« Back

Cognitive Finance Group

Question 6: Are the issues identified by the EBA and the way forward proposed in subsection 4.3.1 relevant and complete? If not, please explain why.
Cognitive Finance Group

Cognitive Finance Group is a consultancy specialised in applied artificial intelligence in financial services, and the views expressed here reflect our company’s views.
We are the trusted adviser to Boards and senior management on scoping, selecting and implementing artificial intelligence for business growth and increasing competitive advantage.
We have a dedicated team of strategists in financial services as well as data scientists and machine learning experts.
We bring practical knowledge of applying artificial intelligence in financial services gained from client work, work which is informed by our over 45 years experience in blue-chip global organisations and our specialist knowledge about artificial intelligence.
This submission was written by Michael Aikenhead. Michael is a Director at Cognitive Finance Group. He completed bachelors degrees in computer science and law followed by a PhD examining the application of AI for automating legal reasoning. Michael headed the European knowledge engineering team for an AI company which was acquired by Oracle and has spent over 10 years working in banking.

Scope of this submission

This submission focuses on the examination of ‘The impact of FinTech on the business models of credit institutions, payment institutions and electronic money institutions’ outlined in section 4.3 of EBA discussion paper EBA/DP/2017/02 (hereafter ‘the discussion paper’).

Submission

Issues identified and the way forward

Question 6 of the discussion paper asks:
Are the issues identified by the EBA and the way forward proposed in subsection 4.3.1 relevant and complete? If not, please explain why.’ and we make a number of observations in relation to these issues and the proposed way forward.
Paragraph 93 of the discussion paper states that:
Credit institutions may need to adapt their business models, for example, by: (a) adjusting their offered products and their interaction with customers and employing FinTech technologies and products, thus increasing their revenues, (b) adopting new technologies through digitalisation of internal processes, thus reducing costs, or (c) a combination of both.
We observe that the focus here is on effects on credit institutions and benefits from increased revenue and lower cost that may arise for them.
We would emphasise the impact of FinTech on the benefit that can be delivered to customers and that when examining business model impacts in that context of the discussion paper, that focus on customer benefit should be maintained.
In this context we would submit that existing banks have been slow to deliver innovation to customers and regulation should not be used to shield banks from competition - banks should not shelter from competition behind regulation.
Paragraph 97 of the discussion paper states that:
FinTech may be able to offer solutions to increase cost efficiencies, address users' complex needs and generate value for the economy, but in order for these solutions to be delivered, appropriate policies on important issues, such as access to technology, data standardisation and security, personal data protection and data management, need to be put in place.
While we agree with the benefits that FinTech may be able to deliver, we do not believe that policies are needed before technology can be developed to deliver innovation to users. We suggest that innovation will happen, unless policy slows that, and that policy should only be used to intervene in the market if there are clear and compelling reasons, made explicit to participants, for this.
Paragraph 99 of the discussion paper states that:
In addition to creating new opportunities, FinTech may also pose significant challenges, as many credit institutions still operate with outdated legacy IT systems that require additional investments for their modernisation. This is considered to constitute a significant disadvantage for these institutions compared with ‘pure’ FinTech firms, which have modern systems and technology and no legacy issues.
To the extent that this statement about credit institutions operating with outdated legacy IT systems is a statement about current status, then we would generally agree with this.
However we suggest that this weight of legacy systems is a function of the behaviour of credit institutions themselves. We suggest that this weight of legacy systems stems from underinvestment in technology by credit institutions for extended periods of time and that when considering ‘disadvantage’ in institutional investments in technology, that investment in legacy system maintenance and incremental upgrades should be separated from investment in technology R&D and technology driven business transformation.
We suggest that while legacy IT itself is currently an issue for organisations, the organisational functions that drive innovation and deliver transformation will also need appropriate culture and skills, and will need board level support to deliver innovation. To that extent we believe that adoption of FinTech will not just be a matter of the technology itself. Adoption of FinTech is a primarily a leadership issue which needs to be driven from the top of organisations by their Boards. In our experience, we repeatedly see innovation founder due to a lack of understanding and education at the Board level. A key component of the work we deliver is therefore Masterclasses on applied artificial intelligence in financial services for Boards’ education.
We do not accept that existence of legacy IT, and the need for organisations to manage this legacy, is a reason for shielding them from innovation. Credit institutions have taken the benefit of periods of underinvestment and that must be considered as the context for the current impacts of FinTech. Existence of legacy IT in existing organisations should not be a reason to deny customers the benefits that can flow from new technology.
Lastly, we would emphasis that existence of legacy IT is not of itself a barrier to delivery of FinTech driven innovation in existing organisations. We point to Goldman Sachs’ launch of their ‘Marcus’ lending platform as an example of this. Similarly, we would point to the Board level sponsorship provided to FinTech driven transformation in Banco Bilbao Vizcaya Argentaria.
Paragraph 103 of the discussion paper states that:
With a view to understanding the impact on credit institutions’ business models, it is proposed that the EBA take the following actions:
a. hold interviews with a representative sample of credit institutions;
b. develop a thematic report on changes to the business models of incumbent credit institutions.
In relation to the way forward we agree that interviewing a sample of credit institutions would be useful. We would emphasise that these interviews need to be of a standardised nature so that a structured view can be formed across institutions.
We would also emphasise that a structured sample of firms must be selected, which at least covers incumbents, including innovative incumbents, smaller, medium and larger institutions, and new Fintech credit institutions.
Most importantly, we would emphasise that holding interviews exclusively with a representative sample of credit institutions if not itself sufficient to inform on business model impacts. We do not believe that credit institutions by themselves have the capability to fully inform on the potential impacts on business models. For example we do not see technological developments in FinTech being driven from within credit institutions and we do not see the analysis of impacts from these FinTech innovations, for example the effects on jobs, society, the economy, and impacts on issues such as privacy, being driven from within credit institutions.
We believe that credit institutions will feedback weighted towards impacts on their own individual organisations rather than analysing wider benefits from FinTech innovation that could materialise to customers or to financial services markets.
We would emphasise that amongst other groups, new and innovative credit institutions, FinTech startups, the large technology groups, and specialist and expert consulting firms should all be able to participate in interviews and contribute to the subsequent thematic report.
Finally we would emphasise that the proposed examination and thematic report should not be a single point-in-time exercise. FinTech is currently undergoing a period of rapid development and we suggest that this will continue and will continue to drive potential change to business models. In this respect ongoing monitoring of potential impacts is warranted to ensure that business model impacts are understood and benefits are realised.
Question 7: What are your views on the impact that the use of technology-enabled financial innovation and/or the growth in the number of FinTech providers and the volume of their business may have on the business model of incumbent credit institutions?
Impact on business models

Question 7 of the discussion paper asks:
What are your views on the impact that the use of technology-enabled financial innovation and/or the growth in the number of FinTech providers and the volume of their business may have on the business models of incumbent payment or electronic money institutions?
We see various pressures on the business models of incumbents - with increasing competition between incumbents, with increasing competition from new FinTech driven entrants, and increasing competition from large international technology companies. Mixed with this are changing customer expectations about service and product delivery. Amongst other things these changing customer expectations and this increased competition will drive increasing reliance on technology in incumbents, increasing reliance on outsourcing in incumbents, more innovative products for customer, and weakened direct relationships between customers and incumbents.
As example of the increasing competition from FinTech driven entrants we would point to the creation and continued growth of ‘digital challenger’ banks in retail banking across jurisdictions (e.g. Monzo, Starling, and N26).
As example of the increasing competition from large international technology companies we would point to the expanding list of financial services that these companies provide (e.g. Amazon Cash, Amazon Loans, and Facebook Payments).
We see the advent of the Payment Services Direct (PSD 2) (Directive (EU) 2015/2366), and the trend to open banking more generally, accelerating the picture outlined above. With these initiatives we see less customer lock-in to existing organisations.
While we see a growing number of financial service providers, we also see potential in some services for increase in the aggregate value of that service. As example we note the growth of automated wealth managers, where the provision of services at lower cost of customer entry than has been traditional, increases the base of customers able to make use of services.
We think that as business models become increasingly founded on technology, that business model change will increasingly be continuous and increasingly linked to the possibilities opened by development in the technology itself.
As an example we believe that Artificial Intelligence is one of the technologies that provides new competitive tools changing the competitive balance. AI is comparatively little understood in financial services and little used, in comparison to all the potential uses discussed in the literature and that are emerging.
Even staying with more conservative assessments of where AI may be adopted we see AI opening ongoing change to delivery of financial services. The current comparative lack of understanding is therefore a cause for concern and long term potentially disadvantages for those organisations which are not moving quick enough in adopting this transformative technology.
One of the characteristics of AI is that it allows mass personalisation - the tailoring of services for individuals or small groups of customers. Tailoring is evidenced in things like customer interactions, and in more personalised products and pricing. Personalised products and pricing are particularly evident in insurance, but are also evident in retail products like savings where data analytics is being provided direct to customers on their accounts, and automated services are emerging for example to collate balances and automatically run wealth management and investment services on those accounts.
In wholesale banking, we are beginning to see automatic construction of bespoke products, with the composition and pricing of the components of those bespoke products being tied to and driven by the characteristics and needs of individual customers.
This personalisation has proved extremely popular with customers as evidenced by the growth of the aforementioned Marcus lending platform in the United States, digital challenger banks across Europe, and the growth of automated retail wealth management platforms particularly in the United States. It is notable that this type of personalised automated service has been particularly popular in China.
Mass personalisation allows easier scaling while also keeping costs under control - and is only possible with the use of artificial intelligence.
With customer expectation for always available, multi-channel, easy to use, and personalised services - all expectations heightened in younger customers, and heightened by experience of receiving such service in other sectors - the expectation is for financial services to be similarly provided (and dis-satisfaction with financial services stemming from them not being so provided).
To reiterate, we see financial services firms becoming increasingly technology driven (with associated need to understand that they are not just financial services firms which use technology). To reiterate, as organisations increasingly have technology at their core, with the ongoing evolution of that technology we see see ongoing evolution of business models, and products.
One observation related the increasing centrality of technology in services and product offerings, is that technology providers will become more central to financial service operations. With this increasing centrality of technology we see increasing scope for technology companies to themselves provide financial services and to increasingly become competitors to incumbents (and we would again refer to the incremental activity of companies like Amazon and Paypal in provision of financial services and products).
The technology companies that will be increasingly relied on within financial services could be multinationals.
We believe that there should be encouragement to ensure that technology innovation is not simply conducted overseas and consumed in Europe and instead that ongoing examination of the strength of the financial system should include a focus on encouraging stability, innovation and customer benefit across the full financial services ecosystem, including not only the stability and regulation of the system itself, of incumbent firms, but also of new firms and technology providers.
Contact name
Michael Aikenhead