Applying data visualization and knowledge discovery in databases to segment the market for risky financial assets

This paper uses the techniques of knowledge discovery in databases (KDD) and data visualization as a methodology to uncover significant clusters in the ownership of risky financial assets. Partitioning by medoids and data visualization identifies two significant clusters among risky asset holders. Cluster one is comparatively young, low wealth, high income consumers, with mortgage debt and regular savings patterns compared with a segment of older, low income, high wealth irregular savers with outright ownership of property. The analysis reveals that previous economic research into portfolio choices may have missed some important interactions and that specific, in addition to generic product needs, can vary with the lifecycle. Copyright © 1999 John Wiley & Sons, Ltd.