Abstract

The Economic Value of Integrated Data Models for the
Financial Industry

Vanini,P.; Glatter, A.; Janssen, M.

Universität Zürich

A necessary condition for successful decision-making, analysis and processing in the banking industry is the availability of comprehensive and consistent data. Basically, there are two approaches for structuring data: Data exploration and data modeling. These two methods are useful under different circumstances and we first discuss the main criteria to determine which approach should be followed. In a second step a dynamic cost analysis for the two methods of structuring data is performed, where the model captures the criteria discussed such as semantic and syntactic consistency, comprehensiveness, cost of data acquisition and data entry, development costs, project risk and transaction costs. After the cost analysis we tackle the question whether a data model should be developed in-house or whether an existing data model should be bought. This turns out to be largely a multi-agent with hidden information. The interaction is modeled and solved first in a general setup and an example then highlights the different investment costs due to hidden information and differing incentives of the agents in a firm.