Discovering Intraday Market Risk Exposures in Unstructured Data Sources: The Case of Corporate Disclosures

Capital markets react promptly and significantly to critical events that have not been anticipated by market participants. Prominent examples of such market behaviour which risk management activities turn their attention to became evident during the last few months. Today, classic risk management tools perform complex calculations on the basis of structured data sources such as historical price series. In contrast, unstructured data sources, such as corporate disclosures, are widely disregarded. However, such data that has been released newly could carry information that is not reflected in the structured data available in such a situation. We aim to utilize such unstructured data sources and present a text mining-based approach to detect intraday market risk exposures that result from the event of newly published information. Our results provide evidence that unstructured data contains valuable information to discover intraday risk exposures promptly, i.e. represent valuable data sources in this context.