Summary
This publication focuses on adjusting relevant slope2 values across instances with significant observed borrowing demand, in preparation for the upcoming launch of the slope2 risk oracle.
Motivation
With the imminent launch of the slope2 risk oracle, we introduce a mechanism that transforms how bespoke markets respond to liquidity contraction. Current piecewise-linear rate curves fail to manage the joint risks of discrete supply shocks and hyper-elastic borrower demand in relatively centralized, high-leverage lending markets. Large withdrawals can instantaneously push utilization to high levels, where abrupt APR spikes accelerate borrower deleveraging rather than restoring liquidity; the asymmetry between supply and demand responses requires substantially different liquidity conditions for the market to return to equilibrium. Thin-margin leveraged positions thus become unprofitable with a minimal negative effective rate delta (e.g., LSTs/LRTs, sUSDe, PTs), amplifying contraction and generating negative second-order effects.
Instead of these abrupt spikes, the oracle stages its response: beginning with a low baseline deterrent when utilization first crosses the kink, compounding convexly the longer stress persists, and decaying predictably once conditions normalize. In practice, the baseline configured slope value shall be defined at its respective minimum, ensuring that borrowers and suppliers enter each stress episode from the most optimal starting point as a function of the underlying risk components that comprise such a market. This approach avoids unnecessary preemptive shocks, provides a transparent and bounded path of escalation, and guarantees that costs only intensify when real liquidity pressure emerges. In effect, suppliers receive robust solvency protection, while borrowers face a fair, time-aligned incentive structure that becomes stricter only if high utilization persists. For more information on how the algorithm operates, please see the initial research paper on this topic.
According to the algorithm described above, the internal parameters that determine the effective growth in slope2 with respect to time and magnitude above the kink vary for such markets. In the plots below, we illustrate the evolution of the market under various stress scenarios, determined by the welfare function and its associated solvency constraints. As market utilization scales upward for extended periods, the protocol aims to gradually align the expected demand response of the supply and borrow sides, minimizing cumulative borrower cost subject to outstanding market risk. Note that such values imply persistent, continuous utilization values that range above the kink to determine its final value according to the algorithm; thus, abrupt spikes are effectively ignored in this sense.
Risk Oracle Permissions
The risk oracle will cover relevant USDC, USDT, USDe, and WETH reserves, with the specification initially setting slope2 to its minimum value in each market, ensuring alignment with the parameterized algorithm of the respective reserve. Based on historical backtesting and the algorithm’s continuous operation under AGRS constraints, which define both the maximum permissible adjustment and the minimum time between changes, the system will be configured to allow a maximum absolute change of 4% every 8 hours. The technical maintenance proposal will follow shortly.
Specification
| Instance | Asset | Current Slope2 | Recommended Slope2 |
|---|---|---|---|
| Arbitrum | USDT0 | 40% | 10% |
| Arbitrum | USDC | 40% | 10% |
| Arbitrum | WETH | 80% | 8% |
| Avalanche | USDC | 40% | 10% |
| Avalanche | USDt | 40% | 10% |
| Avalanche | WETH.e | 80% | 8% |
| Base | USDC | 40% | 10% |
| Base | WETH | 80% | 8% |
| BNB | WETH | 80% | 8% |
| BNB | USDC | 40% | 10% |
| BNB | USDT | 40% | 10% |
| Ethereum Core | USDC | 20% | 10% |
| Ethereum Core | USDe | 30% | 12% |
| Ethereum Core | USDT | 14% | 10% |
| Ethereum Core | WETH | 20% | 8% |
| Ethereum Prime | WETH | 25% | 8% |
| Gnosis | WETH | 80% | 8% |
| Linea | USDC | 60% | 10% |
| Linea | USDT | 60% | 10% |
| Linea | WETH | 80% | 8% |
| Optimism | USDC | 40% | 10% |
| Optimism | USDT | 40% | 10% |
| Optimism | WETH | 80% | 8% |
| Polygon | USDC | 40% | 10% |
| Polygon | USDT0 | 40% | 10% |
| Polygon | WETH | 80% | 8% |
| Plasma | USDT | 20% | 10% |
| Plasma | WETH | 20% | 8% |
| Plasma | USDe | 50% | 12% |
| Scroll | WETH | 80% | 8% |
| Sonic | USDC | 40% | 10% |
| Sonic | WETH | 80% | 8% |
| Parameter | Maximum Change (Absolute) | minimumDelay |
|---|---|---|
| Slope2 | 4% | 8 Hours |
Next Steps
We will move forward and implement these updates via the Risk Steward process. As the Risk Steward constrains slope2 changes to a maximum of 20% every 3 days, updates that require more will be implemented gradually accordingly.
Disclosure
Chaos Labs has not been compensated by any third party for publishing this AGRS recommendation.
Copyright
Copyright and related rights waived via CC0.