MARKET DEPTH AND LIQUIDITY PROVISION PROFITABILITY UNISWAP V2/V3

This paper compares the market depth and liquidity provision profitabil- ity of the automated market makers (AMMs) Uniswap V3 and Uniswap V2 empirically. Due to the complexity of the algorithms, the ETH-USDC pools of both AMMs are used to make this comparison. Specifically, the mar- ket depth is measured following the framework of (Robinson and Liao, 2022) To compare the liquidity provision profitability, the framework de- veloped by (O’Neill, 2022) is used to split returns into Fee Yield, Inven- tory Holding Returns and Adverse Selection Costs (Impermanent Loss). This, mathematical model has been adjusted to remove the unrealistic as- sumption made that no liquidity is added nor removed during the period of the liquidity position. The results show that the mean market depth of Uniswap V3 is a factor of 15x higher than the mean market depth of Uniswap V2 under a price impact of +/-2%. This difference diminishes, the higher the price impact is. This is due to the concentrated liquidity structure of Uniswap V3. Furthermore, results show that the mean Fee Yields of Uniswap V3 liquidity positions are higher than the mean Fee Yields of Uniswap V2 liquidity positions. Nevertheless, due to high Ad- verse Selection Costs associated with Uniswap V3, the average total re- turns of Uniswap V2 liquidity positions are higher. When distinguishing between active and passive liquidity provision strategies, a similar conclu- sion can be drawn taking Inventory Holding Costs and Adverse Selection Costs into account. Although, for passive strategies, we can not conclude that the mean Fee Yields differ significantly. So-called gas fees are ignored in this paper