Stress Test Banks: Selection of key indicators for financial stability

Abstract

Financial crises impact on societies, which could lead to an economic stagnation. A historical example of this dynamic was the bankruptcy of the Bretton Woods agreement, in force from 1944 to 1971, when the president of the United States of America decided to devalue the dollar (Bernstein, 1984), which generated the need to begin to coordinate the finances at international level in a different way to the one known at that moment. The changes implemented after the bankruptcy of said agreement have proved not to be sufficient. Reflecting this is the last major global financial crisis that began in the US in 2007 and extended to other economies as of 2008. Although academics and regulatory institutions do not agree on the origin, scope and conclusions of this crisis (Lo, 2012), work has begun on a global level to try to mitigate the effects in case of similar situations, and to stabilize markets urgently.

Among the macroprudential tools that have had wide implementation and development to analyze the risks and effects of a possible new crisis are the calculation and analysis of the resistance of the entities to different hypothetical crisis situations. The current models implemented to carry out these analyzes are stress testing, which are designed by regulatory and supervisory entities, and in some cases they are even executed, generally using historical scenarios. The definition of such scenarios must consider the comments made by Dees, Henry & Martin (2017), who argue that these should reflect a sufficiently severe, but plausible financial crisis situation, but the more severe or extreme are the defined scenarios, less plausible they become. Despite the great progress made in this type of tests, there is a long way to go.

In this work, reference is made to the stress testing analyzes, more specifically to the definition and construction of the scenarios that allow carrying out this type of analysis. For this, more than 500 factors have been considering (typical of the banking and generic systems of macroeconomic level), associated with the credit risk of the Spanish financial system and the Andon methodology was applied to select those factors that met the SMART requirements (Specific, Measurable, Archivable, Relevant and Time-bond). The techniques and methodologies proposed in this study have allowed selecting five factors that seek to reflect and help the study of the sensitivity of the Spanish financial system against credit risk, from a generic perspective (top-down).