Check against delivery.

It gives me great pleasure to be speaking to you today in Malta. Your country lies at the crossroads of several of the world’s greatest cultures, many of which have left their mark in Malta. And of course many more people are still coming here today. Some of them are tourists who want to make the most of the guaranteed sunshine on your many beaches, while others come to admire the abundant cultural treasures. And there are those – mainly young people – who come to Malta to learn English or to party. Or to do both. But there are very many people who have never set foot on any of the Maltese islands but still know the country – from the tremendously successful “Game of Thrones” television show.

As you may know, one of the recurring themes in this series is the transitory nature of this world. In Game of Thrones, certainties that were long held to be immutable and incontrovertible are cast aside in the blink of an eye.

But what does that have to do with my topic today, the future of financial supervision? The answer is simple: what we are currently witnessing in the financial world does feel a bit like Game of Thrones – everything is changing. Fundamentally and very fast. Of course changes in the world of the supervisors are not triggered by the power struggles of rival clans. We supervisors – and the entities we supervise – are being challenged by a variety of fundamental trends. First, an extended period of time in which financial markets have to cope with rather unusual market conditions, most notably: low or even negative interest rates. The impact ist severe – and the longer such conditions continue, the more severe the consequences will be.

A second fundamental trend is the the emergence and ever growing presence of the European ideal. Financial regulation and, in fact,in parts even supervision is one of most integrated policy areas within the European Union. Needless to say, I very much welcome this trend, but it would be naïve to ignore the many challenges associated with such a fundamental transformation. Continuing our journey on the European route and creating a unique, but highly effective and efficient interaction, between the national and the European level is an ongoing task which requires our attention and care.

A third challenge is obviously digital transformation. This digital change is not wafting like a gentle summer breeze over the beaches of Malta. It’s sweeping over the industry like a storm – and is shaking up business models, companies and even entire markets – including supervision, I suppose.

Our mission as supervisors is to ensure the proper functioning, stability and integrity of the financial markets. And that in itself is a major undertaking. On top of this, however, we must also endeavour to play a key role in shaping this digital transformation within the scope of our mandate. A condition for this is that we understand what is happening and that we make – and keep – ourselves fit for the digital era. We must spend enough time and attention on the digitalisation of the financial world and its repercussions for supervisors – and of course, this is exactly what all of us are doing. Let me just focus on five fundamental issues we need to consider.

1) Our tools may change, but not our craft:

Let me start with the good news. As much as digitalisation may change the financial world, one thing will basically stay the same: our supervisory craft. One of the main focuses of our work in the future will continue to be to analyse data from and about companies and to draw our supervisory conclusions from that data. It goes without saying that we will have other tools at our disposal than we had in the past and today. We are ourselves going through a process of transformation and, little by little, we are becoming completely digitalised. How did former Hewlett-Packard CEO Carly Fiorina put it? “Everything that can be digitized will be digitized.” You already know the buzzwords, Ladies and Gentlemen: “Suptech”, the state-of-the-art technology that we supervisors use to fulfil our mandate, and “Regtech”, the technology that companies use to comply with supervisory requirements. Both of them are becoming increasingly important.

I don’t think I am proclaiming any particularly revolutionary messages if I assume that digital technology – and artificial intelligence in particular – is going to be playing a greater role when it comes to identifying and resisting violations of the rules and financial crime. We are also thinking about how we can make the reporting regime fit for the digital era, and which processes we have to adapt to do this. At the moment, we still have one leg in the analogue world. We distribute templates and the institutions fill them out more or less semi-automatically. Of course a process like this has as much of a role to play in a digital world as a stoker does in a high-speed train. Going forward, the banks must be able to transfer their data to us directly from their systems. This will save them effort and expense and will give us more up-to-date and possibly also more meaningful data. All of this may still be a vision for the future, but we are already on the starting blocks. My banking supervision colleagues are currently preparing a feasibility study on the digitalisation of the reporting regime that we expect to conduct next year.

Another area where experts see a lot of potential to use new technologies is curbing insider trading and market abuse. We have already initiated an “Automated alarm and market monitoring system” (ALMA). Under Article 26 of the European Markets in Financial Instruments Regulation (MiFIR), investment services enterprises that enter into transactions in financial instruments must notify their supervisor of the details of those transactions as quickly as possible – and at the latest by the end of the following working day. Our “Automated alarm and market monitoring system” is intended to comb through these datasets autonomously, looking for conspicuous trading participants and transactions. We use data mining methodologies to detect similarities, patterns and anomalies in the trading data or the trading behaviour of market participants: clustering, for example, or peer group analyses. However, we are not creating our own black box to do this. What we will not tolerate from industry, we cannot cultivate ourselves. It is axiomatic that we, the human supervisors, still take the final decision about which facts are considered to be conspicuous and which supervisory action would be appropriate.

2) What skillsets will supervisors need in future?

The fact that the supervisors’ craft won’t change in principle does not mean that everything can stay the way it is. As I have already mentioned briefly: digitalisation will demand new skillsets from everybody involved in supervision. Above all, this means a thorough grounding in the MINT disciplines, in other words Mathematics, Information science, Natural sciences and Technology. That’s because we must understand the key digital mechanisms and their impact on our supervisory activity.

To enable future generations of supervisors to move swiftly and prudently within a digital world, authorites must invest in suitably skilled staff. A quick look at the manufacturing industry may help. Previously, car manufacturers needed motor mechanics, today they need mechatronics and software engineers. What are the mechatronic engineers in supversion in five, ten or 15 years? What do we have to do today to have such skills available when we need them? At BaFin, we have been thinking about the primary competencies we will need in the future. Analytical skills, for example.

We are developing human resources policy measures to establish and strengthen these skills. That applies equally to our hiring practices and to training and support for our existing staff.

But it is not just digitalisation that is changing the professional profile of the supervisors. Another example, as mentioned before, is the increasing Europeanisation of regulation and supervision, the growing importance of working in international bodies. The Single Supervisory Mechanism, which will turn five in a couple of days, for the first time shifted supervisory responsibilities to the European level, which has significantly changed day-to-day work in banking supervision. The supervisors of today and tomorrow not only have to be technically up-to-date, they must also be capable of operating shrewdly and confidently in the international arena.

3) Which business model will dominate in the future?

Digitalisation is making the financial markets attractive not only for small, innovative fintechs, but also for large online platforms such as Apple, Google, Amazon and Alibaba. Some of these “bigtechs” are already offering payment services in Europe. For instance, Google launched its Google Pay payment service in Germany in summer 2018. Just a few months later, in December 2018, Apple entered the German market with its own smartphone-based payment service offering. These global technology giants are still largely holding back from the traditional banking business. But the picture already looks very different in the USA. And hundreds of millions of customers in China are already using bigtechs to handle their banking transactions and services. Just think of Tencent, the parent group of the WeChat messenger service. When Tencent decided to establish its own bank a few years ago, the institution had several million customers spread across the entire People’s Republic in just a few weeks.

I agree that the Chinese banking market is not comparable with our own. But I do wonder what will happen if the bigtechs get excited about the European banking market. Would we then have to subject such tech giants to our supervision? Certainly not the companies as a whole, but we might have to supervise specific activities and their influences on financial markets. This is not as revolutionary as it might sound at first, as it is something we have been doing for a long time in the field of securities supervision.

Take the example of industrial companies whose shares are admitted to trading on the regulated market. We do not supervise them as companies. At the same time, though, as issuers of securities they are subject to various conduct obligtations like ad hoc and other disclosure requirements. In other words: we are already very much used to supervising specific activities as.opposed to whole entities and I am convinced that it is time to consider expanding the regulatory toolbox along those lines within the context of digitalisation as well.

But there is more to it than that: data management monopolies will emerge in addition to platform operators, and we will then face new systemic questions. If and to the extent new dependencies and hence new risks for financial institutions may emerge , we will have to constantly review and adapt our supervisory toolbox, and possibly, as mentioned before, the scope of supervision altogether.

Next to prudential considerations there are also far reaching challenges from a consumer protection point of view – which is also part of BaFin`s regulatory scope of responsibility.

However, the industry should also have a vested interest in properly handling customer data. If consumers as well as corporate customers start losing trust in the financial institution´s abilitly to manage their most intimate data with integrity and sincerity, that could ultimately damage the financial markets as a whole.

4) Who are we supervising?

Another pretty fundamential question is: who will we actually be supervising in the future? People, machines, or both? Insurers, for instance, are now already generally in a position to manage processes such as risk assessments for new business and claims processing, in particular for standardised cases, but increasingly also for more complex ones, without human intervention. It is obvious that it will also be possible to delegate entirely different processes and quite important decisions to computers in the future. But what happens if something goes wrong and errors are made? Can a board member say “It wasn’t us, it was the algorithm”? I say: No! Ultimate responsibility must remain with management, meaning people. That is also why we cannot accept models or algorithms if these are presented to us as black boxes.

Rather, we will do everything in our power to sustain the principle of human responsibility, even and particularly when innovative technologies and solutions are used.

5) Timing

The last aspect I want to highlight relates to the question: How long will supervisory approvals be valid in future if artificial intelligence and machine learning increasingly determine the pace of events in the financial world as well? Let's explore, to pick an example what this would mean in the context of the models for determining regulatory capital requirements that are subject to approval. Self-learning elements that continuously evolve may change the functioning of a model that we have just approved into something quite different, even while the ink of our signature on the approval letter is still drying.

We will therefore have no choice but to re-define principles we can use to decide whether a modification already constitutes a model change in the supervisory sense, in a world dominated by Artificial Intelligence.

And we have to define how much a model has to change before a bank or insurer has to get straight back to us again. It will not be easy to find answers to these types of questions. At a very general level we mayl have to develop assessment procedures in the future that focus more on the outputs or output corridors of modelas rather than their design on input factors.

5) Conclusion

Ladies and Gentlemen,

Of course, the topic of “The Future of Financial Supervision” is such a broad field that I cannot, in such a short time, address all of the challenges that are facing us. But the big picture is clear. A strong and effective supervisor that intends to continue successfully fulfilling its mandate of safeguarding financial stability must become more digital, faster and more foreward looking, more international and more networked. That’s why we are here. And: The MFSA and BaFin have a distinct advantage: both of us are integrated authorities.

I believe, we can identify new risks and emerging interconnectedness, and assess their impact on the various sectors, faster than single-sector supervisors.

Combined in a single authority, various divisions can provide information to each other without delay, coordinate their policies and act quickly. Additionally, a growing number of phenomena are extending across several sectors of the financial market. These include digitalisation, of course, as well as money laundering problems, Brexit and many more. I fundamentally believe that the holistic approach of an integrated authority is a good thing.

And that brings us back to Game of Thrones, although I have to broadcast a spoiler alert at this point. If you follow this series, but haven’t yet seen the final season, please close your ears. Following countless battles and even more intrigues, it is not the most powerful lord or the most scheming lady who is sitting on the coveted throne, but Bran Stark. He is a rather puny young man, but he understands the past and can look into the future. Qualities that would also stand a supervisor in good stead. So if Bran Stark gets tired of being king, he could comfortably supervise the Iron Bank of Braavos, which plays an important role in the storyline. Of course, we do not have the special talents of a fictional series hero. But if we adopt a goal-driven approach to our responsibilities, we don’t actually need them.

In particular thanks to their integrated structure, I believe that both of our supervisory authorities are very well positioned to master digitalisation and other challenges.