Motivated by the use of high-dimensional data such as data from several hundred risk-factor changes in the realm of quantitative risk management, we raise the following simple question, namely, How can one detect and visualize dependence in high-dimensional data?