Aave's USDT borrow rate heads to 4.19 percent: what it means for lenders

Lenders and borrowers in Aave's USDT market will soon see a different rate. On October 20 the protocol's Risk Stewards recommended cutting the optimal usage ratio for USDT on Aave V3 Core from 94.00 to 92.00 percent. At the current level of usage that would lift the variable borrow rate from 4.10 to 4.19 percent a year and the supply rate from 3.16 to 3.22 percent. Several other caps are coming down at the same time. AAVE itself trades 13.8 percent below its weekly high.

What Aave is changing in the USDT market

Aave is a decentralised lending market: users deposit funds, others borrow against collateral, and the rate comes out of a formula rather than a negotiation. Those formulas are maintained by Risk Stewards, who can move parameters within fixed bounds without a full governance vote each time. For the USDT market, the analytics firm LlamaRisk filed a recommendation in that role on October 20 in the Aave governance forum.

The starting position documented there: the USDT market on Aave V3 Core holds about 2,892.2 million USDT in deposits, of which 2,475.1 million is lent out. That is 85.6 percent usage, leaving 417.0 million USDT unlent. Those figures underpin everything that follows.

The optimal usage ratio drops from 94 to 92 percent

The optimal usage ratio, optimalUsageRatio in the documentation, is the threshold below which the rate climbs gently and above which it climbs steeply. Beneath that mark the borrow rate follows a shallow slope, above it a far sharper one. The protocol deliberately makes credit expensive once too little liquidity is left.

The base rate and both slopes stay as they are under the recommendation. The borrow rate exactly at the optimal point therefore remains 4.50 percent a year, and 14.50 percent at full usage. What moves is only the point at which the steep stretch begins.

Why the rate tracks usage

A lending market without a reserve would be useless to depositors, because nobody could withdraw. The rate is therefore also a steering tool: as usage rises, borrowing gets more expensive and depositing more attractive, until liquidity flows back. Move the optimal point down and that mechanism bites earlier. At today's usage of 85.6 percent the shift is small, 4.19 percent instead of 4.10 for borrowers and 3.22 instead of 3.16 for depositors. The change only becomes noticeable as usage runs towards the new threshold.

The withdrawal buffer grows to 231.4 million USDT

The second consequence is the one that matters to depositors. The optimal point sets how much of the deposits stays available for withdrawals in normal operation. Two percentage points less usage means two percent more of deposits held in reserve.

In numbers: the buffer at the optimal point rises from 173.5 to 231.4 million USDT, shown in the Risk Stewards' table as a gain of 57.8 million. In the same move the borrowing headroom up to the optimal point shrinks from 243.5 to 185.7 million USDT. The arithmetic is a trade, more certainty on withdrawals against less room to borrow. On the Risk Stewards' account, roughly 185.7 million USDT of credit demand can still be met before the steep stretch takes hold.

That trade-off is the heart of the matter, and it can be read against Aave too: whoever has been borrowing large sums there at a flat rate loses part of that room. Which platform offers which terms is set out in our comparison of crypto lending providers.

Lower caps for USDe, LINK and rETH

Alongside the rate model, the Risk Stewards recommend several lower caps. On Aave V3 Core the supply cap for USDe falls from 400 to 300 million, for LINK from 11.4 to 9.0 million tokens and for rETH from 49,000 to 40,000 tokens.

The reasoning rests on user behaviour. USDe deposits fell 57.3 percent within 30 days, and the old cap was only 64.9 percent full. The new limit sits at roughly 1.16 times the balance actually present, which puts cap utilisation after the change at about 86.5 percent. For LINK the cap was 66.3 percent full and deposits fell 9.5 percent over 30 days. A cap set well above what is actually there constrains nothing when it matters.

USDe on Plasma, WPOL on Polygon

The list reaches beyond the main market. On the Plasma instance the borrow cap for USDe falls from 100 to 50 million, and the supply cap for the yield instrument PT-sUSDE-22OCT2026 likewise from 100 to 50 million. On Polygon the supply cap for WPOL drops from 101 to 65 million tokens and the borrow cap from 30.43 to 25 million. Four separate instances of the protocol are affected, each with its own limits.

Aave V3.8 adds a reentrancy lock

A second item sits in the forum alongside the parameters. On October 21 Aave Labs filed the proposal for Aave V3.8, described by its authors as a deliberately small step with a single focus: security.

At its core is a lock against reentrancy on pool operations, active during external callbacks. Reentrancy describes the case where a contract is called again while an operation is still unfinished, and meets accounting that is not yet complete. The lock is meant to keep the pool's bookkeeping consistent even when a single operation talks to the pool more than once.

What transient storage has to do with security

Technically the lock builds on transient storage under EIP-1153. Such protections used to count as too expensive, because every check costs gas. Aave Labs writes that transient storage removes that objection and that the protection can be added with little overhead. The protocol explicitly frames the change as a guard against theoretical attacks, not as a response to an incident.

One clarification that Aave Labs added in the forum on October 23 matters here. A user asked whether the lock also covers the callback of a flash loan. The answer: flash loans are not secured in V3.8. Both flashLoan() and flashLoanSimple() stay out of scope, so a liquidation can still be triggered from a flash loan callback. On Aave Labs' account the upgrade therefore touches no existing integration and no tool built on flash loans.

Which Aave instances the upgrade is meant to reach

The proposal names the target environments one by one: the V3 instances on Ethereum in the Core and Lido variants, plus Plasma, Avalanche, BNB, Monad, Base, Arbitrum, Mantle, X Layer, Polygon, Linea, Optimism, Gnosis, MegaETH and Celo. Only active instances are covered; where every value is frozen and the caps stand at zero, things stay as they are.

Two formal points belong to this. V3.8 ships under a Business Source License whose change date is set by Aave governance and recorded in a text entry under v38.aavelicense.eth; the rights are to pass later to the Aave Foundation through a separate request. And before the vote itself, the code is to go through a review round at Certora and with third parties. The upgrade is therefore not decided, it is at the discussion stage.

The AAVE price sits 13.8 percent below its weekly high

The token itself has had a weaker week than the broader market. AAVE trades at $136.79, up 1.6 percent on the day and down 8.8 percent over seven days according to CoinGecko, as of Sunday afternoon. Market capitalisation is around $2.11 billion, daily turnover around $162.9 million, and the token ranks 48th.

The weekly high dates from October 21 at $158.75, the weekly low from October 24 at $135.06. The price stands 13.8 percent below the high and 1.3 percent above the low, closer to the bottom of the week. On the link between governance business and the price we wrote once already on October 15, then about the push for native bitcoin as collateral.

Levels above and below: $158.75 and $135.06

The two weekly readings are the nearest reference points, and nothing more than that: no price target and no recommendation. Below sits the weekly low at $135.06, just under the current level, and beneath that the round $130 mark. Above lie first the $140 area and then the weekly high at $158.75, roughly a seventh above today's price.

No link between the parameter changes and the price has been established. The rate recommendation concerns the protocol's USDT market rather than the token, and the declines in USDe and LINK ran over 30 days, longer than this week's price move. Pulling the two together is speculation, not evidence.

Access, custody and leverage for investors in Germany

A few practical points sit between the news and your own portfolio. The first is access: under the EU's MiCA regulation a platform may only offer its services in Germany with the appropriate licence, and which assets are tradable differs from venue to venue. Whether your provider lists AAVE, and which supervisor it answers to, is best checked with the provider itself before you place an order.

The second is custody. If the token sits on an exchange it belongs to you economically, but someone else holds it technically. Self-custody means taking responsibility for access, and therefore for securing the key. Either is defensible; the choice should simply be a deliberate one.

The third concerns leverage. With 13.8 percent between the weekly high and the current price, AAVE is not a quiet asset. A leveraged position is liquidated faster at that kind of range than an unleveraged one, and for a token ranked 48th the market is thinner than for the largest assets. Borrowing inside the protocol adds a second form of leverage, whose rate can change, as this episode shows.

Interest income is not a capital gain for tax

Two things need to be kept apart for tax. Selling the token falls under the rules for private disposals, where the holding period matters. Interest or rewards from a lending market are current income and are recorded differently, even when the token itself was never sold. Both often arise in the same month on the same platform, and keeping no separate record makes them almost impossible to untangle later. Which provision applies in your case and which allowances hold depends on your personal situation; a tax adviser settles that bindingly, an article does not.

The next dates in Aave's governance calendar

Neither item carries a fixed date, though both have a sequence. On the parameters, the Risk Stewards' recommendation has been on the table since October 20; it is implemented within the framework provided for it, not through a long vote. For V3.8 the proposal names the review round at Certora and with third parties as a condition to be met before the voting step. When that step comes is open. Both threads run publicly in the governance forum, as of October 25.

Depositors in the protocol will see the effect on their own rate before they see an announcement. A look at the usage of the market in question says more about the coming days than any forecast: the closer it sits to 92 percent, the harder the new threshold bites.

Aave lending market: above 92 percent usage the rate climbs harder

The change is small and its direction is clear: Aave is shifting weight from the borrowing side to the lending side. Three things are worth a look now.

1. Check your own rate. For every deposit and every loan in the protocol, see how high the market's usage currently is and which rate you are charged. What other platforms pay for the same balance is shown in the comparison of staking and yield platforms.
2. Record income and sales separately. Note current interest and sales in two columns from the start, with date and value. A tax tool from our comparison takes over the bookkeeping.
3. Sort out access before it gets urgent. Check whether your trading venue lists AAVE and under which licence it operates. The regulated houses are listed in the exchange comparison.

(As of October 25, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)