💻 computer science

Strategic Analysis of Just-In-Time Liquidity Provision in Concentrated Liquidity Market Makers

This paper presents the first formal transaction-level model of Just-In-Time (JIT) liquidity provision in Concentrated Liquidity Market Makers, characterizing optimal strategies that reveal current providers significantly underperform due to unaccounted price impact while demonstrating that strategic JIT deployment enhances trader efficiency at the expense of passive liquidity provider profits.

Original authors: Bruno Llacer Trotti, Weizhao Tang, Rachid El-Azouzi, Giulia Fanti, Daniel Sadoc Menasche

Published 2026-06-12

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Bruno Llacer Trotti, Weizhao Tang, Rachid El-Azouzi, Giulia Fanti, Daniel Sadoc Menasche

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a bustling digital marketplace where people trade digital tokens. In this world, there are two main types of players: Passive Liquidity Providers and Just-In-Time (JIT) Providers.

The Setting: The Digital Marketplace

Think of the marketplace as a giant, automated vending machine (an Automated Market Maker, or AMM).

Passive Providers are like people who fill the vending machine with snacks and leave them there for days or weeks. They earn a small fee every time someone buys a snack, but they take a risk: if the price of snacks changes while they are away, they might lose money compared to just holding the snacks in their pocket.
JIT Providers are like hyper-alert street vendors. They don't keep snacks in the machine all day. Instead, they watch the machine, wait for a specific moment when a big customer is about to buy a huge amount of snacks, and then instantly stuff the machine with extra snacks just for that one transaction. They grab a huge chunk of the fee, then immediately remove their snacks.

The Problem: The "Sandwich" and the Hidden Cost

The paper focuses on these JIT providers, specifically in a modern version of the vending machine called a Concentrated Liquidity Market Maker (CLMM). In this version, providers can choose to put their snacks only in a specific price range (like only selling "chocolate" flavors, not "vanilla").

The JIT providers use a strategy often called a "sandwich attack":

They see a big trade coming.
They add their liquidity right before the trade happens.
The big trade goes through, and the JIT provider collects a massive fee because they provided most of the liquidity for that moment.
They remove their liquidity right after.

The Catch: The paper argues that many JIT providers are playing this game blindly. They are so focused on grabbing the fee that they forget to check the price impact.

The Analogy: Imagine you are a JIT provider. You see a customer coming to buy 100 apples. You rush to add 100 apples to the stand.

The Fee: You get a commission for helping the sale.
The Price Impact: Because you added so many apples right before the sale, the price of apples on the stand might shift in a way that hurts you when you take your apples back out. If the price shifts against you, you might actually lose money on the apples themselves, even though you collected a fee.

What the Paper Found

The researchers built a mathematical model (a set of rules) to figure out the perfect strategy for these JIT providers. They asked: "If you know exactly how the price will move, how much money should you put in, and exactly where, to make the most profit?"

Here are their main discoveries, translated into plain English:

1. The "Blind" Providers are Leaving Money on the Table
The researchers looked at real data from the Uniswap exchange. They found that current JIT providers are acting like drivers who ignore the GPS. They are making trades that lose them money or earn them very little.

The Result: If these providers used the "perfect strategy" calculated by the paper's model, they could have made up to 69% more money on average. They are currently failing to account for how their own actions change the price.

2. It's Not About the Fee; It's About the Price Shift
Most people think JIT providers make money just by collecting the transaction fee. The paper says no.

The Result: For these providers, the real money comes from price impact (the shift in value of the tokens they hold during the split-second they are in the market). In fact, about 93% of their profit comes from this price movement, not the fee itself. They are essentially betting on the price moving in their favor for a split second.

3. The Double-Edged Sword
The paper also looked at what happens to everyone else if JIT providers start playing perfectly.

For Traders (The Customers): It's good news. When JIT providers add liquidity perfectly, the "slippage" (the extra cost a customer pays because the price moves while they are buying) goes down. The trade becomes smoother and cheaper for the buyer.
For Passive Providers (The Long-term Holders): It's bad news. If JIT providers are too good at their job, they steal almost all the fees. The paper estimates that optimized JIT providers could cut the earnings of passive providers by up to 44% per trade.

The Three Types of Trades

The paper categorizes the trades JIT providers face into three "flavors":

The Overpriced Trade: The market price is higher than the machine's price. The JIT provider steps in, sells high, and makes a profit. This is the "easy money" scenario.
The Arbitrage Trade: The machine's price is wrong, and a trader is fixing it. The JIT provider steps in, but the price moves against them. They usually lose money here unless the fee is huge.
The "Overshoot" Trade: The price moves past the fair value and then corrects. The smart JIT provider only puts money in for the part of the trade where they will win, and stays out of the part where they will lose.

The Bottom Line

The paper concludes that while Just-In-Time liquidity providers are a powerful force that makes markets more efficient for traders, they are currently playing a suboptimal game. By using a smarter, math-based approach to decide when to enter and how much to invest, they could drastically increase their profits. However, doing so would come at the direct expense of the long-term, passive investors who keep the machine running.

The authors suggest that the future of these markets depends on finding a balance where these "sniper" providers can do their job without completely eating the profits of the "farmer" providers who keep the system alive.

Technical Summary: Strategic Analysis of Just-In-Time Liquidity Provision in Concentrated Liquidity Market Makers

Problem Statement

While Automated Market Makers (AMMs) like Uniswap V3 have achieved significant adoption, the incentives and behaviors of a specific class of Liquidity Providers (LPs)—Just-In-Time (JIT) LPs—remain theoretically underexplored. Unlike passive LPs who maintain liquidity over extended periods, JIT LPs are strategic agents who momentarily supply liquidity for a single swap to extract disproportionately high fees. These agents typically operate via "sandwich attacks," inserting an "add liquidity" transaction immediately before a target trade and a "remove liquidity" transaction immediately after.

Despite their prevalence and the emergence of technologies facilitating their operation (e.g., Flashbots, Uniswap V4 hooks), there is a lack of principled, transaction-level models describing JIT behavior in Concentrated Liquidity Market Makers (CLMMs). Existing literature often fails to account for the complex trade-off between fee revenue and price impact (impermanent loss) at the granularity of individual transactions, leading to a gap in understanding optimal JIT strategies and their systemic effects on market efficiency and passive LP profitability.

Methodology

The authors develop a formal, transaction-level model for JIT LPs within the CLMM framework (exemplified by Uniswap V3). The methodology proceeds through four main stages:

Theoretical Modeling of Price Impact: The paper redefines "impermanent loss" as "price impact" to encompass both gains and losses. It derives necessary and sufficient conditions (Theorem 3.1) under which a price movement results in non-positive price impact (a gain) for an LP. This analysis distinguishes between trades that move the pool price away from the external market price (favorable) and those that move it toward the market price (unfavorable).
Utility Optimization Formulation: The decision-making process of a JIT LP is modeled as a non-linear optimization problem. The objective is to maximize a utility function $U(s; \theta) = F(s; \theta) - C(s; \theta)$ , where $F$ represents fee income and $C$ represents price impact (cost). The strategy space $S$ includes the choice of liquidity amount $L$ and the price range $(a, b)$ , constrained by a budget $\rho$ .
Algorithmic Solution: The authors prove the existence of a global optimal strategy (Theorem 4.1) and propose an algorithm (Algorithm 1) to compute it. The algorithm iterates over feasible price intervals and solves a one-dimensional non-concave optimization problem to determine the optimal liquidity level $L^*$ for each interval.
Empirical Validation: Using on-chain data from the USDC/WETH pool on Uniswap V3 (January–June 2024), the authors simulate the performance of real-world JIT transactions against the derived optimal strategies. They analyze 6,829 identified JIT swaps to evaluate current performance gaps and market-level implications.

Key Contributions

1. Characterization of the Price Impact Landscape

The paper formally characterizes the conditions under which liquidity provision yields gains or losses. It establishes that:

Diverging Prices: If a trade moves the AMM price away from the external market price ( $q \to q'$ such that the distance increases), the LP experiences non-positive price impact (a gain).
Converging Prices: If a trade moves the AMM price toward the external market price, the LP experiences positive price impact (a loss).
Limiting Behavior: As liquidity $L \to \infty$ , price impact is governed by the misalignment between pool and market prices; as $L \to 0$ , price impact vanishes.

2. Formulation of JIT Utility Optimization

The authors define a transaction-level utility function that captures the trade-off between fee revenue and price impact. They classify optimal JIT behavior into three strategic archetypes based on trade direction:

Overpriced Trades: The pool price moves away from the market price. These are highly profitable for JIT LPs as they capture fees and favorable price impact.
Arbitrageur Trades: The pool price moves toward the market price. These are generally unprofitable for JIT LPs unless fees are sufficiently high to offset the price impact loss.
Overshoot Trades: The trade crosses the market price. Optimal strategies involve concentrating liquidity only in the "favorable" segment of the trade (where price impact is non-positive) rather than the entire range.

3. Algorithmic Strategy and Existence Proof

The paper proves the existence of an optimal strategy for any given swap and provides a computational method to find it. The algorithm demonstrates that optimal strategies often involve allocating the entire budget to specific tick ranges where price impact is favorable, rather than spreading liquidity broadly.

4. Empirical Insights on Real-World Performance

By fitting the optimal solution to real-world data, the study reveals:

Suboptimal Current Behavior: Existing JIT LPs often fail to account for price impact, leading to participation in unprofitable trades. The authors estimate that adopting the optimal strategy could increase JIT earnings by up to 69% on average over small time windows.
Driver of Profitability: Contrary to the assumption that fees are the primary driver, 93.7% of JIT revenue in the studied dataset comes from price impact gains, not from extracting fees from passive LPs.
Market Impact: While current JIT participation is low (<2% of fees), simulations suggest that optimized JIT strategies could reduce passive LP fee income by up to 44% per trade. Conversely, this increased liquidity reduces slippage for traders, improving market efficiency.

Significance and Claims

The paper claims to provide the first formal, transaction-level model of JIT liquidity provision for CLMMs. Its significance lies in:

Bridging the Theoretical Gap: It moves beyond general LP incentive analysis to specifically model the high-frequency, strategic behavior of JIT agents.
Correcting Misconceptions: It demonstrates that current JIT strategies are often suboptimal because they ignore price impact dynamics, and that price impact, not fee extraction, is the primary source of JIT profits.
Systemic Implications: The work highlights a tension in the AMM ecosystem: while strategically optimized JIT liquidity improves capital efficiency and reduces trader slippage, it acts as a frictional force that erodes the profitability of passive LPs.
Practical Utility: The proposed algorithm offers a concrete tool for agents to maximize utility, suggesting that the current "noise" in JIT profitability is due to a lack of optimization rather than inherent market inefficiency.

The authors conclude that while JIT LPs currently operate with significant room for improvement, their strategic deployment can enhance market structure by reducing price dislocations, provided the trade-off with passive LP returns is acknowledged. Future work is suggested to explore protocol designs that balance these competing interests.