According to the World Economic Forum, “Generative AI refers to a category of artificial intelligence algorithms that generate new outputs based on the data they have been trained on. Unlike traditional AI systems that are designed to recognise patterns and make predictions, generative AI creates new content in the form of images, text, audio, and more.”
Stanford University’s Artificial Intelligence Index Report 2023 places the AI job postings in the finance and insurance industry as the third-highest in terms of percentage of total job postings when compared with those in 2021 and 2022. In the broader banking and finance segment, people are looking at AI to help them in risk assessment, fraud detection, customer experience and answering questions related to financial advice.
In this article, let us explore the potential impact for the investment management industry, which focuses on long-term management of financial assets.
A question that the asset management industry has been trying to answer is how to get better returns. Traditionally, in actively managed funds, fund managers have to understand market signals and build a portfolio that aligns with client goals and gives the best-in-class returns. Over the last few decades, we have seen the rise of passive funds and exchange traded funds.
How does this situation change with the mainstreaming of AI, especially with the advent of generative AI?
The CFA Institute, the global association of investment professionals, has just launched the Handbook of Artificial Intelligence and Big Data Applications in Investments. Here are a few of their learnings:A survey by the CFA Institute asked what kind of people would companies in the sector hire over the next two years. Most respondents wanted to hire finance talent with some AI and big data skills.
The application of AI and big data in investment management is across business functions — core businesses, risk management, sales and marketing, cybersecurity, customer service, back office and compliance.
Natural language processing is a natural application in the asset management industry given that many actions need text data. It has been used to summarise information, extract topics, search information, answer questions, analyse sentiments and recognise named entities.
Some specific and emerging use cases today include:
Analyse new data: For example, can sentiments be analysed from social media posts to provide more real-time insights before they get captured in financial results
Analyse patterns from data: For example, searching for ESG (environmental, social, governance) themes in a corporate social responsibility report or assessing risks in corporate filing documents which are large, bulky documents through a bag-of-words approach
Increasingly, such information can be used to augment data for client insights and with the new advances of multimodal models, there could be more ways to look at text and tone.
Indeed, Bloomberg recently announced BloombergGPT, a 50-billion-parameter generative AI model trained on a wide range of financial data to support tasks within the financial industry. Bloomberg said this could help in improving existing natural language processing (NPL) tasks, including sentiment analysis, named entity recognition, news classification and answering questions, among others.
Morgan Stanley highlighted another market underlying need recently. Given that companies in financial services often want their data to stay proprietary so that they can create their own AI models — as the AI systems learn more with data — trust in data is becoming an even more important requirement.
So how could all of this help the industry?
Deloitte in its recent report — AI: The Next Frontier for Investment Management Firms — talked about 4 pillars of transformation:
Generate alpha: Using alternative data sets, firms can improve their performance and generate more alpha and returns
Operational efficiency: Using this approach, the traditional cost centers can start becoming more of AI-enabled “as a service” offerings
Improve product and content distribution: AI can help customise content for better customer experience
Manage risk: AI could be a key differentiator in how risk is managed in the firms by helping add to data and preparing for unforeseen events
For people entering the investment management sector today and over the next few years, this means apart from knowledge of finance, they might have to learn how to analyse data and use tools such as machine learning and also elements of programming. The most important quality will be ability to think, reason and to ask the right questions.