Receive GCMG News

Send a blank email with SUBSCRIBE in the subject line to to receive notice of risk management ideas and information added to BarrettWells and T3P websites.

Thursday, August 15, 2013

Complex Systems such as ‘economic risk’ demand a new approach to potential modelling?

“Many people in various fields tend to follow the Young method. If they are studying economics, or the human body and health, or the workings of the brain, they tend to work with abstractions and simplifications, reducing highly complex and interactive problems into modules, formulas, tidy statistics, and isolated organs that can be dissected. This approach can yield a partial picture of reality, much in the way that dissecting a corpse can tell you some things about the human body. But with these simplifications the living, breathing element is missing. You want to follow instead the Champollion model. You are not in a hurry. You prefer the holistic approach. You look at the object of study from as many angles as possible, giving your thoughts added dimensions. You assume that the parts of any whole interact with one another and cannot be completely separated. In your mind, you get as close to the complicated truth and reality of your object of study as possible. In the process, great mysteries will unravel themselves before your eyes.”

This quotation from the book ‘Mastery’ written by Robert Greene, addresses the limitations of the formulaic, mathematics based approach to studying what are now called ‘Complex Systems’.

Complex Systems are systems that operate on the basis of so many random variables that they cannot be successfully expressed within mathematical models. You will find a fascinating explanation of Complex Systems, and the technique recently developed in order to understand them, on the TED website – - in a presentation by James B Glattfelder, titled ‘Who Controls the World?’

The global economy is a Complex System since it is created and influenced by the actions and decisions of billions of global citizens. Each action and decision is potentially a random variable.

Credit risk, particularly Probability of Default, is a Complex System as it is impacted by many variables both within a business entity and external to a business. That is the reason why Probability of Default has not been successfully modelled at the enterprise level.

Statistically derived Probability of Default does have a use at the portfolio level – provided the potential failure of one customer or supplier is a sufficiently small portion of the whole portfolio so as to be survivable. In other words, provided the portfolio is large enough and diverse enough.

This supports the contention that credit risk at the enterprise level should not be managed on the basis of Probability of Default, as derived from a mathematical model. It should rather be based on a holistic assessment of the forces shaping the future viability of the enterprise, and therefore on the practical probability that it will pay or fail to pay its debts in full and on time.