John Bates has been posting about issues in the trading community, covering problems for example at one exchange that coincidentally is a customer of a different CEP vendor. Although the use of CEP technology may enable bigger and more dangerous trading “issues” to occur, such issues are certainly not the fault of the CEP technology per se. The need for users to apply basic good software engineering practices, such as the validation and verification of event patterns and business logic, applies as always!
As it happened I noticed John’s blog post at the same time as discovering a 2009 Gartner report (via Google) and hype cycle that covered CEP in financial service applications, and in particular claimed that the “CEP-based trading application” was tipping past the “hype peak” and sliding into the “trough of disillusionment”.
Although I respect the Gartner guys – who no doubt have many more information sources in this area than the CEP community – I don’t find this credible. CEP is not just a vendor technology but an approach to event-based systems. I wonder how trading applications can be built NOT to do complex event processing? If anything, trading applications are becoming more complex (requiring more risk and governance rules, volumes, etc etc) requiring yet more complex scalable pattern matching capabilities, more business user interfaces and control, etc etc – all of which are foci for CEP technology vendors.
Of course, Gartner could be implying that the remaining investment banks and hedge funds are developing their own technologies for CEP and turning away from the CEP vendors. Again, given the impressive and growing talent in the CEP vendor space, I find that unlikely too. But as John Bates says, “we can’t rest on our laurels”.