As Wall Street continues to feel the profit pressure of shrinking trading volumes, financial firms are looking to rein in their technology spending and tap into the economies of scale offered by commercial data centers. Choosing third-party data centers operated by the likes of Savvis, Equinix and NYSE Technologies, capital markets firms can rent servers and license software on demand, pay for access to high-speed market data feeds, and acquire storage space for regulatory purposes -- all without having to build or maintain these capabilities in-house.
Increasingly, Wall Street firms, including brokers, providers of algorithmic trading strategies and exchanges, are deploying a variety of applications within third-party data centers -- where they already might colocate their algo trading strategies -- to take advantage of cloud computing and its cost efficiencies. The popularity of big data also is fueling the migration of analytics to data centers that offer an elastic cloud-computing model. Some firms have even turned to Amazon's public cloud to access cheap processing power for large, complex data sets.
Economic realities are motivating sell-side firms, in particular, to reduce their technology spend. With equity trading volumes declining 40 percent since 2008, infrastructure budgets are squeezed. "We're talking about lower transaction fees, which then means there is less money to spend for infrastructure, which includes equipment, communications and all those things that are required to operate a business," says Kevin Carrai, head of connectivity and member services at Direct Edge, an operator of U.S. equity exchanges.
Unfortunately, some firms are shuttering trading desks or exiting certain securities types -- Nomura, for example, has said it will retrench in U.S. equities. According to Carrai, firms that operate wholly owned autonomous subsidiaries are folding in those affiliates so they can share technology infrastructure and reduce costs. To help clients cut costs, he says, Direct Edge launched in August the Edge Ticker Plant, which integrates 20 market data feeds with direct access to all major U.S. equity exchanges without requiring clients to spend money for other circuits.
"Using the ticker plant can create massive volumes of data, so firms can consolidate the data to support their trading systems, versus having to build their own ticker plants," Carrai explains. To deliver the ticker plant to customers, Direct Edge is leveraging the Exegy Ticker Plant, what it refers to as "an accelerated market data aggregator," and a proprietary network, Edge Connect. "Customers connect to us and they access the ticker plant. It's another way to reduce their ticker plant costs to run their trading systems," says Carrai.
More Than a Cost Play
While most market participants are choosing the private cloud with dedicated resources so they are not sharing machines with other firms, some high-frequency trading firms are using the Amazon public cloud to obtain the computational horsepower they need to run complex calculations. "The cloud is important to HFT because they can run complex computations on hundreds of servers in parallel," according to Manoj Narang, CEO of Tradeworx, a firm involved in proprietary trading and a provider of a high-performance trading platform. "More so than any type of trading strategy, HFT strategies need to examine absolutely massive amounts of data in order to generate predictive trading signals," Narang said during a September Wall Street & Technology webcast, "HFT Under Attack: Using Big Data to Find New Correlations and Trading Opportunities."
Two and a half years ago, Deep Value, a Chicago-based provider of algorithms to buy- and sell-side firms that also has a broker-dealer affiliate, began using big data analytics to improve executions. Despite the depressed environment in U.S. cash equities, CEO Harish Devarajan says, the firm's business has grown well -- it executed nearly 3 percent of overall market volume on its highest-volume day this summer.
"Our sustainable edge is our ability to sift through and work with large amounts of data to try and go after the ever more elusive alphas that are still left in equity markets," Devarajan tells WS&T. Citing the availability of computing as a utility, he says Deep Value utilizes Amazon EC2's elastic computing services to crunch data. "When you need 300 machines, you don't need the cap-ex [capital expenditure]; you can rent processors by the hour."
According to Devarajan, Deep Value has been analyzing the liquidity contributed by high-frequency traders. "Whether you like it or not, the majority of the liquidity that you interact with is from high-frequency traders," he says. "By looking at data thoroughly with a large number of machines, we are able to fine-tune our algorithms to benefit from behaviors of high-frequency traders."
Many of the market players that are offering colocation services in their data centers also are bundling other services in the same facilities. For example, NYSE Technologies, which operates a mammoth data/liquidity center in Mahwah, N.J., launched an industry-specific cloud platform, dubbed the Capital Markets Community Platform, in June 2011.