The Bank of England will publish the stress test results next week (having brought it forward from December).
The big banks face a tough annual cyclic scenario and an additional “lower-for-longer” exploratory scenario. The longer scenario is also interesting as it becomes less dominated by mortgage losses and the longer term will need an increased focus on retail
unsecured and wholesale lending.
This is going to increase the need for better models in some harder to model areas like cards and car loans where there there isn't such good long run loss data to build traditional regression models.
This means that the changes driven by stress testing are far from over and it should drive more use of modern data science and exploration of alternative approaches such as agent based modelling.
Whether the banks will have time to do these also faced with an EBA stress test in 2018 remains to be seen. Interesting times.