Stockholm (NordSIP) – Fuelled by breakthroughs in machine learning and computing capabilities, Artificial Intelligence’s meteoric rise of late has sparked a wave of innovation across virtually every industry imaginable, investment management included. Yet, as AI surges ahead, so too do the questions around the technology’s ethical implications, its environmental footprint, and its impact on society.
For this year’s International Women’s Day feature, NordSIP couldn’t resist delving into the captivating intersection of AI and investment management, as seen through the eyes of some women working right at the frontline of both fields. Whether overseeing investments, risks, or sustainability issues, these ladies offer diverse and valuable insights into the ongoing technological revolution’s multiple potentials and pitfalls.
Putting AI into perspective
Having experienced several technological advances over the past twenty-five years, Hildur Eiríksdóttir, Executive Director at Íslandsbanki Asset Management, sounds unfazed by AI’s rapid rise. “When I started working with investments back in 2000, bonds were traded by handing over a physical paper with a signature on the back,” she reminisces. “Now it takes seconds to execute the same trade electronically. Information that took days to gather in the past is instantly available at the tip of our fingers. High-frequency platforms can execute trades without any human involvement.” That said, Eiríksdóttir admits that adding AI to the technological mix is accelerating a development which was already extremely fast.
As someone who has been professionally curious about AI since its early days, Elena Westerdahl, Senior Quantitative Portfolio Manager at Swedish pension fund AP3, has long been aware of the constraints in effectively deploying the models in investments due to the scarcity of financial data. “Machine learning requires a large ecosystem for training, whereas financial data can be almost overseen with a naked eye or using simple factor models,” she explains. “Applied to investment decisions, many of the models were, and are still stuck in the middle, with an overfitting and underfitting dilemma.”
Large language models (LLMs) are, however, a game changer, according to Westerdahl. Having played around with platforms utilizing LLMs for AI-driven algorithms, she sounds optimistic about such tools’ potential to enhance existing workflows. “The models are getting better fast, but they still have a lot of limitations. Once artificial general intelligence (AGI) arrives, things might get much more interesting, with problem-solving capacity increasingly resembling the human thinking and reasoning able to generate new knowledge,” she adds.
Although still far from perfect, it is obvious that many investment professionals are eager to embrace the power of AI in their day-to-day work. “With the introduction of ChatGPT Plus in early 2023, the question wasn’t any more whether to incorporate AI into our investment process, but rather how to effectively do it while aware of any potential pitfalls,” says Huizi Zeng, Portfolio Manager at Espiria, part of East Capital Group.
An endless catalogue of potential uses
“In the realm of active investments, speed is paramount,” explains Zeng. “This encompasses everything from routine tasks to the generation and validation of creative research ideas. AI tools are invaluable in expediting various tasks, from synthesising and summarising extensive financial data, reports, and news articles to facilitating sector and company research. Among other areas, AI tools are also streamlining editing processes and data analysis, and even enabling us to present visual information in a more engaging manner.”
Julia Sommer Legaard, Risk and Data Manager at Industriens Pension, was an experienced user of AI even prior to joining the Danish insurance company. “At Industriens Pension, we employ AI in several investment processes such as asset allocation, market predictions, outlier detection, rewriting one language to another, code support, summaries of articles, and search optimisation.”
Others, such as Anna-Stina Wiklund, Head of Sustainable Investing and Portfolio Manager at the Evangelical Lutheran Church of Finland Pension Fund, use ChatGPT mostly for inspiration, or to make texts more formal or entertaining. “We are, however, planning to employ the technology to extract data from, for instance, the large amount of text in due diligence questionnaires, or to check portfolio managers’ background with the help of AI.”
Just like her, Eiríksdóttir also sees the potential for using AI in the screening of external funds. “AI simplifies information gathering and analysis, further enabling data-driven decision making,” she says. “In the future, we might even see AI-generated suggestions of changes to portfolio composition based on a set of requirements, especially risk profiles.”
Fondita Fund Management Company is another organisation looking to incorporate AI. “We are learning, for instance, how Microsoft Copilot can make presentations better and graphs even more expressive,” says Janna Haahtela, Portfolio Manager at Fondita. “We have also been looking into data providers who use AI for ESG analysis and company profiling.”
Opportunity or threat?
As an asset owner working with extensive investment data, Danish PKA counts on AI to enhance the organisation’s analytical capabilities. “Given the relatively small size of our team, AI presents an opportunity for us to explore new investment avenues that traditionally required substantial resources,” says Dewi Dylander, CSO, Deputy Executive Director for Sustainability, Policy & International Relations at PKA. “It has the potential to streamline tasks and processes, especially those that are time-consuming or repetitive, improve decision-making, and uncover insights that may not be readily apparent through traditional methods. Successful implementation will, however, require a multidisciplinary approach, involving individuals with diverse skill sets. It is essential to approach it thoughtfully and with caution to mitigate potential risks such as biases and governance issues.”
“In itself, AI at a current stage is an automated solution to solving tasks which can be both an opportunity and a threat,” reflects Westerdahl. “As AI-generated content gets ever more persuasive, we need to find ways to understand when the algorithms mislead us.”
Ellen Andersen, Portfolio Manager at Storebrand Asset Management, is concerned that we are becoming more vulnerable to cyber-attacks and sees the need for companies to prioritise cybersecurity solutions. “However, from an investment perspective, the use of AI can be beneficial for companies and the products and services they deliver,” she says, mentioning the examples of Duolingo, an educational technology company that collaborates with chatGPT to incorporate AI into its learning methods, and Crowdstrike, a cybersecurity company using AI to enhance threat detection and response capabilities.
“Ultimately, whether AI is perceived as a threat, an opportunity, or both, depends on how it is developed, regulated, and integrated into society,” reflects Kiran Aziz, Head of Responsible Investments at KLP Asset Management. “There are concerns about job displacement, privacy invasion, ethical implications, and potential misuse. On the other hand, it is an opportunity to enhance efficiency, productivity, and quality of life and revolutionise various industries, such as healthcare, transportation, finance, and education.”
Mind the gender gap
“AI has the potential to help close the gender gap in both investment management and technology,” asserts Aziz. “It can, for instance, mitigate bias in hiring and promotion processes by using data-driven algorithms to assess candidates based on skills and qualifications rather than gender.” She is, however, concerned that AI algorithms could also inadvertently perpetuate gender bias.
“The problem is that while AI is a new technology, it’s not starting from a blank page,” explains Tulia Machado Helland, Senior Sustainability Analyst at Storebrand Asset Management. “The databases used to train AI reflect existing inequalities and biases based on gender, race, ethnicity, or religion, historically prevalent in society. This makes AI inherently prejudiced and may lead to biased decisions.” She urges everyone to use inclusive algorithms, such as those promoted by < A+ > Alliance. “If we do this, we can prevent machine learning from creating a dystopia where bias and prejudice are encoded into our lives.”
Taking a more optimistic stance, Zeng is hopeful that generative AI can help individuals, irrespective of their prior technological proficiency, enter the tech or investment sectors with greater ease. “As Nvidia CEO Jensen Huang aptly stated, ‘everyone can be a programmer’ now that AI is capable of comprehending diverse inputs, including speech,” she says. “Lowering the barriers to entry in technical and data-intensive analyses may democratise opportunities, while human creativity and unique perspectives remain highly valuable in discerning insights not readily apparent from large datasets or trend analyses.”
Wiklund has also high hopes for the technology’s democratising potential. “For instance, with virtual-reality glasses, you can experience how coral reefs have been damaged through the eyes of a scuba diver,” she says. “Similarly, you can step into the world of someone with a different ethnicity or gender and experience first-hand their encounters with other people. This way, I think AI can help close the gender gap.”
Haahtela, on the other hand, struggles to see how AI alone could rectify the gender imbalance in the two industries. “However, if I simplify it and say that the value of empathy and human skills will be even more valuable than before then that should give women a clear edge,” she adds.
Be that as it may, AI can only help so much, as Sommer Legaard points out. “You as a person need to be interested in finance and technology and the fact that you can write code faster does not in itself change that,” she says. “You cannot write a complete script with AI. You need to know programming, possess critical thinking, and excel at problem-solving. It is the last percentages that AI cannot solve that really matter.”
“Technological literacy is becoming ever more important in today’s investment landscape,” agrees Eiríksdóttir. “Access to information is empowering, but achieving gender equality will require all to embrace the digital transformation.”
Ultimately, addressing the gender gap in both investments and technology starts by having an open discussion about it, according to Eiríksdóttir. “Promoting role models working in the field is a welcome initiative. Making women more visible is exactly what is needed, it is being part of the change,” she concludes.
