by Panos Archondakis, Senior Director, Account Management, EPAM
August 1, 2017
There is a growing focus in the wealth management (WM) industry on applying artificial intelligence (AI) to the advisory process, but making sense of the hype requires agreeing on a few definitions.
Firstly, AI itself means many different things to different people. With machine learning and big data there is a sense that by simply plugging in data, going through some basic training processes and standing back to appreciate the results, wealth management can instantly transform into a nimble, advanced, data-driven enterprise. The truth is that AI, big data and machine learning are fundamentally sophisticated, highly scalable and enormously flexible tools that identify patterns in complex data sets.
Deciphering what these patterns mean, determining the business values and insights that can be derived from them, and how they should influence our response all require a deep understanding of the business foundation and its relation to clients, products and market conditions, and processes. In short, there needs to be a solid business concept in place to create value from the data.
Understanding the Broad Scope of AI
AI methods can be applied across several functional areas within the banking industry and provide a variety of opportunities and benefits. The methods themselves, with their underlying basis on pattern recognition, can be further categorized as follows:
With these different methods in mind, we have compiled a list of potential, high-level use cases for AI in wealth management:
AI Presents a Wealth of Opportunity
Looking at the wider spread of activities in wealth management, we see a number of opportunities where these approaches can be applied, such as:
Client Service & Experience
Client on-boarding is an extremely cumbersome and manual process that could be improved in many ways. For example, implementing RPA can streamline Know Your Customer (KYC) decision-making through more interactive and intuitive information gathering, as well as making areas like client risk and suitability more predictable.
Chatbots could aid improvements in client interaction and service across all channels. Tying these interactions intelligently to the entire customer journey can significantly boost response times as well as the client’s perception of the business.
Biometric security features, such as face or voice recognition, also add value to the customer experience. This can ultimately deepen their level of trust when executing authenticated transactions, from logging in to a system, to placing an order, to initiating a payment.
Ensuring that all relevant and actionable content is personalized and tailored to the context in which it is presented, the client experience and engagement will be positively enhanced.
Back-Office & Operations
Repetitive by nature, operational procedures can benefit greatly from AI, whether that means replacing or assisting the human decision-maker. For example, RPA could be applied in the processing of trades or transaction breaks, on-boarding or other risk-related activities.
Prospecting & Matching
By using modern customer experience and marketing tools to analyze, segment and understand client preferences, an organization can get a far more accurate and granular breakdown of the client base, both actual and target.
Publicly available sources – including social media, news and targeted publications – can be leveraged to generate precise prospecting and acquisition plans with an improved conversion probability.
Additionally, client profiles can be compared with previous results, matching prospects to suitable advisors in a process similar to an online match-making service.
Advisory Process & Recommendations
At first glance, there would seem to be a clear opportunity in using the segmentation and recommendation approach to offer specific product opportunities to target clients. However, very strict regulatory requirements and fiduciary duty prevent full flexibility in this model.
The full-service advisory process is usually proprietary, which is a unique selling point and is often very closely guarded within such organizations. Individual product recommendations based on customer affinity, peer analysis and advisor effectiveness create a transactional environment that contradicts the desire to increase fee-based revenues.
A key part of generating any recommendation is to have feedback on what is successful, but the time horizon of determining a “good” product recommendation could be quite far into the future. The approach may be better suited with a fully automated, hands-off approach for mass market, with a variety of eligible retail products covering loans, insurance and so on.
Enhancement of existing algorithmic approaches and models for portfolio optimization and investment advice are in their infancy, with few real-world examples in active use. Given the regulatory hurdles involved and the need to demonstrate control of the advisory process, banks will be forced to balance cost, risk and potential return: What benefit to performance is feasible above current returns? The approach is possibly better suited to discretionary or institutional strategy management.
CRM & Sales
In this final category, where we consider a kind of opportunity management, we see huge opportunity to apply traditional eCommerce techniques of segmentation and recommendation within the client relationship.
Rather than using the analytic process to identify specific product recommendations, the outcomes can be used to inform advisors of client affinities, as well as recommended content and conversation topics .
By combining content, new product and non-traditional opportunities (e.g. IPO, M&A, co-investment), advisors can more easily find and prospect pre-qualified leads with a higher probability for closure.
In addition, non-financial offers, such as corporate events and other soft opportunities, can be highlighted to create general relationship management opportunities.
What the Future Holds
Introducing AI into wealth management will likely continue to be challenging, and financial organizations will often struggle to identify how it bests suits their particular circumstances. Ultimately, the technology will be used in areas where it can minimize cost, maximize legacy data value and integrate the investment as part of a wider digital strategy.
If defined and implemented correctly, AI can offer many exciting opportunities for wealth managers, but time will tell which ones are most viable when it comes to making the biggest impact on customer experience, technical debt and, perhaps most importantly, portfolio performance.