Incentive Impact Analysis: Kwenta

🔍Monitoring
1

Authors: @raho @tnorm

Note: This analysis extends our earlier work on Synthetix and Polynomial OP incentives. For background, see the analyses on Synthetix and Polynomial.

Program Information

Key Takeaways

  • After seeing multiple protocols divert from their intended distribution, we’d like to gather community feedback on a process/steps that can be taken to prevent this in the future. We’d be happy to assist with developing a process for protocols to obtain approval from the Governance Fund to adjust the distribution method of OP in their proposal. Given the recipient’s obligation to the Governance Fund to uphold their proposed distribution, the Governance Fund or Grants Council should have the option to accept or decline any change requested to distributions (and request the return of allocated OP if terms can’t be met).
    • Additionally, as multiple projects have diverted from their proposed distribution methods due to deprecated protocol offerings, it could be beneficial to implement prerequisites for future applicants insistent on having an implementation method (and any necessary internal approvals) before applying, ensuring the swift deployment of OP upon receipt.
  • The lump sum distribution model is gameable and led to 22% of Kwenta trading coming on the day before the end of the epoch.

Introduction

Kwenta is a derivatives trading platform built on Synthetix Perps liquidity. Kwenta’s platform offers a front end for traders to gain exposure to over 60 assets with up to 50x leverage. Kwenta was among the first 24 projects funded by the Optimism Collective in Proposal A: Governance Fund Phase 0 - Batch Vote with 900K OP on June 09, 2022. As seen with multiple other Governance Fund OP recipients (Polynomial, Synthetix, Rubicon), Kwenta’s distribution was altered after receiving funding. As this has become a recurring theme amongst funded protocols, we believe the logical next step is to design a process for protocols to adjust how they distribute their allocations. It is a great cause for concern that the funded parties feel no obligation to the Governance Fund when diverting from their proposed usage.

1600×907 173 KB

Proposed Distribution

  • Platform Migration Incentives: 600K OP (66%)
    • “Method: 1000 trader addresses who were unable to claim $DYDX and had a minimum deposit of $1000 USD in dYdX will be able to claim OP tokens after making 10 trades on Kwenta, generating a minimum PnL of 10%, and having a total cumulative volume of $5000 sUSD. 1200 wallets are eligible for this reward, the first 1000 people to claim the allocation will receive their OP.” (Source)
  • Competition Prizes: 130K OP (14%)
    • “Method: Kwenta will be hosting 2 competitions to kickoff the launch of the Kwenta token. The OP token will be used to incentivize participation by being added to the prize pool.” (Source)
  • DAO Contributors: 170K OP (20%)
    • “Method: Individuals who have contributed towards governance and development within the Kwenta DAO will receive 50% of the OP based on a Core Contributor-derived assessment of the value provided by the included members. 50% of the OP would put towards Kwenta devDAO bounties, incentivizing development on top of Kwenta and Optimism.” (Source)

Actual Distribution

1600×907 127 KB

Trading Rewards: 600K OP

Concurrent with Kwenta’s grant application in the Optimism forum, Kwenta published KIP 16, which proposed a matching program for any approved OP distribution. This was reversed in September of 2022 when Kwenta passed KIP-29, effectively diverting the retroactive OP distribution to dYdX users due to complexities in implementation. KIP 29 established a user onboarding program utilizing quests. The proposed onboarding program rewarded users who completed quests by doing various tasks within the protocol. These quests provided a boost for traders who had historically used dYdX before the KIP 29 and those who had completed Kwenta’s Optimism (Galxe) Quest.

Due to the ‘deprecation of V1 Cross Margin’ (source), KIP 29 was never implemented and was eventually followed by KIP 61 in April 2023. KIP 61 outlined a plan to provide trading rewards in conjunction with Synthetix OP trading rewards and was implemented from April 2023 to September 2023.

1600×849 198 KB

(Source)

Implementing KIP 61 ultimately resulted in the distribution of 600K/900K of Kwenta’s OP allocation to trading rewards instead of the proposed ‘platform migration incentives.’ As seen with the Rubicon and Polynomial distributions, Kwenta’s OP distribution method was changed due to the deprecation of its V1 Cross Margin and the lack of a plan to implement the proposed distribution to dYdX traders. Again, we’d like to emphasize the need for prerequisites before funding.

Trading Competitions: 120K OP

As proposed, Kwenta directed 120K OP to two independent trading competitions promoting Kwenta product launches.

  • The first competition was the Mobile Launch Trading Competition, which allocated 20K OP to traders utilizing Kwenta’s mobile interface from August 16, 2022, to August 21, 2022. The results of this competition can be found here.

  • The Cross-Margin Trading Competition ran from November 09, 2022, to November 23, 2022, which allocated 100K OP to the top 100 traders in terms of PnL. The results of this competition can be found in this blog post, with a follow-up spreadsheet here.

DAO Contributors: (?)

The latest OP distribution outline shows that 180K OP are still either earmarked for or have been distributed to DAO contributors. We assume that the OP has been at least partially distributed as outlined, but since the distribution addresses are not documented, we could not confirm with certainty.

|476x349.46068001278314
|476x349.460680012783141600×1181 432 KB

OP Remaining: 53K OP

The multi-sig wallet Kwenta sent 896K OP to (and presumably owns) still holds 53K OP.

Program Overview

With most rewards allocated to the trading rewards program beginning on April 19, 2023, we dove into the program’s impact on the OP network. As we’ve seen massive increases with the other programs, the rewards program significantly impacted daily transactions and transacting addresses. However, it remains to be seen how sustainable the trading rewards will be, as Synthetix has extended the rewards with a modified program.

Daily Transactions & Transacting Addresses

Kwenta accumulated over 393K transactions throughout the trading rewards program, averaging 2.7K daily. The average daily transactions during the program increased ~4x the average daily transactions of the 30D before (756). The return on the 393K transactions amounts to 1 transaction for every 1.5 OP spent.

1600×907 191 KB

The number of daily transacting addresses on Kwenta also surged during the program, with the average daily transacting addresses increasing from 332/day 30D before to 512/day during, a 1.5x increase. Overall, the program returned one transacting address per ~8 OP spent.

However, it’s worth acknowledging that (as noted in our Polynomial Analysis) we saw Kwenta’s market share take a significant hit with the launch of Polynomial’s incentive program in July, which we see impacting both daily transactions and addresses in the second half of the program. While the amount of OP rewards was similar to that of Kwenta, Polynomial added a unique leveling mechanism for traders, along with trading competitions, and the small increase in rewards may have been enough to attract traders to the platform.

Network Fees Generated

1600×907 126 KB

Throughout the program, Kwenta generated 89 ETH in network fees, returning 0.0001 ETH per OP spent on incentives. Before the program, Kwenta averaged 0.09 ETH/day in network fees generated, which increased by ~7X to an average of 0.62 during the program.

It’s worth acknowledging that (as noted in our Polynomial Analysis) we saw Kwenta’s market share take a significant hit with the launch of Polynomial’s incentive program in July, impacting daily transactions and addresses in the program’s second half.

Deep-Dive: Tuesdays are for Trading

Also mentioned in our Polynomial analysis, the addition of OP rewards from integrators such as Kwenta and Polynomial caused the weekly rewards to exceed that of which fees were paid (seen below).

We also observed that the increase in Polynomial’s trading volume far exceeded the increase in daily users, causing concern for inorganic trading (i.e., wash trading). As a result, we decided this made for a reasonable starting point with Kwenta as well.

Looking into Kwenta’s volume compared with daily active traders, we see less aggressive spikes in trading volume compared to daily traders. The average daily traders increased from 324 traders 30D before the program to 329 traders during the program. The average daily trading volume increased from $83M to $127M.

1600×870 250 KB

We found more modest spikes in Kwenta’s volume per trader, which averaged $251K the month before trading rewards and $379K during rewards. While this is a 1.5x increase, it pales compared to Polynomial’s volume/trader, which averaged $14K the month before rewards and $335K during rewards, a 24x increase. Insights following the Synthetix extended rewards program, which began on September 13, will provide a clearer view of the program’s overall impact.

1600×907 206 KB

Given the results of our past analysis, we expected large volume spikes along with the launch of rewards and wanted to get more granular on our previous analysis. To further investigate signs of inorganic trading, we found inspiration in @John_TV_Locke’s writing for Messari’s Synthetix quarterly report. In this report, Johnny pointed out that since rewards began on April 19, 2023, 24% of the Synthetix trading volume occurred on Tuesdays.

924×661 215 KB

(Source)

Tuesdays are significant as it marks the last day of the Synthetix (and Kwenta) trading reward epoch. At the end of each epoch, Wednesdays at 00:00 UTC, a lump sum of 300K OP was distributed, divided amongst traders based on their fees. With this information, traders could more accurately gauge the rewards per $1 spent in fees closer to the end of the epoch (ie, Tuesday).

Leveraging this insight, we dove deeper into the daily breakdown of trading volume throughout the trading rewards program. We were shocked to see the trading activity occurring on Tuesdays, in line with Johnny’s report. The cumulative volume on Tuesdays accounted for around 22% of the total volume throughout the trading rewards program ($4.1B).

1600×907 134 KB

We found that the average daily volume on Tuesdays during the program was $189M, compared to an average of $25M before, an 8x increase. While the average volume overall increased from $27M to $121M, marking a 4.5x increase, the volume leading into the end of the epoch doubled that.

1600×907 135 KB

While this does not inherently suggest inorganic usage, it further solidifies points made in earlier reports. Despite this activity, Kwenta has more consistent trading throughout the rewards program than Polynomial, which could indicate the efficacy of either more extended programs (Kwenta) or adding trading competitions on top of incentives (Polynomial).

Kwenta generated $18B in trading volume during the rewards program by distributing 600K OP, making the return for each OP spent ~$30K in trading volume. While impressive, this may prove hard to reproduce without the addition of Synthetix rewards underlying future programs.

Recommendations

  • Distribution Alteration Process: We recommend developing a process for protocols to obtain approval from the Governance Fund to adjust the distribution method of OP in their proposal. Given the recipient’s obligation to the Governance Fund in upholding their proposed distribution, the Governance Fund should have the option to accept or decline the change request (and subsequently request the return of OP).
  • Incentive Implementation Prerequisites: As multiple projects have diverted from their proposed distribution methods due to deprecated protocol offerings, it could be beneficial to implement prerequisites for future applicants insistent on having an implementation method (and any necessary for internal approvals) before applying, ensuring the swift deployment of OP upon receipt.
  • Less Predictable Rewards Schedules: Due to the distribution of the lump sum of Kwenta and Synthetix OP rewards at 00:00 UTC every Wednesday, traders could game the rewards and take advantage of excessive rewards the day before. Perhaps a bit of randomness regarding distribution could have been helpful.
  • Granular Analysis: While we’ve seen with most programs, the protocols having distributed OP to date have proven capable of driving temporary usage metrics to spike but, in most cases, haven’t proven capable of sustaining. While the temporary spike in these metrics isn’t necessarily alarming, we believe the actual return on OP allocated comes from sustained usage, not temporary spikes. Luckily, one of the beautiful aspects of Optimism’s governance lies in the commitment to experimentation. Still, as Phase 0 and earlier programs conclude, the Governance House should seriously probe the efficacy of OP distributed to iterate upon these distribution methods.

Closing Remarks:

The OP incentives provided by Kwenta, in addition to the Synthetix program, ultimately increased the average daily traders and volume throughout the reward program but could have been impacted by the massive surges in Polynomial’s program. However, depending on a grant’s goals, the sustainability of these reward types suggests that future grants might benefit from more experimental and innovative designs.

It’s difficult to assess the long-term effectiveness of ongoing programs. While some have claimed the increase of cumulative volume and fee revenue is an immediate success, we believe the evidence of more permanent trends will better indicate the success of the program for the OP ecosystem. Because fee revenue depends on volume, which was incentivized, it’s difficult to assess the protocol’s long-term revenue sustainability during the incentivized period. Following the extended program’s conclusion (which caps rewards at a maximum of $1 OP reward per $1 in fees spent), we’d like to reassess the sustainability of these rewards over longer timeframes.

Many perceived improvements to Kwenta’s program can be traced to the lack of transparency and accountability requirements of the Phase 0 program rather than to Kwenta itself. We’ve expressed our interest in helping achieve this in any capacity possible and look forward to the community’s feedback on logical next steps.

Resources

10 Likes
2

Awesome write up @raho :fire:

2 Likes