Prediction markets: An emerging form of gambling?

We write to direct the readership to a potential new form of gambling: prediction markets. Prediction markets offer binary yes/no contracts whose prices reflect the market's estimate that a real-world event will occur. For example, a ‘Candidate X to win the election’ contract might be priced at $0.60, reflecting a market view of a 60% chance of victory [1]. Purchasing one contract pays $1 if Candidate X wins and $0 otherwise. These commonly focus on outcomes of political and sporting events. Prediction markets in the United States are overseen by the Commodity Futures Trading Commission [2], meaning they are often regulated as investments rather than as gambling products [3]. This means typical state-enforced harm-minimisation measures common in gambling regulation, such as state self-exclusion registers, advertising restrictions and in-product safer-gambling tools, generally do not apply [4]. However, functionally, these products resemble gambling. Users stake money on uncertain real-world outcomes that resolve in either a full payout or nothing, depending on the outcome. Short event cycles permit repeated re-entry and loss chasing. Risk is heightened by 24/7 mobile access, push notifications, easy deposits and constant availability. The blurry line between high-risk investing and gambling has long existed in financial markets [5], especially for high-risk behaviours such as day trading or margin trading. Gambling through financial markets also has a long pedigree: individuals have used traditional stock trading [6] and, more recently, cryptocurrency trading [7] as alternative ways to gamble. These have been linked to significant harms, including economic loss [8], mental health problems [9, 10] and addiction [11]. Therefore, regardless of the legal label, these products pose a significant risk. Widely used share-trading platforms are now offering prediction markets to the broader population, most notably, the integration into the trading platform Robinhood [12], which is the most used stock trading app in the United States [13]. Robinhood has been found to host users who engage in high-frequency trading [14], implying the presence of day traders on the platform. Prior research finds that people experiencing problem gambling are more likely to engage in day trading [15] and that day traders are more likely to engage in other forms of gambling [16]. Therefore, the embedding of prediction markets into these widely used platforms may directly expose vulnerable groups to these products. With little peer-reviewed literature, action is needed now. We propose a research agenda modelled on adjacent domains such as cryptocurrency: (1) characterise how users perceive and use these products [8]; (2) quantify the prevalence and severity of harm; (3) identify product and interface features that elevate risk; (4) test safeguards for effectiveness; and (5) assess how these products are advertised [8]. Evidence such as this would allow substantiated questions to be asked to legislators about the appropriate classification of these products. None. Benjamin Johnson: Conceptualization; writing—original draft. Gary Chan: Writing—review and editing; supervision. B.J. and G.C. declare that they have no conflict of interest.