eThesis - School of Business electronic theses

In this thesis, I develop a financial stress index (FSI) for the Finnish financial system. The FSI aims to reflect the functionality of the system and to provide an aggregate measure of financial stress in the money, bond, equity and foreign exchange markets as well as in the banking sector. The FSI is a composite index that combines information from these markets and provides a measure of stress in the financial system as a whole. The FSI has obvious benefits for all participants in the financial markets who need a tool for monitoring the functioning of the financial system, as it provides information on systemic stress events which are not as easily captured with the stress measures of individual markets or sectors. The index is based on market data and uses 14 individual financial stress measures and tests three different methods for the aggregation of these measures into a financial stress index. The resulting stress index is a continuous variable, with a monthly frequency and covers the most important parts of the financial system. The main practical motivation behind the study is the demand for a measure indicating general systemic stress in the financial system, emerging from the current issues of macro-prudential policy.

Based on a literature review, I use variance-equal weighting, principal component analysis and an application of portfolio theory to form the FSI for Finland. I continue by assessing the different methods and resulting FSIs based on their ability to identify past financial stress periods as well as their ability to capture standard definitions of systemic financial stress.

Finally, I also examine the relationship between financial stress, measured by the FSI, and the real economy. I use industrial production growth as a proxy for the development of the real economy. Utilizing three different regression models, I find that shocks in the FSI have a statistically significant negative effect on industrial production. Applying a threshold vector autoregressive model, I also find evidence that the relationship between stress and the real economy is nonlinear and the effects of a shock in the FSI depend on the prevailing state of the financial system.