Financial services and technology might be an odd couple but through their unison we saw the birth of Fintech which is defined in the Oxford Dictionary as: “Computer programs and other technology used to support or enable banking and financial services.” In today’s world one could slightly tweak this definition to, technological innovation in financial services. Fintech through Big Data, Blockchain, Internet of Things (IoT), use of Cryptocurrencies, Artificial Intelligence (AI) and a more effective exploitation of digital channels, social networking and mobile devices is disintermediating the financial services industry through innovation.

Within financial services, AI technology could allow companies to establish new business models, reduce risk and expenses, and increase productivity. It can also be used by regulators in Suptech like South Korea’s supervisory authority (FSS) is doing, where AI with big data are used to monitor fintech operators. AI however could also be pivotal to what we call Robo Advisory. Robo Advisory is a fast-growing phenomenon in fintech and has the potential to severely disrupt the global financial market. Even though existing data on Robo Advisory market share varies strongly, all indicate a rapid growth, particularly during the recent years. As a research paper from Deutsche Bank shows, growth was particularly impressive in the US, where assets under management of Robo Advisory start-ups increased by approximately 800% from $ 2.3bn in 2013 to $ 20bn in Q1 2017. As of 2018, the US is by far still the leading market of Robo Advisory, comprising about $ 266bn in assets under management out of a global $ 371bn. Other sources estimate even higher numbers. The potential here is huge as according to Business Insider the global assets under management by Robo Advisors in the year 2020 will start at USD 0.82tn over USD 1tn, or USD 2.2tn and run up to USD 8.1tn.

Robo Advisory is a term used to describe the automation of financial services traditionally provided by human financial advisors, which now can be done via code, and through an automated process. Currently it is largely limited to passive investing methods being personalized, however only to the extent of the profile that an investor is classified into.

With the shift however into a more data driven process through the substantial advancements made in the field of machine learning, and more specifically deep learning, we are moving into the realm of completely personalized investment products via autonomous Robo Agents. Aside from the latter, with legislation like the PSD2 these Robo Agents will be able to consider more information about the client, and design investment advise or portfolios bearing in mind the client’s preferences and constraints at a quasi-individual level. This is what in the industry is known as the Robo Advisor 2.0 or X.0.

The potential for this new Robo Advisor is immense but it also puts many established business practices and regulatory paradigms into question. Consequently, regulators are faced with the question of whether the present regulatory framework is still appropriate. Several jurisdictions including Australia, Canada, China, France, Germany, Italy, Netherlands, Spain, United Kingdom, Hong Kong, Korea and the United States plan to or have already issued guidelines on their use. EBA (European Banking Authority), EIOPA (European Insurance and Occupational Pensions Authority) and ESMA (European Securities and Markets Authority) are also closely monitoring further market developments, with ESMA also issuing guidelines in 2012 and 2018. Albeit at this stage, most guidelines, suitability requirements and positions issued merely touch on technology neutrality and provide static applicability of the existing rules and regime to Robo Advisors, circling round adherence to existing regulation and suitability checks, matching laws like MiFID2 criteria, ESMA Guidelines, access to accounts in compliance with PSD2 or GDPR or requirements on product governance, and documentation, and as yet do not capture fully the new dimension of Robo Advisors 2.0 or x.0.

To this end there might be issues which will not be covered adequately by the existing regulatory outlook. Let’s list some cases. There will be cases where the automated or worst still, the autonomous function where AI is involved would not have a transparent process on how it is structuring its output functions and decisions (the black box syndrome).This limitation is problematic when the pre-determined assumptions or categories are not entirely appropriate to the customers’ personal situation or else if a regulator would want to check that there is no bias, inference or discrimination and if the algorithm is actually working correctly and also using the correct data sets, patterns, analyses etc. The latter instance will become more prevalent as we disintermediate in toto the human advisor with an autonomous algorithm. Another risk that would merit regulatory attention, not only in regard of Robo Advice, but rather concerning the fintech, is the problem of cyber risk and thus cyber security.

As the trajectory of certain aspects of financial services is being disintermediated from the human component and shifting more to automation as well as to autonomous algorithms, the level of trust and regulatory target is also being shifted. Whereas advisors, regulated firms, employees and managers were the capture of regulation, and were heavily regulated with ex ante and ex post obligations, with their disintermediation and with more algorithmic presence, someone will need to make sure that the code is safe and apt for use and that a mechanism is in place for the eventuality when this is not the case.

I believe a different approach should be considered and countries like Malta could be well positioned to add more certainty and trust in Robo Advisors and this fintech innovation by disintermediation. The MFSA (Malta Financial Services Authority) has recently launched its Fintech vision and strategy and one of the tools intended to be used is regulatory sandboxing which could also be used for Robo Advisors. From this angle, there is nothing ground breaking and the MFSA like other regulators abroad would be testing its resolve as it will be venturing more into the technology arena and thus out of its comfort zone and expertise. This also stultifies the use of a regulatory sandbox if the same authority would just be looking at a static option and merely on suitability checks matching existing financial services legislation.

The MFSA however unlike other authorities abroad has a unique ally here. As I had written in previous articles, Malta is one of the few (if not the only country) who also has a dedicated Digital Innovation Authority (MDIA) which aims to create more trust and transparency by certifying certain forms algorithms. I had also opined in previous articles last year on a specific Regulatory Sandbox that can be used to develop an environment in which AI and its underlying logic and code are able to function according to pre-determined functional outputs in a testing environment followed by certification by the Authority. The recently published consultation on AI in Malta takes on board my suggestions and endorses these points. Building on the latter, the solution would thus be to have both Authorities work on a multi-disciplinary regulatory sandbox, looking at the suitability of the financial service components as well as the suitability of the underlying technology, where the desired outcome could be certification of the technology and an authorisation for the financial services. This could be ground breaking as it will also be a catalyst for future regulation and other financial services authorities from other countries could be invited to participate along supervisory authorities like ESMA to monitor and supervise.