1. Introduction
Investment advice is one of a number of services introduced to overcome the information asymmetries that structurally characterise the relationships between professional market operators and investors. The latter, in fact, are considered the ‘weak’ players in the financial landscape and often place their trust in professional operators who, thanks to their experience and knowledge of the markets, can help them invest their savings correctly. This relationship is established by means of a mandate contract between the investor and the intermediary.
From an analysis of the provision, it is clear that ‘personalisation’ and ‘determination’ are constituent elements and essential requirements of the service. In particular, the recommendation of a specific transaction must be formulated taking into account the economic characteristics of the client, the conditions of the contract and must refer to an equally specific financial instrument.
Over the years, financial advice has undergone major changes, both from a regulatory and an operational point of view. With the introduction of the Directive 1993/22/EEC, the service was relegated to a mere accessory activity. Thus, allowing the provision of the service to anyone who was interested in it. However, an unreasonable solution had been reached: only authorised intermediaries were required to comply with the rules related to their status, such as the obligations of fairness and transparency, while other operators could carry out such activities without particular constraints.
Ten years later, the changes introduced by Directive 2004/39/EU (known as ‘MiFID I’) and the subsequent – after another ten years – 2014/65/EU (known as ‘MiFID II’) are of great importance, as they have strictly regulated the obligations of intermediaries and the modalities of providing such a service in order to provide a regulatory framework capable of adequately protecting investors.
2. FinTech and RoboAdvice
As mentioned above, the evolution of the service, far from being limited to the regulatory aspect, also affects the operational aspect. Specifically, the technological evolution that has affected the financial markets has innovated the methods by which services are provided, sometimes to the point of changing some of their essential (or, in any case, structural) characteristics.
The advent of the so-called ‘FinTech’ has radically transformed the traditional paradigms of the markets, introducing an innovative reality that presents ever new questions and challenges from a regulatory, social and economic point of view. Investment advice has been profoundly affected by these changes, undergoing progressive digitalisation and automation over time. In this context, the term ‘Robo Advisor’ was coined to indicate an artificial intelligence (AI) system in charge of processing customer data and thus indicating a recommendation regarding a transaction in financial instruments (and, therefore, representing a form of investment advice). The service thus characterised can be broken down into three further forms, based on its level of automation: (i) robo advice pure, characterised by the complete automation of all phases of the service; (ii) robo advice hybrid, in which there is a collaboration between human and algorithmic operator; and (iii) the so-called ‘robo4advisor’ (or ‘robo-for-advisor’), which is an AI system that is not used by the customer for their own operations, but by the advisor as a tool to support their business.
The results collected in the first years of AI application show how the algorithm is profitably integrated into this activity.[1] However, it is necessary to question the ability of these new forms of advice to respect the requirements envisaged and designed for the ‘traditional’ type of advisory service. In particular, there is debate about the ability of an algorithmic system to comply with the rules developed for assessing adequacy and correctly weighing and balancing the client’s data and preferences.
Furthermore, it should be noted that so far there have been no changes to the sectoral regulations, as it is believed that the current provisions are also suitable for regulating the phenomenon of RoboAdvice. In particular, the European Institutions allow for the full application of the principle of technological neutrality,[2] not imposing, for the moment, more specific sectorial regulations for intermediaries offering their advisory services through AI.
3. The adequacy assessment in RoboAdvice
The term ‘suitability assessment’ refers to the activity of analysing the profile of the investor, in order to understand if the recommended transaction is suitable for that person, based on their characteristics and needs. In the initial phase of the process, some details of the client are collected, in order to outline a specific profile. The data to be recorded relates to the knowledge and experience in investments and markets; the financial situation (including the ability to bear potential losses); and, finally, the investment objectives and risk attitude. This information will then be analysed to understand which transaction can be considered appropriate for the client’s profile. This assessment must necessarily be carried out even if the intermediary uses an AI system to provide the service.
The European Securities and Markets Authority (ESMA)[3] has specified that it must be clearly indicated if and to what extent the human agent is involved in the provision of the service; secondly, in an attempt to make the client responsible for the transmission of his/her data, it must be explained how his/her answers influence the determination of the suitability of investment advice. Furthermore, it is necessary to explain which information is used to choose the most appropriate investment, in particular, it must be indicated whether the profiling process will take place exclusively on the basis of the specific questionnaire or whether additional data collected online will be used. Finally, it is expected that the method by which the information is periodically updated is indicated.
The collection of customer data presents significant challenges when the activity is managed by a Robo Advisor. In accordance with the Know Your Customer Rule and the Suitability Rule, pursuant to Article 25 MiFID II, it is necessary to introduce tools to ensure compliance with regulatory obligations, even in the absence of a human operator. In reality, the issue has not destabilised intermediaries, since, in the case of digitised consultancy, there is provision for the transmission of an online questionnaire that must be filled in completely by the client.
All this is not unrelated to what already happened (and happens) in practical terms: the use of such questionnaires, in fact, has been successfully implemented for some time in order to speed up the adequacy assessment process. Despite this, the absence of a human operator means there is a need to identify a method of making the customer responsible and ensuring that they maintain a high level of attention and care when filling out the form provided, in order to avoid the so-called phenomenon of the too fast click decision[4] or an overestimation of their skills, which could compromise the entire evaluation process. This has led to the introduction of a series of rules of conduct for the intermediary, aimed at the drafting of an intuitive questionnaire, which allows the client to give the right weight to the answers he provides, despite the absence of a human professional who can guide him.
Furthermore, even in the case of RoboAdvice, it is necessary to draw up the so-called suitability report, as required by recital 82 of MIFID II. This document must contain all useful information about the recommended investment and a reasoning as to why the strategy offered is considered the most appropriate for the characteristics of the individual. This information, accompanied by a description of the risks of the operation (which it is the intermediary’s responsibility to make clear), must also be transmitted, as mentioned above, even if the service is offered with an AI system.
4. Development of sustainability criteria and requirements
With reference, however, to the analysis of the information collected, there are no particular problems in the case of digitised consultancy, as it is undeniable that the algorithm is capable of processing a multitude of data in a short time, with performance infinitely superior to that of humans. In any case, to ensure the proper conduct of the adequacy assessment by the AI system, it is necessary for the intermediary to periodically check the software they use in order to verify its correct functioning. In particular, a specific company strategy should be established that provides for continuous checks, updates and verifications of the computer system.
In particular, recent issues relating to sustainability preferences should not be neglected in the design of software . In fact, in recent years there has been a growing trend to orient investments from a sustainable point of view, in compliance with the so-called ‘ESG criteria’ (Environmental, Social, and Governance).[5] In fact, during the phase of collecting information from the customer for the adequacy assessment, it is possible to ask them to express preferences related to sustainability. Although the new movements for environmental protection are ideally agreeable, it is necessary that they be coordinated with the provisions regarding adequacy mentioned above, to guarantee that the investor is properly protected in the choices he makes, regardless of how ‘noble’ these may be.
In its Guidelines of 2024,[6] ESMA, aware of the new environmental issues that are influencing investors’ choices, has established that, in the assessment of suitability, it is important that requests related to sustainability are also taken into consideration, albeit with a different weighting compared to traditional criteria. In fact, ESMA has established that the traditional paradigms that govern the matter and allow us to understand whether an operation can be considered appropriate or not to the client’s profile, must have priority in the evaluations of the intermediary, leaving a merely secondary space to choices related to sustainability. For this reason, in the case of digitalised consultancy, the algorithm must be programmed consciously, so that it respects this order of preference. Therefore, it will be necessary to first of all ensure that the AI system is able to verify if the adequacy requirements are met, based on the information that is traditionally collected, and only secondly analyse sustainability preferences. In this way, it offers to protect the customer through a two-phase activity. The first part of the activities, which does not involve any significant innovations, is aimed at identifying the products most appropriate to the economic profile of the subject and, only subsequently, an additional filter will be applied, carrying out a second screening among the instruments thus identified, to select those that respect the sustainability characteristics required by the investor.
What has been said so far highlights the continuing importance of assessing adequacy to ensure investor protection in financial markets. In fact, even in the face of technological evolution and new demands (such as those on ESG criteria), it is necessary that such assessment maintains its centrality in the regulatory system. In the near future, it may be necessary to adapt the tools (algorithms) and criteria (weighting of client preferences and investment objectives) to enable the issuance of an investment recommendation that respects the rationale of the discipline of investment advice, namely the protection of the investor.
End Notes
- Brian Fabrègue, ‘Algorithmic Trading Regulation in the EU and in Switzerland’ Schweizerische Zeitschrift für Gesellschafts- und Kapitalmarktrecht sowie Umstrukturierungen. ↑
- According to which it is not necessary to intervene to define specific regulations aimed at regulating the new cases that have been introduced by the Fintech phenomenon, but it is considered sufficient to correctly apply the traditional provisions of our legal system, also to avoid constructing an excessively rigid regulatory framework that limits technological development. For an analysis of the principle of technological neutrality, see Committee on Economic and Monetary Affairs, ‘REPORT on FinTech: The Influence of Technology on the Future of the Financial Sector’ (European Parliament 2017) A8-0176 <https://www.europarl.europa.eu/doceo/document/A-8-2017-0176_EN.html>. ↑
- European Securities and Markets Authority, ‘Guidelines on Certain Aspects of the MiFID II Suitability Requirements’ (ESMA 2018) Final Report ESMA35-43–869 <https://www.esma.europa.eu/document/guidelines-certain-aspects-mifid-ii-suitability-requirements> accessed 7 April 2025. ↑
- That is, impulsive statements that are not the result of a considered evaluation and risk distorting the information gathering process, see Michael Tertilt and Peter Scholz, ‘To Advise, or Not to Advise— How Robo-Advisors Evaluate the Risk Preferences of Private Investors’ (2018) 21 The Journal of Wealth Management 70. ↑
- Romain Berrou, Philippe Dessertine and Marco Migliorelli, ‘An Overview of Green Finance’ in Marco Migliorelli and Philippe Dessertine (eds), The Rise of Green Finance in Europe: Opportunities and Challenges for Issuers, Investors and Marketplaces (Springer International Publishing 2019) <https://doi.org/10.1007/978-3-030-22510-0_1> accessed 7 April 2025; Peterson K Ozili, ‘Green Finance Research around the World: A Review of Literature’ (2022) 16 International Journal of Green Economics 56. ↑
- European Securities and Markets Authority, ‘Guidelines on Funds’ Names Using ESG or Sustainability-Related Terms’ (2024) Final Report ESMA34-472–440 <https://www.esma.europa.eu/sites/default/files/2024-05/ESMA34-472-440_Final_Report_Guidelines_on_funds_names.pdf> accessed 7 April 2025. ↑
