Our client is a global PE firm. To improve long-term portfolio performance, they needed a more data-driven, systematic method of analysing risk in the markets they were exploring.

Using machine learning, we analysed 140 factors – such as perceived sentiment towards political leaders, commodity prices or levels of food production – to predict the likelihood of a political, economic or social crisis within a given country. These insights are then visualised and rationalised, supporting their investment decisions.