• Peter Gaffney of O2O Capital Partners argues in CoinDesk (October 21, 2026) that fixed-rate, fixed-term loans are the main missing piece for scaling tokenized real-world assets.
• His cited one-year average Aave V3 Ethereum USDC borrow rate of 4.56% sits against RWA yields of 4.5%–5.2%, a spread of roughly -6 to +64 basis points by simple arithmetic.
• He says Aave's USDC pool stayed at 100% utilization after the April 2026 KelpDAO rsETH exploit, with borrowers paying the 14% maximum for five days. This is his account, not independently checked here.
• The Block reported Morpho Midnight's July 2026 launch on Base and said Midnight-specific rates, volumes and TVL were not disclosed.
• Watch for protocol-reported Midnight volumes and quoted fixed rates against RWA yields; redemption lags of T+7 to T+90 matter as much as the rate.

Fixed-rate, fixed-term lending is the missing piece for scaling tokenized real-world assets in DeFi, according to Peter Gaffney, co-founder of O2O Capital Partners, writing in CoinDesk on October 21, 2026. The argument rests on rate math and on stress events he describes. It is advocacy from a market participant, so the figures are best read as his case, not audited data.

Variable rates leave little room for leverage

Gaffney argues that utilization-priced pools make leverage on tokenized assets unpredictable and confine it to higher-yielding products. He cites a one-year average USDC borrow rate of 4.56% on Aave V3 Ethereum, against a JAAA yield of 4.5% to 5% and a Maple syrupUSDC yield of 5.2%. He puts the spread at about 50 basis points; the cited numbers imply a range from roughly -6 to +64.

The tail risk is a rate spike, not the average

The op-ed's strongest point is about tails. Gaffney says Aave's illustrative USDC curve rises from 4% to 14% as utilization goes from 92% to 100%, and that after the April 2026 KelpDAO rsETH exploit the pool stayed at 100% utilization while borrowers paid the 14% maximum for five days. A looped position yielding about 5% would lose money at that rate. These details are his and were not independently verified here.

Redemption timing is the second mismatch

Gaffney also points to term mismatch: he says RWA redemptions run from T+7 to T+90, and tokenized credit from T+1 to beyond T+90, while variable loans can reprice at any time. A fixed term matched to the redemption window would let a borrower hold the asset to maturity without refinancing risk. He cites Kamino fixed-rate loans against Figure's auto-loan portfolio at 5.3% in rolling 30-day terms as an existing example.

Midnight exists, but its scale is undisclosed

The Block reported on July 21, 2026 that Morpho launched Midnight on Base, a fixed-rate, fixed-term protocol where lenders and borrowers negotiate terms through an offer book. Lenders' funds stay on Morpho Blue earning variable rates until an offer is taken. The report said Midnight-specific rates, volumes and TVL were not disclosed, so the product's traction cannot yet be measured.

The sizing example is a scenario, not a forecast

Gaffney sizes the opportunity with an illustration: moving 5% of $12 billion in Sky and Ethena reserves into fixed-rate looped RWA strategies at 3x would create $600 million of equity and $1.8 billion of RWA demand. That is arithmetic on his own assumptions. It shows what the thesis implies if issuers and treasuries adopt it, not evidence that they will, and no such allocation has been reported.

What to watch next

For practitioners, the test is data. Look for protocol-reported Midnight deposits and loan volumes, the fixed rates actually quoted against RWA yields, and fees and redemption lags that eat into carry. Compare the 4.56% average with a dated Aave utilization series. Until those exist, fixed-rate credit as the unlock for RWA leverage is a credible hypothesis, not a demonstrated market shift.