I know crypto has been getting hammered of late, Sushi included, and it seemed as though there was no bottom for the token. But there are still exciting developments occurring that give one hope that SushiSwap will regain some of its prior status. Sushiswap was able to increase the spending efficiency in the Sushiswap from ~$2,000 in volume per Sushi spent in incentives to $10,000; a 5x increase in the impact of the Sushi incentive spend.
DeFi protocols are spending at ludicrous rates in order to attract users. The sheer volume of both new and established protocols expanding, extending, or announcing incentive programs accounts for billions in annual spending. Since the start of DeFi, incentives have been an essential tool to bootstrap liquidity and users, kickstarting the virtuous cycle of liquidity attracting users, users generating fees, and fees attracting more liquidity. When used correctly, an incentives program can have a dramatic effect on a protocol’s growth and can make or break a protocol’s success.
Despite the history of incentives in DeFi, there has been an ad-hoc approach to how or when these incentives are spent. There are many questions for a protocol setting up incentives to grapple with. What is the optimal total amount of incentives to spend? Which liquidity pools are most deserving of incentives? How should incentive spending change in response to market conditions?
In May, SushiSwap came to Gauntlet looking for more understanding of the effectiveness of their token emissions. Already well into Onsen Round 2, the program had been successful at bootstrapping and attracting liquidity for lesser-known projects. At that time, over 100 pools and countless other project applications were vying for SUSHI incentives.
From the outset, the Onsen program set a tiered structure, based on market cap, to dictate token emissions. Midcap ($25–100M), Lowcap ($5–25M), and Gems ($1M-5M) were each awarded a fixed amount of incentives. During May, approximately ~192k SUSHI per day ($2.4M) was emitted to attract liquidity across all pools. Incentivized pools include permanent large-cap pools, such as ETH/USDC, and small-cap pairings through the Onsen program.
Early Onsen program success and expanding demand meant optimized distributions were no longer a “keen to have” but a “need to have.” The Sushi team needed to know if their target spend was too much or too little. If incentives were directed towards the right pools. If short-term liquidity would translate to long-term liquidity and trading volume.
Gauntlet Platform’s Incentive Optimization solution has two components:
- Competitive Market Model
- Empirical Experimentation Model
The Competitive Market Model consists of a modified Markowitz model which examines the market returns for a given liquidity pool. By examining competitive returns within the DeFi space and the distribution of LP capital the platform comes up with a baseline of incentives required to ensure managed pool returns are in line with that of the market. The goal here is to ensure that protocols both maintain the liquidity that they have now as well as provide a theoretical model for how much liquidity to expect to inflow after an incentives change.
The Empirical Experimentation Model takes this to the next level where the platform empirically analyzes the impact of incentive changes. Some pools may be more “sticky” than others where LPs are less sensitive to changes in incentives. By statistically modeling the impact changes to incentives have on the capital inflows/outflows of a liquidity pool we can directly model the “elasticity” for each of the pools that we are managing. In doing so incentives can be moved from pools where LPs are indifferent to receiving incentives to pools where LPs react strongly, maximizing the impact incentives have.
Since Sushiswap has a fixed token emissions schedule, the goal with incentive recommendations was to maximize the impact. This can be achieved by diverting incentives from low-impact to high-impact pools.
One of the largest Sushiswap pools where the platform identified a large opportunity for incentive savings was the YFI — ETH pool where Sushiswap holds nearly 90% of the dex market’s liquidity and nearly 80% of the dex market’s volume (as of Oct. 12 2021). Given this pool’s relative dominance in the market as well as the relatively poor utilization of this pool in terms of volume this was a primary target for incentive reallocation.
The platform reduced incentives in this pool by 75% over the course of 3 months with no measurable decrease in liquidity saving Sushi ~$22.4M in annual incentives at the current Sushi price.
This was achieved without a significant change in trading volumes in this pool, thus increasing the efficiency of incentive spend in this pool (defined as Daily Trading Volume (USD) / Incentive Spend (Sushi)) by 927% from $173 per Sushi on 7/2/2021 to $1,604 per Sushi on 10/10/2021.
The platform then was able to re-allocate these incentives to alternative pools such as the RGT-ETH pool where the market conditions are much more competitive with Sushiswap only capturing a 62% market share and a very high utilization rate of the liquidity already in this pool.
Immediately after the incentives increased there was rapid capital inflow into the pool increasing liquidity by ~30% over the course of 1.5 months. This increase in liquidity was matched with an increase in trading volume as volume moved to Sushiswap following the liquidity brought in through higher incentives.
In part due to the changes in incentive distribution, Sushiswap was able to increase the spending efficiency in the Sushiswap from ~$2,000 in volume per Sushi spent in incentives to $10,000; a 5x increase in the impact of the Sushi incentive spend.
In absolute terms, this result reflects an increase of ~$300M in daily trading volume without any increases in Sushi emissions. This represents a ~$55M increase in annual protocol revenue to Sushi holders.