Cross Venue Arbitrage and the Risks It Carries

Cross venue arbitrage exploits brief price gaps between linked instruments, and its main risk is one leg filling while the other does not.

Cross venue arbitrage trades the same security, or two economically linked securities, across different venues at the same time to capture a brief price gap between them. The strategy is only profitable when the gap can be captured faster than the market itself closes it, which places speed and infrastructure at the center of the entire business.

A stock trading simultaneously on more than a dozen exchanges and dozens of dark pools does not update everywhere at exactly the same instant, because each venue processes and publishes its own quotes independently and the connections between venues carry their own propagation delay. A price move on one exchange takes a small but nonzero amount of time to be reflected on every other venue, and in that window the same instrument can briefly show two different prices to two different observers.

Latency arbitrage exploits exactly this window: a firm with a faster view of the market, built on infrastructure like a colocated server and a dedicated microwave relay between data centers, can see a price move on one venue and trade against a stale quote on another venue before that venue's own price updates. The edge in this strategy is measured in microseconds to low milliseconds, and it disappears entirely for a firm whose infrastructure is even slightly slower than the fastest participants in the market.

The mechanics of exactly how much faster a microwave path can be than a fiber one, and which strategies that speed actually helps, are worked through in more detail in the piece on microwave and laser links in latency sensitive trading, since the arbitrage strategy described here is the direct commercial reason that infrastructure gets built in the first place.

A cross venue trade typically involves two legs, buying on the venue showing the lower price and selling on the venue showing the higher price, and both legs need to fill for the trade to lock in a profit. If one leg fills and the other does not, because the opposing quote moved or was cancelled in the fraction of a second between the two orders being sent, the firm is left holding an unhedged position at exactly the moment the market has just shown it is moving, which is the central risk of the entire strategy.