Banking with Smart Contracts

Transcript

so we were all talking about before that we got on stage about this panel and potentially something very important very serious smart contracts banking and we we agreed that when you talk about something this serious you're supposed to look at the audience like they're naked so you're not nervous and we gotta say you've got a great-looking group here so as any banking panel would start we have to start with a disclaimer and no one here is talking on behalf of the banks and we're all talking from personal opinion and we're just gonna go through

basically what we have been doing in our past what we see as present opportunities and what can happen in the future so could I start with you Lee can you hear on this mic okay so my name is Lee brain I work at Barclays in the investment bank part of the CTO division my background is primarily computer science many years ago so degree and PhD in that my specialism was in functional object-oriented programming combining the two in the past 25 years since then been working in banking primarily Investment Banking were also

wealth management and corporate banking and years ago some retail banking there's been a mix of technology advanced technologies and also a lot of banking now within the last year or so at Barclays investment bank we've been involved or I've been involved looking at architecture and innovation so from an architecture perspective it's looking at technology standards and there about a few hundred of those in the bank and then if we look at innovation I've been focusing in the last year at blockchain and specifically within about looking at shared Ledger's meaning interbank Ledger's and also

executable business logic so smart agreements smart contracts that's my introduction yes I started life as a mechanical engineer quick quickly switched into into software I used to work with embedded systems technology and and now and joint joined UVs been with us for 13 years always been a technologist my work in our CTO function so I'm a Technical Architect the last few years I've been

working in R&D team so I've I've touched on many many technologies and and now my focus for the last year has been been being blockchain and that's my specific role at the moment is to investigation blue.i I'm my name is James Casas and prior to joining consensus as CFO and head of capital markets I managed a derivative arbitrage book and structured products for a bank that used to be known as Lehman Brothers and subsequently I found myself looking for a new job and ended up running a hedge

fund of life settlements and longevity risk and things that we're doing at consensus along with our partners at Microsoft we have created a simple use case of Total Return swap and basically what it is to do is is really more to have the bank's really good to see lots of the different use cases of what what can be done with the blockchain so the the swamp basically takes you through all the different areas that we've been discussing at the conference if you just imagine prior to entering into a trade you're gonna want to know

who is you know you have to go through a know your customer and so we use an identity type of system and from there we again needing to know what kind of collateral is going to be needed to be held so our reputation systems things that are not to be in place today or even the next year or two but really to give an idea of what what is the future to look like so on entering a transaction where we can highlight obviously eighty plus zero we're using smart contracts as well as I believe balanced threes been mentioned several times with their triple entry accounting

the transaction becomes you do the trade plus your confirmation so it goes out to all of the you know both counterparties as well as moralized on the blockchain and then and then what we do is is basically allow for market fluctuation having price changes and automatic collateral readjustments without having the back office be involved and even employing again the reputation system to for a kind of party to have a basically a default risk or a reduction in in their reputation and then again just summing up the the transaction again

through the triple entry and all of this really just shows kind of different spots where the banks can really look to be able to experiment and figure out what other use cases are out there from using blockchain technology am I on can you hear me great so I'm I'm Rex I guess before I introduce myself I'm gonna steal this line from from Lee because he actually said it just before we became we came on here so I was here on the first day and I've got to say it's amazing to see how much enthusiasm you guys still have towards the end of the fifth day which i think is a kind of a tribute to how powerful

the conference is so kind of well done you guys so to me my background is computer science I guess I should say what I do at stock gen so I guess you could think of me as like the the senior architect in London I work in the investment bank I head up the UK architecture for the investment banking bit so that's not the retail part of Societe Generale which is in France so computer science is my degree I spent about the first 10 years of my career building derivatives risk management systems and the second half of my career has been in software and a bit of hardware architecture as well so my day

job I spend an awful lot of time with fintechs so being based in London and London being one of the major hubs for FinTech activity we kind of take advantage of me being located here so that means that there's a strong streak to innovation and collaboration for my job and as part of that I spend a lot of time thinking and working with other folks on bit tech so things like blockchain so that's my background you know my name is Andrew Keyes I'm a recovering banker I used to work at UBS and build lots of products and distributed them specifically for insurance in longevity capital markets and my view on the banking sector with respect to smart contracts is somewhat

our thuggin all to what we've been seeing in the consortium settings where I believe that there's a standardization request and I've thought that what we really need to do is give the bank's access to tools so they can play with the technology and rather than define standards give them the ability to understand what this technology is which is somewhat pursuant to what we're doing now with Microsoft where we give this testbed for the banks to play so with that said where do you think the future

could hold indeed so you're mentioning one particular Avenue of experimentation and indeed so if we were looking at the multiple avenues so we see actually the two cross-pollinating effectively you have experimentation where you have aetherium and other technologies that we can play with we can experiment we can explore the functional the non-functional behaviors but in parallel with that we also need to do a structured design method so architecture design engineer that so this is a common theme that we've been hearing over the last maybe six months or so so we need to interplay the two of these and learn from each other and I think one of the

things we found in the last few months is the speed at which certain technologies have matured so if we look to say six nine months ago when we were looking at some of the experimental technologies and they kind of fail at the first hurdle and we if we're assessing scalability performance data privacy etc and it's encouraging to see as the months go by and this is incredibly fast that startups are already pivoting making adjustments to their propositions and the technologies in order to come back with things that are more suitable for banking and I'll give one example for this so data privacy is key if we look about a year ago maybe a year and a half ago everyone

was saying well it all needs to be completely transparent in public then there's a bit of a shift and saying well actually we can do some encryption as well and model items within that okay take you the next step further maybe actually you don't want just one chain maybe need multiple chains and then the different chain so you can partition data so you're actually then starting to get inside the heads of the architects within investment banks when they say is the first step will it allow us to do ABCD so in the past it used to be failed failed failed and it was like failed parts of failed paths and it's nearly getting there now so it's quite inspiring as we were saying to actually see that progressing we definitely see the value in experimentation

you know we demoed the small pond experiment you know where we try to model security just you know that that's that's probably very useful and as an educational tool and also to stimulate conversation inside the bank we find you know Express to a line you know we found that experimentation allows us to discover you know areas where we need to focus and you know and that's that's when we will issue the challenges out to the community to to help you know fight find those solutions you know enable scale enable transparency and privacy the right balance of those things they will need need to be ultimately a need for

standards but we want those to be open we want those to you know open access and and and and you know flatten flatten the the current structures we have today so you know I think both we need both sides of the story I agree where we from our different enterprise banking clients it seems like it goes the same way where we have the bank is requesting we want some kind of use case that we want to work with and it may not even be a case that they're going to then implement through it but it it seems like it's giving them much more of an education of how to utilize this type of technology so and yeah I think kind of just a natural flow will

be a use case with the internal in the bank as well as the you know then growing that from that type of private blockchain to more of a concerted consortium blockchain and then grabbing you know partners so that they can match up consortio are a good way to do that validation as well and you know you see it's not it's not something we can we can you know for want of a better term win on our own you know we need we need to we need participation that that involves us as banks but also also the community at large the fintechs startups like I said a lot of the thinking in the

last couple of months for for suction as being around you know what value internally could we generate by making use of blockchain and and so that's kind of figuring out okay so what's the state of the art and as these guys have already been saying that that frontier of knowledge she's moving on so so fast and that was kind of evident from from a lot of what's been talked about its kind of conference so just understanding where are we now where are we going to be projecting ourselves and then kind of playing a mating game which is what are the internal use cases that we've got where do we see that we've gotten efficiencies that we can manage to eradicate how can we create value internally and to our clients from the technology and it's just kind of been

trying to figure that out I think the the dlg consortium which kind of happened you know what three four five months ago now was kind of really a recognition that most organizations figured out that an obvious thing that we an obvious ambition that we have is simplification and acceleration of the back office is that we've got so we've got similar set of financial products we've got a similar set of processing challenges and cost overhead associated with that and that the promise is to be able to make use of blockchain technology for the benefit of you know ourselves our

clients and also for the public in the form of great transparency greater predictability greater proof via construction of correct running now this is the ambition that we have and and the consortium moment was kind of saying well actually if we're going to achieve that no one can do it on their own because you kind of know that there's no point having a chain if no one's on your chain okay you've got to be able to figure out how to move a significant proportion of the industry forward for enough products to be able to get like an economy of scale effect so in the last couple of months it's really been realizing actually let's come together

and figure out how we can manage to have a significant number of the financial participants take a step together and that kind of means figuring out a what can we do with the technology now B where is there whitespace do we see that there are smart folk out there who are looking to occupy that to figure out how to be able to build techniques around scalability or privacy as Li was talking about or alternatively should we as the financial services offer out a rallying cry to to the smart folk to kind of say hey we need a solution here it's going to be of benefit broadly if we can figure out how

to do this and start maybe putting a spotlight on some of the areas where we think if we could answer these problems before maybe these problems it would unlock a pathway for us to be able to deploy the technology quicker and faster and do you have any concepts that are relatively low hanging fruit I I see lots of use cases and and I'm interested in and it's not really a timeline but more of what are the bottlenecks from when we have proof of concepts into real life using the blockchain most banks have listed their use cases

for blockchain they've taken a view of business priorities against them together with the technical debt within the bank meaning how much effort would it take to get off current systems and migrate to this new system there's also the business benefit each Bank has its own products where it has a strategic focus looks to increase revenue etc but one of the things that applies across all the banks is this idea of the cost reduction so all investment banks are under pressure to reduce costs if you're looking from an ops and tech perspective there are clear targets against the bank banks that are expressed publicly and we view innovation could be applied to

reduce cost that idea helps you then focus your priority list for what are your use cases so put in the middle layer the reducing cost above that would be regulatory so imagine if there's a regulatory use case something that's coming up in the next few years a new requirement and a solution to that could involve a shared ledger to help you get there then you'll be able to get additional business sponsorship there would kind of be a win-win for everybody also potentially an equivalent of sponsorship from the regulator to encourage that so I put mandatory regulatory at the top cost saving in the middle and then business growth below

that from an investment bank perspective also need to bear in mind you're saying in terms of roadmaps so there are large incumbents that perform excellent roles at the moment stable reliable so if you think of the major incumbents whether you're looking across effect settlement security settlement the list goes on and on they have a valuable role and over the longer term there may be opportunities to optimize that processing but what I just wanted to highlight and this is really throwing out to the community to come back with a solution a lot of the conscious smart contracts you see that are posed on the board are designed to be solved on something like aetherium bear in mind when you move to models

where you've got for example security of settlements you've got FX settlement you've got netting processes there are things that are a lot harder than very small smart contracts things are pulled together into a netting pool quite often the netting isn't just adding up your credits and debits as a sum much more complicated you need to try lots of different combinations to get a high settlement efficiency so how do you do that currently what you do is you stop time you take a snapshot and then you perform an optimization and at the end you release that back and inform all the banks what has changed quite a few incumbents ask you to transfer more

money than you think you need to perform your settlement as a wire and then at the end of that they transfer the leftover back to you those sorts of challenges need to be incorporated into smart contracts in some way and that's very hard I think it requires not just banking FinTech I think it requires academia as well so one of the things we're looking to do next year is the banks together with academia for example the Alan Turing Institute in the UK is shaped some of these really tough challenges we're not looking to solve the technology for it we can pose the problem and then expose it and look for academics to work together with FinTech and I think at that point when we've got

securities settlement with netting then we know it's come of age so I put that later down the road map has been saying yes we want to improve our operational efficiency and reduce our costs and and these kind of technologies you know they are there is great potential there to do that you know if we can can have you no single authoritative record then good thing for the organization's you know if we can if we can distribute trust again I think that's a great

concept I mean for us one of the key key challenges that you know is is the identity you know know your customer and and those are the key challenges that we have to meet and and we all we all do the same process so so to me that represents a great opportunity so so you know and I think you know there's a lot of great minds in this room and in the community at the margin and I think it's a problem that can be solved you know we've been very public about the kinds of areas that we're looking at you know settlement is a is a great is a great use case and again you know an area

where we can we can achieve khan amis scale and you know our focus is always though is obviously on the client so so everything that we you know we want to do as organizations is around our clients so the things that are beneficial to them are great so so the ease of access of this technology is actually curious on the the cost of the settlement when we look at the different banks we've looked with you know it's been more emerging markets developing areas have seen ways where they could implement this technology easier just because here in the you know UK in the

US and things like they were there's very strong systems in place they main may not be as efficient so it seems like it's going to take a longer time for that type of cost reduction whereas I think you know a kyc type of idea where that can start to be in filled I don't know spread throughout the and bringing on customers might be an easier where I'm just curious on how you address the settlement side of things because it would be such a big change so I was picking the hardest case security supplement if you're gonna pick an easy one okay for something that's asset

registry-based so for example if you're looking at a particular physical commodity that's allocated that may be easier if we're gonna if we're gonna issue other kinds of assets on onto a chain then then we really need to be able to settle on the same same fabric if we have to settle and I separate fabrics and then that creates extra complexity that we don't know don't necessarily need so so I think you can draw some I think you know the importance of being able to do DVP on chain rather than a deliverer immobilize

an asset you know on a chain and then and then have to do a payment of changing it in a traditional system so I knew that kyc was going to come up before they kind of got to me so so just to kind of jive on um what Lee said so probably one of the first kind of like you know base camp for us in in a broader adoption of the technology is going to be trying to figure out how to do some kind of shared know your customer kind of control process so you know we are obligated by regulations to go through control processes to figure out that we are selling our financial products to the right kind of folk every

single market player has to follow those kind of controls and it is a you know woeful woeful situation that we can't manage to figure out how to do something together but the problem is we don't know how to be able to have you know to associate a thing to a digital identity to be able to show you know to be able to be sure of the entity that we're talking about and so whilst this isn't a use case for us to think about in terms of well this is one of our financial products but you want to get the banks on board figure out how to help us with kyc and anti-money laundering controls because everyone has to do it it's a burden for everybody it just increases

the cost base that we have and we've got to figure these things out before we can manage to to do things at scale so as lee was saying it's a challenge help us figure out how to how to do that kind of stuff without lock chain technology I was just curious to know what your thoughts of the future customer of a bank because because I think that there's a generational gap between the 10 year olds the 20 year olds versus the 50 year olds and 60 year olds and how will that affect strategic initiatives well I think ideally you've been talking to retail banking not Investment Banking

on that point however if we were to look at for example within Investment Banking by customers so if we take technology as a part of the bank and the business is another part so there's a customer relationship there so technology is delivering value to the business what type of technology how does that change over time so we already have a position where new graduates join they have certain expectations around the development processes and this isn't just the work environment it's how long does it take to get something live what's the nature of the work environment how quick and agile are we

so from that point of view you're there you need a transformation within technology as a starting point so things like DevOps processes automated testing standardized and disciplined agile as a methodology now it may sound quite process-oriented and not that much to do with smart contracts but actually unless the banks move to that model and they're partway down the road all these tech new technologists are going to take 10 years before they actually get implemented so we need to move to this new model so different parts of the bank move at different paces we've already got parts of many investment banks that are already highly agile on what we need to do is to expand that to the rest of

technology and have a business working with technology so in terms of the customers expectations I would say first of all there's the employees expectations and then then becoming faster and more agile and reflecting faster pace of change will enable us to deliver different products the the Millennials aspect and so on I don't think applies so much to institutional investors who are participants in Investment Banking there may be cases for example of people looking to swap assets and they'd like to do it on their iPad outside of the trading desk but I haven't been aware of that demand yet to agree with what Lisa and yeah it's a slightly different

problem in in some of the more investment banking asset management fields you know for retail I mean I think the you know one challenge I'd throw out there is no the today you you do need to understand the technology you need you do need to be aware of where the problems you know lose your lose your private key and know you've you've lost the asset that's tight to that so so you know I think again it's it's an opportunity for the for the banks and the community to provide systems to enable you know people to carry out transactions on on these platforms in a

secure way with some level of guarantee that that should something go wrong that they have have recourse obviously you know we don't we're not we're not suggesting that we want to completely control these things but we need to you know provide enablers you know I talked about before you know it's its distribution of trust so but but they're still you know there is a need to still provide some of the key services that we do today to clients thank you very much in we have time for one question so the question was was a minimum

transaction volume so I think it depends well how you're partitioning your system in which use cases so there's mention of a blockchain but the reality is there are going to be many chains within that for if you look at classical distributed databases and assume a similar approach is taken then you can imagine these being partitioned so you can look at low volume areas high volume I wouldn't expect it to be used for high frequency trading in appropriate for many years to come however if we're looking at you can look at securities settlement where they may be taking a hundred thousand transactions in a batch pumping through in a fifteen minute period you may yes so but I don't actually think that's the the challenge here I I think there are

different challenges rather than just like because I've seen those being solved and I've seen people running etherium and other smart contracts they're showing really high transaction throughput and we've run this on some nodes I think there's a different challenge and it actually really given the company here it relates to the consensus and it's the granularity of consensus that you're applying to it so I think one of the computer science problems is taking the granularity that you've got from a banks perspective which is the agreement whether it's an an executable agreement the business logic so what consensus if I gave you an agreement what consensus would you need

the parties that are party to that agreement whereas what you're looking here is the next step is how many parties need to agree to process that a bunch of transactions so there's a little bit of a mapping change as it goes down the old model is your central came to Clearinghouse that process it easy one person with that you delegate it so the scalability here I don't thinks the number of transactions I think it's how to work out the model that encapsulate that idea that the consensus is at the agreement and then work out what you need to do underneath and maybe you know if we say boss the combinatorial explosion of every combination of different power

parties and say create a chain for every single one does that scalability make sense so I think that would be a better dimension of scalability to worry about and then work out how to address does that make sense thank you all for your time