Academic journal article Journal of Financial Management & Analysis

Weighting Methods for Financial Stress Indices - Comparison and Implications for Risk Management

Academic journal article Journal of Financial Management & Analysis

Weighting Methods for Financial Stress Indices - Comparison and Implications for Risk Management

Article excerpt


Increased risk in financial markets has raised interest for measures of financial markets' stress (systemic financial stress). Financial stress indices (FSIs) are applied by individual banks to assess and predict aggregate financial risk, while financial supervisors may base macroprudential policy on the level of stress indicated from FSIs. In all cases, the quality of systemic risk management depends on the quality of the systemic risk measure.

Since the inception of the most recent crisis, a number of FSIs have been developed (see the overview in Table 1). While it is questionable if systemic stress can be captured in a single measure', almost all FSIs involve the choice of different indicators for financial stress and their aggregation into one overall measure for systemic financial stress. While constructing FSIs, special emphasis is placed on the question of what indicators and/or markets to choose as a representation of overall risk and how to synthesize them. While the relevance of weighting methods is recognized principally2, there are few approaches to assess them systematically. A major contribution is the empirical comparison of four different weighting methods by llling and Liu3. In many cases, a specific weighting method - often equal weighting4 or principal components - is simply selected without being explicitly discussed and indicators are aggregated without explaining the sensitivity to the weighting mechanism.

However, this simple selection has to be considered critically, as the weighting of indicators affects the level of the FSI and is assumed to further impact the FSI's behavior over time. This impact may be particularly relevant for financial markets with dynamic and individually behaving submarkets. Therefore, the objective of this study is to explore in more detail the relevance of weighting methods for the construction and application of FSIs. Particularly, it is asked :

* from a conceptual perspective, if and how weighting matters for FSIs,

* from an empirical perspective, what are the differences and similarities across time between FSIs constructed on the basis of various weighting schemes, and what is their sensitivity with respect to different sets of indicators used for FSI construction,

* from an application perspective, what are the implications for the selection of optimal weighting concepts in the context of FSIs.

Where the conceptual discussion grounds on evidence from literature and own considerations, empirical aspects are investigated on the basis of stress (sub-) indices that - while referring to the same time series indicators - are aggregated using different weighting methods. Applying data from 1991 to 201 1 and six different weighting concepts, two sets of indicators are implemented to capture the sensitivity of weighting methods to the data series. In the remainder of this study, the relevance and scope of weighting regimes is assessed. Empirical aspects are analyzed while examining the statistical properties of single FSIs and comparing them to one another. While much evidence is found for the relevance of weighting techniques, the final selection of appropriate weightings depends on the overall architecture of FSIs.

FSI Construction Excitation Monitoring in Financial Systems

FSIs are statistical instruments for monitoring the excitation in financial systems. Beyond conveying information about the state of the financial system (e.g., the level of instability or stability), they are capable of shedding light on the origins of financial stress. Knowing the level of stress and its origins is then conducive for actions in risk management, both from the perspectives of the individual agents (e.g., investors) and the institutional agents (e.g. firms and supervisors). However, the quality of these actions depends on the quality of the underlying FSI including its weighting technique as a specific construction principle. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.