PDF Cassandra Brochure

CASSANDRA aims to prevent two forms of market abuse.

  • Insider trading; the trading in stocks and shares using non public information.
  • Market manipulation; the publication of misleading, false, or inaccurate information which drives up or down share prices.

CASSANDRA combines the power of statistical modelling approaches with headline analysis and web search capability to provide both early warning and retrospective analysis of data. Insider trading can occur in many scenarios.

  • Merger or acquisition between two companies
  • Change of senior management (CEO etc)
  • Approval or rejection of a pharmaceutical company’s latest drug.

This information becomes public in many different formats, company web sites, news stories (Bloomberg, Reuters), chat rooms (Yahoo, Motely Fool). CASSANDRA operates in two stages:

  • Identifying “areas of interest” in a financial time series. Using an event date, and then moving backwards over any number of trading days, unusual trading days are identified and stored for further analysis.
  • Using semantic web technology we seek to identify internet based stories that relate to the company under investigation. Using sophisticated artificial intelligence software we aim to provide clients with a Decision Support System which “tells them” when market manipulation may occur by explaining sudden changes in price, volume and trade in terms of activities in a wider world context.

It is this merger of statistical, artificial intelligence and semantic web technologies which gives CASSANDRA its power and offers financial and business organisations a powerful tool in the fight against market abuse and manipulation.