Henry Ford famously remarked that “a customer can have a car painted any color he wants as long as it’s black”. Whether the founder of the eponymous car-maker uttered those words is a question for historians. And in those early days of driving, a one-size-fits-all approach was probably the most efficient way to make the Model T.

A little over a century later, the same assumption prevailed for bank stress tests – until Silicon Valley Bank, Signature Bank, First Republic et al showed that since all bank balance sheets are not alike, the scenarios they ought to include in their stress tests should also be different. But are there really practical ways of tailoring prescribed scenarios and correlations between risk drivers for all 5,000 or so U.S. banks, without creating overwhelming complexity? And if so, what’s the best solution?

For the answer, check out Alla Gil’s latest Global Association of Risk Professionals (GARP) column, which you can read here (HINT: it involves using AI-driven software to generate full-range scenarios and then linking those scenarios to banks’ exposures).