‘Chemistry’ is widely seen as the most important criterion when selecting a financial adviser. This is so even when the personal point of contact is part of a large organisation with a trusted brand. How, then, can financial technology that substitutes humans with robots possibly compete?
It’s a big claim but we think a digital format can be more intimate and more trusted than personal chemistry can achieve.
Suppose technology was used to redefine what the adviser’s role is. If the point was to substitute dependence on an agent with informed self-selection, it would be replacing not one form of agent with another but an agent with yourself. That doesn’t have to be the reason for the technology. It should be an aim in itself, since using advisers as proxy decision makers has quite clearly left customers of the industry less engaged, with a weakened sense of personal responsibility and increased vulnerability to exploitation. With or without technology, the aim of advice should be to educate, enable and empower. In other words, to support personal responsibility by imparting the expert’s information advantage. Making customers smart.
This digital vision does not need chemistry to generate trust or convey support and generosity. That’s because the personality you are engaging with is your own, not a third party. You imbue it with your own characteristics. You are honest with it. And you give it time.
In a digital financial advice market geared to informed self-selection, the smart customer no longer asks the question of a skilled adviser what would you do if you were me but rather what should I do, having myself the skills of an informed adviser. The job of the technology is to embed the skills in the decision making. This will involve interaction with powerful engines, to perform calculations and impose logic, but always supporting, not replacing, your responsibility for the process and its conclusion. The back-end technology has to be the guarantor of the technical integrity of the process but it is front-end design that is key to the illusion that it’s your own knowledge and decision making that is getting the job done. The nature of the interaction is also critical if it is to reward both generosity of time (more and better inputs provide better outputs) and honesty (it’s only yourself you would be fooling).
This sounds fine in principle but it’s worth going through a checklist of the functions a human adviser performs that need to be substituted.
The process always starts with chemistry because the adviser, who at this stage is probably in pitch mode, perhaps even in a non-obligation meeting, wants you to share not just facts and figures but feelings about your relationship with money and how money affects your relationships. Chemistry is there to encourage sharing. Given our natural reserve, this ought to be easier if in the early stages you are sharing facts, figures and feelings only with a temporary database on your device, rather than with an individual or firm with whom as yet you have no relationship. It’s the job of the design of the interface to convey something of the experience of working in the way the process requires, including motivating you to share more or spend more time inputting information, because you see you are getting something useful out of it.
Once established, your relationship with a financial adviser, stockbroker or portfolio manager will be expected to yield some continuous benefits. It can bring order to your financial affairs. It can help you follow through on commitments you make, such as to regular saving. It can help you to keep your own emotions from interfering with rational decision making. It can help you anticipate and plan for events or life stages. It can explain decisions. All of these functions can be performed with greater reliability and consistency by machines than by humans. This already evidenced by the march of quantitative techniques in investment management, where robots already regularly outperform humans in quality and cost.
A further effect of replacing dependency with self-selection is that an agent is no longer needed to interpret or interpolate. That’s a strength. Interpretation is open to error by skilled yet subjective individuals. Informed self-selection relies instead on the user directly revealing their preferences and trade-offs, in response to information feedback. Because the trade-off problems involved in planning life goals typically involve multi-faceted aspects of risk, and because definitions of individual welfare or ‘best interests’ are typically complex, directly-revealed preferences are likely to be more accurate than those arrived at by interpretation by an adviser or by proxies such as psychometric personality tests. Any decision process that can handle this will also contain its own audit trail for how choices were made, reducing the need for compliance as an externality and for ombudsmen to sort out disagreements.
Technology’s advantage also lies in demonstrating that the machine has no agenda other than yours. A computer can of course be programmed to be as exploitative as any human agent but it can also be conspicuously disinterested and complete. It demonstrates that by providing answers to questions that are consistent regardless of the implications for the firm, even to the point of revealing you do not need the service being offered. The machine doesn’t care whether you take more or less risk, or choose to avoid risk altogether, because it is not answering as an agent (what would you do) but as an extension of your own thought process, with your personality and your agenda.
What about regulation? Informed self-selection theoretically does away with the need for a product recommendation. Regulation of advice to buy a retail investment product requires a recommendation, based on suitability given the precise personal context. This advice model assumes that the purpose of the decision-making process is to end up with a product. The process could just as easily be the route in to an investment service, where all implementation decisions are left to the service provider but they implement a plan made by the client using that front-end technology. The value moves from the product itself to the knowledge of how to use it. The regulatory test for the service provider is its initial and continuing suitability, but not a recommendation. It’s worth noting that most robo-advice models rely on continuous discretionary management contracts rather than discreet points of advice.
Our attempt to design a technology-driven process of informed self-selection has been some 15 years in the making. My initial vision was an entirely digital process but that was in the dot com boom. Fowler Drew chose a more gradual development, more cyborg than robo. We still rely on advisers to guide the client through the goal planning using outputs from the same models that we use to manage the goal-based portfolios. The portfolio management, on the other hand, has always been entirely quantitative and, necessarily, on a discretionary basis rather than advised.
You can see for yourself whether we’ve made a good job of it so far, by test driving our Drawdown Planner for a spending goal. Spending is the ultimate purpose of most private-client investments and drawdown is the aspect of it crying out for a better solution. The free-to-use Drawdown Planner is an interface that connects to the mathematical engine used to manage a Fowler Drew client’s spending goal portfolio, delivering the planned outcomes, within defined tolerances, at specific dates. Both the front end and some of the back-end calculations have been slimmed down for ease of use and speed but I hope it gives a good impression of the richness of the scope and the potential for user engagement. Could it be better designed? Of course. Could it be entirely digital, with no human guidance involved? Technically, yes.
So do we think robo advice is the future? Yes. It has the capacity to be more accessible, cheaper and, more importantly, better. All three are socially desirable, if not vital, in a market where advice is the preserve of the affluent yet also frequently mediocre.