Building a resilient credit union, and credit union system, is the one big challenge facing those leading credit unions. The recent pandemic presents a huge economic disruption and a test. At the same time it raises questions about how we manage risks.
In a new book, Radical Uncertainty, two esteemed British economists are critical of conventional approaches to managing business risks and the extensive use of statistical models. They argue that the elegance and complexity of these models too often disguise the arbitrary constructs and probabilities that are built into them. The two also argue that models always ignore the unknown events, events that are simply outside our frames of reference.
For a quick read, John Kay provides a succinct summary of some of the arguments, but ‘updated’ to feature the COVID-19 virus, in this Prospect online article. Mervin King, a former Governor of the Bank of England, introduced the seeds of this new book in his last book, The End of Alchemy – as noted in this Bloomberg Review. The review highlights some of King’s radical proposals for financial services too.
This is from Kay’s recent article:
A global pandemic was a likely event at some point, a known unknown in that sense. But the occurrence of such a pandemic in 2020 was not a very likely event, and we could not in advance do anything more than guess at what form it would take, and even then our guesswork was likely to be limited by mixing and matching between what we know about more familiar pathogens. We could acknowledge the possibility of something new and different, outside the range of past experience, but have only a limited ability to imagine what this might be, still less reckon with the probability of it coming to pass. The question “what was the probability that coronavirus would break out in Wuhan in December 2019?” is not one to which there is any sensible answer.
The authors are tough on the ‘quants’ that have led the way in model building and in debates about financial risks and risk measurement. These are the people who have dominated investment banking for a generation and also paved the road to the crash of 2008. Kay and King propose that statistical methods can be used, but only with some reasoned judgement that accepts the weaknesses and likely blind spots of the analytical tools.
Intriguingly, this week’s Economist plays up a similar theme in discussing models being used to project a pandemic’s path, and its economic impact.
The book’s approach prompted two different thoughts related to our credit union world here in BC.
First, I grow sceptical of the heavy reliance our regulator places upon CUDIC risk scores and various stress tests. These models are not very complex, but they provide a convenient ‘cloak’ on reality, and likely redirect too much attention to that which is being measured. The models, and the metrics they employ, start to become more important than the observable larger world. I will have to write another post about the weaknesses of the CUDIC model itself.
And second, the ‘quantitative’ methods almost always emphasise efficiency and short-term returns. These issues are at the very heart of consolidation trends and operating at scale. But these business strategies obscure the concentration risks and appear to undermine longer term resiliency of the networked ‘node’ or community. Kay’s article closes with the following paragraph; his allusions to manufacturing have parallels in financial services where operational efficiency, liquidity and capital are ‘optimised and leveraged’:
But as the economy is convulsed by the coronavirus-induced lockouts, shutdowns and panic purchases, other (non-financial) modern business fashions—such as lean production and just-in-time inventory management—are likewise exposed as dangerous devices for flattering short-term profits at the expense of long-term business resilience. The vicissitudes of our uncertain world have not only subjected our society to a brief if nasty disease, but also exposed our economy’s susceptibility to, in the parlance of the hour, a serious underlying condition.
For traditional credit unions, ‘mutual aid’ is a larger concept that may have exceptional value in times of stress, as people lean on each other. Institutions from afar, built around a profit-seeking model, are less likely to display the compassion or responsiveness needed.