Uncorrected oral evidence: Regulatory sandboxes in financial services
Oral evidence, 21 October 2026.
The Chair
Welcome to today’s meeting, which is the first oral session as part of the committee’s inquiry into regulatory sandboxes. Thank you to Professor Allen and Professor Zachariadis for attending. The session is open to the public. It is broadcast live and is subsequently accessible via the parliamentary website. A verbatim transcript will be taken of the evidence and will be put on the parliamentary website. Before we get going, I need to declare my interests, which are that I hold investments in a number of listed financial services companies that could be affected by regulatory sandboxes. I invite colleagues, the first time that they speak in this inquiry, to declare any interests themselves. Could I ask each of you to introduce yourselves? We will be going directly into questions after that. Professor Zachariadis, could you just introduce yourself to the committee for the record?
Professor Markos Zachariadis
Yes. It is a pleasure to be here. I am an academic. I work at the University of Manchester at the moment. My full title is professor in financial technology and information systems, but my background by training is in the economics of innovation. I did my PhD at the London School of Economics a couple of decades ago. Since then, I have worked at different universities, including the University of Cambridge, Warwick Business School, and now the University of Manchester at the Alliance Manchester Business School. In terms of my research, I should also give you a bit of a perspective on why I am doing what I am doing. I have always researched the economics of innovation, and of data innovation in particular, within financial services. That is the single thing that I have been looking at. Within that, we aim to understand a bit more the role of technology in changing business models and operational models at the firm level, the overall market architecture at the industry level, and the benefits that we get in terms of productivity, but also consumer welfare and so forth, at the economy level. I have done multiple studies that have looked at many different technologies. Regulatory sandboxes are one of the devices that I have been looking at carefully, because they are very closely aligned with technological development in the sector. We have a minor research paper out called “FinTech Regulatory Sandboxes: When Regulation Meets Innovation”, which touches upon many of the questions that have been shared in advance of this particular hearing. That is my introduction.
The Chair
Thank you. Professor Allen, could you just introduce yourself for the committee?
Professor Hilary Allen
Yes, sure. I very much appreciate the opportunity to be here. I am a professor of law at the American University Washington College of Law in the United States. I have been looking at financial regulation globally for my entire academic career, with a particular focus in the last decade on the regulation of fintech. As part of that, I have authored four papers on regulatory sandboxes, which I believe you have access to, and I look forward to your questions.
The Chair
Thank you very much. If we could start with a general background question, could each of you outline what you see as the role of a regulatory sandbox in the context of financial services? What are they for?
Professor Markos Zachariadis
One could put a definition on them, even though I have to say that definitions have been evolving, because the role of the different sandboxes that we have observed around the world has changed. Even in some cases, their initial intention has been quite different. By and large, a regulatory sandbox is a controlled environment in which firms, particularly financial technology firms, but also incumbent financial institutions and payment infrastructures, can test innovative financial products, services, technologies, or even business models with near real or simulated users, under the supervision of financial regulators. That is the broader definition, if you like. I mentioned that there are different kinds of sandboxes. I do not know whether you want me to go into those now.
The Chair
Yes, why not.
Professor Markos Zachariadis
One of the reasons that we did the paper that we co-authored with colleagues from the University of Manchester at that time is that regulatory sandboxes tend to do two things as devices. The first is to act as learning devices, which are then used by regulators, but also by the firms that participated within the sandbox, to learn more about the regulation as well as how to regulate. They also act as regulatory devices. A stricter definition of the purpose or aim of the regulatory sandbox is, basically, to make sure that the firms that are part of the regulatory sandbox cohort comply with the existing regulations. It is about making sure that there is control and that existing regulations are being imposed on newly founded firms, which often do not necessarily have enough knowledge to do so. There is a lot more to it, but perhaps I can pass on to Professor Allen.
Professor Hilary Allen
I agree that these regulatory sandboxes vary from jurisdiction to jurisdiction. They have been adopted all over the world at this point, following the FCA’s lead. There is an implicit justification that is rarely stated for them, but is very much an important part of the appeal, which is as a signalling device. It is supposed to signal that a jurisdiction is, in essence, open to fintech business. There is that signalling aspect, but the more stated goals—and this is a riff on what you have just heard—are, first, an opportunity for regulatory learning about new technologies, and, secondly, a way of reducing regulatory barriers to entry in order to allow fintech firms to thrive. Those are the two more common stated justifications for adopting sandboxes.
The Chair
Excellent, that is a good background.
Lord Sharkey
I am interested in how the effects of these sandboxes are to be evaluated and measured. I am particularly interested in the absence of any counterfactuals. How is it that the operators decide whether the outcomes are beneficial? When it comes to outcomes, I am also interested in how they are divided. Who typically benefits from these? There is also the question of whether these things are intrinsically unfair.
Professor Hilary Allen
I am happy to take this question because it segues into a lot of the concerns that I have over the business model. I should say that I have been looking at these business models since, basically, their inception. My concerns about them have deepened over time, for the reasons that you have articulated. Basically, when we are judging the these sandboxes, the empirical literature on this generally focuses on whether the innovators have been able to access financing and to patent things. If you look at the FCA’s early report card for its sandbox, it was really also focused on how the innovators are experiencing these sandboxes. I do not think that that should be our metric of judging the success of these sandboxes. We should be judging them, first of all, on whether that innovation is broadly socially beneficial. There is a broad conversation to be had about whether fintech innovation is beneficial for the public at least. I should concede here that most of my experience on this issue is restricted to the United States, but most of the fintech business models that we see are really more about avoiding regulation as their secret sauce rather than technological innovation. A lot of those products are quite predatory in many respects, so I do not think that we should assume that fintech innovation is necessarily socially beneficial. As I said up front, the other thing that these are supposed to do is to inform regulatory learning. There are a number of reasons why that goal is circumscribed. One is that sandboxes put regulators in a really unusual position of being a cheerleader, in essence, for the firms that they have picked to participate in the sandbox. First of all, that is conducive to cognitive capture. It is not surprising if regulators come to identify more with their firms than anything else. It also prevents important regulatory knowledge sharing among other jurisdictions, because there is this sense of wanting “our sandbox firm” to do better than other sandbox firms. I could go on, but the things that we should be judging these sandboxes on are how they are improving regulatory learning and at what cost, because they are very expensive to deploy, as well as whether the public is benefiting from this fintech innovation. Instead, they are largely judged from the perspective of the innovators themselves.
Professor Markos Zachariadis
I would agree. If you look at the FCA’s original sandbox, which was launched a few years back, in 2016, the aims were to reduce the time and cost of bringing innovation and innovative financial products to market, to facilitate access to finance for innovative firms, which is very important for innovation to flourish, and to enable more products. Perhaps the most important pillar for that would be the last one, which would be to improve consumer outcomes by encouraging beneficial innovation and competition. Sadly, for the last pillar, we do not have enough information that this has been done at least in the way that people thought that it would. We do have some information. There is existing research. There is one paper in particular that was done by some colleagues at the Bank for International Settlements and published in a finance journal a couple of years ago. Sadly, we wanted to do something very similar, but they beat us to it. That paper showed that firms that participated within the regulatory sandbox enjoyed better access to finance because of the signal effect that Professor Allen mentioned. These firms had adequate access to how to meet regulatory standards and comply with regulation. The fact that they were part of the regulatory sandbox signalled to investors that they could be more confident with their investment. There is a bit of correlation, although they had access only to the information that the FCA published online, which included the fintechs that took part in the regulatory sandbox, and not the ones that applied but did not get accepted. There is a sample selection bias, which, as per the economics literature, can skew your results. From the way that they did the research and similar pieces around the world in different sandboxes, it is fair to say that participating in the regulatory sandbox does increase access to finance for those firms, hence the innovative solutions being out there a bit quicker. Of course, consumer welfare and benefits for consumers have not been assessed adequately. As an economist, but also as a researcher, it is quite a tricky thing to do. Nevertheless, that does not mean that we should not take future steps in trying to assess that, not just for regulatory sandboxes but for fintech innovation in general and those regulations that seem to be a bit more relaxed for new and innovative solutions coming to the market.
Lord Sharkey
I am just curious as to why there are no agreed metrics for measuring the success of these things, which have been around for a long time.
Professor Markos Zachariadis
There are metrics. Competition, for example, is one thing that one can try to assess. Particularly when it comes to the fintech space, we are trying to do something similar, not for the regulatory sandboxes but for regulations that encourage more innovation, such as open banking, open finance, or open data within financial services. There is a lack of data that we can access to try to establish that. The FCA sits on this data and I have urged it before to share a bit more with us in a safe way, so that we can do some of the independent studies that would try to measure these things. That has not happened yet. We are constantly in conversation with the FCA and would love to collaborate with it in order to do this research and assessment.
Lord Sharkey
I imagine that these processes are quite expensive. Does that not create an obvious bias against very small firms wanting to take advantage of them?
Professor Markos Zachariadis
Running the regulatory sandbox is expensive. For those who take part, there is the potential for some residual benefits, but, for those who do not, that could be unfair or skewed to the benefit of the ones who do. That is one argument that people have mentioned before, but I am not sure it is strong enough to say that we should not have any regulatory sandbox at all. It might be a good argument to tell the regulator to perhaps be more transparent in the way that it then publishes the benefits for those firms that take part in the regulatory sandbox, so that other firms can learn from it. There is definitely room for improvement, but not necessarily to justify that there should not be any regulatory sandbox at all. That would be a bit far-fetched.
Lord Hollick
AI is becoming increasingly important in helping consumers understand financial products, find and locate suppliers, and possibly make a choice between one supplier and another. To what extent can sandboxes help to decide on how best that should be regulated? You made a comment, Professor Allen, about avoiding regulation. It is fair to say that AI is rather shy of being regulated, so how does the sandbox help us to solve that problem?
Professor Hilary Allen
One of the challenges with a sandbox is that it is supposed to be a temporary thing. There is a period within which there is less risk of enforcement and some of the regulatory barriers to entry are brought down. That is supposed to be a finite period for the participants to learn how to operate in a regulated environment, and for the regulators to learn how to regulate going forward. That is the idea. Some experience with the Canadian securities regulators is very important to acknowledge here. The challenge is that often, when you get to the end of that temporary period, the firm is in no better position to comply with regulation than it was before. The regulator is then faced with a very difficult decision: “Do we extend indefinitely the exemptions that we gave or do we shut the business down?” You probably will not be surprised to hear that what happened in Canada, with a lot of the dispensations that were given to crypto firms, is that they were allowed to continue on their exemptive basis indefinitely, because the regulators did not want to shut them down. That is problematic for a number of reasons, one of which is that, because the conditions within a sandbox are bespoke to the individual firm, you have different firms operating in different legal environments in perpetuity. Generally speaking, I am very concerned about the sandbox being the way in which AI regulation is formulated. I would say that you can live without sandboxes quite easily, and I think you should, for a number of reasons. First, when you have a sandbox out there, what it says to firms is, “Don’t bother trying to comply with the laws. We have acknowledged that the laws aren’t right for you. Come on in; the water’s fine”. If the decision is that the laws are not fit for purpose, that can be done through waivers and things that apply across the board, and then you avoid the issue of picking winners. I am very concerned about sandboxes being the space in which AI regulation is negotiated, in large part because there is no public participation in that space. This is a bespoke negotiation, in essence, between the firm and the regulator. The firm often touts its superior technological expertise and says, “You don’t really know what’s going on here. Let me tell you how we should be regulated”, and the public has no voice in that process. This is a phenomenon that I am quite concerned about.
Lord Hollick
The example that you have given in Canada is that, in essence, something that possibly should be regulated is, in fact, being whitewashed, because it is allowed to continue.
Professor Hilary Allen
Yes, because it is very difficult to put regulators in a position where, in essence, they have to perform as executioner for a business model that is not able to comply with the full panoply of regulations, and it is not surprising that they chose not to perform that role as executioner.
Lord Hollick
Do you share those concerns, Professor Zachariadis?
Professor Markos Zachariadis
I do. They are all valid. Having said that, there is still a way for regulatory sandboxes to exist within that context. One can develop certain safeguards, but also a framework that will ensure that the expectations for having relaxed regulations in perpetuity, for example, would not exist. It would also make sure that there is the right environment for organisations to test their products, but within existing regulatory frameworks, so that they do not have issues or challenges in complying with future regulation when they take their products to market. There are a few things that one can do to accommodate that. My concern is mostly around the nature of artificial intelligence as a technology that is qualitatively quite different from other innovations. Regulatory sandboxes tend to be quite contextual, so you try your products based on a very particular number of users. Artificial intelligence, particularly GenAI, relies more on open systems, where a lot of data comes from different sources, basically, which cannot necessarily be controlled within the regulatory sandbox environment. In that case, maybe different techniques should be implemented to be able to understand how AI can be utilised within financial services in a more rigorous but also safer way.
Lord Hill of Oareford
I ought to declare an interest as chairman of ICE Futures Europe and an adviser to Santander and Visa Europe, which could, theoretically, pop into a sandbox one day. In 2016, we start with the FCA doing the first sandbox, and then a number of jurisdictions have followed suit. Do you see any patterns or trends in different jurisdictions across the world that are worth bringing to our attention?
Professor Hilary Allen
Something that I already mentioned was that, implicitly, a lot of these jurisdictions have been adopting regulatory sandboxes as a signalling device to say, “We are welcoming fintech businesses”. A challenge with that is that, as more jurisdictions opt into it, that delivers diminishing marginal returns. Some of the reporting that has been done at the international level suggests a little bit of buyer’s remorse. While this is not true of the UK, there was, in many jurisdictions, a failure to appreciate how expensive these regulatory sandboxes are to adopt. It was thought of as, “We can just say that we will have a sandbox, and that will be enough to attract new business”, but they are, in fact, quite expensive to implement, because the “je ne sais quoi” of these is a lot of regulatory handholding. Some of the research that has been done says that, if you do not have that regulatory handholding, the firms are less likely to succeed. It is not just the dispensations from law that help, but that guidance, which is expensive. We have definitely seen some buyer’s remorse. We have definitely seen diminishing marginal returns as more and more jurisdictions have tried to deploy sandboxes. The UK, as first mover in this space, is somewhat immune from those concerns, but fintech sandboxes had lost some of their lustre around the world. The previous question was about AI, and we are seeing a resurgence of interest in regulatory sandboxes being the solution for many of the AI issues, so it is time for a second round in that regard.
Professor Markos Zachariadis
What we have witnessed is, in some cases, the development of regulatory sandboxes, particularly towards giving access to data. A natural following device would be the digital sandboxes that even the FCA has experimented with. That is giving access to consumer data coming from synthetic databases. That is one of the trends that we have identified, and that has been used as a vehicle by different regulators around the world as well. The other thing is the openness and realisation that, of course, the financial system is an international system by default. You cannot really isolate a particular economy or system by itself, because there are a lot of international markets that big players within a particular economy want to be part of, such as financial markets and payment infrastructures. We have seen a shift towards more international communication and trying to develop standards and innovations that are, basically, on a par with other countries and financial markets, and that dialogue is a very positive development. Some of these have managed to creep into regulatory sandboxes and become a bit more dominant as solutions or discussions.
Lord Hill of Oareford
I have a specific question for Professor Allen. You obviously do not like sandboxes, but what is your view of how one should—or perhaps you do not think that one should—encourage regulatory innovation? How does one get regulators to think differently about their job? Do you think that there are better ways of doing it than the sandbox?
Professor Hilary Allen
I do, yes. A lot of the research has shown that, even if you are just trying to prioritise the innovating firms, the benefits come from that bespoke guidance rather than the regulatory dispensations. You can avoid much of the deregulatory effect and still preserve a lot of the regulatory learning by having regulators simply offer guidance to people. A term that you often hear, and one that the FCA uses, is “innovation hub”. That kind of format, rather than a sandbox, achieves a lot of the goals without those negative possibilities of having firms locked in perpetuity into less regulated environments. It may be that, through this process, regulators discover that there are regulations that do not work for a particular business model. I do not think that you should assume that all business models are valuable, but, if they do decide that that business model is so valuable that it should be allowed to flourish, and that there is some regulation that needs to be pared back in order to let that happen, they can issue a broad-based waiver rather than having bespoke waivers that favour the firms that are able to get into the sandbox. I should note that a real challenge or problem with these sandboxes is that there are rarely any formal criteria for who will be accepted into the sandbox. You might have some high-level thoughts or criteria expressed, but, generally speaking, it is quite discretionary. Given that, as we have seen from the empirical research, admission to these sandboxes is very beneficial for the firms themselves and, therefore, exclusionary for firms that are not selected, broad-based waiver criteria are the way to go if it is concluded that regulation needs to be changed. As a regulator, you can figure out whether that regulation needs to be changed from your conversations with industry through an innovation hub.
Lord Turnbull
What I am missing here are success stories. It has been around a decade. In that decade, has there been anyone who went into a sandbox and developed something that was very successful—lots of other people copied it; the regulator thought that some of the changes that they made were worth while—and who can say, “Look at me now. I’m a household name”? Where are these people who have been through this process, benefited from it, and demonstrated this idea that, by starting small and experimentally, you can go on to greater things? I have not heard of any case where we can now say, “Do you remember them 10 years ago? They were just a tiddly little company, and now they are this”. Are there any?
Professor Markos Zachariadis
To be fully transparent, in the research that we have done, we have not focused on one particular case study that went through the regulatory sandbox and observed it over time. My understanding is that, even if you look at only the UK FCA regulatory sandbox, there are quite a few big fintechs that are now quite reputable and that a lot of people use. Whether that can be attributed to them going through the regulatory sandbox is the question. If we knew 100% for sure that that can be attributed to their experience that they received through the regulatory sandbox, and that the FCA perhaps changed or adapted regulations over time to be able to encourage this innovation, we would be much more confident and positive about the sandboxes, but that is a big question. How do we measure that? What kind of information and data are out there and can be used as research to prove that correlation and relationship?
Lord Turnbull
Professor Allen, are you aware of anyone who says, “My first great break was the sandbox, and it went on from there”?
Professor Hilary Allen
I am not. Quite a number of firms using the FCA sandbox are already established financial institutions, so it is not just new start-ups that are participating, but a lot of established financial institutions as well. I want to also pick up on one of the points already raised. Sandboxes are challenging in an environment where cross-border operations are really the sine qua non of financial services. There have been attempts to create cross-border fintech sandboxes under the auspices of the Global Financial Innovation Network—GFIN—of which the FCA is a part. It tried, once, to run a global sandbox. One or two firms got through. Ultimately, the challenge was harmonisation. Even within a global sandbox, you had to get multiple regulators to sign off on the firm and the conditions under which it would operate, which is just hard. That is the same challenge that you face with any kind of firm trying to operate cross-border, so I do not think that regulatory sandboxes really alleviate that challenge. There are limitations to what these sandboxes can achieve.
Baroness Bowles of Berkhamsted
I started off with a triangle of concern. It is now an octagon. You have dealt with or mentioned some of the issues. Initially, my biggest concern is a lack of transparency, which, in a sense, you have both referenced, and then a lack of a public position in this, to which I could add a lack of parliamentary position. In this untransparent process, another is law modification, whether that comes about by individual or blanket waivers, or by law changes, possibly without going through the usual channels. A broadly cast waiver is one way of changing the law, if you can set it aside. How do we deal with this? Transparency is meant to be disinfectant. How can we cope with this? What modifications can we make while not embedding model preferences? You mentioned patents in some jurisdictions. We do not have business model patents in Europe, but there could be other IP issues that are then embedded, in essence, in legislation, which just seems to end with an oxymoron. What would be your solutions?
Professor Markos Zachariadis
It is for the FCA to think about this and to establish a framework of transparency that can share more information about the process and the lessons learned. I was looking for a report last night, and the only thing that I could find from the FCA was the lessons learned report that is already a decade or so old. There needs to be a more systematic sharing of information and insight as to what is happening within the sandbox and what things other firms within the UK ecosystem, being a fintech or an incumbent financial institution that wants to experiment with new innovation, can learn from. That is one thing. I am not sure how far you can go. I understand the concept of transparency, but I also respect what you mentioned around IP. There are a few things that the regulator cannot necessarily share because they relate, for example, to consumer behaviour or particular consumer information. These are data and information that cannot necessarily be shared with the public, even though we, as researchers, would like to have more access to data to be able to assess and understand the effectiveness of many of the things that the FCA does, or that regulators and policymakers put on the table, including existing regulations such as open banking. Again, there might be a way to try to establish more transparency within that context.
Professor Hilary Allen
If you will humour me, I might answer your question at a 30,000-feet—sorry, I should use the metric system; at a very high level. So much of what we are seeing right now is really a response to a societal conviction that we should never do anything to impede innovation. That is a paradox in many regards, because we are told that innovation is inevitable, and that there is nothing that regulation can do about it, but also that innovation is so valuable and fragile that regulation should never take any steps to impede it. That is the ecosystem in which you have agencies such as the FCA trying to figure out what to do about fintech innovation. It is this real fear of never being, in any way, an impediment to that innovation, and so you get the result that you have articulated, where you have things being done informally and quickly in order to always accommodate innovation. That often peels away laws that were passed, often by Parliament, and that had a decidedly harm-protective function. There is a very big conversation to be had, which goes far beyond sandboxes, about the degree to which long-standing legal precedents established by Parliament and the courts should be peeled away in order to accommodate innovation. If you have that reflexive fear of doing anything that could get in the way of innovation—to be clear, not all innovation is good, or good for the public—you are going to end up with suboptimal outcomes.
Lord Lilley
I am still suffering from a lack of concrete examples of sandboxes, how they have been used, and what the outcome is. I hear two slightly different things. One is that they are there to help the regulators learn how to regulate some innovative process. Both today and in our background research, we have heard that regulators are using it to learn how to regulate this new thing. That seems to be broken down into two things. First, should they change the rules to make it possible if there is an infringement or a conflict between this new process and the existing rules? The second is this business about learning how to regulate some new thing. That is a bit odd to me, and I would like someone to explain it a bit more. What do they have to learn other than what the rules should be? The second thing that we have had referenced today and previously is that it is helping new businesses comply. That seems to me to be more of a concierge service. In our earlier research, we said that regulators should not just say, “We’re not there to help. Those are the rules. If you break them, we will put you in jail” or the equivalent, but that they should say, “These rules are for this purpose. If you do this, it will comply with them. If you do that, it won’t”. That does not seem to me to need a sandbox. That is a concierge. Can you guide me?
Professor Markos Zachariadis
That concierge that you mentioned was also mentioned before with the innovation hubs. The FCA calls them innovation pathways. There is a type of concierge service, in some ways, for new ideas to go there and approach particular agents that the FCA would then assign to that organisation. They would then be in discussion on how that particular innovation can comply. That is one thing to do. The regulatory sandbox is probably a bit more systematic in the way that the FCA will use the sandbox as a learning device, which is what you referred to before, to understand a bit more how new technological developments might affect financial services overall through products and services, but also through modifications in the overall industry architecture, which is something that one of the latest papers pointed out. We have seen a gradual shift to a new kind of architecture within financial services that we did not have before, which is more modular, because of new technologies such as application programming interfaces and communication of data between agents, but also institutions. These are deeper discussions or reflections that the FCA develops through conversations that it has with innovative firms, fintechs and other institutions that want to innovate and experiment with new technologies. Over time, it develops new ideas that would then potentially lead to new regulation. That does not happen automatically. It is not that, through a particular cohort in the regulatory sandbox, the FCA says, “We are now going to do this”, but it gradually builds the intelligence through this dialogue that it is establishing with the fintechs or the organisations that are part of the sandbox in a systematic way that helps them reflect. As a regulatory device, which was mentioned before in the categorisation that we had in our paper, it also helps fintechs or other organisations comply. Some of the regulations are not necessarily prescriptive, where you do one, two, three, and then you are done, but more descriptive. New technologies might be closer to complying with a set of rules or a bit further out, so the FCA helps them or nudges them into a particular direction. It goes without saying that a regulatory sandbox is not always the most effective mechanism that one should adopt. In the last few years, we have seen the introduction of other mechanisms, which is something that we have not discussed. We have seen regulatory tech sprints, for instance, which the FCA has used. Another is challenges. The regulator might come up with a challenge and invite industry and universities to collaborate to come up with a solution. These are also very effective mechanisms for a regulator to get that kind of intelligence and understand future use cases and how new technologies can be used, as well as business models and ideas that it can get from industry and universities. I had the opportunity to participate in but also assess, through the Bank for International Settlements, where I worked as a consultant for a while, some of these challenges. These could also be effective mechanisms. We are not saying that regulatory sandboxes are completely rubbish, but there are vehicles and mechanisms that regulators can use.
Professor Hilary Allen
I have two reflections on your comment. The first is with regard to concierge services. Providing concierge services to firms in a world of plentiful resources is a good thing to do, but it is often a resource question. There should be more interrogation of whether this is the best use of regulatory funding and whether the public could benefit more from a different allocation of resources. Secondly, I wanted to respond to your question about what regulators need to learn about their laws in the context of new technological innovations. The issue here is application. Good law is flexible and can accommodate changes in the world, but it is sometimes not entirely clear how pre-existing laws will apply to a new technology. That is often interpreted as a reason to get rid of the pre-existing laws, which is often a mistake. Law is often developed through analogy, and what the regulators can learn from this is where the analogies lie between what they are used to regulating and how this has changed. If you will humour me once again, there is a broader issue here about the degree of legal certainty that innovators tend to demand. They think that they should be provided an absolutely clear roadmap of what the law is. That is unrealistic. Law has always been susceptible to interpretation and evolution, and there have been plenty of examples of innovation that have worked quite well and flourished in an environment of some legal uncertainty. In fact, there has been a lot of innovation that has flourished in response to legal requirements. In other words, “The law says you have to do this. You figure out the way to do it with your new technology”. While there is scope for regulators to be speaking to firms to learn how to apply existing laws to their technologies, that should not automatically default to those laws being peeled away, waived or repealed in order to accommodate those new technologies.
Lord Griffiths of Fforestfach
I almost feel like being a little flippant, but, if, as Lord Lilley suggests, it may be a concierge service, my first instinct was, “Why don’t we privatise it?” That is being slightly flippant. You are both at the cutting edge of research and so on, and you both know quite a lot about the UK system. Is there anything that we can learn from what is happening in the rest of the world, particularly the US, that could improve our system?
Professor Hilary Allen
I do not think that there are necessarily good things for you to learn from the US approach. Sandboxes have never really taken off in the United States. The Consumer Financial Protection Bureau proposed one at one point, but it was really set up more as a deregulatory mechanism than anything else. That happened at the end of the first Trump Administration and was abandoned by the Biden Administration. We have seen individual states try to come up with fintech sandboxes. Arizona was the first. It has had very little uptake, in recognition of an issue that we have already talked about, which is that you want a big market for your services and, if you are restricted to operating within just one state, that is not a big enough market to really justify the investment. In that sense, there are no regulatory sandbox lessons to draw from the United States.
Lord Griffiths of Fforestfach
If I could just interrupt you there, what can we learn from the fact that sandboxes have not taken off in the US?
Professor Hilary Allen
That is a great question, because most of the funding in fintech is happening in the United States. What that tells you is that you do not need a sandbox. If your goal is to further the fintech industry, and I am not suggesting that it should be, fintech has flourished in the United States without regulatory sandboxes. This is where I was heading. There have been plenty of other accommodations, particularly under this second Trump Administration. All kinds of financial regulations are being rolled back to accommodate different fintech industries, particularly the crypto industries. I do not think that that is advisable, but, again, I would note that that investment in all these fintech businesses really was at its zenith during the Biden Administration, so I do not think that you necessarily have to provide these regulatory dispensations in order to encourage the growth of the industry.
Professor Markos Zachariadis
If regulatory sandboxes were just about increasing funding and entry to fintech organisations, yes, I would agree. It is not just about the fintechs, but also for the regulators to learn. Down the line, while it might be a slightly bigger investment in terms of putting together a regulatory sandbox, one could argue that, in theory at least, during the process of the sandbox, the knowledge that regulators pass on to fintech firms reduces the cost of future non-compliant firms. This clarity would reduce the workload of agents within the regulator who would have to deal with that particular firm in the future. More clarity and information, and reducing information asymmetries, would potentially be good. I do not have data that can assess whether regulatory sandboxes are a good investment. Looking at different countries, they have been funded through different means, such as the industry or the European Development Bank, or different sources that do not necessarily rely on the regulators, so it was a much easier idea to pitch. I will just revert to what lessons can be learned from different countries. One could perhaps spin that around and highlight a few things that can be done to improve regulatory sandboxes in the UK, and I made a short list here. One thing that we touched upon before was transparency. More transparency would be very beneficial, so that we understand a bit more the reflections from the regulators through the regulatory sandbox. Other firms that do not participate at that moment in time can also benefit from more transparency in terms of the outcomes and how they benefit consumers in the end, which is the ultimate aim of any regulator. In that context, independent evaluators such as academics need access to more data, means and resources to be able to assess those aims and whether they have been met. One thing that was alluded to through the discussion around AI would also potentially be sandboxes targeted at a particular technology, or more thematic sandboxes on particular use cases that benefit the public, for instance. That could be another way forward to improve the existing regulatory sandbox model. I do not know whether that helps, but those are some extra thoughts.
Lord Davies of Brixton
Thank you for all this. People need financial services, which are, by and large, provided through a market, but it is quite a strange market because of the inevitable and massive asymmetry of information between those providing the service and those receiving it. Because of that, we have a regulator. If it was not for the asymmetry, perhaps we would not need a regulator, but we do. It seems to me that, from our discussions on sandboxes, it is perceived as a service to providers. They have a product, they want to produce it, and they want authorisation. Financial services regulators tend to end up seeing their clients as the providers, whereas, in truth, their clients should be the consumers. Is there a way of recasting sandboxes with the explicit aim of providing more information and more substance to consumers’ interests rather than those of the providers?
Professor Hilary Allen
To be clear, the FCA has always said that one of the goals of its sandbox is to promote financial inclusion by creating more financial services products for traditionally excluded consumers, so there has definitely always been a consumer benefit justification for the sandbox. As we have seen and have discussed, it is very hard to quantify that and to see whether that is really what is happening. You end up with something that appears very much to benefit the innovators, and then assumptions that that innovation will trickle down benefits to consumers. I am not sure that there is a way of reorienting sandboxes to be truly more consumer-focused. If you think of the whole goal of the sandbox as being to work with the firms themselves, that is an environment in which you would naturally expect the regulator to develop a relationship with the providers. As you said, that is what tends to happen. Those risks of what we call cognitive capture, or identifying with the industry, are much higher in a sandbox where, as a regulator, you have, in essence, chosen a product that you want to see succeed, and so you become a cheerleader for that product and, ultimately, that is how you measure the success. I do not know how you could make a sandbox more explicitly consumer friendly, largely for the reasons that we have discussed. How do you quantify the benefits of that regulation for the public? Again, I speak more from the US perspective here, but we have seen that many of the things that have been billed as democratising finance or promoting financial inclusion have been quite predatory in the United States. Extending credit to people who already were overextended, or facilitating gambling by people who did not have money to lose, make up a lot of the financial products that are being marketed as financial innovation, so I am sceptical on that front.
Professor Markos Zachariadis
I will just point to what I mentioned before about transparency. You are right. It is a bit misleading to look at a regulatory sandbox and think that this is a service just for the providers that take part in it. We should see it largely as a way for the regulators to understand a bit more about innovation and technology, new developments in terms of business models, and how these can, in the end, benefit consumers effectively. That is the missing link there, to make it a bit more obvious.
Lord Vaux of Harrowden
I should start by declaring an interest. I hold shares in Fidelity National Information Services, Inc., which provides technology to financial services companies around the world. I am struck a bit by a contradiction in what I am hearing here. I am hearing that there is very little evidence about how effective sandboxes are in terms of innovation, competition, and even their impact on regulation, and yet, at the same time, we are seeing that the concept has been taken global. Most regulators now have sandboxes, but what has driven their proliferation, given that there is so little evidence as to their impact?
Professor Markos Zachariadis
I do not think that there is no benefit at all. In theory, you can look at things that go through sandboxes and think about the benefits that they then deliver to consumers, but also competition. Having more successful fintech organisations that predominantly go through sandboxes is a good thing in terms of competition, because they can offer cheaper, faster, and more consumer-friendly products that benefit customers. Overall, they impose more competitive pressure on incumbent organisations to then be more competitive and user friendly. We have seen that a lot in the UK through the fintech revolution that started about 15 years ago. You do not need to have academic evidence to look at your banking application and see how it has evolved over the years. I would like to think that this is largely through the pressure that a lot of the incumbent organisations felt from innovative firms such as Monzo and Revolut, which flourished because of the UK fintech innovation ecosystem. I would just say that regulatory sandboxes have not been successful everywhere. There are countries where a regulatory sandbox exists and is then shut down, because they could not really identify a concrete connection to further investment in fintech or the flourishing of innovation. One of these countries happens to be Greece, where I come from. They did a sandbox for a year. There was no resource to support it. There was no evidence that attributed anything to it in terms of fintech innovation and competition, so they shut it down because the money ran out. I am sure that, if we look at the data, there will be more countries that went through the process. Regulators look at one another. They see a regulator innovating, and the FCA is at the forefront. It did that regulatory sandbox a few years ago, and then others picked up on it. There is no doubt about that. That also happens in any innovative firm. That is why we see the flourishing of sandboxes. There is also benefit for the regulators, not just the organisations going through it. While we do not necessarily have a million papers assessing and proving that, using econometric models followed by undisputed results, one could, in theory, line up a lot of arguments for why a regulatory sandbox should exist, so I would not shut it down completely. I would not say that there is zero evidence, that they have done more bad than good, or that they ate the FCA’s money where the opportunity cost dictates that they should be doing something else. We are not entirely sure about that. I would not go as far as arguing that.
Professor Hilary Allen
I agree with a lot of that. Under Occam’s razor, the simplest answer is often the correct one. Regulators emulated each other, in large part because this seemed like an easy win-win. It was a way of signalling to the world that you were an innovative jurisdiction, but there was often no appreciation of how expensive it would be to operate, and so we saw some buyer’s remorse and some countries shutting them down. Other jurisdictions, I am sure, have found some things to like and keep about them.
Baroness Donaghy
I have no financial interest in this study. I am a graduate of English language and literature, so I have an interest in encouraging accessibility, particularly of language, and the elimination of exclusivity. It seems to me that this topic is at the pinnacle of exclusivity and opacity, and really a bit of a con. Like Lord Lilley, I am trying still to get to the bottom of it. Both of you have said that a bit more transparency and evidence from the regulators about this subject might be helpful. You might vary a bit on whether sandboxes are important or worth working on, but what happens if the emperor has no clothes, which is why the regulator may not be willing to give this so-called evidence? What happens if it is just a defence mechanism to impress politicians that the regulator is cutting-edge?
Professor Hilary Allen
As a fellow English literature graduate, I have used the emperor’s new clothes as an analogy many times when speaking about this sector. The key thing to realise about that story is that the reason why people go along with these things does not necessarily have to be venal. It is that crowd mentality and the desire not to seem stupid. There is no data supporting the very widespread narratives about many fintech products and their benefits. Once you start interrogating a lot of these products, it really does fall apart, but it requires people to be willing to go out on a limb and question the conventional orthodoxy. You alluded up front to including people. The narratives about financial inclusion, and democratising finance and accessibility, have been spun so thickly that, when you challenge these products, you are often accused, as I have been, of wanting to cut consumers off from choice and from the tools that they need to build a better future for themselves, and so trying to say that the emperor has no clothes is a very challenging position to be in. There are different versions of that story, but, if you go back to the one from Hans Christian Andersen, the courtiers never admitted that the emperor had no clothes, even once the child called it out, because it was just too hard to admit that they had been wrong for so long. There is a lot of truth in that parable, and I perhaps have gone very off topic, but I do not think that it is necessarily venal that the regulators are going along with this story.
Professor Markos Zachariadis
My answer might be shorter, but there is a discussion to be had. The FCA will have no reason to not supply further information and evidence if it is asked to. Going back to the overall discussion around innovation, particularly fintech innovation, I do not think that anybody is saying that there is no innovation out there or that it does not do anything for the public. I would not go as far as saying that, in the end, innovation is bad, but, across many industries, there are examples where what we think of as innovation did not have any positive effects for the public at all. For instance, social media is a multi-billion dollar industry right now, but I do not see any benefit that it brings to the public. It might be the other way around, with so many hours being lost in doomscrolling through social media that it reduces productivity. That is an example that one could think about. In fintech, I do not see any particular examples that could be so damaging, apart from, of course, creating a new monopoly and then skewing prices to the benefit of providers rather than consumers. The FCA would definitely be open to discussion and then to providing more data and transparency if it was asked to.
The Chair
I would like to thank both of you very much for coming to this first evidence session. You have set us up extremely well for the remaining stages of our inquiry, and many of your comments have been very thought-provoking. Thank you very much for joining us.