Article Detail

A Financial Stress Index for the Analysis of XBRL Data

Journal 34: Cass-Capco Institute Paper Series on Risk

Amira Dridi, Silvia Figini, Paolo Giudici, Mohamed Limam

In this paper we present a novel nonparametric approach to measure financial stress index to analyze financial data arising from balance sheets. We compare the results achieved with our approach with classical methods based on variance equal weight in a cross validation exercise via Monte Carlo simulation. Empirical investigations achieved on a real dataset show that our proposed approach provides a better performance in terms of error rate.

Recently, researchers have become interested in stress periods that measure the vulnerability of a financial system in order to take actions to avoid potential crises. Appropriate methodologies, such as financial stress testing [Alexander and Sheedy (2008); Committee on the Global Financial System (2005); Kupiec (1998)] and financial stress index (FSI) [Dridi et al. (2010); Cardarelli et al. (2009); Balakrishnan et al. (2009)] are used to determine stress intensity of a given system.

The stress level is measured on a scale ranging from tranquil situations, where stress is quasi-absent, to extreme distress, where the system goes through a severe crisis. The interaction between the shock’s magnitude and the system’s fragility determines the stress level. One relevant measure of the latter is the FSI.

Related indices have been developed by Hanschel and Monnin (2003) for the Swiss National Bank and by Illing and Liu (2006) for Canada. Cardarelli et al. (2009) computed an FSI for developed economies, and Balakrishnan et al. (2009) proposed an FSI for emerging countries and compared it with developed ones. Dridi et al. (2010) proposed an FSI based on extreme value theory (EVT) to determine stress periods in the Tunisian banking system. They propose to merge the information available on a measure called FSIj, j=1,...,n, which underlines the relative stress index for the jth statistical unit, using a variance equal weights approach (VEW). The latter computes the FSI first by standardizing the variables and then averaging the standardized scores using identical weights.

The aim of this paper is to extend the methodology proposed by Dridi et al. (2010) to a different context. More precisely, on the basis of a real dataset composed of XBRL balance sheets, we propose to measure FSI for the financial data of small- and medium-sized enterprises (SME). From a methodological perspective, the novelty of the proposed approach is the introduction of a nonparametric FSI based on the empirical distribution function (EDF). We also provide a Monte Carlo computational procedure to compute and assess the Value at Risk (VaR) for FSI models using the EDF and VEW methods. Empirical evidences are achieved on the basis of financial indicators based on liquidity and debt information extracted from XBRL balance sheets.


Leave a comment

Comments are moderated and will be posted if they are on-topic and not abusive. For more information, please see our Comments FAQ
This question is for testing whether you are a human visitor and to prevent automated spam submissions.