French Banking Federation

This paper describes the main characteristics of automated financial advice tools: An automated tool is used directly by consumers, without (or with very limited) human intervention.
Though advice is said to be used in its common sense, automated financial advice tools definition is very restrictive. It sums 3 cumulative characteristics: a consumer-facing tool, which uses personal information provided by the consumer to produce an output, the latter being reasonably perceived by the consumer as financial advice. The algorithm “decides” which products should be recommended to the consumer, meaning recommended transactions on particular financial instruments.
So, the consultation seems to focus on recommendation tools. If this understanding is correct: first, many FinTechs would not be classified as automated financial advice tools because, because let’s say, they classify clients without exhaustive personal information or recommend an allocation but not a specific transaction. Then, a regulation restricted to this scope of tools would let many tools out and would therefore be of little impact.

Although we agree in principle with the main characteristics of automated financial advice tools which are presented in page 12 and 13 of the consultation paper, we are of the opinion that:

- firstly, only fully automated tools should be included in the scope. As soon as there is a human intervention during an important step of the process this should not be qualified as automated financial tools (nevertheless, our working groups show that it is necessary to maintain in a second phase a non automated assistance for any complaints, mistakes or difficulties). For example, if an IT tool is used only to recommend the investment advice previously provided by an asset manager or the research department of an investment firm, whether publicly or not, the process should not be considered as fully automated and therefore should not be included in the scope,

- secondly, with regard to consumers, we think that any regulation should take into account the fact that automation in financial advice is primarily developed to target retail clients. That is why we do think that only customers who really need to benefit from protection should be included in the scope. The focus should be on non-qualified investors such as those who are categorise as retail clients under MiFID.

Furthermore, it is of utmost importance to maintain a level playing field between all market participants proposing automation in financial advice. We would like to draw your attention on the development of new players implementing very innovative approaches for the user experience, especially with regard to risk profiling. Some of these new market participants are submitted to lighter rules allowing them to escape to some extent to regulation such as MiFID and PRIIPS. This situation could be seen as giving an unduly competitive advantage to new players, regardless of the quality of the advice they provide.


Meanwhile, the banking profession points out that the use of this type of tool is currently marginal but could expand in major proportions in the future (see last questions).
Most often, this type of approach is combined with human advice. Automated tools are used to generate adequate “advice” and the customer who receives it can ask for “face to face” or “distant” advice.

As the document stresses, robo-advisors compile data inputted by the consumer and then use an algorithm, which, after analysing the compiled information and using a decision-making tree, determines what product or service to offer to the consumer. However after this point, the consumer is typically asked to provide his contact details so that the financial institution can contact him to arrange a personal meeting. Indeed, the process could go to the end without any human contact.

The used of an algorithm means that the determination model could include information that does not necessary come from the consumer’s replies. Hence, the offer, proposal or advise that will be delivered to the consumer could, in some cases, be based on major client typologies and, hence, not be a fully personalised response corresponding exactly to the consumer’s situation.

You must distinguish between the portfolio construction process and everything that goes around it. The portfolio construction process is done by a machine but everything else requires human intervention (model building, consistency check, sales…) and this means teams of experts are needed. It is crucial to ensure the protection of final investors with properly regulated experts. These experts should comply with all existing requirements applicable to Asset Managers: knowledge, reputation and experience.

As of today, automated advice is far from AI, it is more a question of data mining and optimisation. Over the short term, automated financial tools are not intended to completely replace financial advisors. Indeed, there is still value in human advice (via a portfolio manager or a financial advisor), able to take into account human parameters in the decision process. For instance a human can challenge client’s investment profile, level of knowledge and explain the impact of investment decisions to the client. So, whether we refer to specific individual situation or to the understanding of customers ‘needs, human interaction and professional expertise cannot be suppressed completely in the interest of our clients. Thus, automation in financial tools would create value if it is a decision support tool, helping financial advisors to provide clients with investment advice.

This approach is all the more relevant as players on the same market are sometimes subject to different regulations, which means that they can more or less use very different scoring keys without the client’s knowledge (e.g., scoring could include information from social networks, including “likes”, integrity, etc.). Such an approach could result in the tool’s offering products and services that do not match the client’s situation, depending on the weighting that this factor has been given within the algorithm.

All in all, it is important for the consumer to receive equivalent levels of confidence, security and information for all services or products offered to him through a robo-advisor. For this to happen all players operating on the same market must achieve the right balance between a client experience facilitated by automation and the quality of advice provided. This requires that all market players be subject to equitable market rules that apply to everyone in the same way.

Hence, whenever new regulation is worthwhile, it must apply to the business and not the entity that runs this business and must not apply only to those companies that are already subject to regulation."
As far as financial advice is concerned, we understand that MiFID regulation applies also on automated recommendations and, consequently, robot advisers are subject, among others, to the suitability, the product governance, the financial skills level and the conflict of interest rules as long as the advice is about financial instruments.
It seems to us very important that all the advices given respect such rules including the two following cases (which seem to be exempted from MiFID regulation):

- If robot advisers limited their services to an asset allocation recommendation without any financial instrument directly recommended. In that case, clients would not receive any support in choosing the suitable financial instruments to apply the asset allocation: neither risk assessment nor product analysis. In addition, they couldn’t rely on the assessment of the robot itself.

- If the recommendations are given by persons providing investment advice in the course of providing another professional activity not covered by MiFID, provided that the provision of such advice is not specifically remunerated". That could be the case of some participating platforms or any activity deciding to deliver investment advisory for free that could create an unfair competition environment and no guarantee for the client about the reliability of the advice.

Digital investment tools such as fund screening, model portfolio or back-testing are more relevant if they focus on customer’s needs, objectives and experiences. Such tools are already being developed in Retail networks to improve the quality of investment advice provided by investment advisors. In addition, automated advice is also a way to remove emotion from the investment making decisions, especially in a volatile market. Nevertheless, in adverse market conditions the assistance from a human financial advisor still makes a lot of sense to help clients assessing their options

The fine tuning of the digital and physical factors for each client, or at least, for each segment, should lead to an optimization of the benefits and of the risks mitigation."
The Federation cannot discuss in detail the individual initiatives that its members may take.
The Federation cannot discuss in detail the individual initiatives that its members may take.
This is a very competitive business with low entry barriers as the technology and the investment strategy behind as they show up currently in the market can be considered as key mainstreams. The key challenge remains the ability to generate strong inflows and therefore revenues.
In the financial sector in general and in insurance in particular, the matter of regulations that are in force and regulations that are being implemented appears to pose some potential problems.
The first sticking point is in subscriptions. Regulations now have greater standards (including the provision of documents, know-your-customer due diligence for AML-FT and FATCA/CRS aspects, screening for the freezing of assets), which require interactions with the client prior to any new relationship. These standards are enhanced considerably for remote/online selling, which would be the case for robo-advisors. These constraints do not seem to match the “gamification” approach of the customer experience of current robo-advisors. Will it be possible for market participants subject to regulation to offer an attractive customer experience? In addition, national regulations regarding online services and specific rules regarding the subscription of complex products like structured products could slow down the development of these new techniques.

The second sticking point is the regulations that are currently being implemented (MiFID 2/PRIIP/IDD). During their multi-year development framework no thought at all was given to robo-advisors. Will they be compatible with robo-advisors?
Thirdly, and as mentioned in our answer to point 1), some new actors have the capacity to offer the same service as banks or insurers without respecting the MiFID constraints (gaming questionnaires for instance). Conversely, we would like to clarify our capacities to use the same format, while respecting the MiFID risk requirements.

Besides, the paper states that the robot advisers could facilitate the cross border financial services. We just underline that the problem here is not technical but regulatory (many national regulations).
The analysis makes sense with regard to speed, ease of use, audit trail and controls. The ability to deliver advice at the same time to a large number of clients without waiting for individual meetings is important. It increases the ability to provide service offerings to clients. It appears clearly that the automation of financial advice is more relevant to inform customers about their investment in a continuous way and also to ensure a follow-up on very specific aspects that do not need human intervention.

For the quality of service, it is true that an algorithm has a number of advantages (decrease in the cost of providing advice, broadening access to financial advice, reduce some elements of behavioural biases, very interesting way to formalize all the business rules that are used, and make sure that they are applied in a very systematic way, thus eliminating all the potential biases that exist with a human processing, provides a clear audit trail on why a specific advice has been provided, which is important both from a customer relationship management standpoint and from a regulatory one) provided it has been well designed, and is well maintained. Therefore the question of the expert team that will take care of the creation and update of the business rules is key for providing relevant advises. Once created and live, the algorithm does not suppress the need for human expertise, but it implies a different mode of organization. In addition, robot advisers would probably have the capacity to respect better, in a more efficient way, the requirements of MiFID (suitability, appropriateness, costs, record keeping and other reports, etc.).

Meanwhile, we would stress an advantage – that of consistency in advice (see B5). Systematic and constant compliance with client wishes is a basic issue and makes a big contribution to protecting customers, which is a major channel of current regulations.

However, as question“Q1” states, consumers must be provided with all information allowing them to know and, especially, to understand, that the offer and level of advice they receive are based on an algorithmic analysis and, in the case of advice, industrialised processes. Accordingly, such advice may under no circumstances reflect a personalised and adapted analysis of the customer’s own situation, nor achieve a level equivalent to that of a specialised adviser.

As for the quality of the information provided, the banking profession points out that it is necessarily and closely correlated to the quality of suppliers used. This quality can be judged, for example, in the frequency of updates and the sources used.

We would therefore like to highlight the issue that automated advice could not be adapted to all segments of customers.

Similarly, we are very cautious with the cross border arguments, as there are strict regulations on that topic.

Besides, the cost reduction envisaged in the discussion paper is not so evident because of IT developing, maintaining and security costs. For that purpose specialized teams will be necessary.
All in all, the EBA must nonetheless remain alert to avoid comparing advice delivered by an advisor to that provided by a robo-advisor. In an effort to promote speed of access and advisory, it must not promote low-cost advice using a very limited choice of products or services to the detriment of reliable personalised advice based on a more detailed exploration of the customer’s needs, which allows him to access a greater depth of services and products.

Personalised advice makes is possible to steer a greater portion of household savings.
There is one benefit of “enhanced/additional service” that can be added. As it is fully digitalized and automated, customers could expect to receive more regular updates in order for them to redirect (or not) their investments (advice sent automatically), and manage their portfolio more dynamically. In addition, consumer can start investing in a proper asset allocation product with very low capital vs. current financial advice offer limited in some countries to mass affluent clients.


The harmonization (not standardization) of the customer experience is a benefit, as long as the advice remains personalized and takes into account the risk profile and investment objectives of each individual customer.
Yes we do. Benefits to financial institutions are:
- Cost to serve can be reduced.
- Institutions have access to a wider range of consumers.
- They can deliver a standardized approach, consistent customer experience. Looking at these benefits, we clearly face some challenges when adopting “automated tools”:
- When the cost of providing (automated) advice decreases, pricing F2F-advice becomes even more difficult. The F2F-advisor will have to prove his added value, which requires a high level of accessibility and training. F2F-advisors have to really know the client’s personal situation and can offer products that are not offered by the automated advice tools.
As automated advice tools democratize access to financial advice, current segmentation could be impacted. Clients will probably choose their own segmentation in the future.
- Financial advisors may also rely on automated tools which provides them with up-to-date data, consistent information, logical algorithm, prices of transaction benchmark

The advantage of an audit trail (B11) will not always be clear (best case scenario). As audit trails require ex-ante documentation ex-ante, they must be designed from the start to be properly implemented, but when technological innovations focuses mainly on speed, the issue of documentation is seldom a priority and is postponed on a regular basis.

Even so, the banking profession regards robo-advisors as a new added-value offer making it possible to address client expectations and needs at a more competitive cost, and not as a mere way to cut costs. If cost to serve may be reduced, we highlight the fact that IT and change management investment are required, and many more (marketing, compliance). In other words, the cost reduction is not sure because the investments (initial build and run) are not negligible. The quality insurance of the service is still a strong need. An expert team has to create and maintain algorithms
Benefits are more likely to come from the growth than from the cost reduction in a short term
In cases where the advice provided is not equivalent to what could be delivered by a human advisor they at least allow financial establishments to provide access to advice that is a little more constant and consistent without having to set up a meeting. They can make themselves available to a larger number of customers while providing a broader range of products and services.
So robo-advisors are not in competition with human advisors. They make it possible to ensure a certain level of advice and information availability. Hence, the development of robo-advisors that the human advisor can use but that are not made available to the end customer.

We also think that automation of the advice production is a way to comply with the MIFID2 regulation, as there will be a clear audit trail of the reasons why a specific advice has been provided to a given customer at a given point in time. The robotization of the MiFID2 requirements could be quite interesting in order to ease a very heavy process, which, again, seems not to be shared with some new competitors. We could change the statement and say that the important improvement on customer protection generated by the new MIFID2 regulation, for instance, requires to collect all contextual information as to how, when, and why the advice was produced. The automatization of the process can be part of the answer and help deliver the required audit trail.

These tools can enhance the customer experience of online services while allowing heightened competitiveness.

The consistency of the customer experience is something very interesting, with a gamification of the risk profiling based on the recent works on the behavioural approach. This transforms a painful customer journey into something more attractive. We think that such innovations are also part of the success. Coming back to the purpose of this discussion paper, it might be worth that regulators clarify to what extent such innovative methods are compliant with the existing regulations, to avoid potential distortions of competitions, with banks facing more constraints than new players.
We share the view that automated financial advice tools in the banking, insurance and securities sectors, could be a very useful complementary tool for human advisors.
Banks could also increase cross-selling, and benefit from an increased global relationship with customers.
In France, potential benefits may be higher in the securities sector. Generally speaking, consumers go for these solutions (securities accounts, shareholder savings plans) when they are looking forward to getting higher yields in an autonomous way.
From our perspective, risks are well identified. However, the banking profession feels that, whether the customer is in contact with a robo-advisor or a human advisor, the level of risk is the same, only its nature is different.

Robo-advisors may in some cases be used for industrialisation (low-cost) and in this case, the customer must be aware that the advice he receives is not the result of a detailed and personalised analysis of his situation. But in other cases, it will be part of a broader advisory process that is not necessarily low cost.

An important point is the quality of the redaction of the advice. A basic recommendation like “buy” or “sell” is worthless if there are no explanations on the context, the investment objectives and the risk consumption before and after the proposed deal. This also raises the question of the expiry date of the advice, which needs to be managed. If a customer waits too long before he validates the advice, this one might have become irrelevant with a risk of not meeting anymore its objectives.

There is a risk of misunderstanding of the client. The risk of misunderstanding the advice provided by a robo-advisor is greater than with a human advisor. Even if precautionary questions could be inserted to ensure that the customer has understood, there is nothing to assess a divergence in understanding and prevent the consequences this could have in the customer’s choice of whether or not to subscribe the offer.

This risk can be exacerbated by the terms used, the assumptions used for determining the recommendations, the priorities given in the rules for implementing the algorithm to various criteria (personalised analysis based on data inputted by the customer or analysis based on pre-set customer profiles).

We strongly believe in the necessity to maintain, at least and at a second level, a team of human advisers or an appeal for assistance, in order to solve many problems, such as mistakes, misunderstandings.

Depending on the algorithm, the personalization can be more or less sophisticated. It’s why taking into account the different business models (retail mass distribution and advice or tailor made private banking for instance), the importance of the human advice and its expertise will be different.

To avoid misunderstandings, the quality of the formulation of the advice will be important.

The risk of error in the tool exists, and to mitigate it there must be a team of experts advisors mobilized during its inception, and also once live to detect potential errors, areas of optimization of the algorithm, and update it depending on the changes of the environment. This also applies to the detection of fraudulent modifications inside the algorithm.

Similarly, when it comes to the issue of data protection, the banking profession would like to point out the importance for providing the customer with total transparency and the utmost security with regard to the use that market participants could make of these personal data.

Here, the EBA must ensure that future European regulations apply to all market participants offering advice based on a robo-advisor for the same activity.

Looking at these risks, financial institutions clearly face some challenges when adopting “automated tools”:
- When building an automated tool, we have to inform clients of the characteristics of the tool: how does the tool function?, what can the client expect?, what type of products will be offered in the end?, how will we use the information inputted by the client?, …
- Automated tools can/will be an important part of our advice story, especially in the digital era. But it seems clear that a F2F-approach will have an added value in the future. A personal advisor can put the advice, given by an automated tool, in the right perspective. He can check if the advice given matches the real personal situation of the client. Client may have misinterpreted/misunderstood some questions, the tool did not offer a product that suits the client’s needs better, the advisor has information that was not asked by the tool etc.
- Some risks have to be mitigated by internal organisation: rules regarding selling info to third parties.
The consultation points out the risk of consumers making unsuitable decisions because the tool facilitates them to move too quickly through the process". This type of risk could be exacerbated in case of very volatile financial markets where consumers might panic and sell quickly their automated long term investments. The possibility for consumers to join a financial adviser might help them to understand recent market volatility, review the initial investment time horizon and the percentage of their assets invested in the strategy.
We also question the way scenarios analysis will be used in the tool and how relevant they are. We considered that the scenarios used in the Robot Advice should follow the new rules that will be implemented with MIFID2 (Product Governance Scenario Analysis for Financial Instruments)."
The risks will be higher on financial instruments or strategies that are more volatile or more complex (according to MiFID 2).
We have observed in the past risks of panic decision on classical investments in a context of chaotic markets. We think these types of risks could be accentuated in case of automated investment. It will worth studying recent investors’ reactions to volatile equity markets since last summer when invested through automated advice without any kind of financial advisor support. So it would make sense to apply Standard disclaimer, already used by Asset Managers in their prospectus.
Moreover, the possibility cannot be ruled out of risks from tool malfunctions. We have not observed directly the realisation of this risk but an Internet search does turns up robo-advisor references with surprising results in customer classifications based on answers to customer profiling questions. This risk is clearly already a reality.
Yes, but we would highlight
The risks related to a faulty automation
It can be mitigated through the presence of an expert team during inception and after in the run phase to check and update the algorithm. The generation of advice in natural language is also a protection, as it enables an easy check by a variety of actors (advisors, compliance, legal,…).
Risks related to liability allocation:
• That risk is inherent to disintermediation as a whole and is not specific to automated advice. Taking into account the rising sophistication and personalization of the robot advising, and the quality of the process which is required by MiFID, the responsibility is quite the same between robot and human advice. In case of dual advice: robot and human, it seems to us to be complementary.
• Anyway we agree that in an approach involving different players, with financial institutions and technology or services providers, the split of responsibility must be clearly documented, both from an operational and contractual / legal standpoint.
• Points to consider in enforcing regulations on all entities:
o Regulations regarding on-line selling.
o Cross border rules to make sure that the advice takes into account all parameters, including the fiscal residence and the localisation of the client.
o Specific rules regarding the subscription of complex products like structured products.
As it is the case for customers, the financial institutions are also at risk due to the “widespread usage of financial advice tools”. Paragraph 77.

Financial institutions have to be conformed to strict regulation, to avoid penalties. The development of new tools implies an upgrading in the regulation area:
• On the one hand we have to be sure that we are able to use automated tools that respect existing and future financial regulation relative to asset managers.
• On the other hand, all stakeholders of this new market should be submitted to the same regulation framework (“same level playing field”) in order to avoid distortions between banking/insurance actors and no banking actors and ensure a proper level of customer protection

The viability of the robo-advisor business model depends on how this market develops. The larger the number of players, the harder it could be to ultimately maintain its profitability.

Keep in mind that there is no single answer to the issue of implementation cost. Some players will be low-cost, but for others investment costs will be heavy. There is also the matter of acquiring a customer portfolio for new players which could lead to partnerships. New players have the technology but not the customer base and the cost of acquisition is high. We expect more and more partnerships, or acquisitions of FinTechs, to internalize the know- how and the technology.
Yes, automated advice can bring accessibility, transparency, control, convenience and lower cost.
But at this stage, and taking into accounts our forecasts, it seems to us that at different stages and for different purposes, depending on the business lines, the human and robot advices will remain complementary.
Should increase dramatically the adoption of digital adoption, it won’t probably replace human financial advice.
Ultimately, there will probably exist various models of complementing, to various degrees, the advice provided by human advisors with those provided by financial robo-advisors (some of which will probably come directly from partnerships formed between market entrants and financial establishments, even if only for the cost of acquiring customer portfolios).
The automation in financial advice market will continue to develop in the coming years. However, while the development of automated advice has emerged based on low cost passive underlying vehicles, we are convinced that this trend should and will extend to actively managed vehicles, be they packaged as multi-asset funds or specialized in sub-markets.
We see passive and active offers complementary to fill client needs in the search for the right balance between performance, costs and risk management. Active management plays a fundamental role to an efficient allocation of capital in the overall economy. Professional investors through actively managed funds can also apply sophisticated risk mitigation techniques and implement reactive asset allocation decisions to market developments. Hence the need for having financial automated advice covering also such essential investment vehicles.
For all those reasons, we would like to emphasize the need for “the same playing field” between active and passive moving forward. It is why European authorities must be very careful about competition and regulation issues in order to preserve a good level playing field between the different actors and guaranty a good level of protection and security for consumers.
MiFID 2 forces us to revisit our value chain and the way we advise our customers. Technology can be a part of the answer, but only a part.
In particular this type of approach could be worthwhile for MiFID traceability obligations.

Automated tools can be used to generate adequate “advice” for the advisor to advise his customers. They will be part or combine with the next generation of tools used for Customer Relationship Management (CRM) and Leads/Contact purpose.

Disruption through digitalization is extensively commented in many industries and automated financial advice is often mentioned as one of the many examples of such developments for the financial sector. We agree on all the potential benefits that such development can bring for end investors. However, we consider that providing advice for investment in many aspects is not a standard service and all regulatory developments in the recent years prove how sensitive this service can be. Therefore, a proper regulation need to be applied fully aligned with the existing requirements on human financial advice. The option to access to human contact should be mandatory for such a sensitive matter can be a way to mitigate the risks while securing the benefits.
With the upcoming implementation of MiFID II and PRIIPS, we expect to see an important rise in the offer of automated search engines centered on financial products. Their functionalities will be significantly improved over the existing search engines to the extent that the distinction between these searches engines and other robo-advisers may be partly blurred.
Apart from the improved search technologies that this new generation of engines will incorporate, another key asset that these search functionalities will bring will result from the enhanced granularity in the product information that MiFID II will introduce.
Indeed, as a result from the new product information requirements introduced by the new regulations, product manufacturers and distributors will widely disseminate a product information that will be much more abundant than is currently available to most investors. This product data will incorporate information on the product target markets which notably includes criteria such as the client objectives, its financial situation and its level of expertise which are also key elements in the suitability diligence performed in a financial advice situation.
One more thing: more and more players and more diverse players will enter this market in the future. It is essential that regulation and supervision, including certification processes, are the same for everyone. Otherwise, competition rules will be skewed, and consumers will be left unprotected.
Alain RICHON
F