The long and complex road to CEP

While consolidation in the trendy CEP market is very big news - it is worth taking a step back and looking at the true market penetration for complex event processing the financial services.

CEP collects related and unrelated real-time and historical data and events, business logic is then applied to those events to detect patterns and conduct analysis. It can be used in algorithmic trading to monitor trades and calculate risk on a real-time basis supposedly allowing for real-time hedging.

CEP is nowhere near an off-the-shelf application. The benefits of CEP rely on being tied into a firm's internal business model - allowing for business rules and logic to used in algorithmic trading and surveillance. That process takes an awful lot of investment, architecture design and coding.

With the purchase of Aleri, Sybase itself is dealing four sets of CEP codes - Coral8, Aleri, and the work-in-process Ohio, not to mention Sybase's own CEP initative. (In addition to that there is still debate going on between software engineers on whether Streaming SQL is the best way to develop CEP engines).

All this code, debate and investment does put Streambase's profiteering amnesty programme in a different light. It's not exactly like taking your old clunker in for a trade-in. (or even a Toyota with dodgy brakes)

There is obviously a market for CEP in financial services, especially for finding dealing pairs, risk monitoring and fraud detection.

However, as with any new and trendy technology there is a tendency for its hype to overpower the realities of developing, installing and maintaining a very complex and expensive IT environment.