Volatility and Automation are Straining Systems Capacity, Conference Told

Increased volatility and automation are straining system capacity from a market data and transaction processing perspective. So says a panel featuring Michael H. Boston, Principal, Trade Support Group Operations Manager Securties, Bank of America, Jacob Granek, Managing Directory, Depository Trust & Clearing Corporation and Thomas J. Jordan, President & CEO, Jordan & Jordan at SIFMA yesterday.

The workshop, entitled “Capacity Growth- Evidence, Warnings, and Alternatives” told delegates that there shall continue to be more data, more trades, and more operations to automate.

Numerous statistics outlined the observed trend – the number of quotes continues to grow, the bridge between quotes and trades widens, and the average size of trades decreases. The ratio between quotes and trades on OPRA is as disparate as 4800-to-1 while the average size of trades on NYSE is near 250. Around 90 per cent of all trades are cancelled. Message rates are peaking over 500,000 per second. DTCC processes over USD 2.5 trillion worth of securities everyday.

Panelists focused on observing actual behaviors within an organization to understand the bottlenecks in processing – thus potential failure points – and where they occur. Although a good rule of thumb is to have capacity equal to 2.5 times observed peak capacity, one may find a more satisfactory level of service objective by scaling systems to a desired probability of failure. By plotting instrumented throughput of the weakest components, and determining a maximum acceptable failure rate, IT staff can scale resources as necessary to meet a measurable goal.

Key to ongoing operations are points of measurement and meaningful reports that can be used to establish future capability requirements.