Profitable DeFi Liquidity Provision in 2024 — Myth or Fact?
Introduction:
Sharing research for the community is vital in the DeFi space. The complexity of decentralized finance means that an average investor needs to stay informed to make smart financial decisions. At Datamint, we are committed to this mission, which is why we regularly share our findings.
This piece will give you the up-to-date data-driven view on key questions for every DeFi investor considering to implement a strategy based on concentrated liquidity provision. Is it profitable? How to do it properly? How not to get rekt?
Recently, as the Datamint’s co-founder I participated in the ETHDubai conference, where I delivered a talk on the Dos and Don’ts of liquidity provision on Uniswap v3 (starts at 6:38:25). It was based on our recent empirical research on LP profitability. The discussion was so engaging that couldn’t resist sharing some key insights with Medium crypto community as well. Ok, let’s go!
DeFi Space Today:
In the ever-evolving world of decentralized finance (DeFi), Uniswap v3 remains a hot topic, with many researchers arguing that it may not be profitable for the average investor. Despite these claims, the DeFi sector is booming, with the total value locked (TVL) nearing $100 billion in 2024. Additionally, recent research by Kaiko reveals that Uniswap’s market depth now rivals that of most centralized exchanges, second only to Binance.
This growth is attracting institutional investors looking for higher returns during the current bull run. At Datamint, our focus is on developing effective trading strategies and conducting extensive DeFi research. A significant part of our work involves studying Uniswap and enhancing strategies for optimal liquidity provision, navigating the complexities and exploring the potential benefits it offers.
Understanding Uniswap v3 and Impermanent Loss:
Liquidity provision on Uniswap V3 is indeed controversial. Initial research by Topas Blue and Bancor in 2021 found that nearly half (49.5%) of liquidity providers experienced negative returns due to impermanent loss, and on average, liquidity providers fared worse than simply holding their assets.
Despite numerous studies, creating a clear strategy for profitability remains challenging. One of our goals is to determine if the situation has improved over time.
Impermanent loss (IL) in Uniswap v3 occurs when the value of assets in a liquidity pool diverges from their initial price ratio. This happens because whilen liquidity providers earn fees from trading, the value of their liquidity may decrease relative to simply holding the assets due to the nature of the underlying AMM (Automated Market Maker) value function. The more the price of the assets in the pool changes, the more significant the impermanent loss is.
Our Findings on LP Profitability:
We took on an extensive analysis of liquidity provider (LP) profitability on Uniswap v3, focusing on both Ethereum and Arbitrum networks. Using our in-house blockchain indexers and real-time and historical data warehouses, we have a complete control over low-level data and clear visibility of market activity. Our main task was to calculate the true USD PnL (Profit and Loss) for every liquidity provider using the Uniswap v3 protocol on both Ethereum and Arbitrum. Our findings are based on assumption that players interact with Uniswap exclusively, but it is important to remember the more complex strategies where hedging mechanisms are used (see the remark at the end of the article).
Datamint operates a fleet of in-house nodes for various blockchains, primarily Ethereum and other EVM-compatible networks. This setup allows us to maintain full control over data extraction and processing.
Our analysis allowed us to break down PnL by asset-related profit and fees, giving a detailed picture of LP profitability. We designed our system to allow free exploration, enabling drill-down from pools to individual positions and actions.
We based our study on the following methodology:
- Fetching all Uniswap v3 transactions and event data into our analytical data warehouse.
- Computing end-of-block unmanipulated USD asset prices to avoid discrepancies or manipulations.
- Detecting and labeling manipulative MEV bots using heuristics to exclude them from our research.
- Recreating the Uniswap v3 fee accrual algorithm to compute historical position valuations at any moment in time, as fee accruals aren’t recorded on the blockchain for gas efficiency.
- Calculating daily snapshots for all individual liquidity positions, broken down by asset loss/gain and accrued fees.
Using end-of-block data helps us avoid the impact of price manipulations within the block, ensuring that the prices we use are settled and unmanipulated, providing a more stable basis for our analyses.
This level of detail required processing vast amounts of data and running complex analytics queries, which created significant challenges. Off-the-shelf data providers like Dune or BitQuery couldn’t meet our needs due to performance and metadata limitations. However, with Datamint’s powerful in-house data infrastructure, we overcame these challenges. Our ability to see all market activity and compute profits reliably gave us a massive edge in finding alpha and understanding market microstructure.
Interesting Insights:
Our in-depth analysis of LP profitability on Uniswap revealed some fascinating insights. We found that aggregated LP profits are structured in an intriguing way. Most net profits, both unrealized and realized, come from the WBTC to WETH pool. This suggests that more correlated assets like BTC and ETH, which have a lower likelihood of rebalancing events, tend to be more profitable for liquidity providers.
Profitability structure for all LP providers in top 8 pools by fees accrued (all-time) on Ethereum:
Profitability structure for all LP providers in top 8 pools by fees accrued (all-time) on Arbitrum:
Our research also questions whether actively managing LP positions is more beneficial than using a relatively wide range and ‘setting and forgetting.’ Our findings show that on average, liquidity providers with fewer positions tend to have higher aggregate profits compared to those actively managing multiple positions.
Aggregate profitability for all LP providers in top 9 pools by fees accrued (all-time):
While our focus has been on Ethereum and Arbitrum due to their significant liquidity, we have looked into other blockchains like BNB Smart Chain and Polygon. However, the dynamics are very similar across these platforms. Our methodology can be adapted to study various DEXs and blockchain environments.
Conclusion:
Based on the study results, it appears that staying profitable while providing liquidity on Uniswap v3 is quite challenging. Our comprehensive study reveals that maintaining profitability through unhedged liquidity positions is notably difficult. One might reasonably ask, “What’s the point then?” The answer lies in the persistence of our research efforts: despite the overall difficulty, there are still players in the space who manage to achieve incredible three-digit APYs, although they represent a small fraction of the market.
Top LP by APR on most liquid Ethereum Uniswap v3 pools (active last 3 months, position size >$50k):
Our findings suggest that for most investors, engaging in active and concentrated management of LP positions does not outperform a strategy of maintaining broader, less actively managed positions. This insight holds true regardless of whether the strategy is executed manually or through automated systems. Furthermore, liquidity provision in Uniswap v3 shows higher profitability when dealing with correlated asset pairs such as WBTC-WETH, likely due to reduced volatility and impermanent loss.
Interestingly, factors like maximal extractable value (MEV) and transaction fees, which are significant concerns in blockchain interactions, seem to have minimal impact on the profitability of LPs. This indicates that the core challenges and opportunities within liquidity provision lie primarily in strategic position management and market selection.
Selecting the right backend and nodes is crucial for performance. Although we currently use modified Erigon client for its efficiency, especially in storage, the limited (and even due to phasing out) support for archive functionalities by mainstream Ethereum clients presents challenges. Exploring other tools and technologies is part of our ongoing effort to optimize data extraction and processing.
For this particular study, we update our dashboards once a day, as most liquidity positions last longer than a week. However, for more time-sensitive projects like arbitrage, we use updates with sub-block latency to ensure timely data processing.
Remark on Hedging against Impermanent Risk:
It is important to note that in Uniswap v3, a player’s profitability might not be fully displayed because of various hedging techniques. Hedging can be accomplished by opening a complementary short position or borrowing the volatile asset, known as delta hedging. Another method involves using perpetual futures to compensate for impermanent loss during price drops. However, it is essential to understand how hedging works and what its cost-reward ratio is. We have conducted extensive research on this topic and might publish a separate article dedicated specifically to hedging in Uniswap. Leave a comment if you are interested in this topic.
Continued research in this area is crucial. While the path to consistent profitability is filled with complexities, the presence of highly successful LPs provides a strong incentive to keep exploring this space. By dissecting the strategies that lead to success, we aim to uncover actionable insights that can democratize profitable DeFi participation and foster a deeper understanding of liquidity dynamics in decentralized exchanges.
