(Bloomberg) — AQR Capital Management has appointed a head of machine learning for the first time in its two-decade history.
Marcos Lopez de Prado is joining the $226 billion money manager to fill the freshly created role, according to a Wednesday statement. He left Chicago-based money manager Guggenheim earlier this year after building their proprietary quantitative-strategy business, departing along with his team at the time.
There’s no suggestion the Greenwich, Connecticut-based firm will be moving away from its core as a giant in the factor-investing community, despite recent lagging returns.
AQR has been exploring self-taught statistical techniques, without a dedicated effort until now. Co-founder Cliff Asness has previously expressed skepticism that machine learning will alter the very basics of investing, but has said there’s potential to enhance existing strategies.
Lopez de Prado will serve to complement existing research methods, and eventually create new trading inputs to include in AQR’s products.
“We are always relentlessly looking for new data, signals and techniques to strengthen our investment process and build long-term persistent strategies,” Asness said in the Wednesday announcement. “Our continued investment in machine learning will further advance these efforts.”
When Guggenheim announced Lopez de Prado’s departure in February, the firm said he was leaving to start his own business and agreed to transfer to him the quant unit he started. Guggenheim said at the time it looked forward “enthusiastically” to working with him and his team, and that it had amicably resolved a dispute over ownership of intellectual property.
Lopez de Prado then founded True Positive Technologies out of the $13 billion Guggenheim Partners quant spin-out, before being poached by AQR. It is unclear whether his former colleagues will eventually follow.
Machine learning is a broad category of data analysis, involving everything from traditional statistical methods at an accelerated pace to parsing seemingly random data sets with minimal human intervention to detect tradable patterns.
Lopez de Prado’s prior employers include Tudor Investment Corp. and Chicago’s Peak6 Investments LP. He also wrote “Advances in Financial Machine Learning,” a graduate textbook.