Open Platform: Back to Basics on IT and Risk

As competitive and regulatory forces drive greater adoption of technology from risk management to high-frequency trading, firms are realizing that the next big advantage isn't high-frequency trading but data management, and that any initiative must begin with the ability to handle the most basic of financial building blocks-data. By Irfan Khan, vice president and chief technology officer at Sybase

The "Rise of the Machines" is here: Software and hardware are supplanting humans across industries, and perhaps nowhere more avidly than the securities industry. Technology and the securities business are inextricably linked. Firms rely on technology for capabilities ranging from high-frequency trading to risk management to derivatives pricing. In the wake of the financial crisis, that dependence is even stronger, and - if anything - will grow only deeper in the years to come as companies and regulators seek technological answers to portfolio and systemic risk.

For example, high-frequency trading produced almost 50 days of $100 million in trading volumes for one leading capital markets firm during a 70-day stretch in 2009 (see "Rise of the Machine," The Economist, Aug. 1, 2009), and has triggered a recruiting war for math and computer science talent from academic hotbeds such as Caltech and MIT.

High-frequency trading uses algorithms and distributed computing resources to detect and exploit movements in the market, using computer algorithms to exploit information and make trades seconds faster than human traders, and faster than their rivals, with some algo trading infrastructures boasting 400 or 500 microsecond response times. But some aspects of high-frequency trading have been criticized for giving some market players an unfair edge. Several US exchanges withdrew their "flash trading" capabilities-which let a subset of traders peek at share prices a microsecond before the broader market-while the Securities and Exchange Commission scrutinizes the practice. In addition, the SEC is considering action to bar unfiltered-or "naked"-sponsored access, where brokers let big clients use their platforms to trade directly on exchanges, with no curbs on risk.

Numerous other market practices, such as indication-of-interest (IOI) orders, come into play to link "dark pools" with other pockets of liquidity. These practices function similarly - though not identically - to flash orders, insofar as they are only displayed to a select group of participants. According to Waters magazine's 2009 State of the Trade survey, IOIs-which have existed for decades, and which many believe are an important though imperfect liquidity discovery tool-are undergoing a renaissance, with an estimated 89 percent of respondents seeing value in actionable IOIs.

IOIs, dark pools and a number of other trading practices are under scrutiny but are still in use. As in the case of flash trading, dark pools and IOIs account for large slices of the market but could also introduce systemic risk - a factor that is now heavily driving Wall Street's IT spend. By 2012, IT spending on risk and compliance is forecast to post a compound annual growth rate of about 13 percent, compared to between 3 and 4 percent for overall IT spending, according to TowerGroup. Within the securities business, which is facing demands for greater transparency, spending by hedge funds is expected to rise the most.

Start With Data

Whatever comes out of the current scrutiny of capital markets, one thing is certain: capital markets firms will soon be able to support continuous rates of 500,000 messages per second over a sustainable period. Inevitably, trading speeds will accelerate even more than before and data volumes will soar also.

The total volume and complexity of time-critical, location-specific market data can overwhelm even the largest financial firms' ability to analyze it and act speedily enough to maximize returns. A long list of data-transactions, corporate data, collateral balances, market data, static data-must be collated and consumed on a daily or intra-day basis to calculate the latest margin requirements. Factors such as delivery, timeliness, quality and quantity are also critical for risk managers.

Any capital markets firm attempting to build an enterprise risk management infrastructure must start with data. Companies are increasingly spending critical time and resources trying to deal with inconsistent data, excess latency (today, 20 to 30 milliseconds is considered too slow), redundant data and other issues.

As the financial markets emerge into reform and recovery, the way forward will be shaped by new regulations, methodologies and practices, and new technology. Technology and improved system and process design may help build a single image of data for decision makers to see a holistic snapshot of market activity, capital flows, liquidity and risk.

To cope with the flood of data, securities firms have turned to risk analytics that crunch massive amounts of internal and external market data at lightning speed to give them a picture of their risk profile over the trading day, and complex event processing (CEP) software that monitors disparate datafeeds for patterns and correlations, and makes high-speed trades based on machine-readable news. In February, Sybase acquired CEP software maker Aleri, positioning itself as a leading CEP provider to the capital markets.

Financial firms are developing the capacity to understand, track and quantify risk across all portfolios and assets, while improving agility as volatility continues. These firms demonstrate that mastering the capture and flow of data through faster and more powerful time-series analysis that continuously transforms data into insight and profitable decisions is the key to creating the new competitive advantage.