It’s an alert we’ve seen too much lately: a newsflash that the BATS BZX Exchange was down for almost an hour followed two hours later by Direct Edge reporting a technical glitch causing it briefly to stop accepting orders.
Considering the BATS BZX Exchange and Direct Edge combine to handle about 20 percent of US stock market trading volume, it was unsettling news. Similar outages at the NYSE and CBOE over the past four months demonstrate a troubling trend that threatens the heart of our financial system. As the pace of trading only increases, a one hour exchange blackout can significantly impact trading strategies that depend on making adjustments in milliseconds – and that’s an impact every trader will feel.
Risk management has certainly received a massive amount of attention since the financial crisis, with regulators, market participants and pundits pontificating on it until they’re blue in the face. As a result, financial firms have invested billions on infrastructure enhancements and new risk platforms and applications to safeguard critical systems. Yet even with the regulatory mandates and growing IT spend, these headline-grabbing market failures seem to persist, and with alarming frequency, seriously undercutting trust in the markets.
A recent Global Risk Management survey of financial institutions by Deloitte found that only 38 percent of respondents believe that their risk systems were effective or very effective at managing operational risk – and that number drops to 32 percent when it comes to enterprise risk. Firms have the data they need to identify potential risks; they’re inundated with data that arrives in milliseconds, and from all directions. Access to data isn’t the issue. The struggle for most firms is finding the right solution to cut through the noise and expose the real intelligence in that data that can help them truly manage risk.
[Hackers to Exchanges: You’re Next ]
Data today comes fast and furiously from multiple sources in multiple formats, and firms must be able to take a comprehensive view of data, interpret disparate formats and recognize patterns across multiple streams. Data in a vacuum doesn’t provide detailed, contextual information. But if you can understand it in the context of other data, you can start to examine that data for patterns – and more accurately predict future problems. That means viewing data holistically to truly understand how multiple factors combine to create potentially catastrophic scenarios. This is where case-based reasoning (CBR) shines. Once a CBR solution recognizes the potential for failure, it can proactively recommend a solution based on similar past cases, and drive informed, timely action to prevent escalation.
BATS indicated the blackout last Tuesday was related to an internal network problem. If that’s the case, the right predictive analytics could have warned that their system was on the verge of failure, allowing for immediate action to prevent the outage. But once the genie is out of the bottle, the damage is done, both to your systems – and your reputation. How many more of these headlines do we have to see before firms finally get it right?
Jo Kinsella is CEO of Financial Services for Verdande Technology.