pith. machine review for the scientific record. sign in

arxiv: 2605.02287 · v1 · submitted 2026-05-04 · 💱 q-fin.TR · cs.CY· q-fin.GN

Recognition: unknown

Per-Market Information Leakage and Order-Flow Skill: Two Methodological Lenses on Informed Trading in Decentralized Prediction Markets

Maksym Nechepurenko

Pith reviewed 2026-05-08 01:26 UTC · model grok-4.3

classification 💱 q-fin.TR cs.CYq-fin.GN
keywords informed tradingprediction marketsdecentralized financeinformation leakageorder flowsign randomizationmethodological comparisonPolymarket
0
0 comments X

The pith

Three methodological approaches to informed trading in prediction markets measure distinct layers rather than competing alternatives.

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

The paper compares three approaches to detecting informed trading that emerged nearly simultaneously in studies of decentralized prediction markets. It argues that sign-randomization serves as an account-level test of persistent directional skill conditional on opportunity selection, separate from any direct insider-trading test or per-market measure. A lifecycle heuristic flags potential insiders on a separate population of accounts that the skill classifier excludes by design. Because the data pools politics, sports, crypto and other categories that differ in information access, platform-wide skill classifications become ambiguous about underlying mechanisms. The layers can be stacked into a combined pipeline that gains precision by addressing different dimensions, as illustrated by a 2026 enforcement cluster with an external legal benchmark.

Core claim

These are three distinct layers of detection, not competing methods on a single layer. Sign-randomization is best understood as an account-level test of persistent directional skill conditional on opportunity selection -- not a direct test of insider trading, and not a per-market measure. The heuristic insider flag is separate from the skill classifier, applies to a population the classifier excludes by design, and has unknown precision. The Polymarket sample pools politics, sports, crypto, and other categories with different information technologies, so a platform-wide 'skilled winner' classification is mechanism-ambiguous. A combined pipeline gains in precision because each layer filters a

What carries the argument

The layered detection model separating per-market Information Leakage Score (ILS) quantification of front-loading at article-derived public-event timestamps from account-level sign-randomization for directional skill and lifecycle heuristics for insider flagging.

If this is right

  • A combined pipeline of the three layers increases detection precision by addressing separate dimensions of information advantage.
  • Platform-wide skill classifications remain mechanism-ambiguous when markets from different categories are pooled.
  • The insider heuristic targets accounts excluded from statistical skill classifiers by construction.
  • Per-market ILS scoring can quantify how much information front-loads into each individual contract around external event timestamps.
  • External legal enforcement benchmarks help validate the stacking of statistical and investigative layers.

Where Pith is reading between the lines

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

  • Extending per-market leakage scores to non-political categories could reveal how information technologies vary across event types.
  • Regulators could adopt multi-layer pipelines to prioritize enforcement actions on decentralized platforms.
  • Future studies on other prediction-market platforms could test whether the three layers remain distinct or begin to overlap.
  • Single-method detection studies risk missing interactions among skill, leakage timing, and insider status.

Load-bearing premise

The three methods truly measure orthogonal dimensions of informed trading with no substantial overlap.

What would settle it

Re-running the sign-randomization test within individual market categories and finding that it behaves like a per-market leakage measure, or observing large overlap between accounts flagged by the heuristic and those classified as skilled winners, would show the layers are not distinct.

read the original abstract

April 2026 saw notable methodological convergence in the academic study of informed trading on decentralized prediction markets. Three approaches surfaced almost simultaneously: Mitts and Ofir (2026) apply a composite screen to over 210,000 wallet-market pairs; Gomez-Cram et al. (2026) apply an event-level sign-randomization test to Polymarket's complete transaction history, classifying 3.14% of accounts as "skilled winners" and separately flagging 1,950 accounts as "insiders" via a lifecycle heuristic; Nechepurenko (2026) develops the Information Leakage Score (ILS) framework, which quantifies per-market information front-loading at an article-derived public-event timestamp. This paper provides a methodological comparison. The central claim is that these are three distinct layers of detection, not competing methods on a single layer. Sign-randomization is best understood as an account-level test of persistent directional skill conditional on opportunity selection -- not a direct test of insider trading, and not a per-market measure. The heuristic insider flag is separate from the skill classifier, applies to a population the classifier excludes by design, and has unknown precision. The Polymarket sample pools politics, sports, crypto, and other categories with different information technologies, so a platform-wide "skilled winner" classification is mechanism-ambiguous. The January 2026 U.S.-Venezuela operation cluster, where the DOJ indictment of Master Sergeant Gannon Van Dyke provides a rare external enforcement benchmark, illustrates how the layers stack: lifecycle heuristics identify suspicious accounts; legal investigation addresses non-public-information possession; per-market scoring would quantify how much information was leaked into each contract. A combined pipeline gains in precision because each layer filters a different dimension.

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 manuscript claims that three methodological approaches to detecting informed trading in decentralized prediction markets—Mitts and Ofir (2026), Gomez-Cram et al. (2026), and Nechepurenko (2026)'s ILS—are distinct layers of detection rather than competing methods. Sign-randomization tests account-level persistent directional skill, the lifecycle heuristic flags insiders separately, and ILS quantifies per-market information leakage. The paper argues for a combined pipeline and illustrates with the January 2026 U.S.-Venezuela case where layers stack, using the DOJ indictment as benchmark.

Significance. If the orthogonality of the methods holds, the paper would offer a valuable synthesis for researchers studying informed trading on platforms like Polymarket. It provides conceptual clarity on the different scopes and limitations of each approach, particularly the mechanism-ambiguity in pooled samples and the separation of skill from insider status. The emphasis on external benchmarks is a strength, and the framework could guide future empirical work combining these lenses.

major comments (2)
  1. Abstract: The central claim that the three methods measure orthogonal dimensions with no substantial overlap is asserted conceptually (e.g., sign-randomization as account-level skill conditional on opportunity selection, heuristic as separate from the classifier, ILS as per-market) but supplies no quantitative validation such as pairwise overlap statistics, conditional mutual information, or joint application to the same wallet-market pairs. This assumption is load-bearing for the argument that a combined pipeline gains precision.
  2. The January 2026 U.S.-Venezuela illustration: The example demonstrates stacking in one case but does not test whether the layers systematically filter independent information across the sample, leaving the non-redundancy claim unverified.
minor comments (2)
  1. The abstract refers to 'April 2026' for methodological convergence but the arXiv identifier suggests a later date; clarify the exact timeline and publication status of the cited concurrent works in the main text.
  2. Consider adding a summary table comparing the scope (account-level vs. per-market), data requirements, and known limitations of each method to improve readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

Thank you for the constructive feedback. We clarify that the manuscript offers a conceptual comparison of the three methods rather than an empirical joint analysis, and we address the concerns about validation below by agreeing to revisions that better qualify the claims.

read point-by-point responses

  1. Referee: Abstract: The central claim that the three methods measure orthogonal dimensions with no substantial overlap is asserted conceptually (e.g., sign-randomization as account-level skill conditional on opportunity selection, heuristic as separate from the classifier, ILS as per-market) but supplies no quantitative validation such as pairwise overlap statistics, conditional mutual information, or joint application to the same wallet-market pairs. This assumption is load-bearing for the argument that a combined pipeline gains precision.

    Authors: We agree that the orthogonality of the detection layers is advanced on conceptual grounds, drawing from the distinct scopes described: sign-randomization tests account-level directional skill, the lifecycle heuristic identifies insiders separately from the skill classifier, and ILS measures per-market leakage. The manuscript does not include quantitative metrics such as overlap statistics because it is a methodological synthesis comparing approaches from separate studies, not a unified empirical application. We will revise the abstract to explicitly state that the distinctions are conceptual and that empirical quantification of independence (e.g., via joint application to shared data) remains an important avenue for future research. This change will reduce the load-bearing nature of the assumption while retaining the paper's core contribution. revision: yes

  2. Referee: The January 2026 U.S.-Venezuela illustration: The example demonstrates stacking in one case but does not test whether the layers systematically filter independent information across the sample, leaving the non-redundancy claim unverified.

    Authors: The U.S.-Venezuela case is intended as an illustration of how the layers can be stacked in a setting with an external enforcement benchmark, not as a systematic test of independence across the full sample. The non-redundancy argument rests on the methodological differences outlined in the text. We acknowledge that demonstrating systematic filtering would require broader empirical analysis. We will add a brief note in the conclusion recognizing this as a limitation of the current illustration and proposing joint empirical tests as a direction for future work. revision: yes

Circularity Check

0 steps flagged

No circularity in conceptual methodological comparison

full rationale

The paper presents an interpretive comparison of three detection approaches (sign-randomization, lifecycle heuristic, and ILS per-market scoring), distinguishing them by scope and population without any equations, fitted parameters, or derivations that reduce to their own inputs. References to concurrent independent studies and the author's ILS framework serve as contextual framing rather than load-bearing self-citations that force the central claim. The assertion of distinct layers is conceptual and does not exhibit self-definitional loops, renamed known results, or predictions equivalent to fitted inputs by construction. The provided text contains no quantitative derivations or self-referential reductions.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 1 invented entities

Based on abstract only; the ILS appears to be a new constructed metric whose exact formula and any parameters are not specified here.

invented entities (1)
  • Information Leakage Score (ILS) no independent evidence
    purpose: Quantifies per-market information front-loading at an article-derived public-event timestamp
    New framework introduced to measure information timing per contract

pith-pipeline@v0.9.0 · 5630 in / 1221 out tokens · 32473 ms · 2026-05-08T01:26:06.000051+00:00 · methodology

discussion (0)

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

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Manipulation, Insider Information, and Regulation in Leveraged Event-Linked Markets

    q-fin.TR 2026-05 unverdicted novelty 7.0

    Leverage scales market-price manipulation linearly while shifting outcome-manipulation thresholds and multiplying informed-trading rents in three distinct ways, calling for re-allocated regulatory attack surfaces rath...

  2. A Taxonomy of Event-Linked Perpetual Futures: Variant Designs Beyond the Single-Market Binary Case

    q-fin.TR 2026-05 unverdicted novelty 6.0

    The paper organizes seven canonical variants of event-linked perpetual futures along four design axes, supplying payoff definitions, inheritance rules from prior work, and variant-specific constraints.

Reference graph

Works this paper leans on

41 extracted references · 1 canonical work pages · cited by 2 Pith papers

  1. [1]

    Journal of Economic Perspectives , volume =

    Wolfers, Justin and Zitzewitz, Eric , title =. Journal of Economic Perspectives , volume =

  2. [2]

    and Forsythe, Robert and Gorham, Michael and Hahn, Robert and Hanson, Robin and Ledyard, John O

    Arrow, Kenneth J. and Forsythe, Robert and Gorham, Michael and Hahn, Robert and Hanson, Robin and Ledyard, John O. and Levmore, Saul and Litan, Robert and Milgrom, Paul and Nelson, Forrest D. and Neumann, George R. and Ottaviani, Marco and Schelling, Thomas C. and Shiller, Robert J. and Smith, Vernon L. and Snowberg, Erik and Sunstein, Cass R. and Tetlock...

  3. [3]

    Surowiecki, James , title =

  4. [4]

    Information Systems Frontiers , volume =

    Hanson, Robin , title =. Information Systems Frontiers , volume =

  5. [5]

    The Journal of Prediction Markets , volume =

    Hanson, Robin , title =. The Journal of Prediction Markets , volume =

  6. [6]

    and Servan-Schreiber, Emile and Tetlock, Philip and Ungar, Lyle and Mellers, Barbara , title =

    Atanasov, Pavel and Rescober, Phillip and Stone, Eric and Swift, Samuel A. and Servan-Schreiber, Emile and Tetlock, Philip and Ungar, Lyle and Mellers, Barbara , title =. Management Science , volume =

  7. [7]

    and Stiglitz, Joseph E

    Grossman, Sanford J. and Stiglitz, Joseph E. , title =. American Economic Review , volume =

  8. [8]

    Trading volume statistics , year =

  9. [9]

    , title =

    Kyle, Albert S. , title =. Econometrica , volume =

  10. [10]

    and Milgrom, Paul R

    Glosten, Lawrence R. and Milgrom, Paul R. , title =. Journal of Financial Economics , volume =

  11. [11]

    and O'Hara, Maureen and Paperman, Joseph B

    Easley, David and Kiefer, Nicholas M. and O'Hara, Maureen and Paperman, Joseph B. , title =. Journal of Finance , volume =

  12. [12]

    and O'Hara, Maureen , title =

    Easley, David and Kiefer, Nicholas M. and O'Hara, Maureen , title =. Review of Financial Studies , volume =

  13. [13]

    Journal of Finance , volume =

    Easley, David and Hvidkjaer, Soeren and O'Hara, Maureen , title =. Journal of Finance , volume =

  14. [14]

    Journal of Financial Economics , volume =

    Duarte, Jefferson and Young, Lance , title =. Journal of Financial Economics , volume =

  15. [15]

    Flow Toxicity and Liquidity in a High-Frequency World , journal =

    Easley, David and L. Flow Toxicity and Liquidity in a High-Frequency World , journal =

  16. [16]

    and Bondarenko, Oleg , title =

    Andersen, Torben G. and Bondarenko, Oleg , title =. Journal of Financial Markets , volume =

  17. [17]

    and Ready, Mark J

    Lee, Charles M.\ C. and Ready, Mark J. , title =. Journal of Finance , volume =

  18. [18]

    , title =

    Hawkes, Alan G. , title =. Biometrika , volume =

  19. [19]

    Quarterly Journal of Economics , volume =

    Budish, Eric and Cramton, Peter and Shim, John , title =. Quarterly Journal of Economics , volume =

  20. [20]

    , title =

    Platt, John C. , title =. Advances in Large Margin Classifiers , publisher =. 1999 , pages =

    1999

  21. [21]

    Proceedings of the 22nd International Conference on Machine Learning (ICML) , pages =

    Niculescu-Mizil, Alexandru and Caruana, Rich , title =. Proceedings of the 22nd International Conference on Machine Learning (ICML) , pages =

  22. [22]

    arXiv preprint arXiv:2508.03474 , year =

    Saguillo, Oriol and Ghafouri, Vahid and Kiffer, Lucianna and Suarez-Tangil, Guillermo , title =. arXiv preprint arXiv:2508.03474 , year =

    work page arXiv

  23. [23]

    2026 , howpublished =

    Mitts, Joshua and Ofir, Moran , title =. 2026 , howpublished =

    2026

  24. [24]

    2020 IEEE Symposium on Security and Privacy (S&P) , pages =

    Daian, Philip and Goldfeder, Steven and Kell, Tyler and Li, Yunqi and Zhao, Xueyuan and Bentov, Iddo and Breidenbach, Lorenz and Juels, Ari , title =. 2020 IEEE Symposium on Security and Privacy (S&P) , pages =

    2020

  25. [25]

    and Tetlock, Philip E

    Schoenegger, Philipp and Tuminello, Peter S. and Tetlock, Philip E. and Karger, Ezra and Atanasov, Pavel , title =. Science Advances , volume =. 2024 , doi =

    2024

  26. [26]

    , title =

    Leetaru, Kalev and Schrodt, Philip A. , title =. Annual Convention of the International Studies Association , year =

  27. [27]

    2026 , month = apr, note =

    Nechepurenko, Maksym , title =. 2026 , month = apr, note =

    2026

  28. [28]

    2026 , note =

    Nechepurenko, Maksym and Shuvalov, Pavel , title =. 2026 , note =

    2026

  29. [29]

    Soros, George , title =

  30. [30]

    , title =

    Schelling, Thomas C. , title =

  31. [31]

    , title =

    Manski, Charles F. , title =. Economics Letters , volume =

  32. [32]

    2026 , note =

    Tsang, Kwok Ping and Yang, Zichao , title =. 2026 , note =

    2026

  33. [33]

    and Peng, L

    Ng, H. and Peng, L. and Tao, Y. and Zhou, D. , title =. 2026 , note =

    2026

  34. [34]

    and Huang, Taeyoung , title =

    Clinton, Joshua D. and Huang, Taeyoung , title =. 2025 , note =

    2025

  35. [35]

    , title =

    Murphy, Allan H. , title =. Journal of Applied Meteorology , volume =

  36. [36]

    , title =

    Brier, Glenn W. , title =. Monthly Weather Review , volume =

  37. [37]

    , title =

    Gneiting, Tilmann and Raftery, Adrian E. , title =. Journal of the American Statistical Association , volume =

  38. [38]

    2026 , note =

    Nechepurenko, Maksym , title =. 2026 , note =

    2026

  39. [39]

    2026 , month = apr, note =

    Gomez-Cram, Roberto and Guo, Yunhan and Jensen, Theis Ingerslev and Kung, Howard , title =. 2026 , month = apr, note =

    2026

  40. [40]

    Soldier Charged With Using Classified Information To Profit From Prediction Market Bets , howpublished =

    U.S. Soldier Charged With Using Classified Information To Profit From Prediction Market Bets , howpublished =. 2026 , month = apr, day =

    2026

  41. [41]

    2026 , month = apr, day =

    2026