The September 11, 2001, terrorist attacks gave Silicon Valley a shot in the arm. Executives and engineers there and in other high-tech hotbeds, then reeling from the dot-com debacle, saw growth opportunities through the gloom. They reasoned that if 9/11 was a systems failure — whether of metal detectors or passport databases — then surely new investments in advanced technologies could enhance what came to be known as homeland security while helping to root out enemies of the state.
Technology markets have recovered, with defense and counterterrorism spending playing a part. But notwithstanding data mining and no-fly lists, baggage scans and experiments with fingerprinting and other foolproof identity methods, the war has yet to be won. According to Bruce Schneier, chief security technology officer of telecom giant BT Group, only two measures can be credited with making flying safer: reinforced cockpit doors and passengers’ willingness to resist and subdue evildoers on board.
A similar pattern is unfolding in the financial world. Technologies had fueled the recent boom and bust, first by enabling the creation and trading of highly complex instruments at such speeds and volumes that neither regulators nor, in many cases, bankers and risk managers could track or even understand them. Now, technology is being touted as a means of preventing future crises.
There has been plenty of soul-searching about what went wrong in risk management. In an October 2009 report, a committee of regulators from Canada, France, Germany, Japan, Switzerland, the U.K. and the U.S. known as the Senior Supervisors Group faulted, among other things, “inadequate and often fragmented technological infrastructures that hindered effective risk identification.”
It’s not as if financial institutions were devoid of risk systems and tools before and during the crisis. Analytics and models from companies like MSCI and RiskMetrics Group (which announced their merger in March) have been standard equipment for years.
In this case, however, technologists are part of the critical chorus, conceding that many of their products were not well used. They are freely offering their customers advice that might just be the financial-management equivalent of locking the cockpit door and getting people more involved in the process.
A common refrain is that finance as a whole and risk management in particular became too enamored with and dependent on mathematics, exalting quantitative models and algorithms and crowding out human judgment. “There was too much faith in models,” states Robert Park, president and CEO of FinancialCAD Corp., a 20-year-old Vancouver, Canada–based company that supplies risk and derivatives systems to banks and Wall Street firms, asset managers, insurance companies, central banks and regulators in more than 80 countries. “Tools are only as good as the people who use them,” he adds.
Dan diBartolomeo, president of Northfield Information Services, which he founded in Boston in 1986, says that technology — which he defines as “systems that give you all the information that you need to make decisions and to know and evaluate what you have done” — can’t be blamed for human failures, nor help in cases where models are flawed.
DiBartolomeo, who provided analytical support to Harry Markopolos, the Massachusetts investor who tried to blow the whistle on Bernard Madoff, contends that technology gets undermined when it is not applied rigorously and appropriately, especially because “people tend to take shortcuts.” He is concerned that long-term investors such as pension funds, in their rush to shore up risk management, are buying high-tech solutions that are better suited for the very different needs of sell-side trading desks.
Some cutting-edge technologies show promise for risk management. Virtually all vendors are accelerating their processing speeds to meet growing demand for real-time risk snapshots. Aleri, a Chicago-based company recently acquired by database leader Sybase, developed a liquidity risk management system around complex event processing, a technique designed for high-volume data crunching that has also been applied in algorithmic trading and market surveillance.
Michael Bechara, head of Granite Consulting Group in Brewster, New York, says that neural networks, a form of artificial intelligence that has been used to scan transaction records and detect money-laundering frauds, are beginning to generate interest as an enterprise risk management aid. And Bank of New York Mellon has an ownership stake in and strategic alliance with Berkeley Heights, New Jersey–based Investor Analytics, which builds behavioral economics principles into its risk analysis.
Still, says FINCAD’s Park, “Technology can improve things with better information in a more timely way. But just having better and more timely information doesn’t guarantee decisions will be better.”
Jeffrey Kutler is editor-in-chief of Risk Professional magazine, published by the Global Association of Risk Professionals.
