Comment Text:
Dear Chairman Selig and Commissioners,
I am Alexander Walsh-Kelly, Founder of Vantage, an independent prediction market surveillance and analytics platform. Vantage ingests live data from Polymarket, Kalshi, and traditional financial markets, applies AI-driven causal modeling and on-chain behavioral analysis, and identifies anomalous trading patterns across venues. I respond to select questions where our research and technical development provides relevant perspective.
Response to Question 29,
*Question 29 asks whether there is public utility in allowing participants with asymmetric information advantages to trade on prediction markets, and what factors the Commission should consider in balancing the public interest.*
Informed trading in prediction markets improves price accuracy by incorporating private signals into publicly observable prices. Restricting informed participation would degrade the price discovery function that distinguishes prediction markets from pure wagering. The relevant regulatory question is whether the infrastructure exists to detect when informed trading constitutes manipulation versus legitimate signal incorporation.
On-chain prediction markets provide a structural answer to this question. Polymarket, which settles on the Polygon blockchain via the Conditional Token Framework, records every fill, every wallet address, every position size, and every timestamp on a public immutable ledger. This produces a complete, auditable record of market participation at a granularity that does not exist in traditional derivatives markets.
Building on Mitts and Ofir (2024), who demonstrated statistically significant identification of informed trading using wallet-level composite scoring across prediction market contracts, independent analytics platforms leveraging AI — including Vantage — have operationalized and extended these methods against live market data. Vantage applies aggregate fill filtering at the (wallet, market) pair level, computing anomaly scores across trade timing, size clustering, directional accuracy, and cross-contract correlation
Detection of anomalous informed trading is technically feasible today using publicly available blockchain data. The surveillance challenge in prediction markets is materially simpler than in traditional OTC derivatives, where counterparty identity is obscured and trade reporting operates on regulatory delays. On-chain markets compress this to seconds with full transparency.
The Commission should consider that the public interest is best served by encouraging surveillance infrastructure that leverages this transparency, rather than restricting the informed participation that drives price accuracy. Independent third-party analytics platforms, operating alongside exchange-level monitoring, create layered oversight that strengthens market integrity without requiring the Commission to build or operate surveillance systems directly.
### Response to Question 30 ‚
*Question 30 asks about challenges related to cross-market manipulation and whether prediction markets are more susceptible to manipulation than other DCMs or SEFs.*
Cross-venue manipulation between prediction markets and traditional derivatives is a legitimate concern and a technically solvable one. Our system monitors price divergences across Polymarket, Kalshi, and correlated traditional instruments in real time. When prediction market odds shift significantly without corresponding movement in the underlying reference market, or vice versa, the divergence itself is a detectable signal.
Our proprietary research during the current Iran/Hormuz crisis has produced measurable lag patterns between prediction market repricing and traditional commodity market adjustment. This lag creates a window where cross-market manipulation could theoretically occur, but it also creates a window where surveillance systems can flag anomalous activity in near real time.
On-chain venues offer a structural advantage for detecting cross-market manipulation because the full order flow is observable. A coordinated actor attempting to move a Polymarket contract to influence a related Kalshi bracket or traditional derivatives position would leave a traceable on-chain footprint. Off-chain venues provide no equivalent visibility.
The Commission should consider requiring that registered DCMs operating prediction markets publish sufficient trade-level data to enable independent cross-venue surveillance, or adopt on-chain settlement where this data is inherently public.
### Response to Question 31 ‚
*Question 31 asks whether prediction markets are more or less likely than other derivative markets to be susceptible to manipulative or deceptive devices.*
Prediction markets settled on public blockchains are structurally less susceptible to undetected manipulation than traditional derivative markets. Susceptibility to manipulation attempts may be comparable, but susceptibility to undetected manipulation is materially lower.
In traditional futures and swaps markets, counterparty identity is mediated through clearing, trade reporting operates on regulatory delays, and position-level data is aggregated in weekly reports with multi-day lag. In on-chain prediction markets, every individual trade is attributable to a specific wallet address, observable in real time, and permanently recorded.
Our platform has validated that wallet-level behavioral analysis can identify accounts exhibiting statistically anomalous trading patterns, including pre-event positioning clusters, abnormal win rates on specific contract types, and coordinated multi-wallet activity. The Mitts and Ofir (2024) methodology provides the statistical framework; our proprietary extensions apply this against live market data.
The Commission's regulatory framework should account for this asymmetry in observability between on-chain and off-chain venues when evaluating manipulation risk.
### Response to Questions 2d and 2h ‚
*Question 2d asks about surveillance practices the Commission should focus on for prediction markets. Question 2h asks what factors the Commission should consider regarding blockchain-based prediction markets.*
Three surveillance capabilities are technically feasible today for blockchain-based prediction markets and warrant the Commission's consideration:
First, aggregate fill analysis. By computing fill rates, trade sizes, and timing distributions at the wallet level across specific contracts, surveillance systems can identify accounts whose trading patterns deviate significantly from baseline market activity. This methodology has been validated in peer-reviewed academic research and is being operationalized in production systems including ours.
Second, cross-venue divergence monitoring. Prediction market odds, futures curves, and options-implied probabilities on the same underlying event should converge absent information asymmetry or manipulation. Persistent divergences between venues, particularly when concentrated in specific wallet clusters, are actionable surveillance signals. Our platform currently implements this capability across multiple prediction market and traditional financial data sources.
Third, named-wallet behavioral tracking. On-chain markets enable longitudinal analysis of individual wallet performance across contracts, markets, and time periods. Wallets that consistently demonstrate abnormal directional accuracy or pre-event positioning can be flagged for review without requiring the wallet holder's identity to be known.
Cross-venue divergence monitoring, wallet-level behavioral analysis, and aggregate fill detection are complementary capabilities that are most effective when deployed together. Platforms such as Vantage combine these into integrated surveillance systems operating against live market data.
Artificial intelligence and machine learning significantly enhance these surveillance capabilities. AI-driven causal models can ingest dozens of economic, geopolitical, and market variables simultaneously, identify non-obvious relationships between prediction market price movements and underlying fundamentals, and flag when market behavior deviates from what the causal structure would predict. Machine learning techniques‚ Ai including anomaly detection, clustering, and pattern recognition, Ai can process the volume and velocity of on-chain trade data at a scale that manual surveillance cannot match. Importantly, these systems improve over time as they ingest more market data, creating a compounding surveillance advantage that benefits market integrity. The Commission should consider how AI-enabled surveillance infrastructure, deployed by both exchanges and independent third parties, can serve as a force multiplier for the market integrity objectives of the Core Principles.
Regarding blockchain-based markets specifically: the Commission should recognize that on-chain settlement is not merely a technological choice but a transparency architecture. The public ledger functions as a continuous, real-time trade surveillance feed that any participant, regulator, or independent analytics provider can audit. The Commission should consider whether requiring on-chain settlement or equivalent trade-level transparency for all registered prediction market DCMs would strengthen market integrity.
### Response to Question 10 ‚
*Question 10 asks what role event contracts play in managing price risks, discovering prices, and disseminating pricing information.*
Prediction market contracts on economic and geopolitical events provide price discovery that complements traditional commodity and financial markets. Our research has focused on oil-related event contracts during the current Hormuz crisis, where Polymarket ceasefire probability markets and Kalshi WTI bracket markets provide forward-looking probability distributions not available from any other source.
Our causal modeling system validates directional relationships across the oil macro complex and uses prediction market odds as a distinct input signal alongside statistical correlations, expert sentiment analysis, and speculative positioning data. The combination produces a richer analytical framework than any single source alone.
The Commission should consider that event contracts referencing economic, geopolitical, and commodity-related outcomes serve a genuine price discovery function distinct from gaming or entertainment. Participants in these markets include commodity analysts, macro researchers, and risk managers who use prediction market odds to inform portfolio decisions.
Closing
The Commission has an opportunity to establish a regulatory framework that preserves the price discovery and transparency benefits of prediction markets while addressing legitimate concerns about insider trading and manipulation. The technical infrastructure for effective surveillance exists today, particularly for on-chain venues where trade-level transparency is a structural feature. Independent analytics platforms operating alongside exchange-level monitoring create the layered oversight appropriate for these markets. We encourage the Commission to adopt rules that incentivize transparency and surveillance capability rather than restrict market participation.
We would welcome the opportunity to provide technical demonstrations of our surveillance capabilities or participate in any Commission roundtables on prediction market integrity.
Respectfully submitted,
Alexander Walsh-Kelly
Founder, Vantage
