Opening Cross: Treat Data Management Like An Olympic Competition

As a Brit, my American colleagues feel the need to share their opinions about London’s staging of the Olympic Games with me: how London can’t match the precision of the Chinese opening ceremony, how our athletes can’t match the synchronization of the Chinese divers, or have the staying power of China’s swimmers. But, I say in return, they aren’t superhuman—they’re men, not machines.

In the financial markets, that’s not always the case. Many traders in vanilla asset classes have been replaced by algorithms programmed to seek out pre-defined circumstances and execute against them mercilessly. Like the most competitive athletes, these algos are made to win. But, as Knight Capital discovered last week, machines may be devoid of delays, fear, emotion tor other human weaknesses, but they’re still far from perfect. Some algos are reprogrammed so frequently that they in effect become self-learning and self-correcting, but others execute their instructions with military precision whenever those circumstances are met, whether or not they should still apply or should be updated to reflect an evolving market.

There isn’t much firms can do to completely eliminate rogue algorithms—even with the best intentions and most stringent quality control, unforeseeable glitches can still occur. But they can make sure their algorithms are at the peak of fitness by training them relentlessly against the highest quality data—not just in the “gym” but also out on track.

But the investment required to achieve that can be high, and at present, the West Coast software industry overshadows New York in its ability to attract venture funding. After recent meetings in California, I realized that the East and West Coasts of the US are like fine Olympic competitors, with different cultures, but thrown together on the same country’s team, ultimately aiming for the same goals.

For team events, the athletes must work together, and the team performance is only as strong as its weakest link. For example, Direct Edge and Exegy are working together to deploy a hosted, central ticker plant that creates a consolidated feed—similar to NYSE Technologies’ SuperFeed product—from the exchange feeds Direct Edge already collects to support its core matching business, supports client order-routing to other exchanges, and ultimately makes the exchange an important enabler to participants’ workflow. It also provides another revenue line, hedging against possible fluctuations in trading volumes—giving the “team” a backup strategy if a prolific medal-winner falters.

“Definitely, we are looking at ways to diversify revenue strategies… by offering other products and services rather than just matching buys and sells,” says Kevin Carrai, head of connectivity and member services at Direct Edge. “It creates a deeper relationship with customers, making us more ‘sticky’ and giving us something more to discuss with customers around connectivity.”

Assembling your team and making its members work together is key. CME Group’s CME Direct hybrid listed and OTC derivatives platform has involved the acquisition not only of Elysian Software in 2010, but now also Pivot Inc. as its trade messaging and price discovery mechanism.

And it’s not just about having the best athletes on your team—how and where you use each team member can be critical to achieving the best result. In the data world, imagine a vendor having the most comprehensive datasets possible but making them available in such a way that they were not integrated or intuitively linked. In that situation, even with the best data, clients would not get the best out of the product. Hence, Brazilian data powerhouse Agencia Estado is rolling out a new-look version of its Broadcast terminal, to make it quicker and easier for users to access its wealth of content.

Succeed, and you’ll be crowned a Greek God. But falter on any leg—all essential in the intense competition that is the modern data world—and you could end up a Greek tragedy.