Aave V3.8 proposal pairs security upgrade with proposed risk cuts

Published October 20, Aave’s V3.8 upgrade is in community discussion and framed, explicitly, around security rather than a broad redesign. The Aave Governance proposal calls V3.8 “a non-invasive and very measured release focused on security.”

The headline mechanism is a reentrancy guard for Pool operations during external callbacks. Aave Labs says it uses transient storage under EIP-1153, with a stated aim of making the system “more robust and predictable.” Certora and third parties will audit the code before an Aave Improvement Proposal, or AIP, is submitted. No date for the Snapshot vote or AIP submission appears in the proposal, and audit findings haven’t been published.

For LPs and borrowers right now, that means the event is governance review. Any code-level change waits on audit and AIP steps.

The targeted deployments span Ethereum Core and Lido, Plasma, Avalanche, BNB, Monad, Base, Arbitrum, Mantle, X Layer, Polygon, Linea, Optimism, Gnosis, MegaETH, and Celo. V3.8 will be published under a Business Source License as its own version of the Licensed Work; the proposal doesn’t state the exact “Change Date” for that license.

What arrived a day earlier — and, in the desk’s judgment, more directly relevant to current positions — are the parameter recommendations LlamaRisk published October 19. Those cover supply-cap reductions for USDe, LINK, and rETH on Aave V3 Core, cap cuts on selected Plasma markets, tighter WPOL limits on Polygon, and a USDT interest-rate model change. All of them remain recommendations unless the community adopts them. The LlamaRisk recommendations describe the proposals as based on “user behavior, on-chain liquidity, and position health observed in the latest review of Aave V3 reserves.”

On V3 Core, LlamaRisk proposes cutting the USDe supply cap from 400 million to 300 million, LINK from 11.4 million to 9 million, and rETH from 49,000 to 40,000. Those reductions look material when you run them against current outstanding supply. USDe’s approximately 258.6 million outstanding would sit at 86.5% of the proposed 300 million cap; LINK’s approximately 7.56 million would hit 84.0% of 9 million; rETH’s approximately 33,840 would reach 84.6% of 40,000. That’s 258.6 million ÷ 300 million, 7.56 million ÷ 9 million, and 33,840 ÷ 40,000.

LlamaRisk reports current supply-cap utilization of 64.9% for USDe, 66.3% for LINK, and 69.1% for rETH, with 30-day supply declines of 57.3%, 9.5%, and 4.6% respectively. Balances have been falling, but the tighter caps would leave substantially less room for new supply.

The Plasma recommendations are similarly precise. USDe’s borrow cap would fall from 100 million to 50 million; the PT-sUSDE-22OCT2026 supply cap would fall from 100 million to 50 million. USDe carries 31,058,150 USDe in debt — 31.1% of the current borrow cap, but 62.1% of the proposed 50 million. PT-sUSDE-22OCT2026 shows 25.7% supply-cap utilization with supply down 55.9% over 30 days ahead of its October 22, 2026 maturity; the approximately 25.7 million outstanding balance would represent 51.4% of the proposed 50 million cap.

WPOL on Polygon would move lower too. The recommended supply cap drops from 101 million to 65 million, borrow cap from 30.43 million to 25 million. Current supply-cap utilization is 49.5%, with supply down 34.2% over seven days. WPOL debt sits at 18.26 million against the current 30.43 million borrow cap — 60.0% today, 73.0% against the proposed 25 million.

The USDT change is about rates, not just capacity. LlamaRisk proposes lowering the optimalUsageRatio on Aave V3 Core from 94.00% to 92.00%. The market currently shows approximately 2.89 billion USDT in supply against 2.48 billion in debt, or 85.6% utilization. At that utilization, the proposed model would lift the variable borrow rate from 4.10% to 4.19% APR and the supply rate from 3.16% to 3.22% APR — LlamaRisk’s scenario at 85.6%, not a confirmed post-change market rate. For borrowers: a 0.09 percentage-point increase. For suppliers: 0.06 percentage points. CoinGecko lists USDT at $0.999414, flat on the day.

The practical picture splits cleanly across the two documents. V3.8 would add a security control across the listed deployments if it clears audits and governance. The LlamaRisk recommendations, if adopted, would reduce expansion room for selected assets and modestly raise quoted USDT borrowing and supply rates at current utilization. LPs in capped markets face tighter capacity for additional deposits; USDT borrowers face the proposed higher rate; USDT suppliers would receive the corresponding increase.

The next documented gates are Certora and third-party audits, then an AIP. Snapshot and submission dates remain unspecified.