Tuesday, September 27, 2011

A New Way of Bank Stress Testing

THE BAYESIAN METHOD

In January 2009, during the depths of the financial crisis, the Basel Committee on Banking Supervision published its view that the bank stress-testing which had so-far taken place across Europe and the US was insufficient for a number of reasons.

While the financial market had grown considerably more complex and less transparent since the previous round of financial crises at the end of the 20th century, stress-testing had not kept up with the times. Forward-looking information was not-being taken into account. 

Rather, stress-test so far has been undertaken in the so-called frequentist way, meaning that purely quantitative and statistical approaches have been used. By definition, this way of doing things only examines historical data. Subjective data, which, in expert hands could be used to detect when it is that the current circumstances are likely to change -perhaps even in a heretofore unseen way- is overlooked. Also overlooked is forward-looking information of any kind. Nassim Taleb's famous book The Black Swan swan points to precisely this sort of thinking as being a major weakness in daily wisdom of financial economists prior to the start of the crisis. The idea that the situation might change was not appreciated. Nor was the idea that we may need to update our information to take into account current happenings.

In 2009 and 2010 Committee of European Banking Supervisors (CEBS) implemented the first and second EU-wide stress tests which were conducted along frequentist lines, examining at first only the largest 22 banks in Europe and later all of the systemically most important bank at the national level in Europe. Not surprisingly, the first two European stress-tests were widely considered failures. In 2011, the newly commissioned European Banking Authority (EBA) has thus taken to the task. Because the European stress tests tried to answer the question of what would happen in the event of both current loss expectations as well as a worst-case scenario whereby losses far exceeded current loss expectations but overlooked the possibility of multi-stage economic shocks, they represented the frequentist point of view. Essentially, they do not revise losses in a dynamic way. Thus, they have left something to be desired, irrespective of the number of banks or percentage of GDP investigated. 

Many experts now consider that the key missing ingredient from stress testing procedures is a solid way of capturing the co-relationships between the various stress events. With this in mind, the Bayesian net can help us understand the relationship between some of the key events. Consider for example the relationship between events A and C. 


The Bayesian Net

WHAT IS NEEDED IN STRESS TESTING?
The most important aspect needed in stress testing in order to overcome black swan events is a forward-looking element. Subjective probability is a plausible way in which this can be achieved. A subjective probability is essentially an opinion – hopefully a well-informed one – regarding the probability distribution of an event occurring. While there is no mathematical proof behind the answer, one can expect that in addition to historical probability distribution, a subjective probability might be influenced by expectations, indicators as to what may occur in the future, and indicators as to why this time might be different. The most significant upsides to this approach are the incorporation of a wide range of indicative and qualitative data, as well as the non-reliance on vast and often difficult-to-obtain amounts of data. The latter factor would render this approach particularly valuable during scenarios involving one or more extremely rare (statistical tail) events.

Another important issue which needs to be addressed in stress testing is correlation of shocks. In many stress tests, this has truly come to be a key missing ingredient. The simple fact that a financial institution survives an economic shock may might not be perfectly indicative of bank-survivability when the economic shock in question might also set off (or might otherwise be associated with) a chain reaction of economic shocks.  
-----------------------------------------------------------
About the Authors:
-Max Berre is an economist at the EDHEC-Risk Institute (Ecole Des Hautes Etudes Commerciales du Nord) who has worked as a sovereign debt expert at the Inter-American Development Bank in Washington and has taught financial economics at Maastricht University in the Netherlands.
-Kevin Hoefman is a lecturer of software programming at Hogeschool West-Vlaanderen and a former software engineer and video game programmer. 

No comments:

Post a Comment