Prediction markets are no longer a niche curiosity. As contracts tied to elections, sports, geopolitics and real-world events proliferate, so do questions about market integrity – particularly when it comes to insider trading. Unlike traditional financial markets, these venues often hinge on human actions, discretionary decisions or information asymmetries that are difficult to define, let alone monitor in real time.
To explore how surveillance teams should think about these challenges, Eventus Global Head of Regulatory Affairs Joe Schifano sat down for a conversation with two subject matter experts from legal and regulatory consultancy Tölt Strategies:
- Dorothy DeWitt, Founder and CEO at Tölt Strategies and former Director of the Division of Market Oversight at the U.S. CFTC
- Jason Fairbanks, derivatives regulatory advisor and market consultant at Tölt Strategies
The conversation covered a range of topics related to insider trading in prediction markets, including risk factors, contract design, governance gaps and what regulators can expect from designated contract markets (DCMs) as they continue to scale.
Prediction markets compress real-world events into tradable instruments. Elections, sporting events, geopolitical decisions and regulatory actions can all become contracts with immediate financial consequences. That compression creates liquidity and insight, but it also magnifies information asymmetries and raises new questions about market integrity.
Joe Schifano: One of the fundamental shifts with prediction markets is that insider trading risk is no longer confined to corporate hierarchies. When the underlying instrument is a real-world outcome, the pool of people with potentially material information expands dramatically – and often includes individuals who don’t think of themselves as insiders at all.
Dorothy DeWitt: That’s exactly right. The scope is broader, and in many cases the risk is more acute. In prediction markets, outcomes are driven by people – political actors, athletes, officials, staff – not balance sheets. Discrete non-public facts like a lineup change or the timing of an announcement can reprice a contract immediately.
Jason Fairbanks: I’d add that the incentives can be very different. In traditional markets, even material non-public information doesn’t always lead to a predictable payoff. In prediction markets, the expected value of that information is often much higher because the contract payoff is binary. That certainty increases both temptation and risk.
Schifano: Which forces surveillance teams to rethink some very basic assumptions about how insider risk is identified and managed. You’re no longer just monitoring a narrow set of professional insiders. You’re dealing with a diffuse, sometimes invisible information network surrounding the event itself.
Many prediction markets attract participants through consumer-style platforms. Their users often arrive without prior exposure to regulated trading environments, and the boundaries between “betting,” “trading,” and prohibited conduct are not always clear.
Schifano: It seems that identity, affiliation and context aren’t just KYC boxes to be checked, but essential inputs to a credible surveillance program. What’s different about KYC in prediction markets?
DeWitt: Customer onboarding and identity are absolutely foundational. You need to know who is trading before you can understand why they’re trading. In prediction markets, it’s easy to onboard someone who looks perfectly ordinary until you realize, often after the fact, they’re closely connected to the subject of the contract. If DCMs and intermediaries don’t clearly define the rules of engagement up front, especially around information use and KYC, they risk avoidable compliance issues that could have been addressed through better controls and operational integrity. Exchanges live and die by the integrity of their markets they run, so this not just a regulatory issue but also a business and reputational consideration.
Fairbanks: It’s important to note that a lot of this risk isn’t malicious. It’s behavioral and cultural. People don’t always see themselves as market participants with obligations; they see themselves as users of an app. That’s a governance problem, not just a surveillance one.
Surveillance in prediction markets must function across thousands of contracts, wildly different event types and uneven information distribution. Precision matters, but so does scalability.
Schifano: In conversations on prediction market visibility, I think there’s often an overemphasis on real-time surveillance as the solution. How do pre-trade controls and post-trade procedures augment market supervision?
Fairbanks: Pre-trade controls are critical, but many insider trading cases in prediction markets only become visible after the event. Post-trade surveillance is where you reconstruct what happened: when information existed, who plausibly had access to it and how trading behavior changed as a result.
DeWitt: What makes this a real challenge are the scale and speed of these markets. When you’re listing hundreds or thousands of contracts across sports, politics and geopolitics, you can’t realistically predefine every insider scenario. The answer has to be pattern-based with AI analysis. Instead of asking, “Who could be an insider here?” you look for anomalies – accounts trading outside their historical behavior, sudden clustering around resolution events or participants entering markets they’ve never touched before.
Schifano: That’s an important shift. It seems clear that surveillance needs to become less about exhaustively enumerating risks up front and more about identifying signals that warrant deeper contextual investigation.
Contract design quietly determines how enforceable – and defensible – a market really is.
Schifano: One of the key things to understand about this space is that contract design either amplifies or constrains surveillance effectiveness. Vague contracts don’t just create interpretive disputes – they materially increase integrity risk.
DeWitt: Exactly. If a contract is poorly defined – to cite examples from recent headlines like “Will Country X invade Country Y?” – you introduce ambiguity that makes enforcement and surveillance harder. Clear resolution criteria are a prerequisite for credible oversight.
Fairbanks: There’s also the manipulation angle. If a small group of people can influence or directly trigger the outcome of a contract, insider trading and manipulation start to merge. At that point, no amount of monitoring can fully mitigate the risk. In effect, that means contract governance isn’t just a legal or product decision. It’s also a surveillance decision with downstream consequences.
Regulatory scrutiny may feel distant, but enforcement timelines are long and retrospective.
Schifano: One of the hardest messages for emerging venues is that in any new market, enforcement rarely shows up immediately. What should market participants be doing to prepare for potential examination and enforcement matters down the road?
DeWitt: Regulators won’t expect perfection in the near term, but they will expect a serious dedication to developing new surveillance tools that match novel products and market structures. DCMs need to show that they understand their risk profile and can demonstrate how insider trading and other manipulative trading conduct risks are identified, investigated and escalated in practice.
Fairbanks: The assumption should be that everything happening today is still within the statute of limitations tomorrow. Firms are effectively making decisions now for an audience they won’t encounter for years.
Schifano: That long view is uncomfortable, but unavoidable for any organization that hopes to be in prediction markets for the long haul.
Our thanks to Tölt Strategies for their time and perspective. To learn more, visit their website. And for additional commentary and resources from Eventus on market and product evolution, explore our blog or request a demo.