Just-in-Time Liquidity Providers

• Just-in-time liquidity providers are agents that quickly deploy and withdraw capital for individual transactions on DEXs, minimizing exposure to adverse selection.
• They leverage real-time mempool data to optimize fee income by selecting narrow price intervals and rapidly adjusting liquidity based on trade dynamics.
• Increasing JIT liquidity can paradoxically reduce overall pool depth by crowding out passive liquidity providers unless balanced with innovative protocol incentives.

Just-in-time (JIT) liquidity providers are agents who provision capital to a market mechanism only when and where it is most beneficial—typically for a single transaction, block, or market event—rather than continuously or passively. In the context of decentralized exchanges (DEXs) built on blockchains (especially Automated Market Makers, AMMs, and Concentrated Liquidity Market Makers, CLMMs), JIT LPs leverage protocol and infrastructure features such as the mempool to target fee income and minimize exposure to adverse selection and inventory risk. The emergence of JIT LPs shapes market efficiency, protocol design, fee allocation, and the competitive equilibrium among classes of liquidity providers.

1. JIT Liquidity Provision: Mechanisms and Operational Setting

JIT liquidity provision on blockchain-based DEXs exploits mempool transparency, where pending swaps are visible prior to block inclusion. Mechanistically, a JIT LP monitors the mempool, rapidly deposits liquidity immediately before a swap execution (sometimes paying a gas bribe), earns protocol fees on the targeted swap, and withdraws liquidity—often within the same block—after the trade is executed. This timing enables JIT LPs to participate only in transactions perceived as advantageous, distinguishable from both continuous passive LPing and classical sandwich attacks (Capponi et al., 2023).

JIT LPs can avoid exposure to "toxic" informed order flow by declining to participate when adverse selection risk is present. In contrast, passive LPs must commit assets for longer, cannot perfectly time their participation, and remain exposed to the full mix of order flow including informed trading (Capponi et al., 2023).

2. Theoretical Models and Optimization of JIT Strategies

JIT LP strategies have been studied explicitly through transaction-level models in concentrated liquidity market makers (e.g., Uniswap v3). The core optimization problem for a JIT LP is to select the price interval and liquidity amount $(L, a, b)$ that maximizes expected utility for an observed trade, accounting for fee revenue and the price impact cost (difference between the value of assets deposited and what is withdrawn, i.e., impermanent loss). Formally, the JIT LP solves

$$\max_{(L, a, b)} \quad \mathcal{U}(L, a, b) = \mathcal{F}(L, a, b) - \mathcal{C}(L, a, b)$$

subject to capital constraints, where $\mathcal{F}$ denotes earned protocol fees and $\mathcal{C}$ accounts for price impact (Trotti et al., 19 Sep 2025).

JIT LPs take advantage of the ability to select narrow intervals and move liquidity frequently. Existing implementations, however, often neglect price impact modeling, resulting in realized profits that fall short of the theoretical optimum; incorporating price impact, JIT LPs could increase earnings by up to 69% on average over short time windows (Trotti et al., 19 Sep 2025).

Game-theoretic models of DEXs with asymmetrically informed agents formalize the sequential interaction between passive LPs, various trader types, and JIT LPs. JIT providers enter only for uninformed (non-toxic) swaps, capturing fees without being exposed to adverse selection. Their entry dilutes fee income for passive LPs, impacting the long-run supply of passive liquidity (Capponi et al., 2023).

3. Paradoxes and Impacts on Market Liquidity

A central paradox revealed by recent research is that increasing the number of JIT LPs does not always increase overall liquidity in the pool. This counterintuitive effect arises when trade volume from uninformed agents is not sufficiently elastic relative to liquidity pool depth. If added JIT liquidity does not materially increase the aggregate volume of uninformed trading, passive LPs may see their expected utility turn negative (from adverse selection unbalanced by diminishing fee share) and exit, causing an overall reduction in pool depth and sometimes a "liquidity freeze" (Capponi et al., 2023).

The model formalizes a critical threshold in the private value shock size of uninformed traders, denoted $\zeta_U^*$, such that if $\zeta_U < \zeta_U^*$, passive LPs exit as JIT LPs crowd them out; if $\zeta_U > \zeta_U^*$, trades are more elastic, and JIT and passive LPs become complements, increasing liquidity. Explicitly, for fee rate $f$, the threshold at full JIT probability $\pi=1$ is

$\zeta^*(f, 1) = (\sqrt{f} + \sqrt{1 + f})^2.$

(Capponi et al., 2023).

Empirical evidence in CLMMs shows that widespread, optimally deployed JIT liquidity erodes passive LP earnings by up to 44% per trade (Trotti et al., 19 Sep 2025), with JIT profits predominantly driven by price impact rather than solely fees.

4. Protocol Design and Market Structure: Mitigating Adverse Effects

Several protocol-level solutions can mitigate the negative welfare effects of JIT liquidity:

• Two-Tiered Fee Structures: Implementing a split fee mechanism where JIT LPs receive only a fraction $\lambda$ of their nominal fees, with the balance transferred to passive LPs, preserves incentives for both classes and can prevent liquidity freezes. The effective pool share formulas for fee allocation are

$$\text{Passive LPs' effective share} = \frac{\text{Passive Liquidity} + (1-\lambda) \cdot \text{JIT Liquidity}}{\text{Total Liquidity}}$$

$$\text{JIT LP's effective share} = \frac{\lambda \cdot \text{JIT Liquidity}}{\text{Total Liquidity}}.$$

The optimal split maximizes $\lambda$ subject to the constraint that passive LP expected utility remains non-negative (Capponi et al., 2023).

• Cournot Competition Among JIT LPs: Rather than winner-take-all via gas auctions, allowing multiple JIT LPs to compete simultaneously for pool share reduces their individual aggressiveness, lowers the entry threshold for beneficial elasticity, and increases overall liquidity for a broader range of parameter settings (Capponi et al., 2023).
• Smart-Contract Commitment Differentiation: Fee allocation could be smart-contract enforced using LP minimum commitment durations (i.e., splitting fast and slow liquidity classes), supporting robust, duration-based reward distribution (Capponi et al., 2023).

5. General Theory: JIT LPs as Parallel/Dynamic Liquidity Providers

Theoretical frameworks for liquidity provisioning formalize JIT LPs within a general class of parallel or dynamic liquidity providers. Generalized AMM protocols model each LP as a market maker with an individual cost or generating function, with the combined market cost being the infimal convolution of all LPs' cost functions: $$C(q) = \inf_{\{q^i\} : \sum_i q^i = q} \sum_i C_i(q^i)$$ and the dual (price-side) generating function is additive: $G(p) = \sum_i G_i(p)$ where $G_i = C_i^*$, the convex conjugate of $C_i$. In this framework, JIT LPs can instantaneously modify their liquidity function (e.g., add liquidity at a single price point for just one trade), with the protocol natively handling such dynamic involvement and precise, risk-limited capital requirements (Bhaskara et al., 2023).

JIT LPs directly generalize the model to accommodate instantaneous, granular, and compositional liquidity provision. Continuous versions of concentrated liquidity (smooth "soft" buckets) remove architectural frictions inherent to discrete-tick designs, enabling more efficient and responsive JIT interaction as a subset of the general protocol.

6. Broader Consequences, Empirical Observations, and Comparisons

Empirical studies suggest that in practice, the vast majority of liquidity on DEXs is provided by passive, often low-frequency actors, with JIT liquidity arising primarily in the context of sophisticated participants capable of exploiting mempool visibility and responsive transaction timing (Heimbach et al., 2021). JIT LPs are especially successful in volatile, high-fee environments, but, without carefully designed mechanisms, their aggressive fee capture may decrease aggregate liquidity (Tang et al., 15 Nov 2024).

JIT provision has close analogs to "stimulated refill" and "just-in-time" limit order placement in centralized limit order books, where high-frequency traders replenish consumed queues on short timescales while actively managing adverse selection and queue risk (Bonart et al., 2015). In decentralized venues, unique architectural features—such as block-level transaction ordering and fee allocation algorithms—both enable and constrain the manifestation of JIT LP strategies.

7. Mathematical Summary Table

Feature	Passive LPs	JIT LPs
Timing & Duration	Multiblock commitment	Transaction/block-only
Adverse Selection Risk	Always exposed	Avoided via mempool selection
Fee Allocation	Pro-rata (diluted by JIT)	Pro-rata for target swap
Participation Frequency	Low-frequency	High-frequency, event-based
Welfare Impact (potential, per trade)	Eroded as JIT share rises	Maximized (conditional)
Capital Efficiency	Limited by design/protocol	Near-maximal (targeted)

This table synthesizes key differentiators from empirical and theoretical sources (Capponi et al., 2023, Trotti et al., 19 Sep 2025, Bhaskara et al., 2023).

Conclusion

Just-in-time liquidity provisioning constitutes a distinct class of market participation in DEXs and AMMs, characterized by strategic, highly-timed capital deployment to fee-generating trades while minimizing exposure to adverse selection. Although theoretically optimal for individual JIT LPs and potentially improving on some market efficiency metrics (e.g., reduced slippage), this phenomenon introduces design challenges for protocol architects seeking to preserve deep, continuous pools. The solution space includes fee redistribution, competitive market mechanics among JIT LPs, and flexible, compositional liquidity provisioning protocols. Ongoing research continues to refine both the theory and practical implementation of JIT liquidity in decentralized and hybrid marketplaces.

Key references: (Capponi et al., 2023, Trotti et al., 19 Sep 2025, Bhaskara et al., 2023, Tang et al., 15 Nov 2024, Heimbach et al., 2021, Bonart et al., 2015).