Sign up today and take advantage of member-only content — the kind of timely, cutting edge industry insight that only Money Management Executive can deliver.
  • Exclusive Online Only Content
  • Free Daily Email News Alerts
  • Asset Management Blogs

Managing Risk:Real-Time & On Demand


Real-time. It's a phrase typically used by high-frequency traders on how they execute orders or receive market data before the tick of the most sensitive clock passes.

The distinction was never used in risk management-until now. The impetus: the financial crisis that caught many financial firms ill-prepared and advancements in technology.

Fund managers want to know how a transaction will affect the performance of entire portfolios or compliance with exposure guidelines for investments and counterparties. That often means understanding ramifications on demand, at any time of day and prior to making a trade. It can also mean understanding how market conditions have affected their portfolios so changes can quickly be made.

One of the market innovations: a new type of software now widely adopted in high-speed trading, the complex event processing engine which takes in streams of news, transaction data and other information from internal systems, market data suppliers and trading venues. Risk controls are added in. And trades can get executed, almost instantly.

"Fund managers will often implement real-time risk capabilities as components of advanced trading systems," said Mark Palmer, chief executive officer of StreamBase, a supplier of complex event processing software in Boston.

Case in point: Quantitative fund manager PhaseCapital in Boston, which relies heavily on high-frequency trading and uses StreamBase's engine as part of its front-office trading infrastructure. The engine "scrubs" and normalizes direct market data feeds and helps manage high-speed routing and execution of trades across data centers in Boston as well as Jersey City, N.J.

The cleansed data from the consolidated feed in PhaseCapital's Boston data center prices the firm's portfolio holdings so that profit and risk can be calculated immediately. The data also feeds proprietary algorithms which generate what PhaseCapital calls trade signals, target portfolios and parent orders. The parent orders are broken down into executable orders which are then routed by algorithms running on the Stream Base platform to a individual venues such as NYSE Arca and the BATS X and Y exchanges.

Eric Pritchett, chief executive of PhaseCapital, said his firm's approach, while typical among high-frequency trading firms, is gaining wider acceptance among traditional fund managers.

Although most fund managers are still tackling pre-trade analytics for equities, some are also understanding the need to expand to fixed-income and derivative products. "If the systems don't make it easy to add new products the fund manager is at a disadvantage to its competitors and could make a bad investment decision," said Andrew Aziz, executive vice president of risk solutions at Algorithmics, whose Algo Risk platform is used by multi-asset class fund managers.

Algo Risk calculates net asset value on the fly and provides a range of risk analytics such as value at risk and what-if-scenarios for equities, fixed-income instruments and over-the-counter derivatives such as complex swaps and options. Once the risk calculations are made, Algo Risk will then help the portfolio manager alter his or her trading strategy or rebalance a portfolio.

Risk managers might typically be satisfied with calculating only the value at risk in a trading day or analyzing outcomes using various market scenarios. But portfolio managers are embracing advanced approaches to better capture the presence of fat tails-extreme unpredictable events which can have a major impact.

By contrast, value at risk measures the percentage or dollar value of assets a fund manager might lose for a given probability while stress testing measures the dollar value associated with an extreme loss. Both measures generally rely on predictable market conditions.

"If the model used to calculate these figures is not capable of capturing the fat tails that are present in the markets then things can go wrong as we have seen in both the larger financial crisis and in more isolated fat tail events such as the Greek debt crisis which began in the first quarter of 2010," said David Merrill, chief executive at FinAnalytica. "A fat tail model will rely on the same inputs as more simplistic value at risk models but be more in tune to the potential for a higher upside and lower downside-both tails."

FinAnalytica's Cognity platform can assess the impact of tail events on inflation rate swaps, callable and puttable bonds, floating rate notes and "total return'' swaps on stocks and bonds.

Another use of pre-trade analytics is to meet company policies on credit holdings and credit risk exposures. "Fund managers need to know whether they have violated any limits on either the types of securities, value of securities or corporations in order to corporations to take timely action in normal and distressed market situations," said Else Braathen, chief business consultant and risk management domain manager for SimCorp. The SimCorp Dimension platform will send the investment fund an alert which can be viewed by both the portfolio, risk manager and compliance director.