Inside the Bank-Grade Digital Asset Infrastructure Stack

Transcript

President of the CVA board Jerome Bailey and from Adanov please welcome them on stage. >> Thank you. Hi again everybody. For those who were there uh this morning um I did the opening welcome and now super excited to do this panel. I had the privilege to be able to choose my favorite panel and I choose this one on purpose because as uh the infrastructure is really my my favorite topic as you might know uh beyond CVA my real job I'm working at Adnovoom uh we are software engineering and cyber security company at quarter in Zurich we work with most of the bank um

on traditional uh fintech but also on on crypto and uh really happy to have uh the chance to moderate this panel Okay. Um so for this morning for those who were in the room um myself I really like two things from one from Ilia Vulkov he said that uh digital asset are moving from experience to real product. He says 20 banks in Switzerland are already uh proposing digital asset. So how do we move from this experiment to real product? The second point that I really like was from uh Christian CEO of Franklin Templeton about infrastructure.

remember his pictures uh with the the the containerization of of finance and to sum up he said infrastructure change does don't just improve efficiency they collapse all layers and force institution to justify their existence. So that's what we will discuss today. Um we'll discuss um with uh we'll try to make it as concrete as possible so that when you you exit this room after you you get concrete use case uh we'll walk through you through three three milestone. Uh and to do that I I'm joined by leaders covering the full stack. Um so we have some uh familiar

faces from from the crypto association uh Talos exe uh on execution and connectivity. We have a crypto finance longtime crypto association member uh on regulated custody and settlement and a new one. Thank you for joining Avaloc on core banking and platforms. So I will first let you introduce uh yourself what you're doing in the space uh so that everybody knows who you are. Let's start with you Arena. >> Perfect. Thanks a lot. I'm Arena Castle. I'm product owner of crypto assets platform uh at Avaloc that is known for core banking system but actually we having a bit more to offer.

Phil, >> thank you very much for having me and I must say I'm impressed that still so many people are in the room thanks to you guys as well at this um hour. My name is Phil Dweiler. I'm working um for crypto finance. I've been about 20 years in chatfi and about 10 years in crypto. So I think um what we're discussing today is actually precisely in this overlap of those two industries. >> Um thank you very much. My name is Chris Bach. I would like to echo on that. I also have 20 plus years trafi and then five years crypto experience and I represent Tanos and we are the largest provider worldwide for infrastructure for banks and asset managers in digital assets.

>> All right, thank you. So we we could speak for hours on that topic. Uh we could go super deep, but uh to make it really concrete, I want that we put oursel in the shoes of uh of a Swiss bank uh that wants to start simple with digital asset, let's say custody and trading uh with integration with core banking of a limited set of of top coins. So that's uh that's will be the framework of of today's um panel. So let's start with the milestone one custody for as I see it that's the number one tech solution that uh you have to uh take care of when you are bank um and I really often hear this question about buy versus build um so

let's start maybe with you Phil um as you are more custody focus expert so what is the best argument against building custody in house for example in year one or two uh and >> I enjoy myself discussing this topic here because at crypto finance probably about 90% of our customers are banks so likely each bank is asking themselves should we make or should we buy and I would say in the tech departments obviously the make is prevalent uh because they do like um finding ing out their own stuff. But what we have um observed is that eventually once things

are really put under pressure scalability that it needs to work um institutional grade it's actually difficult also even as a major bank doing your custody on your own. It's not just the technology which you probably could handle. It's the mindset. It's the knowledge across the board. Um and I would say there are many different grades as a bank you can choose entering the space from a simple approach well we are just going via products ETF ETPs etc or you can basically go over traditional rails you currently have established um you go uh you can go via your um CSD via

swift rails even etc. You can um approach a custodian for your custody directly or if you really must you can also insource infrastructure. At crypto finance we could basically offer all of those four um ways but it really is important to us that you as a as a bank as a customer find the right solution because there is not a silver bullet for everyone. Everyone has a different approach that is the best for them. All right. Thank you. So we will also switch between these three milestone uh the the custody, the trading and the integration. So for you Arena when it

comes to custody what from day one what must be reflected in the core banking environment and maybe you can use that uh question to explain us a bit more on how you're working at at Avalog because I have to admit I discovered your your solution. Uh so maybe you want also to share uh everybody knows the core banking uh but tell us a bit more about what you're doing also on the digital asset side at have a look. >> Absolutely I'm happy to do so. So uh maybe to start what you described before right that banks usually try to start very small and then expand with the time and that's the moment when they select the limited use cases they select the providers trading custody and then they

remember about core banking system u only in terms of okay we need to reflect all what we designed in the core banking in order to have the balances updated but that would work in the beginning but then when the volumes grew, you actually will realize that you need to have a proper integration. Otherwise, all the burden is actually on your operations team without having the automations, without having any uh operability um possibilities. you will really have the stretch on that side and that's where the bank finally will start looking okay so what my core banking system offer and at that moment you will discover there are so many uh small but quite painful

pain points um that are usually not uh supported natively by the core banking system could be extensive uh number of the um decimal um points for the crypto assets it's also could the for example management of the gas fees, right? Because that's something that doesn't exist in traditional banking. And probably the most difficult part is that core banking system is not 247 and that's uh potentially where the banks looking now with instant payments. But for example, at the time when we started, that was not yet a requirement.

And crypto was the first use case that really needed 247. And so what Avalog did is we started also thinking that we can design everything in core banking system is just another module inside the core banking system. But very fast we understood that no we need to probably bring it outside. making sure that this standalone application will be fully integrated to the core banking system but still will allow you to offer trading for example 247 for your client and um that's basically our current design. We have um standalone application 247

fully integrated with core banking system as um pre-trade checks as well as for the post trade bookings and also that brought us uh the new value that we can offer to our clients and that's uh actually the operability and uh the orchestration but probably we will want to cover this a bit later. >> Yeah. Yeah. Thank you. So let's switch to uh trading. Uh we'll do back and forth um and let's take the from a starting point. So Christian um should a bank begin with single broker or go straight to multi multi-broker and why and that so might be a question that uh Phil

might add his his view on after but Crystal is this for you? No, of course I'm I'm slightly biased here, right? So, we are selling trading technology for you to be able to actually do the smart order routing and best execution on our platform among several uh trading venues. So, from a perspective time to market of course it's it seems to be a very easy solution to go in and choose the single broker but it gives you of of course and also by definition a single point of failure. That means if you are a bank with let's say million of several million of end customers behind and you rely all your pricing on one single broker and even if it's the best broker in the world if that broker is not access accessible for whatever reason during volatility in the market you

might end up with a awful uh horrible uh uh PR afterwards if your clients won't be able to actually trade in the market. So at least if you go for single boa at least you need some some kind of backup solution some some some some back stop here but also you have to see and that's similar like in FX where basically very few banks only that have been highly specialized will go in and choose a single broker for the FX setup why would you want to do it any different in other OTC markets like for example crypto because uh of course the aggregation of the different prices will give you much better price and of course would also give you uh a basis to actually have a broker revenue where you can argument and say listen you only got 20% of my business but you should should have been

33% perhaps you want to optimize your pricing accordingly things like that and of course you can then also rotate accordingly so from from that perspective I I would always argue if if you have end customers self-directed behind you need to you need to have a certain setup that ensures that you always access liquidity >> I do think it's um no secret that Crypto Finance is um also um cooperating uh with Talos. We're very um happy um with with their services and I can only extend to uh what Christian just said.

What is an issue um at the broker um we as a prime broker. We basically face the same issue. We also do not connect just to one liquidity provider because despite the multi- trillion um market in crypto liquidity can um get quite uh quite thin and as we need to ensure best execution also at the prime broker perspective we also connect to multiple venues multiple um liquidity providers so at our level it's basically the same and needless to say you need to have BCM and backup solutions that if one system is down that also as a prime broker you can swap over to a second approach >> as you're speaking I want to come back

on custody because that's a question I hear for I don't know five to seven years now uh let's double click more on the technology the custody HSM versus MPC so I hear from banks that they go only HSM but then there is some MPC so MPC multi-arty computing it's more software Maybe you can redefine these two. Tell us a bit more about about about these two standards and what are preferred for for banks in Switzerland these days >> with the danger of probably not entirely be politically correct. But I would consider this debate between HSM and MPC probably a little bit um religious. Um I

would say five years ago I would still have been quite a bit um in in the camp of well MPC is interesting um but it has not yet been battleproofed then the argument would come up but um Phil MPC purely the technology has been around almost since the 70s so it's been for a very long time around but it was not used in in the industry that's why at that point of time HSM um based solution solutions were were more prominent. I do think today um you can basically use both approaches. I um was just um talking with uh without uh giving any names away talking um in the in the room

that um they're now looking in um a process of combining MPC and HSM. I think this is fairly interesting. um as well. Um we at Crypto Finance, we feel very comfortable with our HSM solution. Um that said, I would say the focus on this discussion between MPC and HSM is basically flawed because whatever technology you're choosing, the technology is just one aspect. I would argue the even more important um aspect is the governance and the procedure that need to come along. It's um about clear signing. It's about um access rights

etc. that is all on top of the of the tech stack. I would think that the arena um you um you can comment on this as well. Absolutely agree actually and uh earlier we also had a nice workshop where we were discussing that yeah the all the procedures around it all the layers of security that you need to add that's what really matters in order to be um regulated um bank uh with access to crypto >> so I understand it's not one or the other it could be both it should be both in some of the case >> we we see more and more of that that banks actually choose used to have like a like a more trading active and

typically MPC setup and a more like code storage HSM uh setup in the in the back. Um and for that obviously you would need the technology to actually be able to integrate both into your your your tech stack. >> Great. So these two religion can fit together nicely. That's great. Uh Christian uh best execution without overengineering. What does best execution look like for bank trading only top coins as at the start? Let's say >> best execution is is obviously a number of things, right? So it's not only the price, which of course is a very important part of it, but also the likelihood of execution that you have to take into into consideration. For

example, if a broker constantly projects you, we would then put it in in in a penalty box and not consider that uh that broker for a number of seconds or minutes uh until we can actually be insured that that execution is likely again. Um and also what's very important here is of course that you document what you actually what how you actually executing on on behalf of your clients that you have a snapshot of the market and can at any given time prove that this was the best possible for me best possible available price at the crypto market at this very given millisecond.

>> All right. Thank you. Uh Arena come back to you. Uh so you have a standalone model I understood. Uh so where do you draw the boundary? What stays in the core banking and what lives in the digital asset platform? Can you >> actually ultimately uh everything should be reflected to the core banking system. It just not everything is suitable to be uh reflected in core banking system. So therefore usually we um leave outside of core banking everything very much blockchain specific like I don't know um transaction hash uh any uh the balances from custodian and so on and so on. So it's something uh very specific to crypto we live outside. However, all the

balances um all the object modeling uh for the crypto assets as a new type of assets, it's of course in the core banking system. >> All right. Can can you tell us a bit more when did you launch this this service? Uh do you already have some clients running? Because uh Avaloc is really what I think the leader when it comes to private banks in Switzerland core banking and we do at at Anum a lot of integration with score banking.

Everything could be super complicated. So tell us how easy it is to uh to use the standalone version and why did you really choose the standalone? Um >> so um actually avaloc is working on crypto assets since uh five years already. Uh as mentioned we started as a separate module then we converted to the standalone application and the main reason for this is 247 to be really able to offer that to the clients and that's also for example the way for us to look then into the crypto lending because again the risk monitoring should be 247 when we are talking about crypto. Um we have already clients life. That was also a great journey for us because with

every client we on boarded we learned a lot. Uh they really tested our solution in the best way. Um and um you can check actually we have um uh live two clients currently that is BBVA and Zed Cab and as you can imagine Dur Canton Bank is really uh the one who challenged us on the volumes uh on the all the expectations of uh being uh always on top of things making sure that we can deliver deliver on time and also making the client satisfied 247.

All right. >> On on that note, if I may, we as a company, we believe that eventually, and I think that's probably the same belief we have here with many people in the room, eventually all assets will be tokenized. We will all have because we are talking a lot about crypto, but it should more the discussion is really about digital assets. And coming back to Reena's point about having the 247 trading for example and new functionalities and just the normalization of the different symbols and and number of decimals for example. Um it's very difficult to have that built on on on legacy platforms which is actually why Talos came into production in 2018 simply to set up from ground zero and pretty much also uh what Avaloc has done with their with their crypto model. Um and that so what we're seeing

now is we have currently we have two business cases that actually work in digital assets. One is Bitcoin for storage of of of wealth and the other one is stable coins for collateral management and or for crossber payments. Rest of it will eventually come whether it's going to be three years from now five years from now we don't know but eventually will come. But here obviously we are also client driven. So, we don't really see the need to do tokenization of equities just for the matter of tokenization of equities, but only if there's a real need for it and there's actually a peer-to-peer trade like we saw from UBS uh a little bit earlier today and other similar cases. So, I think that that's that's the whole shift in the financial industry that we're seeing right now.

I will probably add um to um what Arena said before that this integration into a core um banking system can be very uh painfully. We uh speak out of experience with crypto finance. We have our own core banking system as well. We see how um how really difficult and painful this um can be. Now this is one aspect of uh the adoption in the industry and we're all here to learn more to see this um industry grow. Um the question I sometimes miss a little bit is it um rather a pull or a push um from a demand in um in in the industry and um that question probably can answered also

along the rails how you bring crypto assets to to this banks and I do think that it can be interesting to start with the old traditional rails. Sorry for the swift representatives um in the room here as well. But this is actually quite interesting because those rails they are connected with all the banks already. Now if you manage um as we did in crypto finance to bring orders um crypto orders on traditional rails you face um much less resistance on the adoption. So um next to APIs I think it's a very interesting um approach also having swift integration.

>> I I know you recently did um a project with clear stream. Clear stream is part of the group. Is that correct? Can you can you tell us a bit more about that? >> That is um uh correct. Uh crypto finance belongs to Deutsche Bur. So does crypto um or so does Clearstream Bank Luxembourg. Um CBL is a international CSD central security depository. Crypto finance is the 61st subcustodian of um CBL which allows us basically with more or less the snip of the finger or let's be a little bit more accurate. um like um a very very light additional paper trail on boarding about 2 and a half

thousand banks um basically within days because the rails they already um exist because the incumbent banks can go via swift to CBL CBL is approaching us and that we are using swift and converters to integrate into our core banking system. All right. I think we didn't cover segregated accounts already. Um, so is it can feel still um what does segregation segregation account uh segregation really mean >> versus versus omnibus? Yeah. Sorry if I'm jumping in here because this is really close to to to my heart.

>> I would say the crypto industry um from a timing perspective is divided into three phases. probably up to 2020 when the argument of a missing uh adoption was um missing institutional infrastructure capabilities where afterwards until like 25 it was missing regulation. Now what do we have in 26 what is missing? I would say the only thing that is missing is the boardroom willingness to adopt. Now that brings me to the first question to back um in 2020 where still um the notion of multiple global regulators was you basically need to have um your um digital assets

segregated onchain. You can't have them um in in an omnibus account. This notion is definitely gone and I'm I'm a big proponent of omnibost accounts for um multiple reasons. Firstly, it is much more efficient because uh you can move assets um more quicker because you don't need to wait um for example for blockchain um blockchain um hops. You don't um need um to go onchain. That means it's additional security if you have lesser transactions. And a lot of the incumbent banks ask, well, can't we have a segregated account, but if we

have payment in and out, please go over a hop um via um a wallet of yours for anonymity. Guess what? You can stay in the omnibus account right from the beginning and you have much more um onchain anonymity. uh what basically um the banks our um clients would like to have. >> If I can add actually I think it's not about choosing segregated omnibus it's just different purpose. So depending what you want to offer you can pretty much offer the omnibus setup for trading for the client and for the same client in case if he wants to store it for long time as a segregated setup maybe even cold wallet setup. So all these can

exist. It's possible to configure. It's possible to offer. So why not? >> I I would like to add if I may. So I' I've had the pleasure of taking a double digit number of banks live and every bank is slightly different and slightly different understanding on how the the target architecture should really look. But I would go with Phil's point. It's only very rare cases actually that that real segregated wallets actually make a lot of sense because if you're doing the the the custody inhouse like you do with equities like you do with well with money as as as such and your end customers might for example just be average retail clients that put away 200 Franks at the end of every month to to to the to the crypto savings plan. They they never actually ask to take delivery

of the coins and they will never deliver coins into their account. So for them it's just a number on their account statement next to the perhaps dollar balance or the BSF shares that they may >> I think um if there are hopefully multiple banks in the room, there is one thing um I would like you to walk away with the understanding that the assets in an omnibus account is not less bankruptcy remote than in a segregated account. >> Okay, good. Uh I see we have less than two minutes but uh you kind of um answer for you for my my my last question already Phil so we asked the question for two other uh so once again we are this bank which wants to launch custody

and trading top coin six to 12 months. What is the one decision they must get right from they want to avoid painful review later maybe question like one one decision that they have to get right from from the beginning from scratch >> where do I want be to be with this decision in five years time that that's really the question to ask yourself >> long-term view and Arena you have few more few seconds to to close you will you'll be the last to answer >> perfect um So I would say make sure that you design your operation system from the very beginning. Make sure you don't forget about core banking system because you can encounter some limitations in a

very wrong time. And then also don't forget about the orchestration that unfortunately we didn't have time to cover but actually yes you can have as many parties as you want uh offering all kind of uh functionalities but you need to orchestrate it end to end. You need to make sure that everything flies smooth that you can extend the providers that you are using that you can shrink if needed. So make sure that you build it modular. Make sure that you think about the workflow end to end and yeah having it future proof is absolutely the key.

>> Excellent. Right on time. Congratulation Arena. So we are at the end of this panel. Thank you very much for this uh discussion. Of course, if you want to go deeper, we only have 30 minutes. Uh join the panelists during the break which is coming soon. But before one last panel uh on layer one and of course we will to keep maximum people in the room we will um at in 30 minutes we will reveal the 2026 digital asset banking report that we've done with the big whale. So please stay and uh see you later for the networking.

Thank you.