title: [Risk Stewards] August 2026 - WETH Interest Rate Adjustment
author: @TokenLogic
created: 2026-08-26
Summary
This analysis evaluates whether the current WETH Slope 1 remains appropriately calibrated to the structural decline in Ethereum staking yields and the resulting weakness in ETH borrowing demand. We recommend reducing WETH Slope 1 to 2.20% on Ethereum Core, Arbitrum and Optimism to realign borrowing costs with the prevailing LST yield environment and support a recovery in utilization.
ETH Borrowing Market Overview
ETH borrowing on Aave remains overwhelmingly concentrated on the Ethereum Core instance, which currently accounts for approximately 1.74M WETH, or roughly $4.1B, in variable debt. This represents approximately 96% of all WETH borrowed across Aave deployments. The next-largest market is Base, with approximately 76K WETH, underscoring the pronounced concentration of ETH borrowing on Ethereum Core.
ETH Reserve Utilization
Following the Kelp rsETH bridge incident on April, 2026 and the forced deleveraging that followed, utilization across the major ETH markets has settled below both their optimal levels and historical operating ranges. While the incident triggered the initial utilization reset, the subsequent recovery has remained limited. The utilization frequency profiles below, measured over the post incident period beginning June 1, show persistent slack across every large market relative to its kink. Ethereum Core has averaged 82.5% since June, with the effective borrow rate remaining around 2.11% to 2.18%, consistently below the 2.35% Slope 1 target.
The time series highlights a clear shift in the utilization regime following the Kelp incident. Through late 2025 and into April 2026, the five largest markets generally operated in the high 80s, with Ethereum Core averaging 85% to 90% outside the December demand trough and frequently approaching its kink during the February and March stress episodes. Following the incident and the associated deleveraging, utilization across the major markets declined materially and has since remained at structurally lower levels.
Ethereum Staking Economics
The economics of leveraged LST strategies have deteriorated as staking yields have compressed, while risk appetite for leveraged positions has also declined. Together, these factors have reduced the attractiveness of leveraged LST strategies and contributed to the persistence of lower ETH borrowing demand and utilization.
This section examines the evolution of staking participation, the resulting decline in consensus layer APR, and how these dynamics are flowing through to LST yields.
Ethereum Consensus Layer Staking Ratio
Ethereum staking participation has expanded sharply in 2026, increasing 19% year to date to approximately 42.4M ETH, or 34.7% of total supply, by late August. This follows a period in which total ETH staked remained broadly flat at around 35.6M ETH, equivalent to approximately 29.5% of supply, throughout the second half of 2025.
The expansion is set to continue over the near term. The validator entry queue holds approximately 2.19M ETH awaiting activation. We project total stake to reach approximately 44.6M ETH, or 36.5% of supply, by early November as the current queue is absorbed.
As issuance is distributed across an increasingly large validator set, the resulting dilution directly compresses the yield available to each staker and, by extension, the yields of the LSTs that account for the majority of ETH borrowing demand on Aave. The following section quantifies this transmission from validator growth to liquid staking yields and, ultimately, borrowing demand.
Evolution of LST Rewards
Consensus layer staking rewards are determined by protocol issuance, which scales inversely with the square root of total stake: APR = base_reward_factor × epochs_per_year / √(total staked), excluding execution layer rewards. Between January 1 and August 23, 2026, the maximum consensus APR declined from approximately 2.77% to 2.52%, a 25 bps reduction driven purely by the growth in total stake, before accounting for protocol fees or validator inefficiencies.
Over the same period, the Ethereum Core WETH Slope1 has remained at 2.35%, apart from temporary stress related adjustments in February and April to May. As a result, the target borrow rate has remained broadly fixed while the staking yield ceiling has steadily compressed, eroding approximately 25 bps, or nearly 60%, of the original spread between the two rates. If the projected validator queue is fully absorbed and total stake reaches approximately 44.6M ETH by November, the consensus yield ceiling would compress by a further approximately 6 bps, further narrowing the spread between staking yields and the current Slope1.
Translating consensus layer issuance into the headline APR of a LST requires several adjustments, some structural and others dependent on the flow of deposits and withdrawals:
- Validator efficiency: Professional operator sets typically achieve approximately 99-98% efficiency, with missed attestations and sync duties reducing realized rewards relative to the theoretical maximum.
- Execution layer rewards: Priority fees and MEV provide an additional source of yield on top of consensus issuance. These rewards are variable and seasonal, fluctuating with network activity and broader market conditions, and currently contribute approximately 20 bps on average. They are passed through to the LST exchange rate.
- Protocol commission: LST protocols typically retain approximately 10% of staking rewards as a protocol fee.
- Redemption buffer: A portion of TVL is maintained in liquid form to facilitate withdrawals. This capital earns no staking rewards, diluting the headline APR.
- Entry queue dilution: This effect is negative during periods of inflows. Newly deposited ETH is immediately reflected in LST circulating supply, with holders beginning to earn rewards, while the underlying ETH generates no staking rewards until the associated validators are activated. Rapid growth during periods of a long deposit activation queue therefore temporarily dilutes the headline APR for all holders.
- Exit queue effect: This effect is positive during periods of outflows. Redemptions reduce LST supply immediately, while the underlying validators continue earning rewards in withdrawal queue until they exit, temporarily increasing the headline APR for remaining holders.
Applying these structural adjustments to current network conditions yields an achievable LST APR of 2.38%, with the yield expected to decline further to 2.33% by November as the validator entry queue is absorbed. At these levels, the current target borrow rate approaches the full sustainable yield of a well run LST, leaving limited room for positive carry after borrowing costs.
Implications for Leveraged LST Strategies
Realized LST yields have broadly followed the decline in staking economics, and in Lido’s case have fallen below it. The wstETH exchange rate APR declined from a monthly average of 2.70% in August 2025 to 2.19% in August 2026, while weETH compressed more moderately from 2.57% to 2.36%. The flow dependent adjustments discussed in previous section help explain this divergence.
Lido has absorbed the majority of 2026 staking inflows, with its pool growing approximately 8% since June alone. The resulting entry queue dilution has pushed realized wstETH yields below even the fee adjusted theoretical ceiling of 2.38%.
The result is a reversal of the historical relationship between the two assets. On a staking yield basis, excluding any additional rewards from EtherFi’s restaking activities, weETH now outyields wstETH by approximately 15 to 20 bps, compared with a typical 10 to 25 bps discount through most of 2025. This comparison is increasingly relevant as weETH has recently transitioned to a pure LST, with restaking rewards no longer contributing to its yield. However, this divergence is expected to be temporary. As the deposit activation queue continues to deplete over the coming months, the associated entry queue dilution should diminish, allowing both wstETH and weETH staking yields to converge toward the theoretical LST APR, represented by the dashed lines in the chart below.
At the current configuration, the relationship has become restrictive for leveraged LST strategies. The 2.35% Slope1 now exceeds wstETH’s realized yield of approximately 2.19%, meaning a plain LST loop carries negative spread as utilization approaches the kink. It is also broadly in line with weETH’s approximately 2.36% yield, leaving restaked loops with virtually no margin at target utilization.
The market’s equilibrium reflects this constraint. Utilization remains below UOptimal because this is the only region where effective borrow rates, currently around 2.10% to 2.14% on Ethereum Core, remain sufficiently below the prevailing LST yield curve to support positive carry. In effect, the reserve can operate near its designed efficiency point only if borrowing costs remain below the yield available from leveraged staking strategies. Unless the target borrow rate is re aligned with the compressed staking yield environment, leveraged demand has limited economic incentive to recover toward pre incident utilization levels.
Revenue Impact of the Proposed Slope1 Reduction
To assess the revenue implications of reducing Slope1, we model annualized WETH interest revenue on Ethereum Core under the current 2.35% curve and the proposed 2.20% curve. A lower Slope1 reduces revenue at any given utilization level, but utilization recovery under the flatter curve can offset this effect through both a larger borrowed base and a higher borrow rate at the resulting utilization level.
All figures use deficit adjusted utilization. The analysis is anchored at the trailing 30 day average utilization of 83.2% for July 26 through August 24, based on approximately 1.74M WETH of borrows against 2.09M WETH of backed supply.
At unchanged utilization, reducing Slope1 from 2.35% to 2.20% lowers the effective borrow rate from 2.13% to 1.99%. This reduces annualized protocol revenue from 5,544 ETH to 5,190 ETH. The resulting worst case revenue loss is 354 ETH per year, or 6.4%. At unchanged utilization, suppliers absorb the corresponding reduction in supply APR of approximately 10 bps.
Revenue is fully recovered at 86.0% utilization, representing a 2.8 percentage point increase over the 30 day average and corresponding to approximately 58K WETH of additional borrowing, or 3.4%.
We expect the required utilization recovery to materialize. Beyond organic looping demand from other venues, substantial idle ETH is already staged in leveraged staking products that deploy through Aave. The Fluid ETH Lite vault currently holds approximately 25.2K stETH of idle liquidity, while the Lido earnETH vault holds approximately 12.2K stETH. Together, these represent roughly 37.4K stETH of undeployed capital awaiting viable loop economics.
Assuming the typical leverage structure of these strategies, this idle capital represents +400K WETH of potential borrowing demand. This is more than six times the 58K WETH of additional borrowing required to reach the 86.0% revenue breakeven. Even partial deployment of this idle capital would therefore be sufficient to clear the breakeven threshold. With the proposed Slope1 restoring positive loop carry, we expect the improved economics to unlock a portion of this existing idle capital and provide sufficient borrowing demand to support the required utilization recovery.
Recommendation
The analysis points to a clear conclusion: the WETH interest rate curves have not kept pace with the structural compression in Ethereum staking yields. Staking growth has reduced consensus APR by approximately 25 bps year to date, with a further 6 bps of compression implied by the validator entry queue over the next 40 days, while Ethereum Core Slope 1 remains at 2.35%. At the current configuration, the target borrowing rate now approaches the full sustainable yield of a well run LST, leaving limited incentive for leveraged staking demand to recover.
We therefore recommend reducing WETH Slope 1 to 2.20% on Ethereum Core, Arbitrum, and Optimism. Given the common underlying LST economics, the same target is appropriate across these markets. On Ethereum Core, this places the target borrow rate approximately 18 bps below the achievable LST APR of 2.38%, providing a modest buffer against the further yield compression expected as the validator entry queue is absorbed. No change is proposed for Base, where existing ETH borrowing incentives help offset the higher Slope 1 and current borrowing demand remains comparatively resilient.
The revenue downside is limited and recoverable. On Ethereum Core, the static worst case reduces annualized protocol revenue by 354 ETH/yr. Under the modelled curve, revenue is fully recovered at 86.0% utilization, requiring only 58K WETH of additional borrowing versus approximately 400K WETH of potential demand already staged in idle leveraged staking vaults. A return to the historical 88% to 90% range would therefore make the adjustment revenue accretive. No change is proposed for Ethereum Prime, which already operates at a 2.14% Slope1.
We will monitor staking yield compression, validator queue absorption, LST yield convergence, and utilization response, and reassess if conditions evolve materially.
Specification
Disclaimer
TokenLogic is an active service provider to the Aave DAO, the beneficiary of stream 100086 and the KPI as outlined in this publication. The scope of this engagement is available via this forum proposal.
TokenLogic supports and maintains an independent delegate voting platform within the Aave community.
TokenLogic and associated entities have no undisclosed material conflicts of interest at the time of submission.
Next Steps
- Implement the WETH Slope 1 reduction to 2.20% on Ethereum Core, Arbitrum, and Optimism through the Risk Stewards, while keeping Base unchanged.
- Track Ethereum staking and LST economics, including consensus layer APR, validator entry queue absorption, and LST yields, to assess whether further yield compression warrants additional curve adjustments.
- Reassess the parameters as utilization and staking conditions evolve, with particular attention to whether the expected leveraged LST demand materializes and whether supplier and protocol revenue improve following the adjustment.
Copyright
Copyright and related rights waived via CC0.