University of Twente Student Theses

Login

An economic capital model for APG Group?

Terpoorten, Hidde (2013) An economic capital model for APG Group?

[img] PDF
1MB
Abstract:APG Group is a financial services company that offers asset management and administration services to Dutch pension funds. Additionally it provides life and disability insurance to pension holders via insurer Loyalis. The Corporate Risk Control (CRC) department of APG wanted to investigate the possibilities to implement Economic Capital (EC) as a tool for risk management and capital planning on APG Group level, also because new regulation requires APG to reassess capital adequacy. EC is a risk measure that represents an economic view on the risks an organisation is exposed to and translates risk into capital buffers needed to survive these risks materialising at a specified confidence level and time horizon. EC can enable APG CRC to translate different risk sources into one common risk currency so that risk levels and risk appetite can be compared between the activities of APG Group. The risk profile of APG Group consists of operational risks originating from asset management and pension administration activities. Additionally insurance company Loyalis is exposed to specific market, credit and insurance risks. Because only a preliminary EC-model for Loyalis is available this research is exploratory and focuses on bringing operational risk into the EC-framework because this is the most important risk category at APG. Methodologies developed by the banking sector are used to quantify operational risk. These methods estimate capital buffers via statistical analysis and simulation of losses. These approaches use information about historical losses as an input. Therefore data on operational losses at APG was collected from different sources, combined and analysed. This analysis showed operational losses at APG have the same characteristics as losses in banking or insurance. This result is only indicative since data on losses is scarce. Because the data driven approach did not produce robust results, structural assessment of experts, is used as input to estimate operational risk via a procedure known as expert elicitation. Using expert opinions on severity and frequency of operational losses lead to a capital estimate 5% off of the internal capital estimate. The EC-model that was already present, but not actively used, for the financial risks of Loyalis was analysed leading to the conclusion that it contains outdated assumptions and methodology to provide useable capital estimates. If APG CRC wants to further implement EC as a risk measure to use in management information, scenario analysis and capital planning the current EC-model has to be updated. Considerable effort of management and experts is needed to bring the model up to market standards. Considering the focus of APG risk management on risk control on process level, EC might provide another perspective on the risk profile of APG. Any new risk measurement methodology should leverage existing procedures and available information so the risk control framework and available data should be part of any new quantitative risk measure, be it EC or other. As follow up from this research the registration and collection of information on operational events is improved. The proposed expert elicitation and loss simulation methodology to quantify operational risk can be implemented in the risk control framework to become a standard part of risk management procedures. This research can serve as a starting point for capital calculation and allocation at APG since this aspect becomes more and more important due to a new regulatory regime APG is subject to.
Item Type:Essay (Master)
Clients:
APG Group
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:85 business administration, organizational science
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:http://purl.utwente.nl/essays/63838
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page