02.05.2013
By Terry Flanagan

Data Centers Muscle In on Front Office Roles

Data centers, once focused on mid and back office functions, are now taking over key front office roles.

“Data centers are where the data lives and where the computing is done,” said Philip Enness, director of markets infrastructure at technology firm IBM.

“Trades are being executed with algorithms capable of making thousands of trades per second, traders are virtualizing their systems into the data centers, and hoot and holler systems have been replaced with social media/collaboration technologies,” he said.

In most respects, data centers have become the nerve centers of capital markets firms, and hubs of e-commerce.

“No longer do we have aggressive traders on the floors of the exchanges, we now have highly educated technologists trading based on the nanoseconds lost or gained by the proximity of their trading engine to the matching engines, the routes taken by their fiber, or the location of their systems inside the data center,” said Shawn Kaplan, general manager of financial services at data center provider Telx, in a blog posting.

In 1971, Nasdaq became the first electronic exchange and began to swing the pendulum between human traders and machines.

Today, electronic traders reside in the highly connected data centers such as 350 E Cermak in Chicago, 111 8th Avenue in New York City, or in the exchange data centers in less prominent locations in New Jersey such as Mahwah, Secaucus, Carteret, Weehawken and Clifton, as well as Aurora in Illinois.

“Ultimately, it is in these locations or in the fiber hung on the poles, or the microwave signals whizzing above our heads, which has replaced the open outcry markets so many still associate with the financial markets,” Kaplan said. “Today, no longer are trades handled by people on noisy exchange floors. Instead, the new trading floors have migrated to chilled data center facilities without a trader in sight.”

The financial ecosystem has come to embrace the entire trading lifecycle, including risk management and accounting.

“Mid-office grids that once ruled the nights are being pulled into daylight with intraday and near real-time analytics capabilities,” said IBM’s Enness. “Back offices are getting smarter as clearing and settlement is being done continuously and across the ecosystem. “

The majority of the world’s electronic matching engines operate on price-time priority in a central limit order book. This means that the first order to arrive at the best price will win that trade.

“Compared to auction-style trading methods, price-time priority places a tremendous emphasis on getting your trade in quickly,” said Kaplan at Telx. “Just a micro-second or less can determine whether your trade will get filled or whether you’ll pay more.”

The systemic latency of the leading high-frequency trading applications can be as little as a handful of microseconds achieved through high performance computing techniques such as hardware acceleration, kernel bypass and massive parallelization.

In relative terms, said Kaplan, “the speed of light travels only about 200 meters per microsecond…a distance easily traversed inside most data centers, let alone between buildings or even states”.

Related articles

  1. Algorithms have become more prevalent in the spot FX market.

  2. Congress Unlikely to Act on HFT

    QB’s Algo Suite for futures market trade execution is also being co-located to HKEX.

  3. Breaking data silos is key to deploying automation beyond 'nuisance' orders.

  4. They can be used on quantum hardware expected to be available in 5 to 10 years.

  5. Regulation and Liquidity Top Concerns in Fixed income

    Streaming blocks change the basis of matching and price discovery so institutions can find new liquidity.