Xavier Desmadryl, Global Head of ESG Research, at HSBC Asset Management considers the growing appeal of ESG investing that has, fortuitously, coincided with the increasing capabilities of artificial intelligence (AI). However, they caution that AI does not herald the end of human participation in investment decisions.
Impact of the rise of ESG awareness and AI on asset management
Concerns over risks to biodiversity, climate change and numerous other Environmental, Social and Governance (ESG) issues surfaced largely in the wake of the 1992 Rio Earth Summit and got a boost with the Covid-19 pandemic crisis. These have triggered the need for finding some relevant information related to these topics which unfortunately happens to be as scarce as it is heterogeneous. Mining such information is therefore a complex process which can be greatly helped by the waft of emerging technologies collectively known as Artificial Intelligence (AI).
What is the biggest contribution AI can make to investment decisions?
AI will enable faster, more accurate and financially relevant investment decisions. Hopefully, the resulting new ESG datasets will help forming a better and more holistic understanding of the company’s investment cases. This process will be largely automated and this will allow portfolio managers, analysts and research teams to focus on higher value-added activities. But human involvement in ESG analysis will continue…and will actually start with the design of supporting research algorithms which will have to be crafted by human intelligence anyway. This nicely illustrates the fact that at HSBC Asset Management we believe that machines should not be left to their own devices. Automated processes require the guidance and oversight of people in order to prove genuinely efficient and not misleading. Notions of harmony, balance and efficiency must be reconciled between the (still unmatched) capacities of humans and the power of machines.
“We also continue to consider the notions of harmony, balance and efficiency which must be reconciled between the (still unmatched) capacities of humans and the power of machines.”
What part does AI play in these trends?
As a means for assembling volumes of data automatically, AI has benefitted from the increasing speed, processing power and storage capacity of computers. Output is bespoke because AI detects and reports issues, whether positive or negative, relating to pre-defined factors. A sub-set of AI, machine learning (ML), allows computers to classify data, solve problems, uncover and predict patterns without being explicitly programmed to do so. AI in asset management is really taking off. Algorithms allow the collection or building of missing information, like data on poorly-recorded companies. For example, the algorithms used to process images are very useful for making financial predictions, as the problem of restoring missing pixels in an image is similar to the challenge of “predicting” future changes in financial markets (see the illustration below). ML detects market anomalies and analyses random datasets. These can establish causal relations and thus forecast financial market changes.