It’s Season 3 of the FSI Deep Dives and on November 7th, we gathered at the Deloitte Tower to discuss how AI and big data influence the integration of environmental, social, and governance (ESG) factors into investment analysis.
Our panelists were Bud Sturmak, Founder of BlueSky Investment Management, who joined us from New York to discuss their approach to investment analysis, Bouchra M’Zali, Professor with the Department of Strategy, Social and Environmental Responsibility at UQÀM, who discussed how AI is impacting her research in the field of social responsibility and sustainable development and Laure Fouin, Associate at McCarthy Tétrault and member of the FSI Board of Directors, who offered her perspective on fintech governance issues. The panel was moderated by Gildas Poissonnier, Senior Manager, Sustainability at Deloitte, also a member of the FSI Board of Directors.
Bud Starmak started the conversation by describing BlueSky’s approach to integrating financial material ESG factors, which can vary across industries, into investment management decisions. There are challenges in using ESG data, most notably, the inconsistency of metrics across companies. Also, data is often backward looking and may not be a true reflection of a company’s current potential. Think of a spill or privacy breach in the past that has led to a tightening of operations and improved performance. Translating ESG data into measurable value creation allows investment managers to identify potential in companies that would have otherwise been overlooked with traditional approaches. Bud believes the BlueSky approach to assessing financial materiality will become more mainstream as investors see that considering ESG factors improves long-term performance and reduces business risk. As more data points become available, the challenge is to develop the right analytical approach and tools.
Next, Bouchra M’Zali talked about making effective use of the wide sets of data we have. Data is continuously being gathered, in real-time, and in a variety of structured and unstructured forms such as text, photo, and video. AI helps companies manage and process data from internal and external sources so it can be utilized for strategy development. If set up to do so, AI technology can handle multiple data formats and can quickly interpret information from various live feeds. This level of automated analysis helps to anticipate social and environmental issues in a community or more globally, which ultimately lead to better decisions. Bouchra also mentioned the challenges that arise with the technology, such as data classification, storage, privacy management, and cyber security risk.
Technological changes have prompted changes to the regulatory landscape of fintech. Laure Fouin helped us understand some of the challenges in governing data and technology. Current laws and regulations are designed to apply to humans, not machines, and this can cause uncertainty when looking for accountability if something goes wrong. For example, if an error is uncovered, we could generally hold a financial advisor accountable for his decisions. With robo-advisors suggesting investment decisions, who would be responsible: the software engineer who wrote the code or the company making it available? Regulators are coping with rapidly changing fintech challenges by allowing pilot project exemptions, known commonly as sandboxes, but a full long-term solution is still unclear.
The Q&A touched on a wide spectrum of topics, from data integrity to intensive energy use in data mining often necessary for AI applications or big data analytics. Bud explained how statistical relevance is necessary to avoid the “garbage in, garbage out” trap when working with data. There are also several fintech start-ups introducing prototypes that remove noise from big data to help with this issue. Another question addressed personal data protection in Canada. Laure explained that Canadian law is relatively strong on protecting personal information, but the issue is whether we have the technical capacity to enforce these laws and protect from cyber-attacks. Despite complex cybersecurity systems, companies remain vulnerable to simple phishing emails, as employees the final line of defense and often fall prey to their click-bait.
Another participant brought up the interesting challenge of significantly high electricity use required for data mining technologies, confirming the often seen phenomenon that when one problem is solved, another is created. It’s important to monitor the electricity use of these technologies and of the operations that thy replace, to understand the net impact, and continuously work to mitigate any externalities caused by data mining or other technologies.
Bud offered a helpful baseball analogy to describe ESG integration as an evolution. He says, “we are in the 2nd or 3rd inning of this”, learning and improving with every play.
Many thanks to the panel experts and to our event sponsor: Deloitte.