By Jon Light, Senior Director of Product Management at Devexperts.
The current surging interest in prediction markets is leading to a general reevaluation of this type of trading, with many financial services firms now questioning whether to offer events-based trading to their own users.
To date, several high-profile firms have moved to incorporate prediction markets into their respective offerings, from brokerages like Robinhood and Interactive Brokers, to crypto exchanges like Coinbase and Gemini, and gaming companies like DraftKings and FanDuel.
This article aims to address some of the common questions, such as design and technology requirements, that trading firms are currently asking about events-based markets.
Platform considerations
One of the first questions financial services firms often raise when exploring prediction markets is whether they ought to be integrated into existing trading platforms or offered through a separate interface.
Designing a prediction markets interface within an existing platform comes with some benefits that should be kept in mind. It allows for consistency across the instrument landscape, helps to avoid user fragmentation, and makes the most of a pre-existing user base.
This means that traditional traders can gain exposure to events-based markets on the same platform they already know, while events-based traders can gradually be introduced to markets they may have never traded before. This potential cross-pollination of users is one of the reasons that many brokers and aspiring “everything apps” are now so focused on events-based trading.
On the design front, events-based contracts are much easier to visualize than traditional markets. The best practices that existing markets have currently settled on involve a simple layout that’s somewhat reminiscent of an online poll, making the event itself clear and the odds very easy to grasp at a glance.
Existing platforms can be adapted to include an events-based interface that’s consistent with these best practices and in keeping with the aesthetic of the platform in question, all while providing visual cues that both traditional and events-based traders would recognize.
Uptime requirements
The perceived simplicity of prediction markets should not be regarded as evidence that setting them up is a trivial affair. Behind the scenes, prediction markets actually have much in common with traditional exchanges and require thoughtful technological implementation in order to perform at scale. This is true of both exchanges and brokers forwarding trades on to exchanges.
For one, prediction markets trade round the clock, which immediately raises the bar as far as uptime requirements are concerned. To this end, system maintenance and market management must be practicable on-the-fly.
The former can be achieved by selecting components that support rolling restarts. The latter by allowing administrators to manage the creation, resolution, and removal of contracts in high-level business logic, rather than requiring these changes to be effected in code.
Reliability and performance
Like crypto exchanges, prediction markets place more of a premium on availability and throughput rather than latency, meaning that they can run effectively in a cloud environment. However, certain replication strategies are necessary to ensure reliability and performance.
These include replication of core system components and synchronization of these duplicates via consensus algorithms, allowing, for example, matching engine and database backups to take over in the event that something goes wrong without missing a step.
The accessibility and relative ease of events-based trading makes this market a prime candidate for mainstream success. For this reason, they ought to be engineered to handle peak concurrent users that can be orders of magnitude greater than the averages the business in question may be projecting.
This can be achieved by a horizontal scaling strategy in which markets are split between many matching engine instances. This can be particularly useful when big events, such as elections, generate massive volumes of trading activity in a relatively short period of time.
Market-specific requirements
Prediction markets also have to have more flexibility in how they’re resolved than traditional derivatives like futures and options. Events may be binary; however, they aren’t always resolved on a fixed date.
A prediction that a certain event will or will not take place before a given date can be resolved at any time prior to that cutoff date. For example, a market titled “Will Trump impose new tariffs this year?” can be resolved any time new tariffs are announced.
Protections against front-running must also be considered as some participants, for example, fans on the front row of a sporting event, could gain knowledge of an outcome quicker than others. This requires latency in trade execution to be introduced in order to level the playing field.
The path ahead?
Prediction markets have a long and storied history of use but have mostly operated on the fringes. This, however, is the first time that a favorable regulatory environment, combined with a huge potential user base from a number of sectors, has the potential of turning events-based markets into a mass market activity.
We anticipate continued growth in this sector for as long as the regulatory climate remains agreeable and expect to see events-based trading mature and complexify as competition forces providers to innovate.
Subscribe to our newsletter
