Bancor and Into The Block found that concentrated liquidity increases loss risk.
Concentrated liquidity in Uniswap v3 improved capital efficiency but increased loss risks.
Here's how liquidity providers can boost profitability.
Bancor just published a paper on impermanent loss when providing protocol liquidity. The study collected data from May's Uniswap v3 launch to September. Permanent losses (-$260.1M) outweigh trading fee returns ($199.3M). It found no indication that active management outperformed passive when readjusting liquidity.
17 liquidity pools accounted for 47% of the platform's TVL in the study. The rest of the TVL was not evaluated since it was in stablecoin or pegged coin pools or pools with insufficient liquidity (less than $10M). Positions, wallets, and pools are the data categories. Many wallets deploy 1.25 to 4 liquidity positions, depending on the pool. Performance is measured by pocketbook or position.
It also measures whether pools had more temporary losses or fees, which can assist alleviate losses and minimize risk. Although gas taxes are measured in the study, these conclusions were made without accounting for them, therefore it is expected that they would worsen the losses.
Positions, wallets, pools
IntoTheBlock has access to the raw data and replicated the results. We compared wallet and position performance. 53.50% of positions were lucrative, while 46.50% were not. Profitable wallets were a minority, with 48.25% versus 51.75 unprofitable. Until the study's date, most Uniswap v3 liquidity addresses were not making money.
The study analyzes risk-adjusted returns for each active position. After studying position activity over different time periods, it's clear that only flash liquidity providers made more fees than transitory losses. Just-in-time liquidity is an advanced and convoluted approach that provides and removes liquidity in the same block as a major mempool deal is going to happen. Liquidity management doesn't outperform specific timeframes.
The study also revealed that wallets often had many positions for the same pool, thus it accounted for all of their positions and segregated them by pools. High correlation pools have most profitable liquidity addresses (BTC-ETH, LINK-ETH, AXS-ETH, FTM-ETH).
BTC-USDC is also uncorrelated. Due to its high Fees/TVL ratio. This shows the need of considering price correlations between assets to limit temporary loss and pool utilization ratios to maximize rewards. Some pools earn more fees than temporary loss, although they are rare.
Solution lessons
After analyzing the research, we may conclude that providing concentrated liquidity has a greater risk but higher reward profile, similar to market making (liquidity by ranges is an abstraction related with central limit order books). New users seeking liquidity may find other protocols where temporary loss is limited or nonexistent. Profitable Uniswap v3 can be achieved by tweaking how the user provides liquidity or by changing the protocol's design.
A user should always remember best practices, such as how strongly connected asset pools would suffer. Those that deploy positions in tight ranges might expect bigger profits but at the cost of more temporary losses, and the period the positions are held open has little effect. In some pools with uncorrelated assets, high utilization and trading fees can offset temporary losses.
Comparing impermanent loss to asset price variation
Uniswap v3 could try some techniques to reduce user impermanence.
The most straightforward would be a liquidity mining operation, offering UNI tokens in popular pools with minimal correlation and projected high temporary loss. Losses can be covered with governance tokens as an impermanent loss coverage or insurance for affected positions (similar to Bancor). Protocol owned liquidity is another exciting issue (Tokemak, Olympus).