18 Jun Harnessing AI and emerging technologies: applications in wealth management
Croesus’ Maxime Dumas on leveraging advanced financial tools to enhance advisor capabilities and gain competitive advantage
Wealth management firms in Canada are engaged in a technology arms race to attract and retain advisors. They are investing in advanced financial tools to enable advisors to conduct more sophisticated financial planning, enhance client service, and gain a competitive advantage. At the forefront of this innovation is the applied research and innovation center Croesus Lab.
In today’s market, portfolio managers rely on assistive technologies to meet and exceed expectations. Automation tools provide competitive advantages that significantly boost the bottom line. Notably, several wealth management firms are leading the charge in innovation, often adopting new wealth management solutions more swiftly than their bank-owned counterparts.
In conversation with WP, Maxime Dumas, principal researcher, technological innovation at Croesus Lab, details how emerging technologies offer advisors the flexibility to customize client portfolios, while improving operational efficiency.
“The Croesus Lab is dedicated to fostering bold and transformative ideas,” explains Dumas. “Our mission is captured by our motto: explore, de-risk, and inspire.”
Launched in 2018, Croesus Lab is a hub of creativity and forward-thinking. Its mission allows the lab to focus on developing innovative ideas without the constraints of production pressures, enabling them to work with emerging technologies that could shape the industry in the coming years.
De-risking through prototyping
One of the core aspects of the lab’s work is de-risking new technologies. This involves building prototypes and validating ideas with clients, leading to the development of new solutions and features.
“We bridge the gap between theoretical research and real-world impact. We’re taking research results or interesting promising emerging technologies and adapting them to our internal projects, addressing our own reality,” said Dumas.
Pioneering Innovation Culture
The lab operates under three main pillars: research and experimentation, innovation culture, and the propulsion of innovative ideas.
“We have a strong innovation culture, where we infuse innovation into the company through workshops, ideation sessions, and even a creativity game. Lastly, we have the propulsion pillar, which accelerates the development of innovative ideas through financing and support,” said Dumas.
The lab’s projects span both applied research and experimental development. “We have two permanent teams in the lab,” said Dumas. “One dedicated to artificial intelligence (AI), where we train models and explore new algorithms, and another focused on operations research, addressing complex optimization problems like portfolio rebalancing.”
Additionally, the lab explores other technologies such as cybersecurity, augmented reality, quantum computing, and blockchain.
AI revolution
The wealth management sector is poised for a transformative shift with the advent of generative artificial intelligence. Banks and wealth management firms stand to gain significant competitive advantages by leveraging AI in areas such as risk assessment, fraud detection, personalized customer service, and algorithmic trading.
Despite the potential benefits, the responsible implementation of AI in wealth management requires addressing key challenges, particularly in the areas of data security and algorithmic bias. Croesus Lab is at the forefront of addressing these challenges, actively researching and developing innovative solutions to harness the power of AI responsibly and securely, ensuring that its benefits are realized while minimizing potential risks.
“Our mission is to simplify and humanize wealth management,” said Dumas. “We’re not trying to replace advisors but to empower them. AI is a tool to enhance the human element, fostering stronger client relationships and trust. By automating administrative tasks, we free advisors to focus on what truly matters: providing valuable insights and building lasting connections.”
For all the potential of artificial intelligence, one of the foremost concerns is data confidentiality and privacy. Ensuring that sensitive financial data is protected from leaks and breaches is critical.
Another significant challenge is the accuracy and biases of models used in financial technologies. While there is substantial investment in developing measures to evaluate and mitigate these issues, it remains a critical area of focus. Ensuring model accuracy and addressing biases is essential for the reliable use of these technologies in production environments.
Ultimately, the wealth management industry is built on trust. Failing to address these risks can severely damage that trust, undermining the very foundation of the industry.
Quantum computing in wealth management
The next technological revolution on Croesus Lab’s radar is quantum computing. Its primary benefit can be summarized as efficiency and speed. For specific types of problems, quantum computers can significantly reduce the time required to solve complex issues compared to classical computers. Quantum computers are designed to address specific problems where better algorithms can leverage their unique architecture.
Quantum computing has the potential to scale to much larger problems than classical computers can handle, allowing for the modeling of more parameters and constraints. This capability is particularly advantageous in finance, where we can create better representations of the market. One key area of interest is optimization. This includes portfolio optimization, rebalancing, and risk management.
Looking forward
Another important research area for Croesus Lab is data protection and security.
“One area we’re heavily focused on is synthetic data generation”, Dumas said. “Essentially, we’re creating realistic, coherent data sets for testing, development, and demos, primarily for system testing. The goal is to produce data that mirrors the characteristics and patterns found in production data but without containing any personal or sensitive information. This approach alleviates concerns about data privacy and security since we’re not using real data.
“However, generating synthetic data is a research challenge. While it’s relatively straightforward to generate data for a single table, handling complex databases with hundreds of tables and ensuring the data remains correlated, especially over time, is much more difficult. This is particularly true for financial time series data. Existing literature offers limited guidance in this area, which is why we’re investing in research.”
Looking ahead, Dumas sees potential in digital assistance and next-best-action systems, which can pinpoint valuable directions and key opportunities for advisors.
By exploring emerging technologies like AI and quantum computing, while prioritizing data security and the human element, Croesus Lab is paving the way for a more efficient, personalized, and trustworthy future in wealth management.