What do you do when you can see a problem building up? For instance, a new idea with good intentions, but that may lead to a mess at some point in the future. And all this is happening in a post-truth world, where the experts can see the problems but nobody trusts the experts.
This is the situation we face with the introduction of the expected credit loss provisioning model under IFRS9. It was generally felt that banks did not have sufficient reserves and their provisions were too low at the start of the financial crisis. The heads of government of the G20 leading nations agreed and asked standard setters to devise an expected loss model. So they did.
The IASB and its US equivalent the FASB put their heads together and came up with two different answers. Each has its merits and problems. The FASB lifetime expected loss model will produce bigger provisions and has a relatively simple measurement principle.
Institutions will recognise accounting losses when they issue loans and then show higher profits as those loans are repaid and the provisions are unwound. So the model provides for losses that institutions do not expect to make. The IFRS 9 model should deliver bigger provisions than the incurred model it replaces while being closer to economic reality than the US model. But it is fiendishly complex with its three stages. In particular, stage two, which requires banks to make provisions when credit quality has worsened.