A spectral visualization system for analyzing financial time series data

  • Authors:

  • Daniel A. Keim;Tilo Nietzschmann;Norman Schelwies;Jörn Schneidewind;Tobias Schreck;Hartmut Ziegler

  • Affiliations:

  • Department of Computer and Information Science, University of Konstanz, Germany;Department of Computer and Information Science, University of Konstanz, Germany;DekaBank, Frankfurt am Main, Germany;Department of Computer and Information Science, University of Konstanz, Germany;Department of Computer and Information Science, University of Konstanz, Germany;Department of Computer and Information Science, University of Konstanz, Germany

  • Venue:
  • EUROVIS'06 Proceedings of the Eighth Joint Eurographics / IEEE VGTC conference on Visualization
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Visual data analysis of time related data sets has attracted much research interest recently, and a number of sophisticated visualization methods have been proposed in the past. In financial analysis, however, the most important and most common visualization techniques for time series data is the traditional line- or bar chart. Although these are intuitive and make it easy to spot the effect of key events on a asset's price, and its return over a given period of time, price charts do not allow the easy perception of relative movements in terms of growth rates, which is the key feature of any price-related time series. This paper presents a novel Growth Matrix visualization technique for analyzing assets. It extends the ability of existing chart techniques by not only visualizing asset return rates over fixed time frames, but over the full spectrum of all subintervals present in a given time frame, in a single view. At the same time, the technique allows a comparison of subinterval return rates among groups of even a few hundreds of assets. This provides a powerful way for analyzing financial data, since it allows the identification of strong and weak periods of assets as compared to global market characteristics, and thus allows a more encompassing visual classification into "good" and "poor" performers than existing chart techniques. We illustrate the technique by real-world examples showing the abilities of the new approach, and its high relevance for financial analysis tasks.