AINL#005 Augmented Intelligence in Investment Management Newsletter
Welcome to the 005 Edition of the Newsletter on Augmented Intelligence in Investment Management (AINL). Every two weeks, we deliver five unique insights tailored to empower investment decision-makers. Our insights are carefully curated by a seasoned team of market specialists. Unbiased, actionable and practical. They will help you navigate through the noise.
AINL#005 SYNTHESIS
What do these recent developments mean for investment decision-makers?
1. Redefinition of Investment Process Workflows and Efficiency Gains
AI’s integration into traditional workflows—such as coding in IT or financial analysis in investment firms—promises to improve efficiency by automating routine tasks. However, professional investors must recognize that AI’s success depends on rethinking workflows rather than merely layering AI onto existing processes. For instance, AI-driven decision-making could decentralize investment strategies, much like electricity redefined factory operations.
2. Personalization as a Competitive Edge
The ability of AI to create personalized content or solutions can revolutionize client engagement in investment services. Professional investors should explore how AI can tailor investment advice or product offerings to individual clients. However, they must balance the human-AI interaction to ensure that AI enhances, rather than replaces, the trust and adaptability required for complex decision-making.
3. Augmented Investment Decision-Making, Not Replacement
While AI can provide insights and identify patterns in financial analysis, over-reliance on AI could lead to errors, particularly in dynamic and uncertain environments. Investors must maintain a critical, human-led decision-making framework while leveraging AI as a tool. Additionally, they should anticipate a redefinition of roles, such as the Chief Investment Officer, to incorporate AI oversight and strategic alignment.
TOP 5 ARTICLES
ARTICLE ONE
Super Mario Meets AI
Experimental Effects of Automation and Skills on Team Performance and Coordination
HUMAN & ARTIFICIAL INTELLIGENCE | Columbia Business School | 12_2022 | Paper
Important Development
This article examines AI adoption’s impact on team dynamics, coordination, and performance. While AI agents often outperformed humans, their inclusion disrupted cohesion, reduced trust, and lowered human motivation and effort. Low- and medium-skilled teams faced greater challenges, while high-skilled teams adapted better.
The study underscores the need for skill composition and collaborative human-AI interaction to mitigate disruptions. Careful planning can prevent the “spillover effect,” where even unaffected team members experience reduced productivity.
Why Relevant to You?
Integrating AI agents is not a walk in the park. Team performance rather suffers than improves through such integration. Net-positive effects for team performance need acceptance, competence and incentives for individual team members and the team at large.
In short, prioritize upskilling and fostering human-AI collaboration to unlock AI’s potential while preserving team dynamics and organizational culture.
ARTICLE TWO
Worldbank Study on AI Agents and Learning Effects
From chalkboards to chatbots: Transforming learning in Nigeria, one prompt at a time
AUGMENTED INTELLIGENCE | Worldbank | 1_2025 | Paper
Important Findings
New RCT (randomized, controlled trial) of students using GPT-4 as a tutor in Nigeria. 6 weeks of after-school AI tutoring = 2 years of typical learning gains, outperforming 80% of other educational interventions. And it helped all students, especially girls who were initially behind. No working paper yet. They used Microsoft Copilot and teachers provided guidance and initial prompts. Learning gains are measured in Equivalent Years of Schooling, this is a pilot study on narrow topics and they do not have long-term learning measures.#
The fact that this is teacher-led is likely very important. We know that independent use of AI as a tutor can harm learning in some circumstances, because it gives the illusion of learning.
In short, the World Bank’s pilot program in Nigeria, demonstrated that integrating generative artificial intelligence (AI) into education can significantly enhance learning outcomes.
Why Relevant to You?
1. The Value of Augmented Intelligence: The study highlights that AI performs best when guided by skilled educators who can contextualise and enhance its use. For investment professionals, this reinforces the idea that while AI and machine learning can process large datasets and provide valuable insights, human judgment remains essential to interpret, validate, and apply these insights effectively.
2. Human Oversight in Decision-Making: Just as teachers ensure AI content aligns with students’ needs, investment professionals can ensure that AI-driven models and tools are aligned with their investment objectives and strategies. They should remain the final decision-makers, recognising AI as a support tool rather than a replacement for their expertise.
3. Customisation and Contextualisation: Teachers tailored AI to suit specific learning environments, making the tools more effective. Similarly, investment professionals can customise AI tools for their specific contexts, such as tailoring predictive models to regional markets, industry sectors, or unique client preferences
ARTICLE THREE
Financial Statement Analysis with Large Language Models
ARTIFICIAL INTELLIGENCE | Chicago Booth | 05_2024 | Article
Important Findings
A research paper published by the University of Chicago Booth School of Business on “Financial Statement Analysis with Large Language Models” investigated, whether an LLM can successfully perform financial statement analysis in a way similar to a professional human analyst. The study design provided standardized and anonymous financial statements to GPT4 and instruct the model to analyze them to determine the direction of future earnings. Result:
Even without any narrative or industry-specific information, the LLM outperforms financial analysts in its ability to predict earnings changes.
Why Relevant to You?
The LLM exhibits a relative advantage over human analysts in situations when the analysts tend to struggle. Furthermore, the prediction accuracy of the LLM is on par with a narrowly trained state-of-the-art ML model. LLM prediction does not stem from its training memory. Instead, they have found that the LLM generates useful narrative insights about a company’s future performance. Also, trading strategies based on GPT’s predictions yield a higher Sharpe ratio and alphas than strategies based on other models. LLMs may take a central role in fundamental analysis and related decision-making.
ARTICLE FOUR
Sustainability Matters: Company SDG Scores Need Not Have Size, Location and ESG Disclosure Bias
SUSTAINABLE INVESTING | Rotterdam School of Management and Robeco | 07_2024 | Paper
Important Findings
The researchers investigated whether SDG scores, which evaluate companies’ alignment with the 17 UN Sustainable Development Goals, exhibit similar biases that affect ESG ratings. The findings revealed that SDG scores are not influenced by biases related to company size, location, or ESG disclosure, unlike ESG ratings.
Why Relevant to You?
Sustainable investors are increasingly using SDGs as a framework for investment strategies to contribute to sustainable development. SDG scores measure a company’s contributions to sustainability goals and contrast ESG ratings, which primarily focus on risk rather than sustainability impact. The study emphasized that SDG scores provide a more comprehensive measure of corporate sustainability performance suitable for constructing investment portfolios supporting sustainable development.
ARTICLE FIVE
LLMs Reflect Ideology of Their Creators
ARTIFICIAL INTELLIGENCE | Cornell University | 10_2024 | Paper
Important Findings
LLMs are trained on vast amounts of data to generate natural language, enabling them to perform tasks like text summarization and question answering. These models have become popular in AI assistants and already play an influential role in how humans access information. However, the behavior of LLMs varies depending on their design, training, and use. This paper uncovers notable diversity in the ideological stance exhibited across different LLMs and languages in which they are accessed.
There is normative disagreements between Western and non-Western LLMs about prominent actors in geopolitical conflicts. Furthermore, significant normative differences related to inclusion, social inequality, and political scandals are found. The results show that the ideological stance of an LLM often reflects the worldview of its creators.
Why Relevant to You?
This paper shows that LLM responses vary ideologically depending on country, language used, company, individual model, and other variables. Not a new insight, but interesting to see it factualised, especially at this stage of the LLM development cycle. Relevant for investors that they will not get an ‘objective’, ‘rational’ response from LLMs in their market analysis, but another subjective one, in need of interpretation..
