Data For the Win: The Power of the Information Exchange with Paul Margarites
Voiceover: if I ran the bank is a podcast hosted by Clayton, we're co-founder and head of product strategy at Fife and I've been Tech does enabling Banks to provide contextualized consumer like experiences to their business pie plate with a thought leader in financial Innovation and hits on the hottest topics in banking Finance in the future payments and he wants to know if you ran the bank. What's the one thing you'd go all in On It featuring in the Spotify app and podcast Google play or wherever you get your podcast and now here's your host clean weird.
Clayton Weir: Hey, everyone, welcome to another episode of if I Ran the Bank. I'm your host, Clayton Weir. I'm super excited to have my guest here today, Paul Margarites from JPMorgan Chase. He is a close personal friend, which is a good way to get an invite onto the show. But beyond that, he is a super experienced financial services professional. I think having seen kind of very different sides of the ecosystem, I think from investment banking through to a little bit on the compliance and regulatory side through to being deep in treasury and wholesale banking innovation realm where I met him and now I think building the future of financial services on the blockchain, if I'm not mistaken. So, welcome to the show.
Paul Margarites: Hey, Clayton, thanks for having me. I'm also super excited to be here and to chat with you.
Clayton Weir: Did I miss anything in your intro that you want the people to know?
Paul Margarites: Only that we've been working together for three years now in some projects, and yeah, I'm now in the blockchain world and part of JP Morgan's onyx team, which is our new business lined under wholesale payments, where we're taking a number of initiatives across our blockchain assets, and changing the future of banking.
Clayton Weir: It's really interesting. And I mean, I imagine whenever you walk into the room and try to talk about this, somebody probably puts up some Jamie Dimond quote about painting Bitcoin from 2016, because he had infinitely many of them. But the irony of that, right, where he was kind of often held up as the public kind of bear on crypto as an asset class, I think JP Morgan has probably been as aggressive as any North American institution on their own proprietary sort of blockchain, crypto sort of investments and experiments over the last like four or five years.
Paul Margarites: Yeah. And what's really interesting and kind of what excited me about the role and the opportunity here at JP Morgan, what I'm working on, is the fact that JP Morgan's approach early on was really to say, "Look. There's kind of two elements here, right? There's the crypto element and the value exchange. For now, let's put that aside. That's something that's going on to the market that was very nascent. Let's put that aside." And I think that's what kind of Jamie alluded to in a lot of those quotes and comments. But there's an underlying technology-- blockchain. It's an underlying technology that when built and deployed in the right way, has a lot of interesting use cases, particularly in the financial services sector, that could really change the way we do a lot of different banking activities, particularly around payments. And so, if you think about payments, there's two elements to a payment. There's the value exchange, which again, we could put to the side, but there's also the information exchange. And the information exchange across banks, across multiple hops, across languages, across currencies across countries, is hard. It's challenging. And that's where the technology and the ability to have a secure information network where banks can share information in a peer to peer manner, can take away all those challenges. And that's what was really interesting, exciting, because we were able to say let's look at the underlying technology first, let's focus on the challenges in the industry and in the information exchange, and then, which we're seeing now, maybe we can layer the value exchange on top of that.
Clayton Weir: Totally makes sense to me as a starting point, and then we had a conversation a few episodes ago on this show, about the nature of correspondent banking and how hard that is and how from a regulatory and compliance perspective it is probably just getting harder day by day. It's not even a technology problem in some ways. It's about the network and how the network interlinks. But on that note, is that a fair way to sum up sort of the spirit of I think it was called IIN and the branding is linked right with IIN in the middle. But actually, is that what you're trying to do is use what the blockchain is good at and stitch together the network of global correspondent banks, the JPMorgan kind of banks, into something that then could exchange data in a very modern way?
Paul Margarites: I mean, you got it. There's the pitch. Thanks, Clayton. But that's exactly how we approach the problem. I mean, JP Morgan's correspondent banking network is tremendous. We work with all the different institutions out there and around moving money across borders. We saw exactly what you're saying, we saw the different regulatory regimes, present different challenges, different banks using different types of data assets, present challenges. And ultimately, it's the end customer that is kind of hurt in the process, right, because you're trying to send money from one place to another, and that could be delayed because one kind of problem within that corresponding network, that could reach a fraudulent account, a number of issues could pop up along the way. And what JP Morgan did, and quite frankly, it's really about the partnerships we're trying to form is we said, "Look, we have the technology." The tech allows us to basically say we agree who's going to be on this network, they're going to be a financial institution or a large corporate, we agree that around data security that the information shared, to not be viewed, only by those asking for the information and receiving the information. And once we kind of had that structure, we said, "What would be our approach to that network?" And what we came up with is that, yes, it's JP Morgan's link network, which was formerly in, but it's not about JPMorgan. It's about the network participants. It's about deploying applications and capabilities across that network that allow the industry to move forward, that allow all the participants and their clients the benefit. And that's really been our approach in terms of how we've established that network and how we're trying to deploy capabilities on that.
Clayton Weir: It makes complete sense to me. And so, if we kind of pause, I guess, on the sales brochure, sort of glossy rhetoric on this and big down underlying, when people complain about international payments or going to have that thing about corresponding banks being broken, it's easy to actually take the opposite approach. It's kind of amazing that it works at all, right? Like, it's kind of actually wild that you can wire money to like the Philippines when you have any concept of how ugly it is underneath. But we could maybe dig into what those issues are or maybe for people that aren't in that, like some of the things that actually happen end to end, kind of get a million dollars from here cross currency into another country.
Paul Margarites: I mean, yeah. Let's take a couple real examples, right? I mean, first of all, let's say we're talking to a corporate treasury team in the United States. We know their processes. We've spoken with those clients, they get an invoice. That invoice might have new banking details. That banking details are to pay a vendor that's in Indonesia, let's use in any way as an example. First of all, is it the vendor that actually sent that invoice? Did they get spoofed via email? Did they actually get it from that vendor? Aside from picking up the phone and calling that vendor or the known kind of contact information on that vendor, no other way of really knowing. Then what we'll find is that they'll enter that information and generate the payment. The payment will go through the various hops, it has to go through the correspondent banking network. What the institution originating the payment, what the treasury team initiating the payment doesn't know are the different rules that exist in the countries that those correspondent banking banks reside, or even the beneficiary accounts reside. Indonesia is a great example because there's a local requirement that the name that the payment is going to actually matches, fully matches, the name on the account. Does every treasury team know that? Does everyone trying to make that payment Indonesia know that? Probably not. Is there a system or service within the financial institution previously that would exist, that would allow them to check that? No. That's why the network is so important. I mean, that's why those pain points exist across that correspondent banking network. There's all these rules. There's different kind of technology assets. And there's no one single way to come together to say, "Okay. Here, we can completely validate that payment, we completely take out the risk of a return. where, one of the correspondent banks may stop the payment, we can dramatically reduce the risk of fraud. So, I know this is going to this account because the bank at the other end actually told me that's their account. So, those are the kind of hiccups that exists and that's why the information exchange can be so powerful
Clayton Weir: I totally get it. Again, on the front end, if we're kind of going to sum that up, there's the idea of, "I'm a payment originator. All the things I need to do that have to be well formed to execute this payment as one thing and the complexity of that market by market is high, that even just a convention of what constituents like the bank account number is fundamentally different in some of these countries than what we're used to. But then there might be these name conventions, I know even some countries, you have to go to the physical head office address to access these payments or different tax IDs or recent codes. So, having access to that effectively, let's say it's an API call, to retrieve or confirm that you're sending a well-formed intent is job number one, right? So, I actually kind of know who I'm transacting with better, I have a better chance of forming this kind of payment method in a way that's going to be successful. But I think, after that, there's also this idea in these sort of correspondent type payments of, like an RFI or request for information, which is what you alluded to, which is this payment may be well-formed, but along the way, everybody that's touching it is nervous for whatever compliance burden they are bearing. And they say, "Hey, I'd really like to know exactly who Company 1132CU is that's sending this and why, because we have capital controls in this country, or whatever?" And so, that's often where these payments then fail is coming ... or not fail, but get held up is going back through the chain to clarify these extra points of information, which it sounds like, again, could be automated enough in a perfect world.
Paul Margarites: Well, it's a giant game of telephone, right? Depending on how many legs in the correspondent banking network that RFI occurred, it is going to initiate where that request comes from. And then it's just going to go back through the network, and that takes time. You have to reach out, email or go through a secure platform to request from the first institution behind you and the network. That institution needs to go behind them. That doesn't add seconds to the transaction or information exchange. That adds hours, that adds days; in some cases, that adds weeks. It's interesting you brought that up because on the link network, that was the actually first kind of application that we rolled out, was called Resolve. And Resolve was around sanctions inquiry. So, an RFI, a request for additional information would come up around sanctions, and our correspondent banking network is kind of that game of telephone I mentioned. By using blockchain and by using a network like the link network, what we're able to do is deploy an application that facilitates exchanging information. So, you could go directly to the requesting institution, directly to the institution initiating the payment, request that information-- that request can be connected to their back office systems, so that in near real-time and via API, they could provide that information. Is it a birthday, looking for birthday-- hits the back office system responds via API, here's the birthday. The institution asking me information could add that to their investigation, can clear it off in a much quicker and seamless matter, and that payment can continue on its way rather than being held up for, what could end up being weeks.
Clayton Weir: So, there's an organization in Belgium that probably either should have done this or feels like they do do this, or are about to do this. How do you see this coinciding, I guess, with sort of the pre-existing kind of industry utilities that do some of this?
Paul Margarites: I'd say what's interesting is the applications we're [JP Morgan] building. They're kind of new and unique, not only in their data, the ability to transfer data, but also in the commercial models that exist for financial institutions. So, what we've built is a model whereby institutions that are providing information through our applications are actually going to receive a revenue stream for that information. We're all helping our mutual clients reduce friction in their payments process. But also, those institutions that are providing that information are going to also receive that revenue. That's kind of a unique model across the world. I mean, most of the models out there are really focused around obviously paying in for the ability and providing your data, but don't really offer that channel of commercializing your assets. Now, what's super, super interesting beyond that is that our network and our application isn't necessarily competing with other initiatives out there. In fact, I would say it's actually quite complementary to those initiatives. You're going to work on those initiatives, you're going to share that information. If there's pre-validation capabilities through some of those global bodies, you're going to do that. But here's this opportunity to use the data assets you've built for those capabilities to also commercialize those data assets, also receive a revenue share, and also potentially resell that capability to your clients, if you so choose.
Clayton Weir: I feel like what you presented there is a really interesting fork into kind of two conversations, right? Maybe both of which are important, but let's kind of try and set them apart. That's not something that you would ever really hear a financial institution say out loud in some ways, which is the idea that I sit on some kind of trove of data... I mean, maybe data isn’t the right word for it, it's really more an asset of kind of knowing about these companies, and who can do what and what's allowed and not allowed. And they're actually being a business model by operationalizing that and letting the kind of whole network use it. As somebody you hear talked about, it's kind of assumed to be a bad word. It's kind of, I guess, just the marketing optics of it kind of sort of sound like the things we don't like about people using data and monetizing it, but it just seems obvious to me as to what the business banks are going to be in in the future.
Paul Margarites: I agree, and sometimes it almost seems like a bad word, if you will. I think what's interesting is that there's a big effort, and rightfully so, around the client experience and the user experience, and part of that isn't just providing banking services that we provided and will continue to provide. Part of that is providing guidance, advice, decision making tools. And decision making tools, the best decision making tools are those that are driven by data. And our corporate clients, and we've seen it, they're making their payments, they're doing their activities, we see certain insights across our payment network, and to the extent that any financial institution is able to take those insights, rationalize those insights, contextualize those insights, and provide them back to the client is honestly an incredibly powerful client experience-- one that could be commercialized quite honestly. And I think there should be an acknowledgement of that as well. I mean, think about our consumer banking experience. And I keep coming back to the concept that us, as a wholesale payments business, we always see that consumer banking experience, what we experience in our consumer banking lives, is really driving what we want in our wholesale banking experience. I go on my Chase Sapphire card, I can see what I've spent money on this month, or last month or last year, I can see that I've spent entirely too much on coffee, which informed me to make my own coffee at home, so I can reduce that expense. This year, I've seen I've spent $0 on travel, which is great. But that experience, that data that the firm has then kind of like rationalized for me and provided insights allows me to make better decisions in my life, and it's the exact same concept from a wholesale banking perspective and that we're trying to provide our customers with the data insights, that they need to make a decision. And I completely agree with you. That's effectively the future of banking. That's the value added service that every financial institution out there is talking about. And this, I don't mean to go back to the sales pitch, though, but this network is the realization and the channel for the value added services.
Clayton Weir: The precedent for this that many people would probably understand, at least short-term the idea of kind of paying for a data type transaction would be something similar to I guess what the account aggregators that the fintechs use today are something like [inaudible 19:25] where realistically, the business model if you don't know is, I'm making something, some kind of financial app, I need to know that Paul ... (Paul, if he says he's Paul), so I asked him to login with his JPMorgan Chase account, the aggregator sends me a thumbs up, that yeah, that is Paul, he is a JPMorgan Chase account and I pay them 75 cents or something for doing that. I mean, that's kind of probably something like what we were talking about, is a bank on the same side says, "Hey, is this person actually this?" whatever you need to know, they are incentivized to fully cooperate because there's like an economic model, and it's good for everybody, the originating bank is happy to pay, they're happy to receive that and fulfill the task, and the transactions were likely to go through, which indirectly probably saves everybody a bunch of costs.
Paul Margarites: There's the secret sauce too because that's, that's a big piece of the commercial model, right? Because any institution asking for information on the network-- so any bank on the network, asking for information and paying for that information, will make their own decision. They'll say, is the operational cost savings that I have and the client value and the market share I may gain or whatever I might do, is that of sufficient value to me? Or do I want to see if there's additional value, if there's a willingness to, like the model you spoke about, is there a willingness to pass along the cost on that. That's a decision every financial institution can make and should make for themselves. But the benefit is, whatever decision you make on that, if you're responding, if you're providing information, you don't care, what decision the requester, for instance, is making, because you're receiving a part of the fee they're paying anyway, you are commercializing your own data asset to help payments go through.
Clayton Weir: So, I think the initial use cases of what you built sound very much like they address the issues, kind of the financial institutions within the value chain, but it seems like inevitably, this is going to extend out, probably not so much directly to retail users, they'll probably experience the value very indirectly. But to large corporates, there's probably a number of ways that this changes their life, and that you expose more and more of this directly to them in the future, right? Because when I think about the problems that they have, in terms of information visibility, inherently being multibank, kind of multi currency, multi country, just the dimension of globalness that all businesses have today, there's probably a lot of challenges that really crop up in their world that this would solve as well.
Paul Margarites: Yeah, in fact, we've talked to quite a few large corporations across different use cases around, let's go with like the account validation piece, right? And really, you should think about it from the, "I'm paying and I'm receiving." When it comes to corporates paying, what we've understood is that a lot of corporates would like to embed account validation capabilities where they're storing their vendor information. So, the ERP is a big one. “Can I validate the accounts that I have in my ERP? Whether it's when I get a new beneficiary, when I get an account updated, or maybe just a general hygiene check? I'd like to run on all my vendors once a year or once a quarter." They'd like to understand to make sure that the account that I'm paying is still open? Is it owned by who I think it's owned? And is there any kind of additional information that would maybe help reduce fraud or ensure that I'm not going to see a payment return? So, from that perspective, just think about it. Again, I received an invoice from a new vendor, I just want to check to make sure that their accounts are open, or I want to make sure they are who they say they are. They could use this network and the capabilities of this network to do that in near real-time.
Clayton Weir: I was going to say, and it's not just I'm guessing them the originating institution, the person that actually holds this bank account or the KYC on some entity somewhere, that they're the only person updating this record going forward, it's going to be the wisdom of the crowd, right? So, another bank might say, "We don't actually have a first party relationship with that account, but it's been flagged like 10 times in our world, because it's always involved in these phishing schemes or whatever."
Paul Margarites: That's where it gets cool, and that's where the banks sharing information amongst themselves gets incredibly powerful, because it doesn't necessarily have to be something on your books or on your records, it could be something you've seen. Here's a real example. We have something on our network in our confirm application, which is our global account validation application called Transaction Activity. It's exactly what you said. It's, "You know what? I don't know that account. It's not in my books. I don't know if that's the account holder. But I've seen that account." And I can tell you, I've seen successful payments going into that account in the last week, so you can feel confident that that account is probably open. Or we could say, "You know what? I've seen that account. I have seen that account, except I've seen fatal returns as the last payments coming out of the account." You might want to do a little more research before making a payment because it's probably going to get returned. That takes the delays, that means I don't make that payment, that means I pick up the phone and that's not wasted time. I picked up the phone and say, "Look. Do I really have your right account information? Because I got some information that says it might not be open." That prevents me from sending the account and seeing a return, it saves me time, it saves me money. And honestly, it's just a better all client experience.
Clayton Weir: Making the world a better place.
Paul Margarites: We're trying to, and I'll tell you what, quite honestly, that's one of our main priorities, right? One of our first use cases that's going live on the network when it comes to account validation, I mentioned, Indonesia. Indonesia is live for us. You can validate accounts for payments going into Indonesia. Our first use cases for workers were working abroad and looking to send money home. What we're allowing is that the branch they go into, for certain branches that are using the application, that when they go into that branch and try to send that money home, that branch is going to run the application, they're going to check and validate the accounts in Indonesia, and they're going to be able to send it and expect a straight through processing. They're not going to have to worry about returns. So, when you're sending money home to your family and you're worried about delays or additional requests for information, or maybe I got something wrong on the account, you don't have to worry about that anymore. It's going to go straight through.
Clayton Weir: Do you have 100% DDA coverage in that market then? Or how does that work?
Paul Margarites: We have an extraordinarily high coverage in Indonesia. There's a local switching network, where our financial institution partner in Indonesia will be connecting to that.
Clayton Weir: Okay. So, that was like for North American Lesnar to be like the equivalent of like the Fed being your partner or something or TCH being your partners, or, in that sense; something that had kind of direct access to the actual clearing wire transactions.
Paul Margarites: Yeah. I'd actually probably equate it to like a confirmation pay in the UK, where the majority institutions connected via confirmation pay. But yeah, there are certain countries and regions that have similar switching networks where one institution is able to respond for the vast majority of the country. I mean, it's just incredibly powerful when it comes to the application and the value it's able to provide.
Clayton Weir: How do you see the operating model for the network evolving over time? Will it always be sort of a proprietary JPMorgan thing or do you think it'd become more of a, I don't know what the word is, but will other banks take a bigger role, I guess, around the world and kind of building it out collectively, or?
Paul Margarites: That's the big question, Clayton. That's the big question. I mean, this is a concept right now that we are incubating within JP Morgan. We are kind of building it out, making sure we're putting the appropriate amount of investment in it. That being said, we look to our banking partners and we are looking at various models where they can really participate and get the ROI out of that investment that they would like. One model that's very clear is that we are building a third party developer program where institutions will be able to create their own applications, they'll be able to manage those their own commercial model of the application, and scale those as they see fit. So, we've got a lot of plans. The main message from me really is that JP Morgan's built this network, but it's not about JP Morgan, it's about the value that all the institutions on the network derive for themselves, derive for the industry, and ways in which we can make sure that that is always the case.
Clayton Weir: To pivot, I guess, I guess slightly. I mean, what's interesting about what you're doing right is we're now. However, many years past the public peak height on sort of blockchain and distributed ledger and all this stuff. And historically, what that means is, generally, now when people are actually kind of building all the applications, it's after you quit hearing about it when the value is generally getting created. And I think we're in a, at least, subsequent Golden Age, maybe on the asset pricing on the crypto side, but I mean, through the real hard enterprise blockchain applications, it seems like that's happening now. And this is one example, what else do you think is out there that should be built or is being built that are kind of valid sort of enterprise financial type applications on the blockchain that are undistributed?
Paul Margarites: I kind of vaguely hint at earlier, but I think in a future state, if you have the information exchange, and you have the value exchange, on the other hand, if you're able to bring them together, that's really interesting. So, if you're able to kind of bring the various coins that other financial institutions as well as JP Morgan are working on, distributed ledger technology together with the information exchange, you're effectively just able to instantaneously or near instantaneously move the money in a very seamless way, move the information in a very seamless way and just dramatically change payments. I think we got some time, though.
Clayton Weir: I was going to ask that question, what time frame are we envisioning that?
Paul Margarites: For playback purposes, I'd be scared of estimating. But I think there's a lot of interesting work really being driven out there in the market. I think we've seen a lot of news come out from JP Morgan too on this case, but I know there's a lot of interesting stuff being done around this.
Clayton Weir: And on that same note on the sufficient time frame of your picking, I mean, how do you see this evolving with banks? Just with data sharing more broadly, of them playing a role of helping you know, I guess, who's accessing your data or safely kind of sharing it with the people that you want them to when you want them to like, Is that something you see evolving, or next five years, 10 years?
Paul Margarites: I think I think that's one of the critical pillars to having something like this be successful, because you need to ensure that the appropriate level of controls and data security and privacy is layered into that system. So, it's all about sharing the information. When it comes to a link, for instance, it's all about sharing information for the context of a payment. So, I want to make sure that a payment is processed, I'm not sharing information about an account to share information or just for the purposes of knowing them, or just for the purposes of making sure I can give them my service. Everything has to do about ensuring the right controls and protections are in place to prevent that from happening. Rather, it's about taking that information to say, will this payment be successful? Yes or no? If it's not going to be successful, what do I need to do? If it will be successful, great, I have everything right and I don't have to worry about anything.
Clayton Weir: Completely makes sense to me. So, I guess just to kind of paraphrase what we've talked about, there is a world where the, I guess maybe not all banks, but a material amount of global banks kind of become connected by this very highly permission, highly secure network where they can communicate to each other, hypothetically, about a wide range of data, but in the short-term really around deep kind of account validation, the kind of data we need to make international payments. And in the future, that's going to massively change probably the ease, the openness, the time and the execution costs of how we move money internationally. That's the close notes of what we're talking about.
Paul Margarites: You got it.
Clayton Weir: And along the way, we're going to massively sort of change the way banks and customers bank, sort of think about the way that data gets exchanged, and how much context we are able to gather about the people that we do business with in a wide range of scenarios.
Paul Margarites: Yeah. And I'd say they're the ultimate beneficiaries of this data sharing capabilities, because now their banks are working together in a way that's going to make their lives and their jobs easier, and more efficient, more productive.
Clayton Weir: Now that is super, super interesting stuff. I think it's a really profound and a really interesting example, where now that we're in the kind of hard work build phase of sort of these blockchain ideas I think is a really interesting example of one. So, maybe just on a lighter note to start to close out, I asked this for a few episodes, but curious what the worst thing you've ever done, worst idea I've ever had, that maybe later turned out to be a good idea that you haven't shared.
Paul Margarites: Maybe, I don't know, working with FISPAN, that one. It turned out to be a pretty good idea.
Clayton Weir: Who would have thought?
Paul Margarites: Who would have thought that meeting Clayton Weir will turn out into a great relationship and a good friendship. That was not a bad idea. That was an incredible idea. Woo! I don't know, worst idea that turned out that turned out really great. I'm hoping it's stumbling into the blockchain space, quite honestly. But I've made a lot of foolish mistakes in my career. But I think, at least my perspective, and how I tried to live is that those mistakes are learning experiences, right? So, as long as I got a learning experience out of anything I've done, I think it's been valuable and worthwhile and hopefully helped me evolve as a financial services professional and as a person.
Clayton Weir: Yeah, I think that's the biggest lie we tell ourselves in the moment is that our careers have some logical kind of arc trajectory to them. It only sounds that way when you stitch it back together.
Paul Margarites: Yeah. When you tell the story and architect the story, sure it makes sense looking back, but I mean, there's jumps all over the place, every career has its own interesting journey that rarely makes true sense.
Clayton Weir: Awesome. Well, thank you so much for coming on the show today. It was great to catch up. This is a little bit different subject matter in a lot of ways from what we're talking about on the podcast before so I hope that everyone finds that kind of as exciting as I do. As always, if you're not currently subscribed, you can always sign up and subscribe on Spotify, Apple Podcasts, I think Amazon Podcasts, which I didn't know there was Amazon Podcasts, and we'd love to have listen to future episodes. If you have any questions, comments, want to be on the show, never hesitate to email [email protected] Thanks for listening. And thanks again, Paul, for coming on.
Paul Margarites: Thanks for having me. Clayton's a lot of fun.
Clayton Weir: Awesome. See you all next time.
Paul Margarites is an Executive Director at J.P. Morgan Chase who founded and led the Treasury Ignition program and possesses over 12 years of experience driving digital transformation at global investment banks and management consulting firms. In this episode, Clayton and Paul discuss data security, blockchain in financial services, and the power of the information exchange.