The financial markets are still experiencing the aftershocks of the pandemic, with volatility remaining above long-term average rates.

Authors

Fill form to download

Emerging stronger Using machine learning to calculate value at risk across asset classes-2

As stability slowly begins to return, banks must be able to make more accurate and timely valuations, meaning that they need to incorporate advanced modelling techniques into their arsenal. Neural networks, a type of machine learning which focus on complex data relationships, can ‘learn’ how to act as pricing engines to calibrate new models, support future-exposure modeling to carry out valuation adjustments, and enable accurate and speedy exposure calculations.

To maximize the opportunities presented by these new machine learning techniques, banks must deepen their engagement with machine learning and develop use cases for testing, before rolling out at scale.  Those banks which act now to embrace machine learning in their mainstream operations through building their capabilities will be able to move more quickly and make decisions more accurately than their competitors. In short, they will be better prepared to face the challenges that lie ahead.

This article was originally published on McKinsey.com on October 29, 2020 and is reprinted here by permission.