Leveraging AI to capitalise on emerging market equities 

The popularity of machine learning algorithms (a branch of AI) has grown significantly as researchers and practitioners explore their potential to enhance investment returns. Robeco’s quant team has conducted a comprehensive study to assess the impact of applying machine learning algorithms to investing in emerging market equities. 

The findings demonstrate the efficacy of these algorithms in detecting complex, non-linear relationships among company characteristics, a task challenging for human researchers.

“We also found that using ensembling, or the ‘wisdom of the crowd’ for machine learning models, could increase expected returns net of trading costs by up to 2% per annum for equity investors,” writes Laurens Swinkels, Head of Quant Strategy at Robeco’s Sustainable Multi-Asset Solutions team. 

Robeco employed various machine learning methods, including elastic net, tree-based methods (random forests and gradient-boosted regression trees), and neural networks. These techniques outperformed traditional linear regression models by uncovering intricate relationships and enhancing expected returns.

Also, to assess investment performance in emerging market equities, Robeco back-tested signals from the machine learning models, forming portfolios based on predicted excess returns. The results indicated that machine learning models consistently outperformed linear models. 

“On average, the returns of the long/short portfolio derived from the two linear models, namely regression and elastic net, were around 0.8% per month,” says Matthias Hanauer, Researcher at Robeco’s Quant Equity Research team. 

“The random forest and gradient-boosted random tree methods generated higher returns of around 1.0% per month, while the neural networks method and a combination of all machine learning models delivered 1.2%. In short, linear models are good, but machine learning models are better,” he adds.

Furthermore, Hanauer acknowledges the influence of conventional quantitative factors but highlights that machine learning models provide additional economically significant insights when it comes to investing in emerging market equities. 

As per Robeco, despite transaction costs and short-selling constraints, using machine learning signals resulted in substantial net outperformance over the market, making it a recommendable strategy for investors.

Read the full insight here.