See Why Abu Dhabi’s Big Sovereign Wealth Fund Wants an AI Group for Investing
Posted on 06/08/2021
As sovereign investors rapidly embrace data as the new oil, artificial intelligence (AI) can be perceived as the refineries.
SWFI reported earlier that both the Abu Dhabi Investment Authority (ADIA) and Singapore’s GIC Private Limited are in a race to construct cutting-edge quantitative investment operations. Sovereign funds have been allocating to quant hedge funds like Citadel, Point72 Management, D.E. Shaw, Two Sigma Investments, and others for quite some time, and believe they can replicate those strategies internally, saving mounds of cash on exorbitant fees. Utilizing artificial intelligence, specifically machine-learning in quant investing, appears to be the holy grail for many institutional allocators, including large sovereign investors and public pension plans such as the Canada Pension Plan Investment Board (CPP Investments). Singapore’s sovereign wealth giant GIC Private Limited formed Kepler FI in 2017 to test out new concepts and strategies in the world of institutional investing. Kepler has a project called Alpha Capture which is a multi-asset class project spanning quantitative research, alpha capture, and optimization. Some parts of Kepler’s adventures include applying machine learning techniques to create a framework for forecasting financial statements’ line items to derive the intrinsic value of a stock. GIC has a sizable buyside equity team that could benefit from greater forecasting tools on stocks.
Witnessing its peers building out quant teams such as CPP Investments and GIC Private Limited, ADIA started to build out its own quantitative research and investment group, hiring former professors, former hedge fund employees, data specialists and investment specialists. The quantitative research group sits in ADIA’s strategy and planning department.
Concepts such as hierarchical risk parity could be potential strategies for sovereign investors. Hierarchical risk parity is a risk-based portfolio optimization algorithm that applies machine learning techniques to seek out underlying hierarchical correlation structure of the portfolio. This would permit groups, or clusters, of similar assets to compete for capital. Assets and factors are moved into clusters with similar characteristics versus traditional asset classes. This can also be thought of as a form of multilevel risk optimization. Earlier, ADIA hired Marcos Lopez de Prado to be Global Head of Quantitative Research & Development. Some of López de Prado’s research centered around hierarchical risk parity.
Some institutional investors are looking to ESG factors and integrating them in a quantitative fashion. For example, Goldman Sachs Asset Management believes some of these ESG signals, especially data from greenhouse gas emissions, could lead to alpha generation.
All in all, sovereign investors don’t want to be left out in the AI race in the financial services industry.