Predicting UK Bank Distress with Machine Learning

Posted on October 20, 2019 by Editor

The UK’s Prudential Regulation Authority (PRA) is working on machine learning techniques to predict bank distress. Using novel data sources, mainly confidential regulatory returns, this still experimental early warning system for predicting bank distress significantly outperformed other models.

This development follows the Danish Business Authority’s machine learning-based Early Warning Europe, which uses large data sets from five European countries to distinguish companies at early stages of crisis from financially sound companies. With more regulators across Europe mandating XBRL based, digital, standardised data, these kinds of large-scale machine learning tools are increasingly feasible.

The paper describes the manner in which machine learning and AI are playing an increasingly important role in modern financial markets. Ensuring a reliable feed of machine-readable data will allow these innovative models to continue to develop. Without high quality structured data from “the horses mouth”, AI projects of this nature expend 90%+ of their effort in attempting to organise and categorise unstructured information.

Read more here.

Other Posts


Would you like
to learn more?

Join our Newsletter mailing list to
stay plugged in to the latest
information about XBRL around the world.

By clicking submit you agree to the XBRL International privacy policy which can be found at xbrl.org/privacy