Robo-advisors and the proliferation of artificial intelligence have been touted as game-changers in wealth management, offering automated investment management with the promise of lower costs and efficient portfolio optimization. Despite their advantages, robo-advisors don’t seem to be winning the battle against human advisors just yet. The reason? Personalization remains the linchpin in the advisor-client relationship, and something that has yet to yield to technological pursuits.
The era of robo-advisors was heralded as a democratization of investment advice, making it accessible to a broader audience with its algorithm-driven, low-touch approach. These digital advisors use modern portfolio theory to construct and manage diversified investment portfolios, often at a fraction of the cost of traditional advisory services. They have become increasingly sophisticated, incorporating tax-loss harvesting, automatic rebalancing, and even socially responsible and thematic investing options to their offerings.
Technology Can’t Do It All
Despite their capabilities, robo-advisors have not managed to capture the market. The potential applications of AI in financial planning are vast and varied. A recent study by researchers from Case Western Reserve University and AIgency, a firm specializing in accounting automation solutions, tested the capabilities of large language models such as Google Gemini, ChatGPT-4, Claude, Mixtral, and Llama-2b against multiple-choice questions from CPA exam preparation materials. The study aimed to evaluate whether these AI models could perform at a level comparable to human professionals.
The results of the study were mixed, with ChatGPT-4 emerging as “the only real option” for high-quality business analysis and reporting automation without human prompting. This suggests that ChatGPT-4 could be effectively used for tasks such as automated financial statement preparation or forecasting. On the other hand, the researchers found that Claude might be better suited for auditing-related tasks, indicating its potential in areas like fraud detection and internal control validation.
While AI applications have proven useful, they have not yet proven to be a universal application in the replacement of human advisors. AI programs that complement the day-to-day tasks of advisors can free up valuable time spent on low-value tasks, and some of the largest wealth management firms are taking notice.
How Major Firms Are Using AI
Large institutions such as Goldman Sachs, Morgan Stanley, and Citigroup have been vocal about their AI deployments and the impact on their bottom lines during recent quarterly earnings calls. This shift towards AI has sparked both excitement and concern within the industry, particularly among human advisors who wonder about their future role in an increasingly digital financial world.
Major Wall Street firms are investing heavily in AI to enhance their advisory and investment services, while still maintaining human involvement. Morgan Stanley, which partners with ChatGPT maker OpenAI, has plans to roll out a real-time chatbot that works alongside advisors in answering client questions. In the past, advisors would have to come back to clients with questions or research they could not answer off-hand. This new technology allows advisors to receive assistance while they have their client’s attention on the phone. The bank said 98.5% of its teams now have at least one member who uses the product once a week or more, and plans to expand AI projects firmwide, reports Business Insider. This custom chatbot can reference only the bank’s internal content and answers questions only related to wealth management.
In May, JP Morgan CEO Jamie Dimon stated that the bank was testing a generative AI “copilot” within its private bank (the wealth management arm of the bank that deals with high-net-worth individuals), drawing similarities to Morgan Stanley’s planned rollout. The tool would work alongside advisors, helping to save time on manual tasks and tracking down information during client calls.
The Future of AI in Wealth Management
As AI models become more sophisticated, there is growing concern among human advisors about the possibility of being replaced by technology services. However, a study co-authored by Andrew Lo, finance professor at MIT Sloan and director of the MIT Laboratory for Financial Engineering, offers a more nuanced perspective. While acknowledging the potential of AI to simulate certain human behaviors, the study suggests that there is still a significant gap between current AI capabilities and the complexities of human financial advice.
When exploring whether large language models (LLMs, the cornerstone of generative AI models) could operate in the best interest of investors, Lo and his team turned to retrieval-augmented generation (RAG), which retrieves domain-specific data from external knowledge bases. Lo’s team created a RAG consisting of financial lawsuits filed as a way to train the model on how to default toward what was considered ethical behavior. His team found that when applying these constraints, ChatGPT 4.0 ended up being relatively fair, but other LLMs came up with unfair biases, gender bias being one among them.
Humans too are vulnerable to biases, but the progress to be made on the technology that advisors compete with is a positive for the wealth management industry. The relationship will likely remain complicated – but unavoidable – for the time being. As AI continues to become integrated into everyday life, RIAs will need to accept AI processes into their businesses to remain competitive. AI technology has the capability to help advisory businesses grow, but firms still have a way to go. According to a 2022 Accenture “AI in Wealth Management” survey, 97% of advisors surveyed believed that AI can help grow their book of business organically by more than 20%, yet 50% stated their wealth management firms are challenged to act on their AI vision with 55% saying their firms’ AI tools and insights are too complicated to use.
Overall, AI will most likely be useful moving forward in helping to deliver more personalized financial advice and scale advisory businesses, but advisors need not worry that these tools will replace valuable human perspective. An ideal harmony for this technology will be an implementation where the AI multiplies the financial insights of the advisor allowing for more valuable human interaction between professionals and their clients.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4788096
https://mitsloan.mit.edu/ideas-made-to-matter/can-generative-ai-provide-trusted-financial-advice
