O2O's Peter Gaffney argues real-world assets need fixed-rate loans onchain
In brief
- Fixed-rate lending is the clearest growth lever for RWAs in DeFi, Gaffney says.
- Onchain RWA leverage runs almost entirely through variable-rate pools, he wrote.
- The spread between RWA yields and Aave's USDC borrow rate is about 50 basis points, he estimates.
- Gaffney cited the April 2026 KelpDAO exploit, when Aave USDC borrowers paid 14% for five days.
The variable-rate problem
Gaffney's column comes with a lot of numbers. By his figures, stablecoin supply sits around $300 billion and aggregate vault assets are close to $20 billion. Morpho alone holds more than $16 billion in deposits, most of it in variable-rate markets priced on pool utilization. He also wrote that Ethena and Sky have committed more than $2.4 billion to JAAA, BUIDL, STAC and USTB.
JAAA yields 4.5% to 5% and Maple's syrupUSDC pays 5.2%, while the one-year average USDC borrow rate on Aave V3 Ethereum was 4.56% (leaving about 50 basis points of spread), Gaffney wrote.
Illustratively, Gaffney wrote, Aave's rate curve lifts the USDC borrow rate from 4% to 14% as utilization climbs from 92% to 100%. He cited Coin Metrics, which found routine daily utilization spikes on a $2 billion pool. He also pointed to the April 2026 KelpDAO rsETH exploit, which pinned Aave's USDC pool at 100% utilization and left borrowers paying the maximum 14% rate for five days.
Gaffney also pointed to redemption timelines. RWA redemptions can settle anywhere from T+7 to T+90, according to his column, and tokenized credit products can run from T+1 to well beyond T+90. He wrote that the October 10, 2025 liquidation cascade became a $19 billion event and marked the start of a nearly 60% drop in DeFi total value locked over the next eight months.
Matching loan duration to asset duration
Gaffney's proposed fix is fixed-rate, fixed-term lending.
He argued it lets an allocator match the duration of the loan to the duration of the RWA, the way banks extend fixed-spread advances against investment-grade paper. Gaffney wrote that this would make a strategy possible: buy a quality income asset, borrow against it at a locked rate below the asset's yield, and buy more of it.
"With the borrow cost and maturity known at entry, institutional counterparties can underwrite an RWA looping vault the same way they approve conventional term financing."
Gaffney also compared $51 trillion in US fixed income with $1.6 trillion in floating-rate leveraged loans in traditional credit markets.
Where it's already showing up
As background, Gaffney noted that Morpho recently opened fixed-rate, fixed-term markets through Morpho Midnight. Coinbase used those markets to launch fixed-rate crypto-backed loans for its customers in September 2026.
Readers should note that O2O Capital Partners may have an interest in the RWA lending market. The thesis here is Gaffney's own, and it isn't LeoDex's view.
Frequently asked questions
Why does Peter Gaffney say variable-rate pools are a problem for RWAs?
Gaffney wrote that onchain RWA leverage runs almost entirely through variable-rate pools, where utilization sets the borrow rate. On Aave, the USDC rate jumps from 4% to 14% as utilization goes from 92% to 100%, and RWA redemptions can take from T+7 to T+90 to settle.
How thin is the spread on RWA looping strategies, according to Gaffney?
Per Gaffney, JAAA yields 4.5% to 5% and Maple's syrupUSDC pays 5.2%, while the one-year average USDC borrow rate on Aave V3 Ethereum was 4.56%. That leaves roughly 50 basis points of spread.
What does fixed-rate, fixed-term lending change?
Gaffney argued it lets allocators match the duration of a loan to the duration of the RWA, as banks do with fixed-spread advances against investment-grade paper. That makes it possible to borrow against an income asset at a locked rate below its yield and buy more of it.
