Crypto-collateral lending as a yield source in diversified pools
Crypto-collateral lending as a yield source in diversified pools
The numbers behind the question
Aave generated $279 million in fees from borrowers across 2024.
Active loans peaked above $29 billion by mid-2025, up from roughly $8 billion at the start of 2024. Morpho, the second-largest on-chain lender, has cleared $170 million in borrower interest over the past year. Across the broader DeFi lending sector, total outstanding loans crossed $51 billion as of mid-2025.
That capital flow has to go somewhere.
The simple version of the story is that lenders supply stablecoins, borrowers post crypto as collateral, and the yield gets paid in real money. The complicated version asks where the yield really comes from, what borrowers are doing with the loans, and what lenders are exposed to when prices move.
For institutional credit funds looking at on-chain yield, crypto-collateral lending sits in an awkward middle. The yields are real and the collateral is over-secured.
The mechanics are also volatile in ways that traditional credit isn't, and most of the textbook risk frameworks don't translate cleanly.
The real question is how much to allocate, in what form, and alongside what other exposures.
How crypto-collateral lending works as a yield source
Strip away the protocol-specific details and the structure is consistent.
A pledger posts crypto worth (say) $1.4M to borrow $1M in stablecoins through a lending protocol. The pledger pays interest on the $1M. The lender (or the pool of lenders) earns that interest. If the crypto drops below a defined threshold, the position gets liquidated and the lender gets repaid from the seized collateral.
What drives the yield is borrower demand for stablecoins against crypto positions.
The biggest demand sources are directional spot positions (borrow USDC against ETH, buy more ETH), trading capital (post BTC for working capital instead of selling), and treasury management (avoid taxable sales by borrowing instead). When crypto prices rise, borrowing demand rises with them, rates climb, and lenders earn more. When crypto prices fall, demand thins, rates compress, and yield drops.
That cyclical pattern matters for portfolio construction.
Crypto-collateral yields are pro-cyclical with crypto prices. They're highest exactly when crypto is rallying (which is when LP risk appetite is also highest) and thinnest when crypto is correcting (which is when LP risk appetite is lowest).
Stablecoin borrow rates on Aave averaged around 4-5% APY through the quieter parts of 2025, briefly spiking to 8-12% during periods of heightened borrowing demand. A book that's only crypto-collateral lending will look different in every regime.
Diversification is the obvious answer.
A pool that mixes crypto-collateral lending with real-world credit (trade receivables, consumer lending, SME loans) smooths the cycle. The crypto leg pays out heavily when DeFi is hot. The real-world legs pay steadily through both regimes.
Together they produce something closer to a stable yield curve than either does alone.
How Rekord uses crypto-collateral lending in its pools
Rekord runs crypto-collateral lending as one of seven deal types that flow into its pools. The other six (Consumer Lending, SMB Acquisition, Luxury Asset, Trade Receivables, Project Financing, CRED) are real-world credit. Crypto-collateral sits alongside them, contributing yield from a fundamentally different driver.
That's by design.
The platform's risk framework treats crypto-collateral lending as a high-volatility, low-default-rate exposure, which is the opposite shape of most real-world credit (lower volatility, higher default rate). Combining the two reduces portfolio variance more than either alone.
The mechanics are tighter than a generic on-chain lender's.
Collateral is hedged at origination through option structures, so a sharp drawdown doesn't push positions into liquidation cascades. Operating LTVs sit well below liquidation thresholds, giving meaningful headroom against intraday moves. Position monitoring runs continuously across the aggregate book.
Liquidations, when they happen, route through pre-arranged execution venues rather than dumping into thin spot markets.
What the pool gets is yield from crypto-collateral demand without the tail risk that pure on-chain lending carries. What individual pledgers get is access to capital with institutional-grade hedging and qualified custody, which most of them couldn't access through a generic protocol.
The allocation across the seven deal types is set by the curator and adjusts over time. Crypto-collateral exposure scales with the cycle: heavier when borrowing demand is strong, lighter when it isn't, with the real-world legs absorbing the difference.
What changes for the LP
For an LP looking at Rekord's pools, crypto-collateral lending is one ingredient in a diversified credit book deliberately constructed to balance pro-cyclical and counter-cyclical yield sources.
The benefit is exposure to crypto-borrowing demand without taking on the volatility profile of pure crypto lending. The cost is that you don't get to choose: the curator decides the mix, and the mix moves with market conditions.
For LPs who want pure crypto-collateral exposure, Aave or Morpho directly is the right venue. For LPs who want institutional-grade credit risk management with crypto-collateral yield as one component, the pool model is the better fit.
For deeper reading, see Portfolio correlation, concentration, and the HHI index and The capital cycle: how stablecoin deposits become real-world returns.