Wells Fargo uses machine learning for performance attribution

Clustering algo delivers speedier and more accurate explanations of portfolio returns

Quants at Wells Fargo Asset Management have built a machine learning model to explain why their strategies are winning or losing.

The project uses clustering algorithms to automate grunt work that has been traditionally given over to junior analysts – assessing day to day why some portfolios are performing differently from others.

“If you have a portfolio within a group of strategies that’s experiencing differential performance, you can feed the weights and your entire factor library into this