Bringing New Technologies to Financial Authorities

By Sean Creehan and Paul Tierno

In this episode, we continued our ongoing series on fintech in Asia by interviewing Simone di Castri, the Managing Director of the RegTech for Regulators Accelerator (R²A). The mission of R²A is to enhance the capacity of financial authorities by harnessing innovative technologies and accelerating promising RegTech (regulatory technology) solutions.

Simone guided us through some of the biggest challenges facing financial authorities, and what technological solutions R²A has been developing to help regulators in emerging markets. R²A has so far been focused on enhancing the capabilities of financial authorities in Mexico and the Philippines to better understand markets and customer needs in data-rich environments. Some of our key takeaways include:

  • A critical challenge in banking supervision is the time and resources required to manage and validate data collected from financial institutions; R²A is solving this using APIs that connect the central bank to financial institutions.
  • In the Philippines, R²A is prototyping a smartphone chat bot powered by machine learning that will enable any citizen to escalate a problem with their financial provider to the consumer protection team at the central bank.
  • Emerging and frontier markets are a natural fit for these new tools and techniques given the importance they place on financial inclusion and resource constraints that encourage innovation.
  • Fast-moving fintech actors have helped spur financial authorities to be more proactive in adopting new technologies in oversight and regulation.

Transcript

Hide this section

Sean Creehan:

Welcome to Pacific Exchanges, a podcast from the Federal Reserve Bank of San Francisco. I’m Sean Creehan.

Paul Tierno:

And I’m Paul Tierno. We’re analysts in the Country Analysis Unit, and our job is to monitor financial and economic developments in Asia. Today, we continue our series on financial technology in Asia.

Sean Creehan:

We spoke with Simone di Castri, a director at RegTech for Regulators Accelerator, R2A. Simone is an expert in RegTech, shorthand for Regulatory Technology. And we spoke with him about how his organization helps countries and regulators leverage technology to supervise financial institutions and promote financial inclusion.

Simone di Castri:

But then there is another solution that we’re prototyping with the central bank of the Philippines, which is a chatbot, that will work on any phone. Smart phone, feature phone, and will allow all Filipinos to escalate a complaint to the central bank. So this will be something that people in the Philippines will see and will use and will change their way to relate to their supervisor and regulator.

Paul Tierno:

Simone explained how R2A has partnered with regulators in Mexico and the Philippines to provide RegTech solutions to use the enormous data flow that comes with financial system expansion while ensuring services are open to still more of their citizens.

Sean Creehan:

Yeah and it was interesting to hear how regulators, who may be assumed to be a bit risk averse, have really embraced a lot of these new technologies. And he challenges the view that regulators are just reacting to financial technology startups, but instead are really experimenting with technology for their own purposes.

Alright, let’s get to the conversation with Simone.

Paul Tierno:

Thanks for being with us, Simone. Could you tell us about your work in the field of regtech? To start, what is regtech? And what is the focus of your current work at R2A? And what are your target markets?

Simone di Castri:

Hi thank you for having me. I am leading this new initiative called The Reg Tech For Regulators Accelerator or R2A. Which has been launched 20 months ago in October 2016 by USAID the Omidyar Network and the Gates Foundation. It’s implemented by BFA. With the aim to bring new technologies to the financial authorities that are involved with market, financial sector supervision and regulation, especially in emerging and frontier markets. So we work mostly in Asia, Latin America and Africa.

We focus on regtech for regulators, so, solutions to help improve the way markets are regulated, as well as Sup Tech, technology solutions that are for supervisors that engage in banking supervision, payments oversight, AML supervision and market conduct supervision. So there is a range of solutions that we have been considering and working on in the last 20 months.

The two core markets we are working right now are the Philippines and Mexico where we are testing 3 different prototypes. And when I say prototype it’s because we are trying to break some of the frontiers in this new field which is the work that several financial regulators are doing to bring machine learning and natural language processing and other quite sophisticated technologies that are available to create solutions for these financial authorities.

So we do proofs of concept. We do prototypes before we create what we do for production of these solutions.

Sean Creehan:

So what are the challenges that you’re trying to solve? Particularly in these developing countries?

Simone di Castri:

Some of the challenges are the traditional ones that are banking supervisors, or people working on payments oversight have been facing. The most critical one is the one around banking reports. The format that is used for the reports is mostly Excel files. And this creates new challenges for financial supervisors. And also the validation of the data reported by banks. It’s something that takes a lot of time and resources.

So that’s probably the main challenge in core banking supervision. But we’re also trying to resolve problems, for example around consumer protection and how to make consumer protection universal in the sense of giving all the people that use financial services an avenue to effectively escalate complaints to the central bank.

Or looking to new data that financial authorities can use to better understand the marketplace. For example, scraping data from social media, using sentimental analysis to understand better what are the needs and the usage of the needs of financial users. And others using financial services. This could help to identify problems right now in the mortgage sector or identify frauds on the schemes when people start talking about it ahead of time.

Right now they’re always quite late.

Sean Creehan:

So just off the first challenge you mentioned, a lot of data in Excel. Is this a matter of, it’s unstructured or in terms of conducting analysis of larger data sets it’s just an unsuitable format for these regulators. And you’re trying to help them with that kind of back end data collection, and then once you have it in the right format then you can help them with some of the more advanced techniques you mentioned earlier. Like maybe machine learning, artificial intelligence techniques to sort through all this information? Is that right?

Simone di Castri:

Absolutely. So Excel is the format that is most commonly used to collect data from banks and other deposit taking financial institutions around the world. And this is true for markets like the Philippines and Egypt, but also Canada, Norway. It’s a really common problem. And in addition, Excel was created to be a computational file, not to gather all this type of information from the financial sector. And the limitations are, in terms of management of the data, once they are collected, the validation. Because there are often too many duplicates. The banks need many different reports on a daily, weekly, monthly, quarterly, annual basis. And sometimes there are different departments in the banks reporting the same data to different departments in the central banks.

But there might be human mistakes or there might be some other reason. It takes so long to clean this data, to validate them, and then a big limitation is the capacity to run complex algorithms on Excel files. So there is a limited capacity that those formats can host. And sometimes we’re being, discussing with supervisors about the need to have a completely different type of tool to do that type of complex analysis.

So in the Philippines for example, we have substituted Excel with an API system that connects the central bank with the banks. And so now the banks are providing the data through the channel and through the APIs there is also the validation that happens in seconds and milliseconds instead of days. Or in some central banks actually it takes months to validate quarterly reports.

This transforms I think the way a central bank is working. Because now a supervisor who needs to decide where to put their resource capacity for risk based supervision can make decisions in a very timely manner with good data, instead of waiting for weeks or months to have the data to decide to understand where the risks are. And where there should be onsite supervision for example.

On top of this, the data are hosted in data lakes or in other systems that really give this capacity to run complex analysis. And then the last thing that’s important is to build on top of that, smart dashboards. Cause right now if you’re limited to using Excel files, you’re really visualizing data with the few charts and files and other parts of the organization that are parts of the office suite but if you think now of the quantity of data that are analyzed over time, it’s almost like you really need tools to let you see the financial sector in 3-D with heat maps and with other tools that are more sophisticated that would get us this far.

Paul Tierno:

It’s interesting because it sounds like some of the problems that regtech tries to solve are very complex but then some of them are a bit more basic and fundamental and at the root of banking supervision and regulation in general. On the other side of the equation you have Fin Techs that are expanding and advancing very rapidly. And I think that that often forces people to look at regtech and Sup Tech as constantly playing catch-up, versus being a bit more proactive in terms of shaping the sphere of the financial industry.

Can you talk about some of those competing interests? Right where you have the Fin techs that are moving very quickly, and the regulatory bodies that are seemingly a bit more reactive?

Simone di Castri:

Sure. So I believe that even if the regulators and supervisors were not feeling the pressure from the Fin Tech sector to innovate the way they do market or oversight regulation, we would nevertheless have the regtech and Sup tech. And that’s because there are these problems that are at the core of financial sector supervision.

Now the fact that Fin Tech is there makes the financial authorities a little more audacious because they see the private sector using all these technologies. And some regulators are, why we’re not? Why we’re not adopting the same technologies for our need, which is exactly the point of R2A way. We don’t build new technologies, we package technologies in a way that are useful for financial authorities.

So there will be probably regtech and Sup Tech even without fintech addressing some core problems of financial authorities. But I think it’s important that these two evolutions are happening somehow at the same time because the fact that fintech is there creates a sense of urgency to innovate within financial authorities.

In several countries the financial authorities are facing the challenge of licensing now over 800 new fintech actors. And that means that they need to resolve the problem with the core banking supervision but also being able to see their data in a different way. Granular data, real time supervision for some of these market players. Also I think the important thing is any better tools for supervision creates more comfort for financial authorities to open the gateways for fintech firms. So to create windows for fintech firms to experiment, to test like the sandboxes and others.

If you don’t have the right tools in place to supervise this new market, you’re readiness for fintech is at a different level. So regtech can really help to accelerate financial innovation.

Sean Creehan:

So our focus here in the country analysis unit is Asia. I wonder if you could talk specifically about the work that you’re doing in the Philippines central bank there and what kind of solutions you’re delivering. And also relatedly, I’m just wondering, for an individual that uses banking services in the Philippines, are they seeing some of these innovations? Or is it really just all on the back end? Is it changing any behavior of the banks because it’s more efficient? I’m curious as to how that is interpreted at the end of the chain.

Simone di Castri:

So the first solution that I described is the API system for bank reports. It’s quite on the back end on the side of the central bank itself. It’s hard for the consumer to see that type of innovation. They will benefit from that because the quality of supervision will definitely increase. It’s very important also for banks because that will really lower the cost of compliance on their side. Because to prepare all of the reports and to update the report in place once the regulation change.

Right now as soon as you want new data from a bank, as a regulator, as a supervisor, there is quite some time it takes for the bank to change the way they’re reporting to add this new data into their reports. Right now this will be almost immediate with the system.

But then there is another solution that we’re prototyping with the central bank of the Philippines, which is a chat bot, that will work on any phone. Smart phone, feature phone, and will allow all Filipinos to escalate a complaint to the central bank. So this will be something that people in the Philippines will see and will use and will change their way to relate to their supervisor and regulator.

So the idea of the chat bot comes because as we are pursuing financial inclusion, financial access all over the world, I think we need to be very committed to provide the records to all the customers that are included in the financial sector. And right now around the world the main avenues to access the central bank are emails or walk in if you have access to a branch, fax or letters or phone calls.

We saw that in the Philippines that even if people have all these channels available many Philippines are not reaching out to the central bank if they have a problem with their financial provider. And so what is available to all the Filipinos is a mobile phone. And so we decided to use that channel to make consumer protection universal in the country.

But at that point there was another problem. How do we create the capacity for the central bank all of a sudden to address effectively all the possible requests for support that we receive. And that’s where machine learning comes to play a very important role. Because with a chat bot you can build on the back end a system that starts working on the different queries that are submitted by the consumers giving some answers, classifying customer complaints and then giving them to the right person within the central bank, consumer protection team.

So really augmenting in a significant way the capacity of the central bank to provide effective recourse to these customers.

Paul Tierno:

So you’re talking right there about work that you’ve done with the central bank of the Philippines, you’ve also worked with Mexico, also you’re talking about additional projects you’ve worked on with Kenya. So it sounds like a lot of these are emerging markets. What makes emerging markets more suited for these applications or for some of these developments? Or is there something that makes emerging markets more amenable to them?

Simone di Castri:

So we have decided to focus on emerging markets because we think that’s in Europe or the U.S. or in countries like Singapore, the central banks have already the resources to go through this innovation a dozen times. When I speak with the regulators from the U.S. or from Europe, they’re also like can you engage us in some of this conversation and keep us updated. Because if that is going to work in the countries where we’ve tested them, the solutions that we are interested in deploying are the same.

But in some cases I think that’s it’s an advantage to work in these markets because there are less legacy systems. So sometimes we can build some of the systems from scratch, which is quite convenient. And also because I think that there’s this important issue of financial inclusion that’s created in many emerging countries. A mentality of innovation and embracing technology to change the financial sector. The mobile money sector has grown first in the Philippines and in Kenya and in Tanzania and so many other emerging markets and there are also some specific tools that we need to develop to be able to supervise those type of products that are not adopted widespread in Europe or in the U.S.

The next project that we might be starting in August is actually a project for agents supervision in Bangladesh. If you think that in the last 10 years, in several countries, the financial sectors change with telecommunication companies entering the financial market. And now they have over 100,000 agents that are deployed to serve customers. It’s a huge challenge for a central bank to find a way to get the right data from these agents, understand where they are, what type of liquidity problems are affecting the cash-outs. And so we need to develop some tools that are specifically dedicated to the needs of these type of financial authorities.

Sean Creehan:

I’m wondering the role of regtech in know your customer verifications. Would you consider what’s going on in India with Aadhaar biometric identification. We heard recently about the first ever KYC database in Singapore promoted by the Monetary Authority of Singapore. Do you consider that regtech?

Simone di Castri:

Definitely, I mean KYC, know your customer, utilities can be considered a regtech solution. Addressing the problem of customer identification and customer due diligence, it’s extremely important in many of the markets where I’ve worked. People lack an official I.D. The process for identification is inconsistent across different financial sector providers or across different agents. There is the problem of regulation, they require everything to be still on paper. Photocopies to be made. Even if you are the middle of nowhere. Without electricity you should have a way to identify a customer. There’s a revolution in the customer due diligence process. So with all the biometric systems, some of those have been deployed in India with Aadhaar, so I would definitely consider that to be a technology solution that is helping regulation to become more effective. And then it’s important for supervision because it plays an important role in AML supervision, the fact that customers can be identified, there is a reliable data base with the I.D.s of the different people.

Paul Tierno:

I wanted to pick up on something you said a little while ago. You said that the presence of Fin Techs has made regulators more audacious. That Sup Tech and regtech would have existed without Fin Techs but it’s made them more audacious. Audacious isn’t necessarily an adjective we think of when we talk about bank regulators or supervisors. And so what has been your experience with regulators? Have they been open to these ideas and you’ve been suggesting things? Or have you had to convince them? Because I think there’s generally the reputation of being interested in the tried and true methods that they’ve always experienced. So I’m wondering what your experience has been like.

Simone di Castri:

I think you make a very good point. The role of central banks is to mitigate and manage risks first. But in the last decade we’ve really seen some of the leaders in the central banking in emerging markets to be pioneers for innovation. I’m thinking of Governor Ndung’u in Kenya, the former governor of the central bank of Kenya, Gordon Espenilla in the Philippines. They understood that there was the problem of financial exclusion needed to be addressed and only bring in some drastic innovation in the financial sector. And so they gave the opportunity to new market players, the telecommunication companies to create subsidiaries to offer financial services.

They were innovators. They found that the gray areas in the regulations. They tested their approach at a time there was not the word sandboxes, but they created the space for this product to be tested and for them to learn from the relationship with the providers and from customer adoption.

How to regulate the services instead of regulating them first. With the risk of killing the possible innovation. And now we see millions of clients that are using mobile money. All around the world. So they are innovators, and when we started discussing with them about this opportunity to join us, I didn’t need to convince them. Particularly with Governor Espenilla in the Philippines, or the President of the CMBV, the Chairman of the CMBV Mexico, they understood immediately what we were discussing, what type of opportunity was there. They asked us to do a diagnostic to understand possible different interventions within the different departments. And then they decided to start addressing some of the priorities. Which are, for example, what we mentioned, this reporting issue in the Philippines and consumer protection and then Mexico is more related to anti money laundering.

And so when the Chairman of the CMBV felt the urgency because they just passed the fintech law they’re about to license new fintech firms, and you need the new system to be able to scale AML supervision without rebuilding the AML supervision teams. So something to make their work more effective.

So to answer your question I think that there are innovators and there are many. The Alliance for Financial Inclusion, it’s a network of over 100 central banks around the world and I think they have really played a catalytic role together with the Gates Foundation, Omidyar Network in the last years to help central bankers to become more comfortable allowing innovation and becoming innovators themselves.

Sean Creehan:

So over the last couple years in a series of conversations we’ve been having with people in the fintech and regtech space, it’s become clear that a lot of the innovation and audacious moves that you’re talking about, it’s coming out of Asia or other emerging markets because of these challenges that you speak of. Do you see a lot of these experiments and these new efforts to develop regtech coming back to more developed markets and influencing the way that we regulate from here the technology that we use? That’s not to say that the regulators here in the United States or in Europe aren’t doing a lot of experimentation themselves. But I’m just curious do you see it coming back to more developed markets?

Simone di Castri:

Absolutely so two advisers to R2A the regtech to regulators are the Bank of London, and the FCA from the UK. Both financial authorities are developing proof of concept and prototyping some regtech sup tech solutions but they also are working with us because they are interested in learning from some of the implementations. And possibly to replicate them in the UK. But when I speak with regulators in the U.S. or in Canada, I have the same type of interest from them.

The chat bot for example, that we discussed before, it’s something that has never been tried by a central bank. And if we were able to make it work to make it smart enough to talk with customers and to give them good answers, then I think it’s something that we will see being adopted everywhere in the world. API is, I believe in the next five years will become the main infrastructure for market supervision. I think we will see perhaps also the World Bank investing massively in helping different central banks around the world to build these API systems. Because once you have the system in place, all the information can really travel on this channels that are secure because they’re encrypted and they really allow to save so much time and so many resources for banking supervision.

Sean Creehan:

To facilitate two way flow of information between central banks and the regulated entities?

Simone di Castri:

Yes, indeed. In Mexico what we’ve been working on is to move all the data for anti-money laundering supervision from CD-ROMs, which is the format that’s been used for many years for the banks to report to the CMPV. The regulatory and supervisory authorities for financial sector in Mexico. Move that from CD-ROMs to APIs and then really create this big data lake to run the AML algorithms. And that is going to save days for the supervisors when they do on-site inspections. Right now it takes three to five weeks for each inspection. For the supervisor of the CMPV. And that time will be cut by 50% so probably they will be able to double the number of on-site inspections that they do. Just because they are able to collect and validate this data through the APIs instead of going there in person and start collecting the CD-ROMs, uploading the data, throwing the CD-ROMs into their laptops, and then running all their calculations and their verifications in Excel. So doubling the capacity of an AML supervisory team, I think it’s quite a significant achievement. With a solution that was developed as a prototype, we’ll take to banks in 3 months and it will take another 12 months to scale to the whole financial sector in Mexico.

At that point once the APIs are there they can use the APIs also to collect the normal bank reports as we are doing now in the Philippines. So really the APIs become the railroad for financial sector data, I believe.

Sean Creehan:

So maybe we can just finish with one last question I asked you to put on your forecasting hat and say what you think the regulator of the future will look like in 10 years, 20 years? Just kind of given these developments and all the exciting stuff you’re working on?

Simone di Castri:

Probably financial authorities will develop more capacity in the field of data science. There will be more technologists, they will be working at central banks. I think the approach toward innovation will definitely be interesting. There will be more and more embracing innovation in the financial sector. And being less responsive and more proactive, using data to drive the development of the financial market. We’re working on a different project, for example, now in Nigeria, with the central bank of Nigeria and the NPS, the payment processor. And one of the goals of the project is to understand what data, what open data platforms can be created by the central bank to drive financial innovation. Publishing data around payments, using demographic data, data about agriculture and solar, can really help Fin Techs to design new products, to test new products and be ready to launch them.

So I think the financial regulator will be in the driving seat on building the future of their economic systems and financial systems in their countries.

Sean Creehan:

Okay well thank you so much for joining us today, Simone.

Simone di Castri:

Thank you for having me.

Sean Creehan:

We hope you enjoyed today’s conversation with Simone. For more episodes like this, you can find us on iTunes, Google Play, and Stitcher. If you like what you hear, please do leave a review. Feedback from listeners like you will help more people find us. And for even more content, look up our Pacific Exchange blog available at frbsf.org. Thanks for joining us.


Opinions expressed in Asia Program publications do not necessarily reflect the views of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. All Asia Program content (audio, text, photographs, graphics, and videos) is protected by copyright. Permission to reprint or reuse content must be obtained in writing.

Please send requests for reprint and reuse permission to
Asia Program
Attn: Asia Program, FISC
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120-7702