Early warning system for insurance sector

• Thesis title in Czech: Early warning system for insurance sector
• Thesis title in English: Early warning system for insurance sector
• Academic year of topic announcement: 2016/2017
• Thesis type: diploma thesis
• Thesis language: angličtina
• Department: Institute of Economic Studies (23-IES)
• Supervisor: doc. PhDr. Ing. et Ing. Petr Jakubík, Ph.D., Ph.D.
• Author: hidden - assigned by the advisor
• Date of registration: 08.06.2017
• Date of assignment: 08.06.2017

References

Billio, M., et al. (2011). Econometric measures of connectedness and systemic risk in the finance and insurance sectors. Journal of Financial Economics, Vol. 104, No. 3, pp. 535-559.

Holl_o, D., Kremer, M., Duca, M. L. (2012). CISS - A Composite Indicator of Systemic Stress in the Financial System. Macroprudential Research Network, ECB, Working Paper Series, No. 1426.

Duca, M. L., Peltonen, T. A. (2011). Macro-Financial Vulnerabilities and Future Financial Stress, Assessing Systemic Risks and Predicting Systemic Events. Macroprudential Research Network, ECB, Working Paper Series, No. 1311.

Gramlich, D., Miller, G. L., Oet, M. V., Ong, S. J. (2010). Early warning systems for systemic banking risk: critical review and modeling implications. Banks and Bank Systems, Vol. 5, No. 2, pp. 199-211.

Maddala, G. S. (1987). Limited Dependent Variable Models Using Panel Data. The Journal of Human Resources, Vol. 22, No. 3, pp. 307-338.

Preliminary scope of work in English

Motivation:
An early warning system (EWS) plays an important role as one of the pre-crisis
instrument in a regulatory and policy framework. EWS should detect instability
in a financial system to mitigate a potential negative impact of accumulated risk
timely. Since most of the EWS developed are focused on banking crisis, there is lack
of literature dealing with EWS that would be able to anticipate _nancial instability
of an insurance sector. The aim of this thesis is to create EWS that will be suitable
for the European insurance sector where EWS is not satisfactorily developed.

Hypotheses:
1. Hypothesis #1: Indicator of future financial instability of an insurance sector
providing an early warning can be developed based on available Solvency data.

2. Hypothesis #2: Model of EWS that would be able to predict financial insta-
bility of an insurance sector could be developed.

3. Hypothesis #3: Signal of financial instability of an insurance sector can be
given by the EWS model timely to provide sufficient time for supervisory
mitigating action.

Methodology:
The thesis will deal with an insurer's distress by de_ning a proxy indicator based
on a solvency position of individual insurance companies. This indicator will be used
as a dependent variable in models presented in the thesis. The overal solvency po-
sition will be explored under Solvency II framework.
Limited dependent variable models using panel data will be introduced for the
purpose of the thesis. As a cross-sectional dimension of the dataset individual insu-
rance companies involved in the European insurance sector will be employed.
The output of the model will be a probability of an outcome happening which
will be an insurer's distress. The distress will be explained by both variables descri-
bing _nancial environment and variables describing speci_cs of individual insurance
companies. Including those independent variables in the models, both outside and
inside real circumstances inuencing the insurance company performance will be
caught. Then, the optimal threshold for interpreting models output will be calcula-
ted. That will be done by a prediction accuracy and false positive/negative rates.
Data of individual insurance companies needed for the thesis will be obtained
from the Bloomberg database. In addition, publicly accessible databases such as
statistical o_ces, Eurostat, World Bank etc. might be used as a data source.

Expected Contribution:
I expect to contribute to the research literature by proposing the early warning
system particularly designed for the insurance sector. Since EWS for European in-
surance sector is not fully developed, I believe the thesis might play a role of a basis
on which further research would be build on.

Outline
1. Introduction - motivation, literature review
2. De_nition of an insurance company's distress - indicator as proxy variables
approach
3. Data overview, de_nition of variables
4. Model - discrete choice models with panel data
5. Results, emipirical analysis - optimal treshold for the model evaluation
6. Conclusion