Reward Locking Parameter Changes: Increase Base Emissions, Decrease Max Multiplier

Author(s): Risk Labs Related Discussions: forum discussion

Summary:

Recent data shows Across’ Reward Locking program does well in retaining loyal liquidity providers (LPs); however, the growth of new LPs has been stagnant. As Across volumes grow, the total liquidity pool size is starting to become a constraint. The Across DAO should make modifications to the Reward Locking program to increase base emissions by 50% and decrease the max reward locking multiplier to 2x (from 3x). This should help attract new LPs without negatively impacting existing LPs nor increasing total protocol emissions.

Motivation:

Recent data shows Across’ Reward Locking program does well in retaining loyal liquidity providers (LPs). The Risk Labs data team found that of all existing LPs, 99% of them have earned a 100 day NFT and 91% of ETH LPs, which accounts for 70% of TVL, have never unstaked and claimed their rewards (for over 250 days!). This affirms reward locking is succeeding in its objective of accumulating a set of loyal LPs. However, the flip side to this is the growth of new LPs has been stagnant. Volumes on Across continue to grow as it steadily gains market share and with new destinations such as zkSync and Base. This has led to more frequent bursts of high utilization of the liquidity pools which increases bridging fees and dampens volumes. In order to ensure Across has the capacity to continue growing I propose making modifications to the emissions and parameters of the Reward Locking program. There are three objectives:

1. Attract new LPs to increase the total TVL of the liquidity pool to facilitate higher volumes
2. Reduce any negative impact to existing loyal LPs
3. Ensure total ACX emissions do not increase

I propose making the following changes:

1. Increase base emissions by 50%

• ~~50,000~~ → ~75,000 ACX per day for Across ETH LP tokens
• ~~50,000~~ → ~75,000 ACX per day for Across USDC LP tokens
• ~~5,000~~ → ~7,500 ACX per day for Across USDT LP tokens
• ~~5,000~~ → ~7,500 ACX per day for Across DAI LP tokens
• ~~5,000~~ → ~7,500 ACX per day for Across WBTC LP tokens
• ~~20,000~~ → ~30,000 ACX per day for Across ACX LP tokens

2. Decrease the max reward locking multiplier to 2x (from 3x)

The resulting higher base APY would attract new LPs and help to increase TVL. Yield hunting LPs who look at websites like DeFiLlama and Nanoly would notice the opportunity more easily. Some of this may be mercenary capital, but the goal is to expose Across to more people in our ecosystem and capture new sticky capital.

Existing loyal LPs would initially see their APY unaffected (ie. 150% x 2 / 3 = 1). However, as the number of LPs and TVL increases, existing LPs could see their total APY decrease as they share emissions with new LPs. On the other hand, the higher TVL will help to increase bridging volumes and the attention from new entrants could bring more awareness to the bridge. Both would act to add more value to the protocol and ultimately help the ACX token (what LPs earn).

The net result should be positive for everybody involved with Across as people “share a bigger pie”.

Specification & Implementation:

The Across DAO wallet controlled by ACX holders has admin rights to change the parameters of the Accelerating Distributor contract that powers the Reward Locking program. The transaction to execute these parameter changes is included in this oSnap proposal.

There is one complication regarding the decrease in multiplier. Base emissions for each LP are accumulated as each block builds; however, the multiplier for each individual LP is applied only when a LP claims rewards. The net result of reducing the multiplier would thus also lower accumulated rewards for existing LPs. To resolve, I propose taking a snapshot of LP accumulated rewards prior to the modification, comparing it to their accumulated rewards after the modification, and sending them the difference as an ACX token transfer on mainnet. This makes existing loyal LPs “whole” and has an added perk of allowing LPs to claim some ACX rewards without having to lose their multiplier. The Across DAO will fund this distribution from its treasury.

Specifically, to reduce operational overhead and gas costs for the distribution, I propose:

Conditions for LPs to receive Distribution

• Only LPs with a reward difference (accumulated rewards prior to change minus accumulated rewards after change) >= 100 ACX (to minimize gas costs of the distribution) will receive a distribution
• LPs with < 100 ACX difference will not be made whole, due to gas costs of making this many individual transfers vs. the value of the reward difference

Funding and Executing the Distribution

• As of Sept. 120th 2023, for the ~1625 LPs with an estimated difference >= 100 ACX, the total estimated distribution is 16.5MM ACX (which will continue to grow up to the final snapshot date)
• The estimated gas costs to send ACX to these ~1650 LPs is ~1-2 ETH (at 10-30 gwei)
• The number of qualified LPs, total ACX distributed and gas costs for distribution will change up to the snapshot when this proposal is executed.
• As the parameter change will occur after this proposal is posted, we cannot yet know the actual total required ACX distribution or include these transactions in this proposal.
• Instead, I propose the Across DAO treasury to the [Risk Labs treasury] transfers 19M ACX (to account for increased distributions at final snapshot in + gas costs) to the Risk Labs treasure, and Risk Labs will perform the distributions to LPs, and then return if any remains after paying for gas and distributing.

Downside (Cons):

There are a couple potential downsides to these parameter changes:

1. Existing loyal LPs may feel that losing their 3x multiplier is unfair even though higher base emissions compensate for this decrease. This could lead to some existing LPs exiting the protocol all together. On the other hand, the anticipated growth of Across protocol and the ability to claim some ACX without affecting their max multiplier may appease these loyal LPs.

2. The higher APY attracts a lot of mercenary capital that could immediately dump any farmed tokens. If the community believes the protocol has a lot of value this higher APY may also attract new community members and any selling pressure will be absorbed by willing buyers.

Voting:

A "yes" is a vote to:

• Reduce the Reward Locking max multiplier in the AcceleratingDistributor contract from 3x to 2x
• Increase Base emissions for all tokens as specified above
• Send 19MM ACX from Across DAO treasury to Risk Labs treasury, who will then perform distributions to eligible LPs subject to the conditions specified above

A "no" is a vote to:

• not modify the Reward Locking program in any way