Paper: arXiv 2503.22502
Authors: Alif Aqsha, Philippe Bergault, Leandro Sánchez-Betancourt
Abstract
We find the equilibrium contract that an automated market maker (AMM) offers to their strategic liquidity providers (LPs) in order to maximize the order flow that gets processed by the venue. Our model is formulated as a leader-follower stochastic game, where the venue is the leader and a representative LP is the follower. We derive approximate closed-form equilibrium solutions to the stochastic game and analyze the reward structure. Our findings suggest that under the equilibrium contract, LPs have incentives to add liquidity to the pool only when higher liquidity on average attracts more noise trading. The equilibrium contract depends on the external price, the pool reference price, and the pool reserves. Our framework offers insights into AMM design for maximizing order flow while ensuring LP profitability.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 4.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper employs advanced stochastic control and game theory (leader-follower stochastic game) with integral transforms and intensity processes to derive approximate closed-form equilibrium solutions, indicating high mathematical density. However, it lacks backtesting results, real-world data implementation details, or trading performance metrics, relying instead on theoretical modeling and synthetic simulation descriptions.
Research Flowchart
flowchart TD A["Research Goal<br>Find equilibrium contract for AMM<br>to maximize order flow & LP profit"] --> B["Methodology: Leader-Follower Stochastic Game"] B --> C["Game Setup & Data Inputs"] C --> C1["AMM (Leader): Fee & Reserve Policy"] C --> C2["LP (Follower): Liquidity Decision"] C --> C3["Noise Trading: Arrival Rate & Size"] C --> D["Computational Process"] D --> D1["Backward Induction: LP's Stochastic Control"] D --> D2["Forward Optimization: AMM's Contraction Mapping"] D --> D3["Fixed Point Iteration for Equilibrium"] D --> E["Key Findings & Outcomes"] E --> E1["Equilibrium Contract<br>Depends on: External Price, Pool Reference, Reserves"] E --> E2["LP Incentives<br>Liquidity added only when it attracts order flow"] E --> E3["Design Implication<br>Optimal AMM design balances fees and liquidity depth"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper derives the equilibrium contract that an automated market maker offers to strategic liquidity providers, formulated as a leader‑follower stochastic game and solved with approximate closed‑form solutions.
Source sentence
We find the equilibrium contract that an automated market maker (AMM) offers to their strategic liquidity providers (LPs) in order to maximize the order flow that gets processed by the venue. Our model is formulated as a leader-follower stochastic game, where the venue is the leader and a representative LP is the follower.
Assumes 6
- The external price process W and B are Brownian motions and the order arrival processes are independent Poisson processes with intensity a0>0. model
Source sentence
We introduce a probability measure P such that W and B are Brownian motions and ˆN −, ˆN + are Poisson processes with intensity a0 > 0. In this probability measure, W, B, ˆN −and ˆN + are independent.5 2Cartea et al. (2023c) is an earlier version of this paper where arithmetic liquidity pools were introduced.
- The model assumes a2=0 for simplicity, so order arrival intensities do not depend on the pool inventory Y. model
Source sentence
In what follows, we work under the assumption that model parameters are such that one is always in scenario (i), which helps us to obtain approximate closed-form solutions. First, we justify (with market data) that the above assumption is a good proxy for reality.
- Parameters are such that both λ− and λ+ are always above the baseline a0 (scenario i). model
Source sentence
In what follows, we work under the assumption that model parameters are such that one is always in scenario (i), which helps us to obtain approximate closed-form solutions.
- The LP risk‑aversion parameter is set to γ=10⁻¹⁸ and the venue’s risk‑aversion to ζ=10⁻⁶. model
Source sentence
The LP’s risk aversion parameter is γ = 10−18 and the venue’s risk aversion is ζ = 10−6.
Show 2 more
- The fee collected per transaction is approximated by r = 0.01 × ξ × Z0. execution
Source sentence
The initial exchange rate is S0 = Z0 = 2820 in the ETH-USDC pair and the daily volatility in Binance is σ = 0.0569 × S0.11 The average traded quantity in a ten minute window is ξ = 300, and the fee collected by the venue per transaction is approximated by the constant fee r = 0.01 × ξ × Z0.
- The temporary price impact parameter in Binance is set to a = 10⁻¹⁴ (negligible). execution
Source sentence
Lastly, for the figures below we let the temporary price impact parameter in Binance be a = 10−14 (negligible walking the book costs in the external venue).
Data
- Uniswap V2 ETH‑USDC pool data crypto-onchain public
Source sentence
We employ market data from Uniswap V2 and from Binance for the pair ETH-USDC between 1 January 2022 and 30 April 2022. During these four months, we monitor the exchange rate in Binance (the midprice St), the instantaneous exchange rate in the Uniswap pool (Zt), and the trading activity (liquidity taking orders) in the pool.
- Binance ETH‑USDC market data crypto-onchain public
Source sentence
We employ market data from Uniswap V2 and from Binance for the pair ETH-USDC between 1 January 2022 and 30 April 2022. During these four months, we monitor the exchange rate in Binance (the midprice St), the instantaneous exchange rate in the Uniswap pool (Zt), and the trading activity (liquidity taking orders) in the pool.
Costs
Modeled
Fee per transaction r = 0.01 × ξ × Z0 (with ξ=300, Z0=2820) and temporary price impact a = 10⁻¹⁴.
Source sentence
The initial exchange rate is S0 = Z0 = 2820 in the ETH-USDC pair and the daily volatility in Binance is σ = 0.0569 × S0.11 The average traded quantity in a ten minute window is ξ = 300, and the fee collected by the venue per transaction is approximated by the constant fee r = 0.01 × ξ × Z0. Lastly, for the figures below we let the temporary price impact parameter in Binance be a = 10−14 (negligible walking the book
Finds 5
- Under the equilibrium contract, LPs add liquidity only when higher liquidity on average attracts more noise trading. theoretical
Source sentence
Our findings suggest that under the equilibrium contract, LPs have incentives to add liquidity to the pool only when higher liquidity on average attracts more noise trading. We find that if higher liquidity does not attract noise trading, even when the venue employs the optimal contract, the representative LP does not have incentives to add liquidity to the pool.
- Approximate closed‑form solutions to the HJB equations are obtained via a series of simplifications. theoretical
Source sentence
4.3 Approximate closed-form solutions The HJB equations above are difficult to solve. In what follows, we work under the assumption that model parameters are such that one is always in scenario (i), which helps us to obtain approximate closed-form solutions.
- In simulations with calibrated parameters, the representative LP adds roughly 8,000 ETH to the pool and pays about 500 USDT in external venue fees. performance
Source sentence
With these model parameters, the representative LP adds roughly 8000 ETH to the pool for which she pays 500 USDT in fees; for comparison, the spread in Binance for ETH-USDT is typically 0.01 which implies that the trading costs in Binance are in the correct order of magnitude.
- Only 0.37 % of the observed data points violate the assumption that a1 ± a3(St − Zt) > 0. statistical
Source sentence
Thus, 0.37% of the data points violate the assumption that in our model ˆa1 ± ˆa3 (St −Zt) > 0.
- Higher transaction costs in the external venue reduce the amount of liquidity added to the AMM. performance
Source sentence
As expected, all else being equal, the higher the fee charged to LTs, the more the venue collects and the better the contract that the venue offers the representative LP, which translates in more liquidity added to 20 the pool. Higher transaction costs in the external venue diminish the activity in the pool.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 26 related papers). Coverage is partial, so this is neither replication nor refutation.
5 papers in the archive
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (95)
- Lectures on BSDEs, stochastic control, and stochastic differential games with financial applications (2016)
- Fundamentals of Stochastic Filtering (2009)
- Blockchain and distributed ledgers: Mathematics, technology, and economics (2021)
- Price discovery on decentralized exchanges (2024)
- Model-free hedging of impermanent loss in geometric mean market makers — → 2303.11118
- Impermanent loss in uniswap v3 (2021) — link
- An analysis of Uniswap markets (2021)
- The adoption of blockchain-based decentralized exchanges (2021) — link
and 87 more
- Mean field games with partial information for algorithmic trading (2018) — link
- Martingales `a temps discret (1972)
- Pathwise stochastic control with applications to robust filtering (2020)
- Portfolio choice, portfolio liquidation, and portfolio transition under drift uncertainty (2019)
- Parameter uncertainty in the Kalman–Bucy filter (2019)
- Liquidity in competitive dealer markets (2021)
- Optimal execution of portfolio transactions (2001)
- Liquidity provision with adverse selection and inventory costs (2023)
- The Financial Mathematics of Market Liquidity: From optimal execution to market making (2016)
- Algorithmic market making in dealer markets with hedging and market impact (2023)
- Convex analysis and variational problems (1999)
- Internalisation by electronic FX spot dealers (2019)
- A mean-field game of market-making against strategic traders (2023)
- Optimal incentives in a limit order book: a SPDE control approach (2024)
- Automated Market Making: the case of Pegged Assets — → 2411.08145
- Price-Aware Automated Market Makers: Models Beyond Brownian Prices and Static Liquidity — → 2405.03496
- Optimal make-take fees in a multi market-maker environment (2021)
- Optimal trading strategy and supply/demand dynamics (2013)
- Blockchain-based settlement for asset trading (2019)
- Dynamic programming approach to principal–agent problems (2018)
- On the impossibility of informationally efficient markets (1980)
- Liquidity and market structure (1988)
- Detecting Toxic Flow — → 2312.05827
- Nash Equilibrium between Brokers and Traders — → 2407.10561
- Unwinding Stochastic Order Flow: When to Warehouse Trades — → 2310.14144
- Strategic insider trading equilibrium: a filter theory approach (2012)
- Double-execution strategies using path signatures (2022)
- Trading foreign exchange triplets (2020)
- A Mean Field Game between Informed Traders and a Broker — → 2401.05257
- A survey of nonsymmetric Riccati equations (2002)
- Non—Blow—Up Conditions for Riccati—type Matrix Differential and Difference Equations (2000)
- Optimal signal-adaptive trading with temporary and transient price impact (2022)
- Optimal trading: The importance of being adaptive (2021)
- Mean-field games with differing beliefs for algorithmic trading (2020)
- Trading algorithms with learning in latent alpha models (2019)
- Merton’s optimal investment problem with jump signals (2021) — link
- Stochastic calculus and applications (2015)
- Enhancing trading strategies with order book signals (2018)
- Option Volume Imbalance as a predictor for equity market returns (2022) — link
- Unwinding Toxic Flow with Partial Information — → 2407.04510
- Optimal execution with stochastic delay (2021)
- Brokers and Informed Traders: dealing with toxic flow and extracting trading signals (2022)
- Liquidation in target zone models (2018)
- Closed-loop N ash competition for liquidity (2023)
- Continuous-time Stochastic Control and Optimization with Financial Applications (2009)
- Portfolio liquidation games with self-exciting order flow (2022)
- Liquidity in competitive dealer markets (2021)
- Investing for the long run when returns are predictable (2000)
- Optimal Trade Execution under Endogenous Order Flow (2023)
- Algorithmic trading with model uncertainty (2017)
- Continuous-time stochastic control and optimization with financial applications (2009)
- Incorporating signals into optimal trading (2019)
- Incorporating order-flow into optimal execution (2016)
- Algorithmic and high-frequency trading (2015)
- Buy low, sell high: A high frequency trading perspective (2014)
- Matrix Riccati equations in control and systems theory (2012)
- Fast and Slow Optimal Trading with Exogenous Information (2023)
- Market microstructure theory (1998)
- Mean-variance hedging for partially observed drift processes (2001)
- Solvability of differential Riccati equations and applications to algorithmic trading with signals (2022)
- Automated market makers: Mean-variance analysis of lps payoffs and design of pricing functions — → 2212.00336
- Optimal make–take fees for market making regulation (2021)
- Uniswap v4 core [draft] (2023)
- Uniswap v3 core (2021)
- Automated market makers designs beyond constant functions (2023)
- am-amm: An auction-managed automated market maker — → 2403.03367
- Automated market making and loss-versus-rebalancing (2022) — link
- Automated market making and arbitrage profits in the presence of fees — → 2305.14604
- Strategic Bonding Curves in Automated Market Makers (2024)
- Decentralized Finance and Automated Market Making: Predictable Loss and Optimal Liquidity Provision (2024)
- Optimal Portfolio in Partially Observed Stochastic Volatility Models (2001)
- Equilibrium in Functional Stochastic Games with Mean-Field Interaction — → 2306.05433
- High Frequency Trading in a Limit Order Book (2008)
- Execution and statistical arbitrage with signals in multiple automated market makers (2023)
- Predictable losses of liquidity provision in constant function markets and concentrated liquidity markets (2023)
- Improved price oracles: Constant function market makers (2020)
- Liquidity provision of utility indifference type in decentralized exchanges (2025)
- Option market making under inventory risk (2009)
- Decentralised finance and automated market making: Execution and speculation — → 2307.03499
- Informed speculation with imperfect competition (1989)
- Continuous auctions and insider trading (1985)
- Buy Low, Sell High: A High Frequency Trading Perspective (2014) — link
- Market making with alpha signals (2020)
- Riccati Differential Equations (1972)
- PORTFOLIO OPTIMIZATION IN A DEFAULT MODEL UNDER FULL/PARTIAL INFORMATION (2015)
- Optimal Market Making under Partial Information with General Intensities (2020)
- Mean Field Game Systems with Common and Latent Processes (2018)
13 of 95 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.