The payments point of view

The payments point of view

Kaspar Loog

Head of Payments and Operations at LHV

Michał Grela

Relationship Manager at Future Processing

Contact me

The landscape of the payments industry has changed drastically in the recent years: from cash and checks to e-wallets: real-time, global, secure digital payments. Nonetheless, the disruption is still only ahead of us and it’s worth understanding what payment trends will shape the way the market goes. New technology, customer behaviour, ever-growing competition force the industry to revisit its fundamentals.

Together with Kaspar Loog, Head of Payments and Operations at LHV, we discussed the going cashless change and its potential domination over the traditional means of payment. We also touched upon blockchain and decentralisation as well as minimising payment integrations to stay on top of risks and costs.

Michał Grela (MG): Hello, and welcome to yet another episode of IT Insights by Future Processing. Today we’re going to talk about the payments point of view. From cash and checks to e-wallets and real-time global secure digital payments, the landscape of the payments industry has changed drastically in recent years. Nonetheless, the disruption is still only ahead of us as far as I believe and it’s worth understanding that new payment trends will shape the way the market goes. As I said, new technology, customer behavior, ever-growing competition, all these aspects force the industry to revisit its fundamentals and it’s definitely an interesting and meaningful conversation to have around the state of art when it comes to payments and the trends that’s going to shape it. And a very relevant guest indeed, Kaspar with me today, head of payments at LHV, an entrepreneur fond of technology and new businesses around it, that’s what he said about him himself and I’m really happy to have you here today with me, Kaspar. Thank you for joining me.

Kaspar Loog (KL): It’s a pleasure to be with you here.

MG: Would you like to tell a few words about yourself and your experience?

KL: Yeah. In short time relatively a new face in banking and more coming from IT or technology background. So I’ve done various roles in IT and technology type of undertakings and I’ve been in banking for the past, something like five years working for Scandinavian banks before joining LHV, which was about two and a half years ago and that is kind of my little adventure towards the more business side of payment technology and I’m product management in that area and I’ve been managing the payment products team and the payment operations team, which is mainly serving the fintech customers. So we are on one hand the third largest retail bank in Estonia but on the other hand, we are kind of like infrastructure payments provider for fintechs in Europe and in UK. So, we provide real-time access to both Faster Payments in UK and SEPA in Europe plus a whole host of other services around it but, in short this is it. More of a technology background but a couple of years in payments now.

MG: Thank you. I’m really forward to understanding your insights then and let’s jump straight to it, starting with the trend of going cashless. I see that it’s not perhaps something that goes cross continent, there’s definitely countries that are more, I would say spearheading this trend but I feel like Coronavirus and the pandemic that we’re currently under only accelerated the move towards cashless. What’s your point of view on that? From the payment’s point of view, with the new normal displacing the traditional means of payment because right now the vast majority of B2B payments in the fintech industry anyways is based on cashless. Would you feel that this trend is something to be extrapolated even in more broader context? What’s your take on it?

KL: I think the March towards a cashless economy or realtime economy is inevitable because there are so many issues with cash that people don’t really talk about, especially from a business perspective and if I look at the overall picture in Europe, then you can see very clearly some trends around being cashless or being more cash oriented and usually it’s connected to a few other factors which are in place. So, the first and most typical factor it’s the gray economy in short. How much of the economy is I wouldn’t say legal because most of the cash economy is also legal but it’s more about avoiding or evading taxes and hiding your income and the less issues a country has with collecting the taxes or the less gray economy there is, the greater the incentive is towards going cashless because handling cash transactions is actually more expensive for the businesses than handling non-cash transactions.
Even if you take into account the all the card interchange fees and the fees that usually the merchants take on, in reality cash is more expensive to handle than card payments or whatever the payment method is in use but usually we’re talking about card payments. The second factor I see, I think that helps a great deal is making the non-cash payments more intuitive, more easy and I think the tap to pay type of approach here has done a great deal to accelerate this because it’s more convenient, I mean, if you look at London for instance going to the Tube, the more terminals you have around you that enable not even cashless but I’m talking about contactless payments. The bigger the they’ll take is the more easy it is to make these small little transactions that are cumbersome with cash. So that’s the consumer perspective of it.

MG: I completely understand and agree with both of these driving factors you mentioned, the first one related to gray economy is I think definitely to the point. There’s a strict correlation between the size of the gray economy and the overall acceptance of society towards going cashless and I really liked the second bit, you mentioned that touches on the user experience side of things and that’s actually related to my second question I was going to ask because there’s this concept maybe not only payments world but the fintech world in general but the concept touches payments as well, meaning the trend of unification of different platforms because from the user experience perspective, this concept is also really convenient to have a unified centralized mean of payment but how does it look like from the perspective of the business one super app to rule them all, is that doable? Is that where we’re going towards?

KL: I think a really good parallel world that we are encountering at the moment is the fight between Google and Apple in that sense because Android and iOS are the big a mobile ecosystems that are basically fighting for existence. Probably the situation is a little bit more colorful when you move towards China or India or some other emerging economies but basically or Southeast Asia which is more dominated by the Chinese as well but all in all, you can look at the platforms from the Apple versus Google perspective and a question whether we will have a unified centralized mean of payments, it’s too early to call the shots yet because there’s so many things which are on the move at the moment. I mean, we’re talking about cryptocurrency which hasn’t really picked up as a proper payment instrument yet.
It’s more of a speculative instrument and used by still a small fraction of people. When we look at payments schemes in general, they are very fragmented and the core issue there is not technology, the core issue there is the legal perspective of it and the complexity of that part, the complexity of settlement and having the legality around it especially when things go wrong and from the UX perspective, it’s unlikely that anyone else at the moment rather than Apple or Google can kind of fix this, that you will have a unified front and even if they try to fix this, there will always be room for local solutions especially in the local markets, smaller markets. If I look at it from a UX perspective I mean, there’s new fintechs now on the rise for instance, TransferWise that started with us who have kind of tried to build a unified experience on making a payment, for instance.
So it doesn’t matter from where to where you make a payment, by and large it looks more or less the same and I think this is a good trend. It makes things easier for the customer but in everyday life, in a typical scenario, you usually pay in the same currency and you’re used to doing it in your own typical banking app. So in that perspective maybe open banking opens up a few doors but still that’s a really early stage of technology at the moment and knowing the speed of development in banking in general, I’m pretty sure that there’s a lot of inertia in the systems. So we have a lot of payments schemes which have been around for literally decades and a big shakeup of the market perhaps was the introduction of the SEPA and the schemes in it the Euro payment schemes but even these haven’t really shaken up the global order, so to speak.

MG: That’s also a very interesting take, starting with the Android and the iOS comparison. I do feel it’s relevant actually here but the surge towards an end to end customer experience, that would be contextualized and efficient. That sounds really promising and to be honest, kind of looking at especially the Asian markets, China, or Malaysia or Vietnam, where they do have this super apps, now I see the trend moving to South Africa too that’s really promising but as you said to be able to introduce similar to in, I would say the Europe or Western continents or the US, I feel like you do have to be at least on a Google or Facebook level to be able to do something like that and still they would be, your nemesis on the other hand, who will be trying to do the same and would perhaps make your efforts pointless.
But what’s also interesting when it comes to the super apps is the integration aspect and from here, I’d like to move the discussion towards the integrated payments meaning the situation where there’s a payment system communicating with other critical software that the business uses and there’s definitely, maybe not a trend, but something I observed recently is the move towards minimizing the payment integration in order to stay on top of the risks and costs, is that also something you notice on the market?

KL: Jumping back to the super apps perspective, I think if let’s say WeChat or Alipay just put their mind to it and decided to conquer a specific market, they probably could do it. Another question there is the geopolitical situation because in the mobile world, we’re seeing this backlash on Huawei for instance, where various countries are raising concerns about the Chinese technology, I’m sure the same concerns will be raised in a typical fashion if their payment apps or super apps would make a move on these markets. Going into the integrated payments area then, I’m not sure I fully follow you on the minimizing payment integrations because payment is usually just one part of the whole float and what I do see is that when merchants are picking their solutions essentially what they’re after is trying to make it convenient for their customers to pay or making it convenient to collect the money in a convenient fashion.
Then in Europe, the focus has been really much on the European Union or European Commission trying to regulate things and make sure that everything is integratable rather than being market-driven, which it is in China and I think that makes a huge difference because right now for instance, open banking is like a bird with a couple of broken wings because of the implementations that have been picked by different banks and the standards that have been picked are not developed from a customer perspective, they are developed to serve a regulatory demand and in the long run it’s a problem because these solutions are not built to customer’s life easier, they are better to tick a box and from merchants perspective, what the merchant is really after is how can I integrate the payment method into my checkup flow with minimum IT effort and the highest conversion possible and of course time delay of the merchant actually getting the hands on the money is also a factor.
If you’re able to tie these three things into one good experience, then it doesn’t really matter how complex it is because in the end, if you’ve got the volumes, the merchant will go for the solution that provides the highest conversion.

MG: Okay. I think you definitely grasp what I meant here. Perhaps I was not super clear here but what you said was really interesting.

KL: Maybe I can just follow up one concrete example. So with LHV, we have built a money collection or e-commerce solution for merchants to collect payments using open banking. So currently we support pretty much the whole of Baltics but we’re planning to expand it for the rest of Europe as well and it has been a really useful learning curve because we’ve actually had live merchants on it from last summer and we actually see what the conversion rates are and how this thing is performing and secondly, we have locally here, a popular payment scheme which has kept the card payments away from the e-commerce game at least here in the local market for a long time, which is called Banklink and one of the typical issues with open banking or payment initiation in general is that you can get information about payment being initiated.
So you can see that from a merchant’s perspective, what you care about is that, do I get the money? And with open banking quite often, you get the feedback that, okay the customer initiated the transfer. It doesn’t mean that the transfer will make it. For low value transactions or digital goods you can pull back from the customer or a subscription type of services, that’s not an issue but if you’re about to ship a 65 inch TV based on this information, you might find yourself in a situation where the money doesn’t arrive but you’ve shipped the goods anyway and these aren’t aspects that haven’t been properly addressed when it comes to open banking for instance. We were trying to alleviate these and we’re talking to actual merchants or customers or fintechs, these are the actual issues that need resolving and whenever a new payment scheme comes along, these are the issues that they need to resolve when rolling out new APIs or solutions.

MG: That’s a really interesting sneak peek into tangible challenge business face. Thank you for for explaining that. The next question I was about to ask you is kind of around the buzzwords or a song called the sexy technologies that move around the payments meaning AI, machine learning or I don’t know blockchain, but let’s stick with the first two for now, AI and machine learning they’re already almost a building block of for example, fraud prevention or security fortification. So what’s the overall adoption level of new technologies such as these across the payments industry from your perspective.

KL: This is a really a funny topic because to be honest, in my opinion, about 95% of people who are talking about AI or machine learning know absolutely nothing about it.

MG: That’s definitely a vertical agnostic sentence to be said. I would say.

KL: Just to give a bit of background, I wouldn’t say I know much about AI or machine learning myself but I’ve run a startup which used machine learning and computer vision algorithms to detect issues in websites using basically these technologies mentioned and they were not even called AI at that time. AI was more brought into the game a little bit later because essentially what we’re still talking about is machine learning and neural networks and then these being used in fraud prevention. Now, when I look at the actual application of these technologies then all these hype words quite often do more harm than good because it feels at some points that there we’re putting out a lot of hammers in the hands of people and everything starts to resemble a name and it doesn’t matter what problem we are actually addressing then you just throw the hype word AI in there and it’s like magic, it will make all the issues go away or it will automatically detect all sorts of stuff but it’s still a tool and it’s still a statistical tool and when talking about statistics people their demands or expectations about AI performance are much much higher than these technologies can quite often deliver. Hence the most well-known example of AI being used is probably for instance, Tesla’s self-driving technology hype. I mean, I’ve lost count how many times this rollout has been pushed down the line and there’s a good example of the Waymo incident, which was basically a car driving on a unlit highway doing something like 100 kilometers or more per hour and it ran over a pedestrian which was pushing a bicycle across the highway. I mean, most people were saying that, “This is AI and they have sensors and whatnot, how didn’t it detect it?”;
But 99 people out of 100 were driven over the same pedestrian just to say there was no way a human being could have reacted that fast and this is actually me trying to say that we’re measuring technology at a higher benchmark than we are actually measuring people and people as a benchmark is actually quite a poor benchmark. Now, if we go now back to payments and fraud and that area, I would say AI or let’s be more precise, machine learning has probably pretty good uses when it comes to fraud prevention and the reason is pretty simple, you’ve got more data, especially on the fraudulent side because what you need to train a machine learning model, a good model is a lot of data and actually what people quite often forget, you need some good people as well to train these models and pick the right machine learning algorithms to train these models.
Once you build them all, you need to keep updating it and so on and when it comes to car payments where fraud is unfortunately quite common and more common, there these approaches, using machine learning can help to build on top of the typical algorithms, the boring good old rules based approaches. So quite often actually machine learning is not a hammer that will solve all the solutions or give you all the solutions, it’s more like you have a bunch of good rules that do most of the work and the machine learning on top of it can help to make these rule-based decisions even more accurate and this is how it usually works. So people tend to talk about machine learning but usually you can go 80% of the way using pretty basic and useful algorithms and only the last mile, so to speak, you can like polish the surface. It’s the last bit of sandpaper you use, it’s the frosting on the cake, exactly.
But there are areas where machine learning doesn’t work really well, in my opinion, sanction lists let’s say some of the AML related topics, especially related to sanctions, these are unfortunately areas where machine learning can’t really help you. It might be able to help you if you’ve got large data sets and when you need to whitelist a lot of false positives which have been picked up by sort of rules but it depends again on your risk threshold because all in all, you might say that, okay, machine has decisions to make and got 100 payments that should be sanctioned or even 1,000,000 payments that should be sanctioned and one of them gets through then you as a bank, you might be in trouble because the regulatory risk approach is that you should have zero and getting to zero is even just as hard or even harder using people. So sometimes these benchmarks are not accurate. My belief is not always the most efficient way to actually tackle the actual problem.

MG: That’s a very interesting take and I also to some extent agree with you when it comes to this. I think skepticism or at least distance that I felt from your side regarding AI and this new tech and how it’s over-hyped that’s definitely true, there’s definitely more hype in it.

KL: I don’t believe it works. I think it’s like a really useful and good tool and it has brought in leaps and bounds of useful innovations specifically in computer vision area, specifically in speech recognition area and so on.

MG: I couldn’t agree with you more. What I meant is it’s not a magic wand and I agree here with you and people do treat it as a magic wand that will magically cure all the diseases and will help us do this leap forward towards every other innovation and that’s not true. It’s a tool like any other and you have to use it right, but there’s another tool other than AI or other technology that sparks even more conversations and it might be even more controversial meaning blockchain. That’s definitely a tool that is over-hyped, it’s not there yet when it comes to adoption but when it comes to payment, there is some chat about turning towards distributed ledger technology and decentralization in order to enable cross border payments, for example and that might be, or is actually kind of a good use of blockchain. What do you think?

KL: Well, I mean, there are players in the market already who are trying to use the technology in order to make the cross borders, I mean making cross border payments is not that difficult when it comes to clearing, the hard part is the settlement actually transferring the money or the value behind the scenes and by and large cryptocurrency can serve this purpose and I think Ripple for instance is trying to build such a platform using their cryptocurrency as the carrier of value or basically using the cryptocurrency to settle between counterparties and it might work but what is the hardest part in my opinion about cross border payments is finding a universal agreement how the issues are clarified. Secondly is the trustworthiness of the settlement value career, so to speak, having a cryptocurrency which is hopefully backed up by a fear or it doesn’t have to be it can be demand driven just like Bitcoin.
It needs to be stable enough that it can be used as a kind of a backup or a conversion tool in the long run but when I think about the de-centralization, it comes with its own issues. So why money systems work overall, it’s because there are rules and regulations and more importantly, there is somebody to enforce them. When we talk about decentralization, essentially instead of rule of the law, you’ve got the rule of the algorithm, with decentralization the rule of the algorithm means that the one with the most or 51% processing power when it comes to blockchain can essentially override this. So there have been attacks on the cryptocurrency systems as well, trying to kind of overpower the processing and then turn the ledger your way essentially.
You can call it hacking, you can call it shaking the system, whatever but the rule of law is based on the amount of processing power. If the decentralization is wide enough. So there’s enough players in the market that no single player can top it, it might work but these algorithms always need time to mature, like our money system has had centuries and decades to mature, I think blockchain needs to follow the same pattern in that sense and all the loopholes that different players try to use, these loopholes will be filled and you will go to the next iteration. So there’s no secret that it’s still a early stage technology that’s why people don’t trust It yet and you need to build trust every shock that the system gets and there is resistance to it, the more trustworthy this technology will get.

MG: So, patience and trust is what we need when it comes to blockchain adoption but I feel like at least from your perspective there is some sort of sentiment towards this option.

KL: It’s one option. I’m not entirely convinced it will be the winning option but you can clearly see both from the price hike that we had again recently and there seems to be more institutional adoption towards blockchain-based cryptocurrencies as well. So I think people are getting more used to the idea and trying to find more ways to use it because quite a few years, it was a nice toy that you can play with but I think decentralized finance is still the Wild Wild West and you can lose a lot of money there or you can make a lot of money there and I’m sure that the regulators will start making their moves on this market. They are already making their moves and this is where it gets complicated because the whole idea behind decentralization is that there are no regulators. The market itself is self-regulating and once you introduce a government sized player in this market then the whole basic point of it happening kind of goes away. So it will be very interesting.

MG: Definitely. I’m also going to observe how it develops with much interest. Last but not least, perhaps a topic that’s also, that is static technology but it’s not that controversial as the previous ones, big data and being data-driven analyzing big data to add valued, to personalize services, to understand customers who track behavior when it comes to payments, hugely adopted already from both the business and the user point of view, is it a chance or threat from, what would you say?

KL: Well, like with any tool, I mean data is a useful thing. You can decide to do good things with it or bad things with it. It’s like a knife. So if you cut the right things with it serves everyone, it serves the customers, it serves the merchants, it serves the banks that are using it or the fintechs that are using it. So I think the issue is that when there are some perverse motives in play, so I think it was a very interesting move recently by Google who announced that they will forbid using these tracking cookies or make it available as a whole due to privacy reasons. So they didn’t say that they won’t use it but what they did say was that they won’t make it available to others, which is a big difference because they didn’t say that they won’t use it themselves. They’re were just saying that they won’t make it available to others. Something less than 20 years ago I was involved in the Estonian Genome Foundation Project, which was mainly about data privacy as well and the big question there was that, you collect all these gene samples, you collect people’s medical history you anonymize them and you put them in a big database and back then it was all driven by special law for that genome fund and the idea was that let’s say you discover a fatal disease which can be tracked back to a certain set of genes, how do you inform people or how do you send these people to screening if you discovered this kind of issue?
And for that purpose a decoding mechanism was devised, which is secure enough so that once a scientist makes a discovery, you can backtrack to the people in a secure way and control and go to the people and tell them that you need to get screened because you have a tendency to have this and that disease. Now being data driven means that you have the same kind of responsibility and you have responsibility in front of your consumers or the people in layman’s terms not to abuse the privacy that they have but regulators have not really done a splendid job here because I mean, talk to anyone in European Union, how much do you like the clicking for a cookie consent on every website that you see? It’s a noble cause with a perverse outcome. Instead of making our lives better it ended up making our lives a lot worse in some sense because you carry on clicking around the web, clicking okays just to read a little bit of news because that particular new site is using cookies to track you or try to get a hint on what you were doing. So data driven stuff, when it comes to banking is the same as, I mean, you can read a lot about people’s lives when you look at their bank statements and with great power comes great responsibility and I think banks are so much more regulated from that regard compared to these advertising monsters. You can already say, because they’re so big. I mean, Facebook and Google has loads more data about a person compared to any financial institution in that sense.

MG: That’s definitely true. Initially, I thought about disagreeing with you because the fragility of data that your bank possesses is on sort of different level than Facebook possesses but on the other hand before I said it, I already thought, that’s not true because people share everything on Facebook anyway even more fragile or vulnerable things, things that will make them more vulnerable than their bank account statements.

KL: Yeah. And we started off with talking about cash. So you can always go and take out some cash and do your dirty work with that. If you want to sort of hide something and try doing that in your browser. I mean, you have to be a pretty well-educated IT security expert in order to leave no tracks when you’re browsing around but it’s much easier to fool your bank compared to fooling Facebook.

MG: That’s true, and yeah I guess when it comes to data privacy online nowadays, I think it’s a myth. There’s no data privacy anyway.

KL: And usually it’s not a problem if you have nothing to hide.

MG: I mean, if it’s used to add some extra features and personalization on top of it to increase my customer experience, I’m fine with it.

KL: I can actually take it one step further, it’s quite funny to see. I mean, Estonia is one of the digitally most advanced countries, at least it’s considered to be and the basis of it all is our national ID. So every person has their national ID number and basically every citizen has their ID cards which can be used to sign documents and get yourself authorized in all sorts of services but it boils down to digital identity and it’s quite funny for us to see countries to the likes of UK or most recently Switzerland that voted against having a national ID because not having a national ID in a secure manner it’s much much worse and not having it means that compared to a citizen in UK, let’s say you want to go to the government and I’ll ask the government on who has made inquires about me and what data have you given to which people or institutions then in Estonia, I can do that because the infrastructure is there and it is secure and it is trusted.
If I would do that in UK, I would end up in, basically it’s impossible because they are not able to track it in any way and in Estonia for instance my medical data is actually accessible to doctors. Basically any doctor can open up my medical history but on the other hand, I can see who has opened up my medical history which means that I can always go back and see but I see there as a person opening up my medical history, why did he do it? He was not authorized and if the person is not authorized, I mean, there have been cases in Estonia where I believe our former President’s medical history was accessed when he was in the hospital and a few people lost their jobs after this because they abused their access but try doing that in any other ecosystem, it’s unlikely that you’re able to track these kinds of abuses.

MG: Thank you for sharing your thoughts on that and thanks for going through all the previous aspects as well. We touched on cashless and how gray economy and UX pushes us towards it. We talked about the unification and super apps and how it affects customer experience. We talked about new tech such as AI, machine learning, blockchain. We talked about data and patience and trust that’s needed to make use of even the most innovative technologies that touches on payments. I believe the conversation was really broad but it was really meaningful at the same time and touched on the most important aspects of the payments industry. That was IT Insights with Kaspar Loog. Thank you.

KL: It was a pleasure talking to you. Thank you.