Abstract

Time Parameter-Variable Analysis and Visualization of Financial Data
Investment Process Management Methods for Fund-Based Investment Forms

When investing in securities, many active decisions are made on the basis of information gained with the help of empirical analyses of historical financial data. Unfortunately, most of the results of these analyses are highly susceptible to any change in the time parameters, such as the starting point and end point of the data used for the analysis, periodicity, and the width of the time window in the case of a rolling analysis. Slight modifications in these can lead to contradictory findings. Even professional asset management institutions are affected by this problem complex. Within a structured investment process, there are numerous decision points at which the results of empirical analyses influence the process. Problems related to the assessment of investment strategies, e.g. by means of backtests, problems related to the evaluation of optimization parameters such as variances, covariances, correlations, and risk premiums from historical financial market data, problems related to the selection of investment instruments, such as the selection of funds, or problems related to the consideration of investment decisions arise either consciously or unconsciously because supposedly definite calculation results fluctuate substantially due to the variation of the examined time period. In all areas it is possible to detect a large variety of different problems by varying the underlying time period used for the corresponding analyses. In addition to the problems that occur in connection with an investment process relating to fund-based types of investment, the dissertation also presents approaches to solutions for systematic variation of time parameters and their analysis. Thanks to the visualization, the complexity of the calculation and especially the substantially grown result set remain in the background, while the decision and assessment base of the analysis is expanded considerably. Thus, the visualization represents the decisive key to the inclusion of the insight gained. The paper also serves to create a new bridge between financial analysis and visualization, which is one of the most dynamic sciences of the recent past.
The combination of the concepts of time parameter variation and visualization lead to a large number of new methods and produce surprising results, which can be profitably utilised in the academic field as well as in the area of financial analysis in practice.