Challenge #2 aimed to build off of the insights from challenge #1, this time exploring the potential impact of gas subsidies on L2s as opposed to mainnet. The program generated a ton of interesting insights for this challenge (see full list of analytics at the bottom); this week's Insights are brought to you by Unit Zero Labs.
Overall Impact of Gas Subsidies
Beginning with the winner of Challenge #2, Brandyn Hamilton suggests in his research that gas subsidies might not significantly influence user activity on L2 networks, similar to findings from previous studies on mainnet. He notes that "the already low fees on these platforms lead to more gas-agnostic behavior compared to L1." The weak statistical correlations observed between gas fees and target variables such as protocol fees and volume indicate that other factors, like market sentiment and broader economic conditions, may play a more significant role.
Layer 2 Chain Comparisons
An analysis by mrwildcat reveals interesting differences among L2 chains:
Arbitrum outpaces all other chains by a factor of two, with an average daily volume of 300M USD, followed by Base, Polygon, and Optimism.
Base and Polygon lead in user numbers, with around 70,000 users making exchanges on Uniswap daily. Arbitrum shows an interesting pattern: highest volume but fewer users compared to Base and Polygon.
He explores Rabby, a wallet that offers a gas subsidy for new users without enough to cover the cost of their first transaction. A useful comparison for exploring the user effects of certain types of subsidies.
Impact of EIP-4844 Upgrade
The EIP-4844 upgrade, implemented on 2024-03-13, has had a significant impact on L2 networks:
Analyst wyeeeh shows above how gas costs per transaction have notably decreased across L2 networks.
User engagement on Base was particularly incentivized by the gas fee reduction, showing a -0.18 correlation and experiencing the largest growth in Uniswap trades and user engagement before and after the gas fee reduction.
Below, analyst abbas_ra_sb shows a closer look comparing activity 30 days after the network update with the 30 days prior:
Token Pair Analysis
Charliemarketplace.eth points out that popular pools like ETH-stable and BTC-stable pairs witness substantial trading volume and are less sensitive to gas price fluctuations compared to niche pools. This suggests that gas subsidies might be more effective when targeted at pools with long-tail assets, which show heightened sensitivity to gas price fluctuations.
Transaction Size and Gas Fees
Kida's analysis shows that transaction fees tend to only affect the bottom end of the transaction size distribution. This implies that gas subsidies might have a more significant impact on smaller transactions, potentially benefiting retail traders more than large-volume traders.
Volatile vs. Stablecoin Pairs
Matiasac.eth, publishing in OurNetwork, notes that Uniswap V3's concentrated liquidity model drives efficiency, as seen in high volumes for USDC/WETH and WETH/USDT pairs. However, V2 pools remain resilient, particularly for meme token pools. This suggests that gas subsidies might have different impacts on V2 and V3 pools, as well as on volatile-volatile, volatile-stablecoin, and stablecoin-stablecoin pairs.
Analyst Whaen states, "Fully or partially subsidizing the gas fee would not bring significant benefits to traders, as transaction fees on L2 are already low, especially after the Dencun upgrade. Indeed, we are seeing a significant increase in collected swap fees and transaction volume, but looking at the transaction fee today, it is already low enough that traders will probably ignore it."
Retail vs. Professional Traders
Brandyn Hamilton's trader classifier model reveals interesting patterns in how retail and professional traders respond to gas fees:
Both retail and professional traders show a positive correlation between gas fees and average order size (0.26 and 0.31 respectively), suggesting that traders may consolidate their trades during periods of higher gas costs.
Other variables like trader volume, transaction count, and the number of unique token contracts showed much weaker correlations with gas fees.
Conclusions
While gas subsidies on L2 networks might not have as significant an impact as initially hypothesized, there are nuanced effects across different chains, pool types, and user categories. The research suggests that targeted gas subsidies, particularly for long-tail assets and smaller transactions, might be more effective than broad-based subsidies. Additionally, the impact of recent upgrades like EIP-4844 demonstrates that continued technological improvements can significantly influence user behavior and engagement on L2 networks.
See full list of analyses:
brandynham: Analyzing the Impact of Gas Subsidies on the Uniswap Ecosystem
wyeeeh: Impact of Gas Subsidies on Uniswap's Ecosystem: A Case Study of EIP-4844 Upgrade on L2s
sage_olamide: Uniswap Gas Subsidy On L2s
mrwildcat: Impact of Gas Subsidies on Uniswap
@tubaecci: Effectiveness of Gas Subsidies
Yuki Li: Uniswap V3 Gas Subsidy Analysis -- On Arbitrum Chain
@kida_sollinked: Uniswap Gas Fee Analysis Part 2
Abbas_ra_sb: Gas Subsidies Effect
berg: Beyond the Fees: Uniswap's Growth on Arbitrum & Optimism
@whaen_n: Uniswap Gas Fee Subsidies Research
Greatafonso: The Impact of Gas Subsidies on Uniswap: A Comprehensive Study
Uniswap Community Research Program
