[Direct to AIP] MKR and USDtb Oracle adjustments

Overview

Recent oracle behavior and evolving market structure warrant simplifying valuation for MKR and USDtb to preserve accurate accounting and reduce exposure to thin-liquidity distortions.

  • MKR: As liquidity and price discovery have migrated from MKR to SKY, the standalone MKR feed is increasingly less stable and directionally weaker. We recommend adopting a derived valuation approach that anchors MKR to the deeper-liquidity SKY market.
  • USDtb: Since USDtb is not used as collateral, short-lived oracle dislocations can introduce unnecessary liability-side accounting volatility. We recommend hardcoding USDtb to $1 to remove transient oracle noise while remaining conservative for the protocol.

MKR

Motivation

Following the MKR to SKY migration, secondary market liquidity and trading activity have shifted toward SKY, leaving MKR with significantly reduced depth across the venues typically relied upon for oracle formation. In this regime, the MKR oracle is more likely to exhibit stale updates, greater deviation under small notional flows, and greater sensitivity to venue-level dislocations.

These risks can translate into significant oracle deviation from the price of the underlying MKR asset and introduce insolvency risk for the Aave market if collateral is overvalued during stress.

However, MKR’s economic value is currently anchored by an open conversion path into SKY at a fixed rate. This conversion path was initially announced as indeterminate; however, that definition has recently been brought up for discussion on the SKY forum. Following the most recent change, which took effect in September 2025, the amount of SKY received per MKR upgraded will be reduced by 1%. The reduction will increase by an additional 1% every three months thereafter, until it reaches 100% in 25 years.

The initial conversion rate specified under the path described is 1 MKR to 24,000 SKY. As such, under a functioning conversion mechanism, any sustained divergence between MKR spot and 24,000 Ă— SKY minus the penalty rate is expected to be arbitraged away (net of operational frictions), meaning that the fair-value reference for MKR becomes the SKY price, scaled by the conversion factor. In this context, continuing to rely on a standalone MKR oracle constructed from increasingly illiquid MKR markets is both unnecessary and directionally inferior from a risk standpoint.

Current Exposure Overview

MKR as collateral remains present but at a low absolute scale. The borrowed-asset distribution against MKR collateral totals approximately $164.52K, in quick reduction over the past 90 days, composed predominantly by stablecoin debt. The largest components are USDT (~$84.82K) and USDC (~$67.47K), with a smaller LUSD component (~$11.5K) and negligible residual balances in DAI (~$723) and USDe (sub-$1). This composition implies that the principal oracle risk surface is the valuation of MKR collateral supporting stablecoin liabilities, where an overstated MKR price can delay liquidations and increase bad-debt probability during drawdowns, while an understated price can generate unnecessary liquidations.

The current highly conservative distribution of MKR-collateralized debt positions means that an MKR price change of up to 30% would trigger minimal liquidations.

Oracle and Price Feed Changes

This proposal is intentionally scoped to Oracle configuration only. The deprecation mechanics that eliminate new exposure are already being handled via caps set to 1, proposed here, and the scheduled move toward a near-zero LTV configuration under the v3.6 recommendation here, which together prevent new leverage from being created while allowing existing users to unwind.

While MKR is being sunset from a risk-parameter standpoint, the protocol must continue valuing MKR collateral and any residual MKR borrow positions using a feed that is operationally maintainable and resilient to low-liquidity artifacts.

The recommended implementation is a derived MKR/USD oracle defined as:

MKR/USD = (SKY/USD) Ă— 24,000 * 0.94

This methodology leverages the fixed conversion rate of 1 MKR = 24,000 SKY and the fact that conversion is available indefinitely; however, it includes a 6% reduction, which is intended to represent the gradually increasing penalty until December 2026.
Further Oracle adjustment recommendations will follow when the introduced reduction is no longer sufficient to fairly represent the MKR conversion price.

USDtb

Motivation

The core motivation for hardcoding USDTb to 1 USD stems from observed mismatches between USDTb’s reported on-chain price and its underlying economic value as a stable asset, occurring alongside relatively thin liquidity on the venues that inform the oracle feed. In particular, we have observed short-lived negative deviations in the USDtb/USD market feed, which appear consistent with the feed being skewed by low-volume prints, transient transaction-level distortions, or other aggregation effects rather than reflecting a true market-clearing price. As liquidity thins, even modest and temporary dislocations can lead to oracle updates that are not representative of the asset’s factual value, while protocol-native whitelisted mint/redeem arbitrage activity has seemingly been inefficiently latent.

From a risk perspective, it is essential to note that USDTb is not used as collateral on Aave V3 Ethereum. As a result, negative oracle moves do not introduce liquidation shortfalls via impaired collateral exits. Instead, the exposure is concentrated on the liability side: if the oracle briefly reports a meaningfully sub-par price, the protocol is forced to treat a transient, low-liquidity oracle print as economically binding, even when it diverges from USDTb’s factual value.

By hardcoding USDTb’s price to 1 USD, Aave explicitly encodes the assumption that short-horizon on-chain price noise is not relevant for risk management for a non-collateral asset. This removes the downside tail risk introduced by transient sub-par prints and prevents brief, low-depth dislocations from translating into protocol-level accounting risk. Moving to a static price effectively neutralizes oracle-driven tails on both sides, while remaining consistent with USDTb’s intended stable design.

In a true fundamental depeg scenario, where USDTb’s backing, redeemability, or market structure is impaired and its market price trades below $1 for a sustained period, the static oracle maintains a conservative posture for the protocol. Debt continues to be booked at 1 USD while the asset’s market value falls. From the protocol’s perspective, this improves resilience: liquidators can source USDTb more cheaply to repay a 1 USD unit of debt, increasing liquidation profitability and reducing the likelihood of bad debt. Borrowers are effectively “overcharged” relative to spot during a depeg, but that asymmetry is protective for the protocol compared to a dynamic oracle that would mark liabilities down in line with distressed, low-liquidity pricing.

Disclosure

Chaos Labs has not been compensated by any third party for publishing this recommendation.

Copyright

Copyright and related rights waived via CC0.

1 Like

Summary

LlamaRisk supports the proposal by Chaos Labs to adjust the Oracle configurations for both MKR and USDtb.

For MKR, as liquidity migrates to SKY, the legacy MKR feed has become susceptible to volatility and manipulation. Transitioning to a derived price based on SKY (factoring in the conversion rate and penalty) provides a more robust and accurate valuation of the underlying asset.

For USDtb, we support hardcoding the price to 1 USD. Given that USDtb is not currently enabled as collateral, the primary risk stems from liability valuation. Transient downward volatility in the oracle (caused by thin liquidity) forces the protocol to mark liabilities down. Therefore, hardcoding the price to $1 is a conservative approach that ensures liabilities remain fully valued. However, it is important to highlight that the root cause of these oracle dislocations is extremely low secondary market liquidity. To make Aave more defensive, LlamaRisk recommends lowering the borrow caps for USDtb . This will prevent additional exposure while market depth is insufficient to support large-scale liquidations or exits.

MKR Specific Considerations

We support the transition to a derived MKR price. As the transition from MKR to SKY continues, the MKR token effectively becomes a claim on SKY (subject to conversion mechanics). Relying on the deeper SKY liquidity ensures the oracle reflects the true economic value of the collateral.

The proposed formula calculates the MKR price as:

image

(Where 0.94 accounts for the 6% conversion penalty currently in effect).

The chart below compares the current Spot MKR market price against the Derived SKY price, illustrating that the divergence of the scaled MKR price (without the 6% discount) has been fluctuating, with the discount of MKR secondary market price vs. MKR-to-SKY exchange rate trending between 1% and 3%:


Source: LlamaRisk, January 22, 2026

The proposed 6% discount would cover the observed discount dislocations, although it is important to realign the discount rate continuously. We will also monitor for potential rule changes in the handling of MKR to SKY conversions as they are defined by Sky governance.

USDtb Specific Comparisons

As noted in the summary, the volatility in the USDtb feed is a symptom of thin market depth rather than fundamental de-pegging. Therefore, it is important to analyze the excessiveness of the underlying issue.

The current utilization of the USDtb pool is heavily skewed by single-party deposits. Ethena remains the dominant supplier (98% of total USDtb supply on Aave Core), creating a unique risk profile where the “market” on Aave is not reflective of broad adoption but rather protocol-level integration.


Source: Etherscan, January 22, 2026

The $1 hardcode is evidenced by the fragility of the secondary markets. Overall, there is 12.2M USDtb deposited in Curve’s USDtb/USDC pool, as well as a small part of 2M USDtb deposited on Fluid. Ethena’s Reserve Fund is the sole depositor in the USDtb/USDC pool, making liquidity extremely concentrated. Current liquidity depth is insufficient to absorb trade sizes of more than $10M without significant price impact, which creates the noise seen in the current oracle feed.


Source: Curve Finance, January 22, 2026

Given this liquidity profile, hardcoding the price prevents oracle upticks from affecting borrower solvency. Nonetheless, given that this instability stems from low liquidity, current USDtb borrow cap (240M), which is more than 8 times larger than the available liquidity TVL (including 10M USD primary redemption liquidity), should be revised down.

Recommendation

In addition to the changes listed above, we recommend the borrow cap for USDtb to be reduced to 100M. This change will be executed by the Risk Stewards in the subsequent batch of parameter changes as agreed with Chaos Labs.

Disclaimer

This review was independently prepared by LlamaRisk, a community-led decentralized organization funded partly by the Aave DAO. LlamaRisk serves as Ethena’s Risk Committee member and an independent attestor of Ethena’s PoR solution. LlamaRisk did not receive compensation from the protocol(s) or their affiliated entities for this work. The information should not be construed as legal, financial, tax, or professional advice.

I’m supporting USDtb price hardcode to 1 USD as asset is not collateral. If some day asset is consider as collateral on general pool or on an e-mode this must be reconsider.
P.S: Notice case of USDtb collateral on e-mode will depend of e-mode type

From the technical side, we support the changes suggested for both assets, with the following remarks:

  • Instead of applying a fixed 6% “discount” on the new MKR/USD calculated from SKY/USD, we think it is better to fetch the discount dynamically from the MkrSky migration contract (called rate there). This way, the feed will automatically apply the live discount there and will not require further maintenance.
    This doesn’t change any security assumptions, as the MkrSky smart contract (immutable) includes proper sanity checks (e.g., the discount is always <= 100%) and increments to the discount are made via Sky governance proposals.
  • Given that MKR at the moment is already effectively frozen (both supply and borrow caps configured to 1), we think for consistency it should be LTV0 too, as that is really the desired effect. Considering it doesn’t change anything fundamental on the listed asset, only adds extra protection due to low liquidity, we will recommend the Aave Guardian to call setReserveFreeze(), just to make effective the LTV0.
3 Likes

We can confirm that the Aave Guardian has approved the transaction setting LTV0 on the MKR reserve of Aave v3 Ethereum Core.

2 Likes

We have created the governance proposal for these price feed adjustments.

Voting will start in approximately 24 hours, participate :ghost:
https://vote.onaave.com/proposal/?proposalId=441