Where to look

Many banks are making good progress in self assessment, and collectingloss data and key risk indicator information. But some industry experts are starting to ask whether institutions are looking in the wrong place for clues to operational risk. Formal op risk data represents only a small proportion of the information that flows daily through organisations in the form of emails, phone calls, word processing documents and so on. It is in this so-called unstructured data that early warnings and evidence of fraud and breaches of compliance and corporate governance rules is to be found. But searching this data in real time takes sophisticated intelligent technology and the ability to handle massive data flows ( see page 4 ).

In the end, managing risk always comes down to having the appropriate data. Insight and innovation only take you so far, then the model, structure or methodology requires timely, clean and reliable data to produce meaningful results. In this respect, the credit markets have been running ahead of themselves, with burgeoning volumes not yet backed by comprehensive and open data sources. But this is changing with the expansion of credit derivatives data services and new products such as