In recent years, the decentralized exchange (DEX) has shown an explosive growth trend in recent years. Uniswap, as one of the largest decentralized exchanges, invented Uniswap, a decentralized financial protocol of the same name based on exchanging cryptocurrencies. Automated transactions between cryptocurrency tokens on the Ethereum blockchain. Currency trading participants create a liquidity pool based on this agreement, and this paper mainly studies the factors that affect the liquidity of currency trading.Specifically, this paper explores the factors that affect monetary liquidity through three hypotheses. The first is the impact of rates on liquidity, exploring the differences in liquidity under various rates. Then, the impact of differences in Uniswap protocol versions on liquidity was studied. Compared with the V2 version, the Uniswap V3 version added centralized liquidity，which changed the unpaid loss. . This feature makes Uniswap V3 the most flexible and efficient protocol. Finally, the influence of different active users on liquidity is compared, and the change trend of liquidity under different numbers of users is explored. Based on the above three assumptions, this paper adopts GARCH and OLS regression analysis to explore and analyze the collected currency transaction data, and draws the following conclusions for the three assumptions: (1) Researcher may conclude that rates are correlated with trading volume, and volume growth impacts liquidity capacity and increases rates. Therefore, the fee rate has a significant impact on liquidity. (2) Compared with V2, V3， quantitative analysis was carried out using unpaid loss. It was found that impermanent loss has a more significant impact on the liquidity of the V3 version but has little correlation with the V2 version. (3) According to the analysis and comparison of the model, there is no obvious ARCH phenomenon among active users, so it is believed that there is no significant correlation between the two. (4) Combining conclusions 1 and 3, researcher further analyzed the impact of several independent variables on liquidity and found that the fee rate has a more significant influence than active users