Set Target Risk Tolerance Ratio - October 6, 2025

Spark's Operational Facilitator has placed a proposal into the voting system on behalf of nested contributor Phoenix Labs.

Review

The Spark community can hereby express support or opposition to the following changes, as described by the author of the proposal:

Summary

This proposal is submitted by Phoenix Labs in its role as a nested contributor, as defined in section A.AG1.2.2.P6.2.1.2.1.1.1 of the Spark Artifact. The proposal recommends that Spark governance adopt a 90% target Risk Tolerance Ratio (as defined in the Sky Atlas) for capital allocation activities.

Background

As part of the Sky ecosystem, Spark is required by Sky to maintain adequate capital to cover potential losses that may occur from its asset allocation activities. This is referred to as Required Risk Capital (RRC), and it helps ensure the integrity of the Sky ecosystem and protect USDS holders from losses, while also incentivizing Prime Agents like Spark to limit the riskiness of their Allocation System exposures (A.3.3.1.1).

There are severe penalties levied on Prime Agents when they have insufficient Total Risk Capital (TRC), either from their own Internal Junior Risk Capital or from approved external sources, to meet their RRC requirements set by Sky (A.3.3.2.7.2). Because the composition of Prime Agent Allocation System exposures can shift from external factors (such as users borrowing funds from a lending market), and the RRC needed for each exposure can also shift based on user behavior and market conditions, it is important for Prime Agents to maintain a buffer of excess TRC vs their current RRC to avoid penalties.

The Risk Tolerance Ratio (RTR) that is the subject of this proposal defines the ratio of RRC to TRC for a Prime Agent (A.3.3.2.7.2.1.1.1). Numbers further below 1 indicate lower risk of exceeding RRC and facing penalties, but also reduce the capital efficiency of a Prime Agent’s capital allocation activities as they are not fully utilizing their available risk capacity to seek out high yielding allocation opportunities.

Currently, the Sky risk framework is still undergoing refinement and there is a short term exemption in place for financial penalties related to breaches of risk capital requirements (A.3.3.2.7.2.1.1.7). As part of this exemption, Prime Agents are expected to maintain a Risk Tolerance Ratio of less than or equal to 90%. This proposal would implement a long term Target Risk Tolerance Ratio that persists beyond the period where this exemption is in effect.

Proposal Details

We propose that Spark governance set Spark’s target Risk Tolerance Ratio at 90%, meaning that Spark will seek to maintain an Allocation System composition where Spark’s current RRC is no more than 90% of its available TRC. This provides an adequate buffer against sudden increases in RRC (due to shifting market conditions or user actions), while also ensuring Spark can make efficient use of its capital and maximize risk adjusted return.

This governance proposal would reaffirm the existing 90% risk tolerance ratio specified in the Spark Artifact at A.AG1.2.6.P15.2.1.3.1.

Justification

As mentioned earlier, we would submit that a 90% target Risk Tolerance Ratio strikes the appropriate balance between capital efficiency and risk management.

In current market conditions, Spark can earn significant increases in protocol net revenue by allocating to higher risk, higher return opportunities such as Ethena, which has recently been yielding between 3-5% higher than the Agent base rate. However, these assets incur significant RRC, so Spark derives direct benefit from being willing to use more of its available TRC to allocate to high yielding investments.

From a risk perspective, the main drawback of using a higher target Risk Tolerance Ratio is an elevated probability of accidentally exceeding the Sky enforced RRC limit and incurring penalties. This can be mitigated by having RRC intensive investments that can be unwound on demand to help avoid a breach and return the Allocation System and RTR to healthy levels.

In Spark’s case, Ethena direct exposure is held 50% in USDe which can be unwound almost immediately, which allows Spark to reduce its RRC quickly. Other significant levers to reduce RRC including sUSDe direct holdings (withdrawable within 1 week), Ethena indirect exposure via PT looping positions (variable time to unwind but Spark can encourage repayment by immediately increasing borrow costs by a factor of as much as 4x), and Maple syrupUSDC (typically redeemable within 1-2 days in moderate size, up to 1 month maximum for very large redemptions). As long as Spark retains enough agility to scale down RRC intensive investments on demand, using a higher target RTR should be manageable without excessive risk of a breach and incurring penalties.

We believe this proposal adheres to the requirements set out in the Spark Artifact and Sky Atlas and is aligned with their purpose.

The proposed Spark Artifact changes can be found in the following pull request: https://github.com/sky-ecosystem/next-gen-atlas/pull/69

Outcomes

• If "For" receives the most votes, the proposal will be approved. The associated pull request will be merged into Spark's artifact. When required, the listed changes will proceed to implementation by the relevant actors.
• If "Against" receives the most votes, the proposal will be rejected. No changes will be implemented.