pith. sign in

arxiv: 2604.15973 · v1 · submitted 2026-04-17 · 💻 cs.CR

Where Does MEV Really Come From? Revisiting CEXDEX Arbitrage on Ethereum

Bence Lad\'oczk , Mikl\'os R\'asonyi , J\'anos Tapolcai This is my paper

Pith reviewed 2026-05-10 08:24 UTC · model grok-4.3

classification 💻 cs.CR
keywords MEVCEX-DEX arbitrageAutomated Market Makersprice jumpsEthereumstochastic processesarbitrage profits
0
0 comments X

The pith

CEX-DEX arbitrage requires trading volumes on the scale of major liquidity pools and produces profits comparable to total MEV.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Standard models of automated market makers assume asset prices follow continuous paths with only small changes, which underestimates profits from trading between centralized and decentralized exchanges. The paper replaces this with a discrete-time model that adds stochastic jumps of arbitrary size to the price process. This change shows that CEX-DEX arbitrage opportunities arise mainly at jumps, demand trading volumes comparable to the total activity in large liquidity pools, and generate returns on the order of observed MEV. The model computes the stationary distribution of mispricing through function iteration and proves the mispricing process is an ergodic Markov chain. When parameters are fit to Ethereum price data, the resulting estimates align closely with empirical measurements.

Core claim

By modeling the asset price as the sum of a diffusive component and stochastic jumps, the extended AMM framework yields arbitrage volumes on the order of major liquidity pool activity and profits comparable to MEV, with estimates that match Ethereum blockchain observations and supply a theoretical account for the scale of MEV revenue.

What carries the argument

Extended discrete-time AMM model in which the price process is the sum of a diffusive component and stochastic jumps of arbitrary distributions, with the stationary mispricing distribution obtained by function iteration on an ergodic Markov chain.

If this is right

  • CEX-DEX arbitrage accounts for a substantial share of MEV revenue without requiring harm to users.
  • The trading volumes needed to capture these opportunities are comparable to the total activity of major liquidity pools.
  • The mispricing process forms an ergodic Markov chain whose stationary distribution can be computed by iteration.
  • The model supplies a natural explanation for several fundamental questions about MEV origins in the Ethereum ecosystem.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If jumps are the dominant source, then systems that reduce price discontinuity (such as faster oracles) would shrink arbitrage profits.
  • The same jump-inclusive framework could be used to estimate arbitrage scale on other chains or in other DeFi venues.
  • Empirical MEV studies that rely on continuous-price assumptions may systematically understate CEX-DEX contributions.

Load-bearing premise

Fitting the parameters of the jump distribution to the same Ethereum price data later used for validation is enough to show that the model quantifies the true scale of CEX-DEX arbitrage rather than simply reproducing its input statistics.

What would settle it

Direct on-chain measurement of aggregate CEX-DEX arbitrage trade volumes and realized profits across multiple Ethereum blocks or days; a large and persistent shortfall relative to the model's predictions would falsify the claimed scale.

Figures

Figures reproduced from arXiv: 2604.15973 by Bence Lad\'oczk, J\'anos Tapolcai, Mikl\'os R\'asonyi.

Figure 1
Figure 1. Figure 1: Flowchart of stochastic model evaluation on real data from an AMM. [PITH_FULL_IMAGE:figures/full_fig_p012_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Comparison of the SPD for five different values of [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Cross-validation measurements on two ETH–stablecoin Uniswap V2 pools [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Convergence of the SPD of the misprice process with ( [PITH_FULL_IMAGE:figures/full_fig_p023_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The average absolute errors of our numerical integrator with respect to [PITH_FULL_IMAGE:figures/full_fig_p024_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: The daily volatility in % vs. the number of DEX transactions per day for [PITH_FULL_IMAGE:figures/full_fig_p026_6.png] view at source ↗
read the original abstract

A central question of the Ethereum ecosystem is where Maximal Extractable Value (MEV)revenue originates and to what extent it stems from harming unsuspecting users. It is acceptable if MEV arises from arbitrages between centralised and decentralised exchanges (CEX-DEX). Yet theoretical models have significantly underestimated the scale of these arbitrages, while empirical studies have highlighted their importance - though these remain conservative estimates, constrained by numerous debatable heuristic assumptions. Revisiting the theoretical model, we found that CEX-DEX arbitrages require trading volumes on the order of the total activity of major liquidity pools and yield profits comparable to MEV. Most prior AMM models utilised the Black-Scholes (BS) stochastic differential equation (SDE) - i.e., geometric Brownian motion - and assumed continuous price trajectories where asset prices move in small increments only.We argue that BS underestimates arbitrage profits by ignoring price jumps, which are precisely the points at which arbitrage opportunities tend to arise. To address this gap, we present an extended discrete-time AMM model in which the price process is the sum of a diffusive component and stochastic jumps that can have arbitrary noise distributions. Although mathematically more involved this framework allows us to employ a general discrete-time SDE and compute the stationary probability distribution via function iteration with geometric convergence. We further prove that the resulting mispricing process is an ergodic Markov chain. We implement our model in C++, collect spot prices and AMM exchange data from the Ethereum blockchain and fit the model parameters to the observed prices. The estimates derived from our model closely match empirical observations and provide a natural theoretical explanation for several fundamental questions in the blockchain ecosystem.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper claims that prior Black-Scholes-based AMM models underestimate CEX-DEX arbitrage because they assume continuous price paths, and that an extended discrete-time model incorporating stochastic jumps (with arbitrary noise distributions) yields arbitrage volume and profit estimates on the order of major liquidity-pool activity and total MEV; these estimates are obtained by fitting jump-distribution parameters to Ethereum spot and AMM price data and are reported to match empirical observations while also proving that the mispricing process is an ergodic Markov chain.

Significance. If the calibration and validation steps can be shown to be independent, the work would supply a mathematically grounded explanation for why CEX-DEX arbitrage is large enough to account for a substantial fraction of observed MEV, addressing a long-standing discrepancy between theory and measurement in blockchain economics. The explicit proof that the mispricing process forms an ergodic Markov chain and the use of geometrically convergent function iteration for the stationary distribution are concrete technical strengths.

major comments (2)
  1. [Abstract] Abstract: the central claim that 'the estimates derived from our model closely match empirical observations' is load-bearing for the paper's contribution, yet the manuscript supplies no description of the fitting procedure, no out-of-sample hold-out set, and no sensitivity analysis with respect to the jump-distribution family; without these, it is impossible to determine whether the reported agreement constitutes an independent test or merely reproduces statistics already present in the calibration data.
  2. [Implementation and Results sections] Implementation and Results sections: because both the jump-distribution parameters and the empirical arbitrage-volume/profit metrics are derived from the identical collection of Ethereum spot prices and AMM pool data, the agreement between model and observation risks being an artifact of the calibration step; a concrete test (e.g., fitting on one time window and validating on a disjoint later window, or using an external price source) is required to substantiate the claim that the extended model quantifies the scale of CEX-DEX arbitrage.
minor comments (2)
  1. The paper would benefit from a short table or paragraph explicitly listing the free parameters of the jump distribution, their fitted values, and the convergence criterion used in the function-iteration procedure.
  2. Notation for the discrete-time SDE and the stationary distribution could be introduced earlier and used consistently to improve readability for readers outside the immediate subfield.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claim rests on the modeling choice that price dynamics are the sum of diffusion and jumps, on the mathematical tractability of the resulting Markov chain, and on parameters fitted to the very data whose statistics are later reproduced.

free parameters (1)
  • parameters of the stochastic jump distribution
    Chosen by fitting to observed Ethereum spot prices and AMM exchange data so that the stationary distribution matches empirical mispricing statistics.
axioms (2)
  • domain assumption Price process equals diffusive component plus stochastic jumps with arbitrary noise distribution
    Core modeling assumption introduced to capture the jumps at which arbitrage opportunities arise.
  • standard math Stationary distribution obtained by function iteration with geometric convergence
    Mathematical technique used to compute the long-run probability distribution of the mispricing process.

pith-pipeline@v0.9.0 · 5613 in / 1468 out tokens · 73484 ms · 2026-05-10T08:24:29.946831+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

31 extracted references · 31 canonical work pages

  1. [1]

    Papers 2309.13648, arXiv.org (Sep 2023)

    Adams, A., Chan, B.Y., Markovich, S., Wan, X.: The Costs of Swapping on the Uniswap Protocol. Papers 2309.13648, arXiv.org (Sep 2023)

    work page arXiv 2023

  2. [2]

    Annales Scientifiques de L’Ecole Normale Supérieure17, 21–88 (1900)

    Bachelier, L.: Théorie de la Spéculation. Annales Scientifiques de L’Ecole Normale Supérieure17, 21–88 (1900)

    work page 1900

  3. [3]

    Bichuch, M., Feinstein, Z.: DeFi Arbitrage in Hedged Liquidity Tokens. Tech. Rep. 2409.11339, arXiv.org (Sep 2024)

    work page arXiv 2024

  4. [4]

    Journal of Political Economy81(3), 637–654 (1973)

    Black, F., Scholes, M.: The Pricing of Options and Corporate Liabilities. Journal of Political Economy81(3), 637–654 (1973)

    work page 1973

  5. [5]

    Papers 2108.06593, arXiv.org (Aug 2021)

    Boueri, N.: G3M Impermanent Loss Dynamics. Papers 2108.06593, arXiv.org (Aug 2021)

    work page arXiv 2021

  6. [6]

    Cartea, Á., Drissi, F., Monga, M.: Decentralised Finance and Automated Market Making:PredictableLossandOptimalLiquidityProvision(November2022).https: //doi.org/10.2139/ssrn.4273989, sIAM Journal on Financial Mathematics

    work page doi:10.2139/ssrn.4273989

  7. [7]

    Applied Mathematical Finance30(2), 69–93 (2023)

    Cartea, Á., Drissi, F., Monga, M.: Predictable losses of liquidity provision in con- stant function markets and concentrated liquidity markets. Applied Mathematical Finance30(2), 69–93 (2023). https://doi.org/10.1080/1350486X.2023.2277957

    work page doi:10.1080/1350486x.2023.2277957 2023

  8. [8]

    Quantitative Finance1, 1–14 (2005)

    Cont, R.: Empirical properties of asset returns: Stylized facts and statistical issues. Quantitative Finance1, 1–14 (2005)

    work page 2005

  9. [9]

    tankov financial modelling with jump processes (2004)

    Cont, R.: P. tankov financial modelling with jump processes (2004)

    work page 2004

  10. [10]

    Springer (2006) Revisiting CEX–DEX Arbitrage on Ethereum 17

    Delbaen, F., Schachermayer, W.: The Mathematics of Arbitrage. Springer (2006) Revisiting CEX–DEX Arbitrage on Ethereum 17

    work page 2006

  11. [11]

    SSRN (2023)

    Dewey, R., Newbold, C.: The pricing and hedging of constant function market makers. SSRN (2023). https://doi.org/10.2139/ssrn.4455835, https://ssrn.com/ abstract=4455835, available at SSRN: https://ssrn.com/abstract=4455835

    work page doi:10.2139/ssrn.4455835 2023

  12. [12]

    Springer, Berlin (2018)

    Douc, R., Moulines, E., Priouret, P., Soulier, P.: Markov chains. Springer, Berlin (2018)

    work page 2018

  13. [13]

    Springer (2019)

    Eberlein, E., Kallsen, J.: Mathematical finance. Springer (2019)

    work page 2019

  14. [14]

    In: Pasik-Duncan, B

    Hanson, F.B., Westman, J.J.: Stochastic analysis of jump-diffusions for financial log-return processes. In: Pasik-Duncan, B. (ed.) Stochastic Theory and Control, pp. 169–183. Springer, Berlin Heidelberg (2002)

    work page 2002

  15. [15]

    The Journal of Prediction Markets1(1), 3–15 (2007)

    Hanson, R.: Logarithmic markets coring rules for modular combinatorial informa- tion aggregation. The Journal of Prediction Markets1(1), 3–15 (2007)

    work page 2007

  16. [16]

    Theyieldcurveandpredictingrecessions

    Hasbrouck, J., Rivera, T.J., Saleh, F.: An economic model of a decentralized exchange with concentrated liquidity (Aug 2023). https://doi.org/10.2139/ssrn. 4529513, https://ssrn.com/abstract=4529513

    work page doi:10.2139/ssrn 2023

  17. [17]

    In: IEEE Symposium on Security and Privacy (SP)

    Heimbach, L., Pahari, V., Schertenleib, E.: Non-atomic arbitrage in decentralized finance. In: IEEE Symposium on Security and Privacy (SP). pp. 3866–3884. IEEE (2024)

    work page 2024

  18. [18]

    Milionis, C

    J. Milionis, C. C. Moallemi and T. Roughgarden: A Myersonian Framework for Optimal Liquidity Provision in Automated Market Makers. In: 15th Innovations in Theoretical Computer Science Conference (ITCS 2024). LIPIcs, vol. 287, pp. 81:1– 81:19. Schloss Dagstuhl – Leibniz-Zentrum für Informatik, Dagstuhl, Germany (2024)

    work page 2024

  19. [19]

    Milionis, C

    J. Milionis, C. C. Moallemi and T. Roughgarden: Complexity-Approximation Trade-Offs in Exchange Mechanisms: AMMs vs. LOBs. In: Financial Cryptography and Data Security. pp. 326–343. Springer Nature Switzerland, Cham (2024)

    work page 2024

  20. [20]

    Journal of Financial Economics3(1-2), 125–144 (1976)

    Merton, R.C.: Option pricing when underlying stock returns are discontinuous. Journal of Financial Economics3(1-2), 125–144 (1976)

    work page 1976

  21. [21]

    Cambridge Uni- versity Press, 2nd edn

    Meyn, S., Tweedie, R.L.: Markov chains and stochastic stability. Cambridge Uni- versity Press, 2nd edn. (2009)

    work page 2009

  22. [22]

    In: International Conference on Finan- cial Cryptography and Data Security

    Milionis, J., Moallemi, C.C., Roughgarden, T.: Automated market making and arbitrage profits in the presence of fees. In: International Conference on Finan- cial Cryptography and Data Security. pp. 159–171 (2024), arXiv preprint arXiv: 2305.14604

    work page arXiv 2024

  23. [23]

    arXiv (2024)

    Milionis, J., Moallemi, C.C., Roughgarden, T., Zhang, A.L.: Automated market making and loss-versus-rebalancing. arXiv (2024)

    work page 2024

  24. [24]

    Milionis, J., Wan, X., Adams, A.: FLAIR: A Metric for Liquidity Provider Com- petitiveness in Automated Market Makers (2023)

    work page 2023

  25. [25]

    arXiv preprint arXiv:2505.05113 (2025)

    Nezlobin, A., Tassy, M.: Loss-versus-rebalancing under deterministic and general- ized block-times. arXiv preprint arXiv:2505.05113 (2025)

    work page arXiv 2025

  26. [26]

    Öz,B.,Sui,D.,Thiery,T.,Matthes,F.:Whowinsethereumblockbuildingauctions and why? In: Advances in Financial Technologies (AFT). p. 1 (2024)

    work page 2024

  27. [27]

    SSRN Electronic Journal (01 2021)

    Park, A.: The conceptual flaws of constant product automated market making. SSRN Electronic Journal (01 2021). https://doi.org/10.2139/ssrn.3805750

    work page doi:10.2139/ssrn.3805750 2021

  28. [28]

    Rao, R., Shah, N.: Triangle fees (2023)

    work page 2023

  29. [29]

    Uhlenbeck, G.E., Ornstein, L.S.: On the Theory of the Brownian Motion. Phys. Rev.36, 823–841 (Sep 1930). https://doi.org/10.1103/PhysRev.36.823

    work page doi:10.1103/physrev.36.823 1930

  30. [30]

    In: Advances in Financial Technologies (AFT) (2025), arXiv preprint arXiv:2507.13023

    Wu, F., Sui, D., Thiery, T., Pai, M.: Measuring cex-dex extracted value and searcher profitability: The darkest of the mev dark forest. In: Advances in Financial Technologies (AFT) (2025), arXiv preprint arXiv:2507.13023

    work page arXiv 2025

  31. [31]

    In: IEEE Symposium on Security and Privacy (SP)

    Yang, S., Nayak, K., Zhang, F.: Decentralization of ethereum’s builder market. In: IEEE Symposium on Security and Privacy (SP). pp. 1512–1530. IEEE (2025) 18 Bence Ladóczki, Miklós Rásonyi, and János Tapolcai A Proofs Proof (Proof of Lemma 1).IfZn+1 = 0then the density ofµ+σεn+1 is 1 σ f y−µ σ as easily seen. IfZ n+1 = 1then we need to calculate the convo...

    work page 2025