Taming the Market Data Beast
Don't get Michael Levas started about how important market data is to his fund.
As founder, senior managing principal and director of trading for Olympian Capital Management in Fort Lauderdale, Fla., the hedge fund manager says he needs a wealth of market data every trading day in order to execute his global macro strategy. "We get our market data from our prime brokers in the various platforms to which we have access," Levas explains, stressing that the firm depends on multiple sources of data.
"We have a prime broker and then four other platforms that we utilize as well," he continues. "The prime broker is a multi-asset platform, which is great; we can get our foreign exchange, commodity futures -- everything we need, and right up to the minute. But we also get other data and information from the various agency brokers: the people on the sell side that give us any information that we need."
Without that breadth of global market data, Levas insists, he and his partner, Arun Kaul, could not trade. "That market data is worldwide," Levas says. "We trade ordinarily on the Swiss Exchange in Zurich, or we'll trade Singapore, and we'll vet that data and information right from the prime brokers."
[Read about how Olympian Capital Management and other hedge funds are using historical market data to sharpen their strategies: "Market Data: Looking Back to Get Ahead."]
In the same way that fighter jets need kerosene-based jet fuel to get off the ground, investment firms large and small rely on trade and quote-related data to power their strategies -- from timing executions to analyzing the deals after the fact and performing long-range and deep-dive analysis for future strategies. Without market data, a trading entity -- whether it's Goldman Sachs, State Street or a proprietary trading shop -- is flying blind.
But even with its vaunted status, market data brings its own unique set of management challenges. If a hedge fund has a global strategy, for example, it needs multiple data feeds from exchanges in various markets. Further, hedge fund managers need sophisticated, low-latency tools to parse and analyze every morsel of data before, during and after a trade.
Meanwhile, the colossal wave of market data continues to hammer away inside order management systems despite the dip in trading volumes since the 2008 mortgage meltdown and the ensuing credit crisis. In fact, the overwhelming volume of market data shows no signs of leveling off.
Trading Strategy Calls the Market Data Shots
When it comes to market data infrastructure and the types of market data that hedge funds need, the funds' underlying trading strategies call the shots. "If you're trying to compete at the bleeding edge of speed and you haven't implemented the appropriate infrastructure that has that kind of speed, or it has errors in it, you're not going to be competitive for very long," warns Paul Rowady, senior research analyst for Tabb Group, the market research firm based in Westborough, Mass.
Nearly every capital markets firm trades electronically via algorithms these days, and the need for fast market data is almost universal among hedge funds and prop shops. "You want to have as low-latency data as you can have, even if you're not a high-frequency fund," says Scott DePetris, COO of Portware, the financial solutions provider. Without low-latency data feeds, he adds, "you're making trading decisions on data that's 3 seconds old."
The types of market data needed are as varied as the asset classes traded and the locations of the exchanges on which they are executed, suggests Phil Lynch, CEO of Asset Control. According to Lynch, new classes of data have emerged in recent years to address these needs. "We're starting to see all kinds of specialized data providers crop up, such as index providers, real-time fixed income prices, mortgages or swaps," he relates. "As more of these complex instruments become exchange-traded, there will be more options for bringing data into the firm."
In the past, obtaining market data was fairly straightforward. Working inside an established sell-side firm such as J.P. Morgan or Goldman Sachs, traders had at their disposal Bloomberg terminals or a Reuters data-feed and an OMS that downloaded market data from the New York Stock Exchange or Nasdaq. Now, with smaller firms trading in multiple asset classes inside worldwide bourses, start-up hedge fund managers want the same quotes, access and data sorting tools that they had at their larger, better-outfitted former firms.
Start-up hedge funds and prop shops must get their market data strategies in order before they open shop and print their business cards and prospectuses, argues Tabb Group's Rowady. "In terms of strategy implementation, it's absolutely important. It's one thing if you go into the lab, develop a strategy and paper trade it for a while," he says. But once a hedge fund manager moves from research to production, "The market data has to be meticulously developed and monitored."
The majority of hedge funds obtain their market data from broker-dealers that have access to feeds directly from the exchanges. But colocation service providers also offer direct exchange feeds of market data, and their pipes are located near the exchange data centers to cut down on latency and offer the fastest data to accelerate the trade life cycle.
"Larger firms traditionally go to Bloomberg and Thomson-Reuters for market data. But for hedge funds, depending on the frequency of their strategies and how high speed they are, they usually get data from exchange direct feeds," says Portware's DePetris. "This can be provided by the hosting provider, and in some cases by the sell side, but in general they are using exchange direct feeds."
According to Olympian Capital's Levas, his prime broker offers a list of exchanges and other trading venues for his strategic requirements. "If we need it from an exchange or a particular market, we can get it from the prime broker; and if they don't have it, they have a list of what they can get," he says. "But nine times out of 10 we can get the data from the brokers in the platforms we have."
Riding the Wave of Market Data
Managing the veritable torrent of market data was once a nightmare for the buy side. But after attempting to manage multiple feeds, solve contrasting and conflicting standards, and store the burgeoning volume of data, hedge funds appear to have all but given up handling market data by themselves. Good thing, too.
The buy side's market data situation remains a relic of the old-fashioned "people and terminal" business, according to Asset Control's Lynch. "It's not really in line with the electronically driven and high-speed front office;whereby fewer people do more through leveraging technology," he adds.
When it comes to validating, managing and consuming market data, Lynch says, the buy side is highly inefficient. "It's difficult for firms to obtain a centralized and reliable view of their positions and risks in order to provide any level of consistency and depth in their reporting."
And managing data is costly. According to experts, the price tag for reliable market data at capital markets firms comes in behind only IT costs and salaries -- even at large buy-side firms, the cost of data often exceeds the cost of Wall Street real estate. "This is something that has to change because the buy-side no longer has the luxury of the margins it once had," says Lynch.
Alex Tabb, a partner with Tabb Group, suggests that the challenge of getting the most up-to-the-second information to the right trader, analyst and portfolio manager from various sources across the globe can be a nightmare. "The logistics of market data are complicated, especially in a multidimensional strategy where you are laying all these different market data feeds on top of one another," Tabb says. "Hedge funds have to focus on the market data question."
And despite the global slowdown in trading volumes, market data volumes continue to climb, driven largely by high-frequency trading. "Even when markets are down, the message rates are up," says Tabb. "Any time a new capacity upgrade is scheduled, the high-frequency trading community will then fill that new capacity." Tabb compares the scenario to automobile traffic on highways: "There are new lanes outside of Dulles Airport [outside Washington, D.C.]," he says. "They're always full."
In fact, some firms simply can't keep up with the rush of market data. "Time windows for processing the availability from the supplier may not give the buy side enough time to revalue funds and perform compliance checks before the next trading day commences," according to Paul Charie, head of buy-side sales, EMEA, for Fidessa. "Some data types are very difficult to source accurately. A classic case is the number of shares in issue, which is key to managing threshold ownership levels." Phil Albinus is the former editor-in-chief of Advanced Trading. He has nearly two decades of journalism experience and has been covering financial technology and regulation for nine years. Before joining Advanced Trading, he served as editor of Waters, a monthly trade journal ... View Full Bio
Tweet This