Written by Paul Nixon, Head of Behavioural Finance, Momentum Investments
Paul Nixon, Head of Behavioural Finance, Momentum Investments has been diving in the world of AI, to explore whether – and if so how – it presents a significant threat to the role of professional advice
ChatGPT can probably teach us all a thing or two about our jobs and go as far as to suggest some valuable ideas you’ve not previously thought of. In my brief experience it’s already suggested some machine learning variables in predictive behavioural models that have been extremely useful.
What this technology has already done very effectively is reduce search costs. As a behavioural scientist I can ask ChatGPT to suggest a framework for incorporating anthropology into changing behaviour and in seconds it will give me five summarised points for me to consider. If I Googled that, I would have had to have sifted through academic papers, online articles and blogs, with the majority being irrelevant.
The question is whether it can replace you as a financial adviser? The answer to this likely rests in what your value proposition as a financial adviser actually is. The value proposition includes two elements which are relationship focussed and technical skillset. In the past, the debate on how much of a threat robo advice was, was compared on these elements. This is an important distinction; AI is not the same as a robo-advice. Robo advice is far simpler – an advanced calculator of sorts (depending on the inputs and outputs). An AI can learn financial behaviour from data it collects and make an investor objectively better compared to a version of themselves without AI. It could potentially offer in-the-moment financial planning guidance and accurately nudge that person in respect of what they want to buy, how much they need it and what it will cost them in retirement income. This is simply because it’s plugged into (or could be) their bank account, credit card and investments and learns from their financial behaviour, both good and bad.
Both AI and robo advice are significant threats to the adviser. A robo-adviser or AI is already capable of figuring out, with limited inputs, which investment product is suitable for an investor’s needs and to make a cost-effective (low fees) recommendation that compounded over time will save them a lot of money – in fact this is equally as important as the investment fund they select. From a behavioural science perspective this advice has zero “noise” or random human variation in advice that is rooted in the many contexts in which advisers make judgements. Any human judgement in fact is subject to inconsistencies that can be as simple as being in a bad mood when making that judgement.
Now let’s perhaps consider the opposite end of the spectrum. A family office offering wealth management for high net worths. These are likely multi-banked clients across regions with both business and personal assets wrapped in trusts and complex structures. The value the wealth manager offers here is a holistic view of the portfolio across these structures where different jurisdictions don’t necessary “talk” to each other or there is no data feed between the institutions. In my experience with such clients, the wealth manager sits on the family “council” and plays a ‘trusted adviser’ role in matters that extend beyond finances. These wealth managers are technocrats but also trusted members of these councils offering advice even on new ventures. Robo advice or any AI is not a threat here at all.
The bottom line is that there are two ways to plan for the encroaching of AI in financial planning and that will be:
1. Deepen your relational skillset. This is basically betting on your humanity as the differentiator. Psychology now forms 7% of the CFP® professional syllabus and the empathy economy will value these skills more and more.
2. Deepen your technical skillset. Developing and specialising in a substrate of financial planning for example places you in a more shielded position when it comes to AI.
Doing just one of the above two keeps you at moderate risk of an AI taking your job. Particularly on the technical skillset angle where an AI is easily capable for example of analysing income statements and balance sheets and making recommendations but perhaps less adept at assessing different tax jurisdictions where information is not shared freely. Similarly, an AI is capable already of diagnosing personality and identifying root causes for behaviour, but humans prefer these diagnoses from other humans that can empathise.
In a nutshell humans and AI are currently good at different things. For now, if financial advice is your career choice you will need to think about how your choice of career would add value in a world where AI is ubiquitous.
About Paul Nixon
Paul Nixon is Head of Behavioural Finance at Momentum Investments. He focuses on behavioural elements for clients and businesses value and develops and applies psychometrics and machine learning data analytics to target behaviours.