The Structural Evolution of Prediction Markets: Lessons from Professional Trading

In this conversation, former SIG trader Andrew Courtney explains the systemic shift as prediction markets move from niche gambling to institutional financial infrastructure. The hidden consequence of this evolution is a liquidity trap where current participants chase high-fee volume while ignoring the inevitable compression of margins. Readers who understand this shift gain a structural advantage: they stop chasing retail trades and start building the resilient, automated infrastructure required to survive the professionalization of these markets. This is about recognizing the transition from an inefficient, human-driven game to a hyper-competitive, machine-dominated system.

The Hidden Cost of Easy Liquidity

Most participants in prediction markets view high take-fees and fragmented liquidity as a barrier to entry. Courtney argues the opposite: these inefficiencies protect human traders from the full force of adverse selection. When markets are in disequilibrium, such as the World Cup example cited, liquidity providers can join large, slow-moving queues and capture spread without significant risk.

However, this creates a dangerous feedback loop. As these markets mature, they will move toward smaller tick sizes and lower fees to attract volume. This shift will compress the dead zone where market makers currently thrive. The consequence is that the edge moves from simply being there to knowing when to leave.

"The pace of that taking flow that's paying a spread to execute, the fact that you only have to wait there for minutes before getting filled on good trade. Again, not in equilibrium. I think in the future... you could smaller tick sizes and lower take fees, and that would promote more price competition."

-- Andrew Courtney

The Fragility of Automated Bonding

The industry reliance on bonding bots, which are automated systems that trade based on final scores or contract settlements, creates a new class of systemic risk. Courtney notes that these bots are structurally fragile because they are coded for normal conditions. When a rare event occurs, like a ground-rule double in baseball that forces a score correction, these bots fail to account for the rule-book nuance, resulting in massive, automated losses.

This reveals a systems-level insight: automation is not a substitute for deep domain expertise; it is a force multiplier for the logic you embed. If your logic does not account for the edge cases of a specific rule-book interpretation, your automation simply ensures you lose money at machine speed.

"A lot of trading is like being there to capture whatever the normal spread is and then knowing enough to either a not get blown up when something like that happens. And you just selling tails in general, there's like an infinite amount of things that you can lose too."

-- Andrew Courtney

Price Discovery and the Inversion of Flow

A non-obvious implication of the current market structure is the potential for an inversion of price discovery. Historically, sports books have been the primary source of truth. But as liquidity shifts to prediction markets like Kalshi or Novig, these platforms may become the sharpest books.

When the sharpest traders migrate to these exchanges, the flow of information reverses. Traditional books will eventually become followers of prediction market prices. For the savvy trader, this means the competitive advantage lies in monitoring the structure of the exchange, such as fee changes, tick sizes, and API access, rather than just the sports data itself.

Key Action Items

• Audit Your Automation: If you are running automated systems, move beyond happy path testing. Over the next quarter, stress-test your code against rule-book edge cases like scoring corrections or voided events that occur outside standard data feeds.
• Shift from Volume to Resilience: Stop optimizing for the highest immediate P&L. Invest in infrastructure that allows you to cancel orders instantly when price jumps occur. This creates a 12-18 month advantage as the market moves away from slow, queue-based trading.
• Monitor Fee Structures: Watch for changes in maker/taker fees and tick sizes. These are not just administrative updates; they are systemic shifts that will kill off entire strategies. If fees rise, prepare to pivot from market-making to opportunistic taking.
• Focus on Contract Literacy: Dedicate time to reading the specific terms of prediction contracts. Courtney notes that people often do not read the contract details closely enough, and this is where the most significant, non-obvious losses and wins occur.
• Prioritize Data Infrastructure: Instead of chasing picks, look for the picks and shovels. The most valuable, underserved area is the monetization and cleaning of raw market data. There is a long-term opportunity for those who can provide reliable, low-latency data feeds to the market.