Robert Bartlett · Maureen O'Hara (Cornell SC Johnson + Stanford Law) · Apr 16, 2026

The most prestigious paper to drop on Kalshi. O'Hara is one of the most-cited microstructure economists alive. 41.6M trades across 478,167 Kalshi markets. Adapts Kyle's λ and Glosten-Harris decomposition to event markets. Headline: single-name markets exhibit greater informed price impact than broad-based markets; "yes" prices systematically exceed actual settlement rates. Directly bridges prediction markets into the modern adverse-selection canon.

Huaiyu Jia · Luofeng Zhou · Wentao Zhang · Lin William Cong (Cornell) · Siguang Li · Shuo Sun · Apr 22, 2026

Lin William Cong is one of the top crypto-finance economists. Releases a unified research dataset: 770K+ Polymarket markets, 943M trading fills, ~2M oracle events spanning Oct 2020-Mar 2026. Demos NBA outcome calibration and CPI expectation reconstruction. This becomes the canonical research dataset for downstream PolyMarket work.

Joshua Della Vedova · Feb 7, 2026

Reframes the smart-money debate. 222M trades. Retail traders pick winners 51.3% of the time but lose money. Automated traders earn $133M via execution edge (2.52¢/contract average), NOT information edge. Companion to Gomez-Cram et al. — together they argue the "wisdom of informed minority" is actually wisdom of better-executing traders.

Joshua Clinton · Huang (Vanderbilt) · 2025/26 · OSF SocArXiv d5yx2

The accuracy bombshell. Head-to-head comparison of IEM, Kalshi, PredictIt, PolyMarket over final 5 weeks of 2024. PredictIt 93% accurate, Kalshi 78%, PolyMarket 67%. The smallest, capped, retail platform was MOST accurate — inverts the "more liquidity = more accuracy" intuition. Direct policy implication for CFTC position-limit debate.

Joyce Berg · Forrest Nelson · Thomas Rietz (Iowa) · 2024/25 · PS: Political Science & Politics

The original IEM team weighs in on 2024. Reports IEM's Dem-leaning signal (85.7% Dem popular-vote-win price on Sep 29, 2024) vs PolyMarket / Kalshi's Trump-favoring signal. Natural experiment on capped-stakes ($500 IEM) vs uncapped (PolyMarket) markets. Pairs with Clinton-Huang on the stake-cap-vs-accuracy puzzle.

Flynn · Tarkom · Finance Research Letters 2025 · SSRN 5073492

The first published "PolyMarket leads stocks" paper. Daily PolyMarket Trump odds vs DJT (Trump Media) returns. VAR + shock identification + Granger causality. Headline: DJT returns lag PolyMarket odds by 3-5 days. The asset-pricing application this lit has needed — and a direct template for the Snowberg-Wolfers-Zitzewitz redux.

Hsiang-Chieh (Alex) Yang · Apr 10, 2026

First rigorous maker-vs-taker skill analysis on PolyMarket. Skilled traders (top 5% by rolling accuracy) earn $121/market as makers and $63 as takers; extract $228M over three years on zero-fee PolyMarket. Sister paper to Bürgi-Deng-Whelan for Kalshi.

Jonas Gebele · Florian Matthes · Jan 5, 2026

First systematic LOOP-violation study across 10 PM venues. 100K+ events; ~6% appear simultaneously across platforms with persistent 2-4% deviations. Attributes deviations to STRUCTURAL barriers (semantic mismatches, capital controls, KYC) not informational. Effectively closes the "cross-platform arbitrage as regulatory wedge" research direction.

Roberto Gomez-Cram · Yunhan Guo · Theis Ingerslev Jensen · Howard Kung (LBS / Yale) · Dec 10, 2025

Sister paper to the team's "Crowd Wisdom or Informed Minority" SSRN 6617059. Constructs a high-frequency, stake-backed measure of earnings expectations from PolyMarket prices vs analyst consensus and IBES. The corporate-finance application of the team's methodology.

Carson Dudley · Reiden Magdaleno · May 11, 2026

Important negative result. Tests PolyMarket / Kalshi forecasts for influenza hospitalizations and measles cases during 2025-26. PMs underperform CDC FluSight and simple statistical baselines. Counterpoint to the "Kalshi macro beats consensus" narrative — domain-heterogeneity matters.

Pu Cheng · Juncheng Liu · Yunshen Long · Apr 3, 2026

38,666 PolyMarket binary markets paired with CLOB and real-time news streams. Evaluates 7 frontier LLMs with execution simulation — only 2 of 7 generate positive financial returns. First contamination-proof, financially grounded LLM benchmark using PMs.

Arcada Labs / Harvard team · Mar 28, 2026

First live-money cross-platform AI benchmark. AI agents trade autonomously with real capital on Kalshi + PolyMarket for 57 days. Frontier models post -16% to -31% on Kalshi but only ~-1% on PolyMarket — significant platform-design heterogeneity for algorithmic traders.

Agostino Capponi (Columbia) · Alfio Gliozzo (IBM) · Brian Zhu · Dec 2, 2025

Capponi is a top mathematical-finance economist at Columbia. LLM-agent pipeline clusters PolyMarket markets into topics and discovers correlated / anti-correlated outcome pairs. Backtested strategies earn ~20% weekly returns. Academic gravitas for agentic-AI trading on PMs.

Lukas Nel · Dec 17, 2025

300 Kalshi questions with verifiable post-cutoff outcomes. Tests Claude Opus 4.5, GPT-5.2, DeepSeek-V3.2, Qwen3-235B, Kimi-K2. Systematic overconfidence across frontier models; reasoning variants do worse. Uses Kalshi as epistemic-calibration ground truth — relevant for AI-econ literature.

Rajat M. Barot · Arjun S. Borkhatariya · Apr 4, 2026

Swarm of 50 LLM personas with confidence-weighted Bayesian aggregation, KL/JS divergence inefficiency detection, and quarter-Kelly sizing for latency arbitrage on PolyMarket. CEX-implied probability arbitrage on stale PM prices.

Atanasov · Witkowski · Mellers · Tetlock · International Journal of Forecasting 41(2) · 2025

Multi-year geopolitical tournament. Small elite crowds outperform larger sub-elite ones; markets vs polls statistically tied. Suggests Kalshi/PolyMarket's mass-participant model may be matched by smaller elite forecaster panels — Brier-score relevance.

Schoenegger · Tuminauskaite · Park · Bastos · Tetlock · Science Advances 10(45) · Nov 8, 2024

12-LLM ensemble forecasts 31 binary Metaculus questions. LLM crowd statistically indistinguishable from 925 human forecasters and beats no-info benchmarks. Foundational for AI-vs-PMs accuracy literature.

Halawi · Zhang · Yueh-Han · Steinhardt (UC Berkeley) · NeurIPS 2024

Retrieval-augmented LM system reaches near-crowd accuracy on out-of-distribution post-cutoff forecasting questions. Aggregates Metaculus, Good Judgment Open, INFER, PolyMarket.

Karger · Atanasov · Tetlock · et al. · Forecasting Research Institute · 2023

169 experts/superforecasters forecast existential risk. AI experts assign 6% AI x-risk vs 1% from superforecasters by 2100. Documents huge sub-population calibration heterogeneity — a phenomenon real-money markets would aggregate via prices but may underweight expert reasoning.

Joyce Berg · Thomas Rietz · International Journal of Forecasting 35(1) · 2019

Canonical IEM longshot result. Standard "longshot bias" (racetrack) does NOT appear on IEM; overconfidence biases prices only at intermediate horizons. Useful priors for what efficient PM behavior looks like absent whale incentives — pairs with Berg-Nelson-Rietz 2024 PS for the IEM-vs-PolyMarket 2024 comparison.

Hardhik Mohanty · Bhaskar Krishnamachari · Apr 2026

KXFED (Fed-rate) repricing predicts BTC realized volatility (p<0.001); KXCPI (inflation) repricing predicts altcoin volatility (ETH, SOL, ADA, LINK). Connects Kalshi macro contracts to cross-asset volatility forecasting — first formal "Kalshi → crypto" empirical channel.

David Krause · Apr 8, 2026

Daily PolyMarket odds for CLARITY Act passage (Jan-Apr 2026) as a regulatory-uncertainty regressor on Bitcoin returns. Multiple regression + Granger causality. Direct application of PolyMarket prices as a regulatory-uncertainty index — methodology will proliferate in macro-finance.

Alex Travin · Nov 30, 2025

Evaluates PolyMarket and Kalshi against IC expert forecasts using Brier scores and ICD-203 standards. Argues for IC integration; discusses insider-trading and manipulation risks. Fills the "PolyMarket vs CIA/State Dept" comparative gap.

Bridget Smart · Ebba Mark · Anne Bastian · Josefina Waugh (Oxford INET) · Jan 30, 2026

ABM simulations characterizing how high-budget agents introduce price distortions in PMs. Analyzes persistence under herding and stubborn-agent regimes. Complements Rasooly-Rozzi's Manifold field experiment with structural ABM methodology.

Maksym Nechepurenko · Apr-May 2026 (5 papers)

Methodological arms race in insider-trading detection. Develops the ILS framework to detect informed trading on PolyMarket. Companion papers on: (a) deadline-resolved extension; (b) population-scale evaluation across 12,708 markets (Oct 2020-Apr 2026); (c) per-market vs order-flow skill reconciliation. Synthesis paper argues Mitts-Ofir + Gomez-Cram + ILS are distinct detection LAYERS, not competing methods. Three competing methodologies all arrived in early 2026 — this direction is now crowded.

Robin Hanson · Overcoming Bias · Nov 25, 2024 + Dec 7, 2025

First-person commentary from the field's intellectual founder. Positions Kalshi/PolyMarket inside his decades-long futarchy vision. Warns "a new prudish temperance movement may shut them down." Argues futarchy now seeing "substantial trials."

Tyler Cowen · Bloomberg Opinion · Feb 27, 2026 + multiple MR posts 2024-2025

Cowen's most-cited 2026 op-ed. Compares PMs to "a new space telescope for social science." Plus MR coverage of: Kalshi-CFTC win (Sep 2024), Kalshi 4.05% cash interest, defense of PolyMarket vs whale-manipulation thesis (Nov 2024), Fed-chair-prediction-market analysis (Jul 2025).

Justin Wolfers · X.com (Nov 3, 2024) · PBS NewsHour (Feb 19, 2026)

Wolfers' most cited 2024-2026 PM interventions. X.com: flags Trump's PolyMarket odds drop (66% → 54%); raises "whale grip diminishing?" question. PBS NewsHour: historicizes PMs (1800s Wall Street curb), endorses informational value, warns about Venezuela insider trading + gambling addiction. Wolfers has NOT published an academic paper.

Andrew Gelman · Statistical Modeling · Nov 10, 2024 + Sep 25, 2025

Most-cited statistician's most explicit PM commentary. First post argues 2024 PM accuracy is N=1 "lucky bounce" analogous to 2008 poll-aggregator performance. Second engages Molly White; flags Ackman-Adams-Cuomo NYC mayor scenario as crossing a line; argues PMs symbolically normalize political corruption.

Robin Hanson · Eric Zitzewitz · Decrypt · Oct 11, 2024

Direct Hanson-vs-Zitzewitz debate. Hanson: insiders make prices more accurate. Zitzewitz (Dartmouth) counters that reputational damage from unregulated insider trading would scare off liquidity-providers and undermine accuracy. The clearest academic-debate framing of the 2024-26 insider-trading policy fight.

Matthew Yglesias · Slow Boring · Feb 25, 2026

Distinguishes prediction-market value from sports-betting harm; describes Kalshi as "basically a sports-gambling company these days." Calls for federal policy preserving PMs while letting states regulate sports gambling. The clearest moderate-Democratic policy position on the boundary.

Eric Zitzewitz · CNN Business (May 9, 2025) · Undark (May 5, 2026)

CNN: papal-conclave markets "one of the ones that you'd expect to be the least well-calibrated since they only get a data point every decade or two." Undark: long-form feature with on-record Rothschild + Gelman + Sethi quotes — argues PMs won't replace polling but complement it.

Scott Alexander · Astral Codex Ten · Sep 17 + Nov 2024

Most rigorous online ex-post calibration analysis of the 2024 PolyMarket signal. Argues Trump price was off by ~10¢ vs Metaculus; attributes inflation to "Theo" the French whale; updates trust in PolyMarket vs Metaculus from 90% to 88%.

Anthony M. Diercks · Jared Dean Katz · Jonathan H. Wright · Jan 2026

The closest thing to a Top-5 paper in the modern era. Documents that Kalshi's modal forecast had a perfect record for FOMC rate decisions since 2022, beating both Bloomberg consensus and fed funds futures. Establishes Kalshi macro markets as a high-frequency nowcasting benchmark. Companion FEDS WP 2026-010.

Mikhail Chernov · Vadim Elenev · Dongho Song · Jan 2025

Time-series econometric framework combining PolyMarket prices, polls, and economic fundamentals; identifies a two-factor structure driving state-level voter preferences and clusters of electorally similar states. Best example of using prediction-market data as a macro/political econometric input.

Roberto Gomez-Cram · Yunhan Guo · Theis Ingerslev Jensen · Howard Kung · Apr 2026

Headline finding: across 1.72M PolyMarket accounts and $13.76B in volume spanning 98,906 events, only ~3.14% of accounts are "skilled" and they drive most price discovery. Settles the wisdom-of-crowds vs wisdom-of-whales debate squarely on the "informed minority" side. Likely the single most-cited 2026 PolyMarket paper.

Pat Akey · Vincent Grégoire · Nicolas Harvie · Charles Martineau · Mar 2026

1.4M users, $20B in volume: top 1% capture 84% of trading gains; fewer than 30% earn positive returns. Longshot betting drives retail losses; market-making is the strongest predictor of positive returns. The closest analog to the household-finance "active management" literature applied to prediction markets.

Rajiv Sethi · Julie Seager · Fred Morstatter · Daniel M. Benjamin · Anna Hammell · Tianshuo Liu · Sachi Patel · Ramya Subramanian · Jun 2025

Compares PolyMarket to three statistical models (Economist, Silver Bulletin, 538) across 2024 races. PolyMarket beats models on high-profile races (presidency) but underperforms on Senate/House contests where liquidity is lower. The clean academic version of the "PolyMarket beats polls" claim.

Felix Reichenbach · Martin Walther · Dec 2025

124M+ PolyMarket trades. Market prices closely track realized probabilities and slightly outperform bookmaker odds. Documents a "default Yes" bias but no general longshot bias — important deviation from racetrack/sportsbook patterns.

Nam Anh Le · Feb 2026

292M trades across 327K binary contracts on Kalshi+PolyMarket. Decomposes calibration into 4 components explaining 87.3% of variance. Persistent underconfidence in political markets, overreaction in weather, near-ideal calibration in sports — a clean domain-heterogeneity result.

Laurie E. Cutting et al. (Vanderbilt) · Jul 2025

PolyMarket vs traditional polling for 2024. PolyMarket superior in swing states (AZ, GA, NC, PA), faster to adjust, but the authors caution about manipulation and demographic skew limiting broad adoption.

Allen Sirolly · Hongyao Ma · Yash Kanoria · Rajiv Sethi · Nov 2025 · 80-page paper

Iterative network-based procedure detects wash-trading clusters on PolyMarket. ~25% of historical volume flagged, peaking at ~60% in December 2024. One cluster contains 43,000+ wallets. Widely covered by Fortune/CoinDesk — the headline integrity-of-market paper of late 2025. Methodologically important: builds a reproducible wallet-clustering pipeline for prediction-market on-chain data.

Joshua Mitts · Moran Ofir · Mar 2026

First systematic empirical-and-legal study of insider trading in prediction markets. 93,000+ PolyMarket markets, 50,000+ wallets. Flagged traders show 69.9% win rate and ~$143M anomalous profits. Headline cases: Iran strike (Trump 2025), Taylor Swift engagement, Maduro raid. Sets the doctrinal framework for treating prediction-market insider trading as commodity fraud.

Siyang Liu · 2026

Forensic blockchain analysis identifies Sybil clusters of PolyMarket accounts funneling winnings to single addresses. Market makers lose more per dollar in compromised markets — a clean Kyle-Glosten adverse-selection result. Companion to Mitts-Ofir but more methodologically focused on the on-chain attribution problem.

Aaron Rasooly · Mahmood Rozzi · Mar 2025

First model tracing the full price path after a manipulation shock combined with an 817-market field experiment. Effects persist ~60 days but decay; markets with more traders, volume, and external anchors are more robust. The new generation's empirical+theoretical complement to Hanson-Oprea-Porter (2006).

Yasaman Rohanifar et al. · Feb 2026

Qualitative sociotechnical audit of PolyMarket. Proposes the "prediction laundering" framework — four stages (structural sanitization, probabilistic flattening, architectural masking, epistemic hardening) by which subjective bets become authoritative-looking probabilities. The first STS-flavored critique of decentralized prediction markets.

Hunter Ng · Lin Peng · Yubo Tao · Dexin Zhou · 2025

First systematic comparison of PolyMarket, Kalshi, PredictIt, and Robinhood during the 2024 election. PolyMarket leads Kalshi in price discovery when liquidity is high; economically meaningful cross-platform arbitrage exists. Natural extension of Makarov-Schoar (JFE 2019) crypto cross-exchange methodology to event markets.

Oriol Saguillo et al. (IMDEA Networks) · Aug 2025 · LIPIcs.AFT.2025.27

April 2024-April 2025 PolyMarket arbitrage analysis. Identifies market-rebalancing (intra) and combinatorial (inter) arbitrage; estimates ~$40M total profit extracted by arbitrageurs over the year. Best taxonomy paper for the cross-market arbitrage landscape.

Jiaxin Yang · Guang Cheng · HaoXuan Zou · May 2026

Reconstructs 75M+ limit-order-book snapshots across 173 NBA games. 7 single-market arbitrage episodes (median 3.6s) vs 290 combinatorial inefficiencies concentrated in the final minutes of live play. The clean sports-market parallel to the election-market arbitrage work.

Barry Eichengreen · Ganesh Viswanath-Natraj · Junxuan Wang · Zijie Wang · 2025

Uses PolyMarket data on Powell-removal contracts cross-referenced with rate-decision contracts via wallet-level tracking. Tests whether beliefs about central bank independence link to interest-rate expectations. Eichengreen brings macro-finance gravitas; first paper to use wallet-attribution as macro-econometric input.

Daniel Rabetti · Jiaqi Shao · Che Zhang · Mar 2026

469 PolyMarket firm-quarter earnings-beat contracts (Sept 2025-Feb 2026). PolyMarket outperforms analyst consensus on US stock earnings predictions. Provocative finding — pushes prediction markets into the analyst-forecasting literature.

Philipp Dubach et al. · Apr 2026

30B WebSocket events over 52 days joined to on-chain trades. Documents 8 stylized microstructure facts including longshot spread premium, depth concentration, wash-share medians. Surprising finding: WebSocket-inferred trade direction agrees with on-chain ground truth only ~59% of the time — a methodological warning for the empirical literature.

Constantin Bürgi · Wanying Deng · Karl Whelan · 2025/26

300,000+ Kalshi contracts. Documents favorite-longshot bias. Takers lose ~32% on average vs makers losing only ~10%. Develops theoretical framework for Kalshi's quote-driven microstructure — the dominant academic treatment of regulated US event markets.

Nahid Rahman · Joseph Al-Chami · Jeremy Clark · Oct 2025

Systematization of knowledge spanning DePMs back to 2011. Presents an 8-stage modular workflow (infrastructure, market topic, share structure, initialization, trading, resolution, settlement, archiving) analyzing decentralization / expressiveness / manipulation-resistance trade-offs. The state-of-the-art primer for anyone entering the literature.

Caleb Maresca (NYU) · Feb 2026

Agent-based simulation with LLM traders. Finds interest-bearing positions eliminate ~83% of the horizon effect on accuracy and triple market participation (17%→62% of wealth). Major mechanism-design implication: long-horizon markets need yield-on-capital to attract liquidity.

Shaw Dalen · Oct 2025

Proposes a logit jump-diffusion with risk-neutral drift treating traded probability as a Q-martingale. Provides calibration pipeline and a derivatives layer (variance, correlation, corridor, first-passage instruments). Positions itself as the prediction-market analog of Black-Scholes.

Kwok Ping Tsang · Zichao Yang · Mar 2026 (v1) · May 2026 (v2)

The single most-cited Polymarket-2024-anatomy paper. Transaction-level on-chain Polygon analysis. Naive volume aggregation overstates the October Trump market ($958M vs $391M actual). Arbitrage half-lives fell from hours to <1 minute. Crucially, the "whale episode" showed two-sided flow consistent with heterogeneous beliefs, not manipulation — the rigorous answer to the Fredi9999 controversy.

Kwok Ping Tsang · Zichao Yang · Mar 2026

Companion to Anatomy paper. Studies how PolyMarket processed three 2024 shocks: Biden-Trump debate, Trump assassination attempt, Biden dropout. Uses Kyle-style price-impact and Glosten-Harris decomposition. Debate effect reversed; assassination effect persisted; dropout produced two-sided trading — clean event-study evidence on prediction-market price discovery.

Maksym Nechepurenko · Apr 2026

Argues prediction markets are not just forecasts but coordination devices producing common knowledge. Uses 2024 transaction data to show that "social force" depends on persistence, trader-type breadth, and cross-platform consensus rather than market size. The cleanest articulation of the self-fulfilling-prophecy framework for PolyMarket.

John T. Holden · Matthew C. Turk · Marc Edelman · Mar 2026

Legal analysis of Kalshi/PolyMarket sports contracts as CFTC "swaps" vs state gaming jurisdiction. Argues for concurrent state-federal regulation given the post-CFTC v. Kalshi court split. The leading legal-academic treatment of the sports-prediction policy fight.

Yicheng Yang · Mar 2026

Structural selection framework for prediction-market pricing as incomplete markets. Latent Gaussian threshold factor with exponential tilt produces a Wang transform as a variance-preserving selection rule — a clean asset-pricing-theory contribution.

Sahil Puri · Jan 2025

PolyMarket English Premier League betting analysis — examines odds vs realized outcomes and post-goal odds reactions. Single-author retail study but useful early data point on sports-prediction calibration.

U.S. Court of Appeals D.C. Circuit · Oct 2, 2024

The legal pivot point. Denies CFTC's emergency stay motion, holding the agency failed to show irreparable harm. Permits Kalshi's Congressional-control election contracts to go live the same day — re-opening regulated US election prediction markets after a 12-year freeze.

CFTC · Sept 22, 2023

The 3-2 CFTC vote prohibiting Kalshi's self-certified political event contracts as "gaming" under CEA §5c(c)(5)(C). Defines the regulatory theory the courts later rejected.

CFTC · Jan 3, 2022

$1.4M civil penalty + cease-and-desist + wind-down of non-compliant markets. CFTC finds PolyMarket's contracts are "swaps" requiring DCM registration. The foundational ruling that drove PolyMarket's 2022-2025 offshore-for-US-users posture.

DOJ Office of Public Affairs · Apr 2026

First federal criminal indictment treating prediction-market trades as commodity-fraud insider trading. MSgt. Gannon Ken Van Dyke (Maduro raid). 5-count felony: unlawful use of confidential information, commodities fraud, wire fraud, unlawful monetary transaction, theft of nonpublic government information.

Congressional Research Service · 2025/26

Neutral CRS framing of CFTC jurisdiction, state-federal preemption, consumer protection, and election-integrity concerns. What Congressional staff actually read — sets the policy vocabulary for committee hearings.

Congressional Research Service · 2026

How STOCK Act §4c(a)(3) of the CEA applies to event contracts. References the Maduro and Iran-strike trading episodes. Reviews the proposed PREDICT Act (H.R. 8076) and End Prediction Market Corruption Act (S. 4017). The legal-theory primer for current policy debate.

Brookings Center on Regulation and Markets · Apr 14 + May 12, 2025 (two-part series)

Aaron Klein + Sen. Jeff Merkley on whether event contracts are innovation or repackaged gambling. Documents Brookings' shift from "interesting innovation" to "needs regulation" framing over a single month — the moderate-Democratic policy position consolidating.

Council on Foreign Relations · 2026

National-security framing of geopolitical contracts: Iran strikes, Maduro, Khamenei succession, Taiwan, Ukraine ceasefire. Notes the ~$550K "Iran trade" and Israeli IDF reservist charges. CFR argues some contracts effectively price national-security secrets.

Rep. Ritchie Torres House Office · 2026

Legislative response from a pro-crypto Democrat — frames insider trading as the failure mode that could kill prediction markets if unaddressed. The moderate alternative to Merkley's ban approach.

Lowenstein Sandler LLP client alert · Oct 2024

Lawyer-grade synthesis of the Sept 6 district-court ruling + Oct 2 D.C. Circuit denial. The best one-stop legal explainer for an academic reader unfamiliar with CEA §5c(c) terminology.

Epstein Becker Green · 2025

Tracks Nevada, NJ, Maryland, Massachusetts, and Ohio federal-preemption suits — the second wave of litigation after the CFTC battle was won. Documents the federal court split on CEA preemption of state gambling regulation.

Wall Street Journal · Nov 1, 2024

The canonical sourcing for the manipulation-vs-information-trader debate. WSJ identifies "Théo," ex-bank trader, behind Fredi9999/Theo4 wallets. He claims his $30M+ Trump bet was a bet against polling bias via a "neighbor poll" methodology, not political manipulation. Frames almost all subsequent academic work.

Chainalysis (X thread + Bloomberg writeup) · Nov 7, 2024

The only public on-chain methodology + dataset for the "single whale or multiple traders" question. Blockchain forensics: funding-pattern clustering + exchange-deposit address overlap. 9 confirmed + 2 suspected wallets ≈ $85M profit estimate.

Rajiv Sethi (Columbia/Barnard) · Imperfect Information Substack · Oct/Nov 2024

The highest-quality real-time economist analysis of the whale episode. Walks through Fredi9999's ~7M-contract position, average entry price, mark-to-market dynamics. Companion post documents an actual failed manipulation attempt (~$11K accumulating ~140K contracts at $0.08) absorbed by arbitrageurs — the cleanest empirical illustration of Hanson's "manipulators subsidize informed traders" claim. Heavily cited by subsequent SSRN papers.

Slate · Apr 2026

Long-form on MSgt. Van Dyke's $33K-to-$409K Maduro-raid trade. Documents DOJ's 5-count felony indictment and the cover-up attempt. Best narrative synthesis of the case that's now the proof-point for prediction-market insider-trading reform.

CoinDesk · Jul 7, 2025

$200M-volume market resolved against media consensus due to UMA oracle voting. Top-10 UMA holders control ~30% of average participation. Cleanest case study of oracle-governance failure in a decentralized prediction market.

Orochi Network · 2025

Technical writeup of the March 2025 governance attack on the "Ukraine agrees to Trump's mineral deal" market. A single actor controlling 25% of UMA voting power flipped resolution to "Yes" despite no agreement. Best technical accounting of how 25%-vote-share + slashing dynamics produced an objectively wrong resolution on a $7M market.

Matt Levine · Bloomberg Opinion · Nov 2024 + Apr 2026

The finance-trader-lens treatment. Two newsletters book-ending the modern era: the November 2024 column establishing that prediction markets had "arrived," and the April 2026 column on algorithmic / LLM-driven prediction-market trading.

Alex Tabarrok · Marginal Revolution · Oct 2024

Academic blog noting Trump-contract price gaps between Kalshi ($0.49) and PolyMarket ($0.53). One of the first observations of the regulatory-wedge phenomenon that Ng-Peng-Tao-Zhou (SSRN 2025) later formalized.

Fortune (Robin Hanson interview) · Apr 26, 2026

Hanson interview arguing insider trading is "the feature, not the bug" — informed traders make markets more accurate. The intellectual counterweight to the Slate/Torres reform framing. Important to read alongside Mitts-Ofir (2026).

Justin Wolfers · Eric Zitzewitz · Journal of Economic Perspectives 2004 · ~3,000+ cites

The canonical survey. Establishes that prediction-market prices aggregate disperse information into forecasts that typically outperform moderately-sophisticated benchmarks. Every prediction-market paper since cites this.

Justin Wolfers · Eric Zitzewitz · NBER WP 12083 · 2006 · 600+ cites

Handbook-style follow-up synthesizing the theoretical case for information aggregation and the empirical record of market-generated forecasts.

Justin Wolfers · Eric Zitzewitz · NBER WP 12200 · 2006 · 600+ cites

Provides analytic conditions under which prediction-market prices correspond to mean population beliefs. Direct theoretical response to Manski (2006).

Charles F. Manski · Economics Letters 91(3) · 2006 · 800+ cites

The key theoretical caveat to "prices = probabilities." Shows that under heterogeneous beliefs and risk neutrality, the price only partially identifies central tendency and reveals nothing about belief dispersion.

Erik Snowberg · Justin Wolfers · Eric Zitzewitz · NBER WP 18222 · Handbook of Economic Forecasting 2013 · 400+ cites

Handbook chapter advocating integration of prediction-market prices into macro forecasts. Reviews accuracy across political, sports, and economic event markets.

Arrow · Forsythe · Gorham · Hahn · Hanson · Ledyard · Levmore · Litan · Milgrom · Nelson · Neumann · Ottaviani · Schelling · Shiller · Smith · Snowberg · Sunstein · Tetlock · Varian · Wolfers · Zitzewitz · Science 320 · 2008 · 600+ cites

The multi-Nobel manifesto in Science arguing that US regulators should liberalize prediction markets. The rallying-cry policy paper of the era.

Robert Forsythe · Forrest Nelson · George Neumann · Jack Wright · American Economic Review 82(5) · 1992 · 1,000+ cites

The founding paper of the field. Shows the 1988 Iowa Political Stock Market outperformed opinion polls in forecasting the presidential election despite biased individual traders.

Joyce Berg · Robert Forsythe · Forrest Nelson · Thomas Rietz · Handbook of Experimental Economics Results · 2008 · 400+ cites

Comprehensive review of 49 IEM election markets from 1988-2000. Documents election-eve forecast errors of ~1.4% for US presidential races.

Joyce Berg · Forrest Nelson · Thomas Rietz · International Journal of Forecasting 24(2) · 2008 · 500+ cites

IEM markets beat 964 polls 74% of the time, with the gap widening at longer horizons (>100 days out).

Joyce Berg · Thomas Rietz · Information Systems Frontiers 5(1) · 2003 · 400+ cites

Argues prediction markets serve as conditional decision-support tools (e.g., IEM 1996 conditional-on-nominee markets). Establishes the "decision markets" use case that later motivates Hanson's futarchy.

David Rothschild · Public Opinion Quarterly 73(5) · 2009 · 500+ cites

State-level 2008 data: debiased prediction-market forecasts beat debiased poll-based forecasts on probabilities of victory, especially early in the cycle.

Robert Erikson · Christopher Wlezien · Electoral Studies 31(3) · 2012 · 200+ cites

The leading dissent from "markets-dominate-polls" consensus. Argues that once polls are adjusted for known biases they beat raw market prices.

Bo Cowgill · Eric Zitzewitz · Review of Economic Studies 82(4) · 2015 · 300+ cites

Internal corporate-market data: 25% lower MSE versus expert forecasts (with an optimism bias at Google). The definitive intra-firm prediction-market study.

Atanasov · Rescober · Stone · Swift · Servan-Schreiber · Tetlock · Ungar · Mellers · Management Science 63(3) · 2017 · 300+ cites

Randomized 2,400-forecaster Good Judgment Project comparison. Properly-aggregated team polls beat continuous double-auction markets by ~12% in Brier score — the strongest empirical pushback on the "markets beat polls" claim.

Erik Snowberg · Justin Wolfers · Eric Zitzewitz · Quarterly Journal of Economics 122(2) · 2007 · 500+ cites

Uses intraday prediction-market moves on election day as exogenous shocks. Estimates that a Republican win raises equity valuations by 2-3 percent. Methodological template for using PM prices as identification.

Erik Snowberg · Justin Wolfers · Journal of Political Economy 118(4) · 2010 · 300+ cites

Massive horse-racing dataset: the favorite-longshot bias reflects misperceptions of probabilities (probability-weighting), not risk love. Key pricing anomaly result applicable to all event markets.

Paul C. Tetlock · SSRN WP · 2008 · 100+ cites

TradeSports data: more liquidity does not consistently improve price accuracy because limit-order traders are systematically picked off by informed flow.

Robin Hanson · Ryan Oprea · David Porter · Journal of Economic Behavior & Organization 60(4) · 2006 · 400+ cites

The classic manipulation-resistance result. Lab experiment showing would-be manipulators fail to distort prices — and may even improve accuracy by giving informed traders better counter-betting opportunities. Foundational for modern Hanson/Fortune debate on insider trading.

Robin Hanson · Economica 76(302) · 2009 · 200+ cites

Theoretical companion to Hanson-Oprea-Porter. Argues manipulators effectively subsidize information acquisition by others, sometimes raising market accuracy.

Paul Rhode · Koleman Strumpf · Journal of Economic Perspectives 18(2) · 2004 · 500+ cites

Documents large, well-organized Wall Street election-betting markets (1868-1940) — the historical precursor to modern prediction markets — and their forecasting accuracy under manipulation attempts.

Paul Rhode · Koleman Strumpf · NBER WP 14377 · 2008 · 250+ cites

Field-experimental and historical evidence that large speculative attacks on political markets have only transient price effects, with arbitrage absorbing the trades.

David Rothschild · Rajiv Sethi · J Prediction Markets 10(1) · 2016

Account-level Intrade 2012 data classifying 6,300 traders into directional vs arbitrage strategies. Documents evidence of a single large would-be manipulator. Methodological template for the Mitts-Ofir (2026) PolyMarket analysis.

Ned Augenblick · Matthew Rabin · Quarterly Journal of Economics 136(2) · 2021 · 100+ cites

Bayesian test (movement = uncertainty reduction in expectation) applied to Betfair prediction-market prices. Finds rough consistency with rational updating once noise is filtered.

Robin Hanson · J of Political Philosophy 21(2) · 2013 · 200+ cites

The "futarchy" proposal — vote on metrics of national welfare, then let conditional prediction markets choose the policy expected to maximize them. Foundational for governance-by-markets discussions.

Rajiv Sethi · Jennifer Wortman Vaughan · Computational Economics 48(1) · 2016 · 50+ cites

Proves convergence properties of LMSR-style AMMs with risk-averse, budget-constrained traders. Bridges Hanson's theory and trading-dynamics empirics.

Robin Hanson · Information Systems Frontiers 5(1) · 2003 · 1,000+ cites

Introduces market scoring rules (MSR) — the theoretical basis for the LMSR automated market maker that underpins most modern prediction markets. Solves the thin-market problem.

Robin Hanson · J of Prediction Markets 1(1) · 2007 · 800+ cites

Formal exposition of the LMSR — the workhorse market-maker mechanism for thin markets with bounded subsidy loss. Only the logarithmic version preserves conditional probabilities under modular bets.

David M. Pennock · ACM EC · 2004

Dynamic Parimutuel Market (DPM) mechanism: informed traders have incentive to reveal information early while subsidy is bounded. The main alternative to LMSR.

Yiling Chen · David M. Pennock · UAI · 2007

Introduces constant-utility cost-function market makers. Proves Hanson's LMSR is equivalent to a negative-exponential-utility market maker — establishes the cost-function duality with proper scoring rules.

Yiling Chen · Jennifer Wortman Vaughan · ACM EC · 2010

Formal equivalence between convex cost-function market makers and Follow-the-Regularized-Leader online learning algorithms. Bridges prediction-market design and no-regret learning theory.

Abraham Othman · David Pennock · Daniel Reeves · Tuomas Sandholm · 2010/2013

An AMM that simultaneously runs at a profit and adapts price impact to liquidity by relaxing path-independence. Overcomes two key practical limitations of LMSR.

Abraham Othman · Tuomas Sandholm · ACM EC 2010 / Review of Economic Design 2013

Field-tested implementation of a liquidity-sensitive AMM at CMU forecasting building-opening dates. Demonstrates combinatorial market design at scale.

Jacob Abernethy · Yiling Chen · Jennifer Wortman Vaughan · 2013

Convex-optimization framework for combinatorial market design. Proves any market satisfying natural axioms must use a convex cost function — generalizes the LMSR/no-regret-learning equivalence.

Chen · Reeves · Pennock · Hanson · Fortnow · Gonen · WINE 2007

MSR and DPM mechanisms, while subsidizing trade, are not generally incentive compatible. Characterizes when traders may profitably bluff or strategically withhold information.

Chen · Fortnow · Lambert · Pennock · Wortman · ACM EC 2008

Proves #P-hardness of LMSR pricing for several combinatorial bet languages (permutations, Boolean). Fundamental computational barrier.

Chen · Dimitrov · Sami · Reeves · Pennock · Hanson · Fortnow · Gonen · Algorithmica 58(4) · 2010

Characterizes PBE in MSR markets. Truthful betting is the unique WPBE under conditional signal independence. Proposes discounted MSR for convergence.

Michael Ostrovsky · Econometrica 80(6) · 2012

The Kyle (1985) extension for event markets. Defines "separable" securities. Proves that with finitely many partially-informed strategic traders, information aggregates in the limit as closing time approaches for separable securities (including standard prediction-market Arrow securities).

Marco Ottaviani · Peter N. Sorensen · J of the European Economic Association 5(2-3) · 2007

Traders who can directly influence the predicted outcome. Characterizes the equilibrium amount of outcome manipulation when prediction markets are used for corporate decisions.

Chakraborty · Das · IJCAI 2016

Game-theoretic model where participants can both trade and influence the resolved outcome. Manipulation incentives differ from standard Kyle-style results.

Charles R. Plott · Shyam Sunder · Econometrica 56(5) · 1988

The foundational experimental result. Double-auction markets aggregate dispersed information to the rational-expectations equilibrium when traders trade a complete set of Arrow-Debreu securities or share preferences, but not otherwise.

Robert Forsythe · Russell Lundholm · Econometrica 58(2) · 1990

Trader experience together with common knowledge of payoff structure are jointly (but not separately) sufficient for laboratory double-auction prices to converge to rational-expectations equilibrium.

Miller · Resnick · Zeckhauser · Management Science 51(9) · 2005

Peer prediction — scoring agents by applying proper scoring rules to the posterior over another agent's report. Truthful reporting forms a Nash equilibrium even without external verification. Foundational for self-resolving markets.

Siddarth Srinivasan · Ezra Karger · Yiling Chen · arXiv 2023 / ACM EC 2025

Resolves markets without ever observing the outcome — pays each agent based on cross-entropy against a randomly selected later predictor. Truthful reporting is a PBE.

Peterson · Krug · Zoltu · Williams · Alexander · arXiv 1501.01042

The canonical mechanism design for fully decentralized oracle resolution: Reputation-token-staked resolution with escalating dispute bonds and a forking endpoint. Conceptual ancestor of UMA's Optimistic Oracle used by PolyMarket.

Frongillo · Papireddygari · Waggoner · ITCS 2024

Formal equivalence between Constant-Function Market Makers (DeFi AMMs) and cost-function prediction markets. Market-making axioms ≡ information-elicitation axioms. Bridges DeFi mechanism design with prediction-market theory.

Dudik · Frongillo · Wortman Vaughan · UAI 2014

Designs adaptive cost functions for combinatorial markets whose value-for-information decays over time. Preserves previously elicited information while eliminating reward for already-public information.

Steven Gjerstad · Univ. of Arizona WP · 2005

Reconciles Manski's risk-neutral wedge with observed market accuracy. With empirically plausible risk aversion and spread-out beliefs, prediction-market equilibrium prices closely approximate the mean belief.