There has been a lot of talk about robo-advice recently, but what exactly is robo-advice? Wikipedia defines robo-advisors as “a class of financial adviser that provides portfolio management online with minimal human intervention. … currently most robo-advisors employ algorithms“. Investopedia says “A robo-advisor is an online wealth management service that provides automated, algorithm-based portfolio management advice without the use of human financial planners“.
When the Financial Services Board (“FSB”) published the Retail Distribution Review (“RDR”) document in 2014 they asked for comment on the proposals contained in the RDR document. The industry provided comments on matters such as adviser classification, services provided to the customer and forms of advice, which included “low advice” and “no advice” models. Commentators highlighted the need to include or address robo-advice as part of the RDR efforts.
In its General Status Update: Retail Distribution Review document published in December 2015 the FSB acknowledged this need to consider robo-advice models and state that robo-advice tools are an important focus area as this is a potentially important tool to mitigate the risk of an advice gap following the implementation of RDR proposals. The FSB indicated that appropriate standards for such tools need to be considered and will require debate about when such tools should be regarded as providing “advice” and, if so, what safeguards will be reasonable to balance customer protection against the benefits of innovation and convenience.
During October 2016 the FSB published the Draft Determination of Fit and Proper Requirements for Financial Services Providers (“the Determination”) under section 6A of the Financial Advisory and Intermediary Services Act, 2002 (“FAIS”) for comment.
The Determination introduces the following definition for automated advice: “means the furnishing of advice through an electronic medium that uses algorithms and technology without the direct involvement of a natural person“. This definition aligns with what is generally termed robo-advice, which is the giving of advice by using algorithms and technology with limited or no human intervention.
Automated advice would obviously be limited to instances where advice as defined in FAIS is provided.
Chapter 3 of the Determination sets competence requirements where a financial services provider (“FSP”) provides automated advice in that the FSP must have at least one key individual who:
(a) meets the competence requirements applicable to a key individual of a Category I FSP; and
(b) has technological knowledge, skills and experience to:
(i) understand the technology and algorithms used to provide the automated advice;
(ii) understand the methodological approaches and assumptions embedded in the algorithms and the rules underpinning the algorithms;
(iii) identify the risks to customers arising from the automated advice; and
(iv) monitor and review the automated advice generated by algorithms to ensure quality and suitability of the advice and compliance with the Act.
The Determination furthermore sets the additional operational ability requirements for an FSP that provides automated advice in that the FSP must:
(a) have adequate and appropriate human resources that have the required competence to-
(i) understand the technology and algorithms used to provide the automated advice; and
(ii) understand the methodological approaches, including assumptions, embedded in the algorithms;
(iii) understand the preferences or biases that exist in the approaches referred to in (ii);
(iv) understand the risks and rules underpinning the algorithms; and
(v) are able to monitor and review the automated advice generated by algorithms to ensure quality of the advice and compliance with FAIS;
(b) establish, implement and maintain adequate policies and procedures –
(i) to monitor, review and test the algorithms and the advice generated by them;
(ii) to monitor, review and test the filters implemented to ensure clients for whom the automated advice is not suitable are filtered out; and
(iii) that set out the level of human review that will be undertaken on the advice generated;
(c) in relation to the monitoring and testing of the algorithms and filters referred to in (b), –
(i) have appropriate system design documentation that sets out the purpose, scope and design of the algorithms and filters;
(ii) have a documented test strategy that explains the scope of testing, including test plans, test cases, test results, defect resolution, and final test results;
(iii) have appropriate processes for managing any changes to an algorithm and filters that include having security arrangements in place to monitor and prevent unauthorised access to the algorithms;
(iv) be able to control, monitor and reconstruct any changes to algorithms or filters;
(v) review and update algorithms whenever there are factors that may affect their relevance (such as market changes and changes in the law);
(vi) have in place controls and processes to suspend the provision of advice if an error within an algorithm or filters is detected; and
(vii) be able to frequently monitor and supervise the performance of algorithms and filters through an adequate and timely review of the advice provided;
(d) have adequate and sufficient technological resources to-
(i) maintain client records and data integrity;
(ii) protect confidential and other information; and
(iii) meet current and anticipated operational needs, including in relation to system capacity.
One of the widely publicised advantages of robo-advice and also the reason why it was seen as a potential way to mitigate the advice gap, is the fact that it is perceived to be a tool where the customer can get free or much cheaper advice.
It is questionable whether automated advice could remain a free or cheaper alternative in light of the competence and operational ability requirements proposed in the Determination. These proposals will certainly provide customer protection, but compliance with these requirements will come at a cost to the FSP, especially from a technology and human intervention point of view. As it is, the search for an individual meeting the FAIS requirement to perform the role of a Key Individual can be a challenge. With the additional competence requirements applicable to such a person, the ongoing employment of such a person will certainly come at a cost to the FSP. The proposals around operational ability seem to require a lot of expert human intervention for a tool that is generally perceived to operate without human intervention. Will the proposed requirements fill the advice gap for those that cannot pay for advice and need to be given access to financial solutions?
The requirements set out in the Determination are rules based. FSP’s would need to incur the expense to comply with each of these rules. One must ask whether outcomes based regulation where the FSB would set the customer protection outcomes that an FSP must achieve, could ultimately pave the way for robo-advice solutions to find traction? While the FSP implements the required processes to achieve those outcomes, would it not be more effective to find a robo-advice solution that can balance customer protection against the benefits of innovation and convenience, especially the costs involved in providing robo-advice. It could ultimately fill the advice gap for those in desperate need of appropriate advice to build financial freedom for them and their families.