By Jason Milionis, Austin Adams, and Xin Wan
Read the full paper here.
The Uniswap Protocol, which has processed $1.5 trillion in lifetime volume since its release in 2018, has pioneered the field of automated market makers (AMMs), serving as a cornerstone of decentralized finance (DeFi). One of the hallmarks of AMMs is that anyone can create a market for any token and provide liquidity without permission – ensuring an equal playing field among all market participants.
The success of these AMMs depends critically on the participation of liquidity providers (LPs), who provide liquidity in exchange for trading fees. Today, we are publishing new research to propose a new metric, fee liquidity-adjusted instantaneous returns (FLAIR), to account for LP competitiveness. Fee Liquidity-Adjusted Instantaneous Returns (FLAIR) Economically rational LPs will consider the cost and benefit of being an LP. The benefits can be quantified through trading fees. Costs are a bit more difficult. In addition to general market risk, LPs must also consider how sophisticated their counterparty is. If a trader has better market information than an LP, the LP risks being on the wrong side of the trade and losing money. One popular metric to measure LP performance is loss-versus-rebalancing (LVR), which quantifies the information imbalance between LPs and their counterparties. However, LVR does not factor in a critical component of AMMs: the intra-pool competitiveness over other LPs that are in the same pool.
FLAIR is a new metric for LP competitiveness that supplements LVR, aiming to capture the dynamic behavior of LPs within a pool. FLAIR reflects reasonable economic intuition that LPs increase competition within pools by allocating capital to pools with higher fee returns, rebalancing in-range liquidity, and timing liquidity deployment to high fee periods. In short, LPs that frequently rebalance will earn more fees on average — a factor that is measured with FLAIR.
Figure 1: Current models would equate the green and blue pools with low costs (LVR). However, FLAIR separates these by LP competitiveness, showing an important second dimension.
FLAIR is designed to measure any number of LP positions over an arbitrary period of time, including a single point-in-time. This metric allows both existing LPs to measure historical performance and prospective LPs to optimize future capital deployments with backtests. Researchers can also categorize different strategies and LPs based on their placement on the quadrant plot shown in Figure 1. Solving for the “optimal frontier” will be a portfolio optimization problem.
Conclusion
FLAIR is not specific to Uniswap v2 and v3 style pools, but is described in our work as a generic formula for most typical cases of AMMs. With some modifications, it can also apply to traditional exchanges – allowing for comparative studies between different market structures and influencing future market designs.
The DeFi landscape is constantly evolving as liquidity provisioning strategies grow increasingly sophisticated. We are excited about this future, and our contribution provides a more holistic view of LP performance. We invite interested readers to consult the full paper for a more detailed understanding.