active management. The vault simply allocates marginal capital to whichever leg has spare capacity, and the system rebalances itself based on where wBTC supply is binding. The Aave pooled market is what makes it possible: only a market with deep, durable, low-rate wBTC borrowing can host one side of this trade reliably. An isolated USDC/wBTC market could in principle exist, but it would be a fraction of the size, the wBTC suppliers would lose the ability to use that wBTC as collateral elsewhere, and the borrow rates wouldn't be structurally low — they'd reflect the narrow demand of whatever niche the isolated market served.

The Real Lesson from rsETH

The rsETH episode is a useful reminder that pooled markets are only as safe as their most aggressive collateral. A blue-chip-only pooled market is a public good. A pooled market that lists every long-tail asset that asks to be listed inherits the risk profile of the worst asset in the set, because losses on any one collateral propagate through the shared pool.

The right architecture is layered. Pooled markets handle wholesale liquidity for blue chips — wBTC, cbBTC, tBTC, ETH, stETH, rETH, major stablecoins — where the use cases are diverse, the suppliers are many, and the borrowers benefit from the cross-subsidy of collateral-only depositors. Isolated markets handle everything else: long-tail collateral, aggressive looping strategies, anything where the use case is concentrated enough that fragmenting capital is acceptable in exchange for ringfenced risk.

Bitcoin Dollar uses both, in the roles each is best suited for. The vault's user-facing position lives in an isolated Morpho market because sBTCD is new and its risk should not bleed into anyone else's portfolio. The wBTC supply that makes that market function comes, in part, from a carry trade against Aave — because that's where the wBTC liquidity is, and that's exactly the use case Aave V3's pooled market design exists to serve.

The critique of Aave worth taking seriously isn't that pooled markets are broken. It's that Aave needs to be more disciplined about what it lists. The architecture is the part worth keeping.