I’ve spent the last couple of years working on, broadly, an optimal model for digital assets – for the issuance of digital assets / and an operating model that would work across digital assets. In the process, we have run up a website, digitalissuance.com, that has the papers we recently published at the Investment Association a couple of weeks ago. You have cards on your seats, which have a QR code, which you can [use to] get access to [the website], and please download the papers. And let us know what you think about it. I’ve been lucky enough over the last two years to have some sponsors, making this possible, those sponsors are primarily [inaudible]. We’ve got Meagen [Burnett, Chief Operating Officer at M&G Investments] here today and also some [inaudible] representation in FIS and digital distributed ledger representation from GFT. We also had the launch at the Investment Association and, of course, today, the Future of Finance with this event. I’m not going to say who it was among our sponsors, but one of the sponsors made this statement: “To remain relevant in how we deliver value to our clients, we need to think differently, and not just tweak the current model: we need to create a new one.” And what this work over the last two years was about was creating and validating and then writing up that model. So what we’re looking at today is the result of that. But before going on to that, I just want to put some facts out there – some facts that I want you to bear in mind in the whole of this discussion. So if we take a big step back from finance, and all the noise and all the people, all the intermediaries and all the processes and all the rest of it, and we look at what this is really about, about what all of our industry is about, why finance is there, it’s really just one thing. It’s about re-engineering future flows. And that’s it. Either you have a current store of value, and you want to re-engineer it into future flows – and that’s called an investment. Or [you] expect to have future flows, and you want to re-engineer it into a current store of value. So that’s all that it’s about. Everything else, in a formal sense, [is] just [inaudible]. It’s just stuff around the edges. It’s stuff in the middle that’s not really what this is about. And then if you think about who is important, then for people with a store of value, who want to re-engineer into future flows, it’s asset owners or investors. The people who have an expectation of future flows, who want to re-engineer it into current value, are capital issuers or borrowers. So those are the people who matter. That’s what matters. It’s about engineering flows in future in the way that we want. And I’ll admit that swaps may be [used to] engineer other flows into future flows. But, basically, it’s about that.
So this is a screaming, shouty man that is going to interrupt periodically. And his first interruption says, `Surely we’re talking about digital assets. Isn’t investment already digital? We’ve got these computer systems and stuff?’ Well, I contend that, yeah, we’ve computerised. We all have representations of assets and transactions. But, essentially, what we’ve done is computerise the physical process, It’s as if we’re still exchanging goods for gold. It’s the processes still separating the agreement of a transaction from the settlement of a transaction. [Inaudible] from the agreement of the transactions. It’s a process that we’ve modelled on, which is an old physical process, and that gives us a lot of issues.
So investment isn’t already digital. Investment is computerised – computer records on entry to process. So digitisation is not a replay. It’s not about a sexier way of just doing the same thing that we do now. It’s a real opportunity. It’s an opportunity for this generation to achieve something which hasn’t been achieved in the past. It’s the chance to create a new model of investment, not just tweak the current model – create a new model of investment. And that needs fundamental change. And if we’re going to have fundamental change, then you can’t just hang on to your familiar, comfortable ideas. You have to get rid of those ideas and replace them with a new set. And that always hurts.That’s always uncomfortable.
So we’re going to have a look at what those comfortable ideas are – very familiar ideas that we’re going to have to part company with, [that] we’re going to have to give up. So I’m dividing the world into the physical world – I don’t mean just literally nuts and bolts, but when we deliver things we deliver assets, where we deliver payments, where we move stuff around between parties – and the genuinely physical world where we have physical assets that are definitely not digital – where we have buildings and things and objects [inaudible], and the digital world. And in the digital world all that we have is a digital ledger and some tokens moving around. And that’s it. So in the physical world – and these are meant to be statements of the blindingly obvious, trades move assets and cash between buyers and sellers.
That’s what happens. And when we trade the thing that we’re trading – the object we see as an asset – we’re going to move that. So, after the trade, what happens is that the buyer of the asset has an asset and usually puts it in a portfolio. It’s like buying something and putting it on the shelf. And as a result of that, the things we’re worried about mostly – so the stuff we worry about with risk, the stuff we worry about with valuation – is about assets and portfolios. We don’t want our assets to lose value; we don’t want portfolios to lose value. So risk is the potential loss of value in those things. We have business systems of many and varied [kinds] which are usually complex, which control and manipulate that data. So they are the things that move assets around, or the representation of assets around, and they embody the intelligence that we have about our business operating models. And when it comes to actually managing and settling trades, then buyers and sellers are involved in that, as are the many, many agents within the current process. So that’s kind of obvious. Nobody’s going to argue that that’s where we are at the moment. Digital world? Forget it. None of those statements is true, not a single one of them. So we have to be prepared to get rid of some very comfortable ideas. Because in the digital world, the only thing which is happening is that value is moving as a token flow on the digital ledger, and nothing else. What we are trading is not assets. What we are trading are tokens, and those tokens can be traded as a whole, as a single token, as a fractional token, as a cluster of tokens, not assets. After the trade, we hold tokens at nodes, not assets in portfolios. And those tokens are the [inaudible] digital tokens. They are not about assets; they are about pledges. They are about pledges to future flows. Value and risk in this context apply to those future flows. I keep coming back to future flows because, remember, that is what the finance industry is for. So this is being true to our real objective. Because what we’re worried about is, `Are those future flows going to happen? Are those token flows going to happen?’ That’s our risk, and our valuation takes that into account. We don’t need business systems to manage this in a proper digital world. Tokens can be smart. Tokens can carry intelligence. Tokens can know what they can do, and what they can do to themselves. And they can manage themselves.
So what does that mean? What can tokens do? It comes from the nature of the digital ledger. The only thing you can do on a digital ledger is move tokens. So it comes, by definition, that what a smart token can do is move tokens, can move other tokens, or can move itself. And one thing which that allows smart tokens to do is to manage and settle trades. So all of those things are true in the digital world. All the things on the left are true in the physical world. And we have to be prepared to say goodbye to them.
And, fortunately, there’s a very, very good track record in history in saying goodbye to familiar ideas, giving us huge advantage. So one very, very familiar set of ideas was the astronomy that was created by Ptolemy in 150 AD. This was so familiar, and so comfortable, that it lasted one and a half millennia, so it lasted 1,500 years. And what Ptolemy said was that the Earth is the centre of the universe and the Sun, among other random objects, rotates around the Earth, orbits around the Earth. Which is kind of fine, except it’s not a very good model. And it’s not a good model because what models are meant to do is to explain and predict the behaviour of the objects within. And Ptolemy’s model says that the planets should be in completely different places from where they are, so it’s not a particularly useful model. But it was very popular because it put the Earth at the centre of the universe, and therefore it was very flattering to us. In order to explain why the planets behave the way they did, rather than the way Ptolemy said they should, you have to invent some really weird objects – invisible planets that acted on the other planets that you could see that made them behave the way that they were observed. And this was to preserve Ptolemaic astronomy. So it’s a set of patches, really. It’s making stuff up that makes the model fit reality rather than moving on to a new model. But eventually this got out of hand. And in 1543, Copernicus published a different astronomy – a different view of the universe, in which the Sun was the centre of the universe, and the Earth, modestly, along with the other planetary bodies, orbited around the Sun.
And that allowed us to do a number of really useful things, like, get rid of the invisible planets, and predict where planets would be accurately. I’m not pretending that Ptolemaic [Copernican?] astronomy is the best model there’s ever been. We know that it’s been superseded by later models. But what it did was to enable us to improve our view of the universe, improve our understanding, and it led to absolutely powerful developments in the science of astronomy. So it led to an explosion of progress. And that’s not just true in science and in astronomy. That’s also true in our world, because something fairly similar happened to option trading. Option trading was interesting and fun and a different kind of thing from just investing in bonds and equities. But it dribbled along, for quite a long time, as something which was just an interesting aside, a sideshow; it wasn’t front row central at all. Then in 1973, it went skywards.
So why did that happen? Well, I’ll suggest to you that two things made that happen. One was Fischer Black and Myron Scholes coming up with what we know is a flawed, but still better model of options than was there before. And then the Chicago Board of Exchange built market infrastructure to allow the trading of standard options on an organised exchange. And the combination of those two things led to that explosion. So the point I’m making is the better model, and better market infrastructure, leads to revolutionary and very, very dramatic change. And we should expect, if we have a better model of digital issuance, a digital operating model, and we build market infrastructure to support that, then we could expect a similar explosion of benefit.
So how do we do that? We’re talking scientific papers. We’re talking about Copernicus and stuff. So I’ve said axioms, but these are just rules really. How should we do this? Well, first of all, all flows of value are token flows on a digital ledger. I can’t emphasise that strongly enough. Ownership is an interesting thing in the digital ledger. There is no kind of notion off-ledger of contracts that bestow ownership or rights or whatever. If you’ve got a token at your address or at your node, it’s your token, so ownership changes when address or location changes.
Everything that can be only digital, not part of the nuts-and-bolts world, not part of the physical world, should exist in digital form only. So we’re not creating something which is a reflection of the physical world: we’re creating something which is a different world. The digital tokens are either digital cash and [we can [inaudible] what that means, or they are pledges of future flows. It’s, again, back to future flows, the whole thing’s about [inaudible] future flows. And that means that one token doesn’t equal an asset or an ownership of an asset. One token is one pledge of one flow from one issuer to one recipient. That’s what this is built out of. And that leaves us owning the physical assets. The stuff that is off-ledger is the stuff that we tokenise. Everything else is just tokens – it’s not tokenised; it’s just tokens.
So our screaming, shouty man says. `What everyone’s talking about is tokenisation.’ With tokenisation, the key is in the word. Tokenisation implies that there is something else out there that is not a token, that is not digital, that is being tokenised. So tokenisation was like the last refuge: if you can’t have a purely digital asset, you have a tokenised asset. So what does that mean, in terms of the tokens? What are these things that populate a digital ledger?
Well, they come into two categories. The things that you can’t have as purely digital are referenced to off-ledger assets. Those assets might be a company; they might be a building; they might be all kinds of things. But they are off ledger.
And then you have digital tokens, which exist only on the digital ledger and the native to it. So we then add to that the fact that some of those are cash and some of those are assets. And we get to a kind of classic consulting quadrant, where we’ve got four kinds of tokens. So cash title tokens- pretty obvious. They are title tokens that exist on-ledger, pointing at the pool of cash off-ledger. If they are fully collateralised, then they are a very, very stable Stablecoin.
Digital tokens? Wel, we’re familiar with most of them: Litecoin, Bitcoin, Ethereum. These things don’t have the pool of cash off-ledger, even though Bitcoin is represented as this gold coin with [inaudible] there aren’t pools of those off-ledger that Bitcoin represents. Bitcoin is only on-ledger, as is Litecoin, as is Ethereum. And hopefully, they are going to be joined by Central Bank Digital Currencies (CBDCs), which will be the same: it will be only on-ledger cash; it won’t reflect off-ledger value anywhere; it will be on-ledger, real cash.
Then, physical assets. We talked about buildings, Picassos, bananas, whatever, companies. These are things that exist off-ledger. And the best we can do is to create the title token on the ledger to represent that ownership.
Digital assets? Back to future flows. So the digital asset is purely a pledge of a future flow. And that’s all. It only exists on-ledger.
So, we talked about tokens being smart, and they’re pretty central to this model. So what is a smart token, what does it do, what does it look like? Well, it turns out that smart tokens are very regular; they’re very common. So every smart token looks much the same. It’s a pledge to a future flow, so you need to know who that’s from. It triggers at a certain point. So that tells you when it’s going to make its flow happen. Or it may be that it triggers on an event or a combination of the two. It has to know what it’s going to deliver – so what are the tokens that it’s going to move around? – because that’s what smart tokens do, because it’s the only thing that’s happening on the ledger, and how many of them there are going to be. So it has some terms that tell it how to calculate the number of tokens to be moved. And then, because we work in a regulated environment, then there will be constraints and rules that need to be [incorporated into the] process.
So that’s a smart token. And it’s called smart because it does three things itself. It self-actuates so it knows where to start. It starts itself. It self-executes, so it does the thing it’s there to do. It does the delivery of tokens and it self-constrains, so it runs its own compliance. And this is a pretty powerful construct, because it allows us to represent a huge number of financial objects.
First of all, instruments. So bonds – it’s a purely financial instrument made of future cash flows. Then you could represent it, so a bond is a string of these things. One for an income event, and one for a redemption event. Swaps are strings of smart tokens, which match each other: one issuer, one recipient for [inaudible] inverted. That’s all a swap is. So any complexity of instruments, as long as it’s made up in that way, we can represent with smart tokens.
But you can also represent things you don’t think of us as instruments, we think of as almost processes, like income and corporate actions. They are pledges as well, as are the other side of the balance sheet – liabilities and securitisations. Every smart token has an asset side and a liability side: someone issues it and that’s their liability, and someone benefits from what’s there is secure that’s the asset. So things that you don’t think of as instruments can also be represented.
And even more surprisingly, things that you don’t think of as kind of part of that picture at all can be represented as well. So an order is just a back-to-back set of pledges. An indication of interest (IoI) is a pledge that you haven’t made yet – it’s a pledge you want to make. So all of these things can be represented in smart token form by clustering smart tokens.
So that’s how we represent instruments, and the trading of instruments then becomes just the smart tokens doing their thing, just acting as they are able to act. So the smart token creates self-execution. So screaming, shouty man says, `But surely we’ve, you know, we’ve spent the last 30 or 40 years automating trading, and [no-one] will think this is a good idea. How does this make a difference?’ Well, yes, we have. And I’ve been involved in that process as much as anyone else. And if we go back to the two key parties who really matter in this, the asset owner and the capital issuer – let’s call them, for simplicity, a buyer and seller – in our current market and our current market infrastructure, if these two persons want to transact, then there is a blizzard of other objects that get involved in that process, many of them regulated entities that are mandatory in the process.
And this is a kind of equity example, but there are plenty of others. And even if you try and make that organised, you’re trying to turn it into almost a flow diagram, look at the interactions between them, and then overlay the processes that these parties carry out, this is almost chaos.
This is this is just not anything you would ever build from scratch deliberately. It all comes from patching, just like Ptolemaic astronomy patching in invisible planets. What we’ve done is patch in entities and objects and processes and controllers in order to make up for the weakness in the model in the first place.
And what’s really scary is that we don’t just have one of these. We have different ones of these for every asset. So we’ve got ourselves into a very substantial, complex mess, where we’re constantly trying to catch shortcomings in the model. And we’ve got multiples of these. And the effect of that is a very heavy load of regulation.
And the reason why that’s a heavy weight of regulation is because the regulation – the complexity and scale of it – is multiplied by the number of entities, the number of interactions and they’re huge.
So that’s where we’ve got to. How on earth did we get there? That looks mad. But the people who got us there, and I’m including myself in that, were well-intentioned, trying to do the right thing, not stupid, trying to improve the process.
And I’d suggest to you that we got there for three reasons. One, because we’ve patched the process and we’ve patched about more, so we’ve just created more and more things like central settlement depositories (CSDs), central counterparties (CCPS), clearing houses, things that mitigate the bad effects of the model that we’ve got, the model that we’ve created. And they usually mitigate risk or reward trust.
Second thing we’ve done – and this is useful and sensible – [is] we’ve improved intermediary processes and technology. So we have made improvements in technology. And the third thing we’ve done is – absolutely in our industry and on the buy-side – manically reduce resource costs through outsourcing everything that moves and offshoring every resource that we can.
So we’ve done those three things. And these three things are well-intentioned; they are practical things that you can achieve; and they’ve been largely successful. The problem is they have run out of steam. This is [the] Mallard [steam locomotive]. It’s a very beautiful object, if you like trains.
It is an incremental improvement on what went before. An A4-specific [locomotive] was better than an A3 specific because it had a better smokebox, better cylinders, I don’t know, a better something-or-other, and it showed its superiority by getting the world speed record for steam engines. Which is great. So it’s obviously a high quality piece of kit. The problem is, before it was built, it was obsolete. And that’s what’s happening. That’s where we are now with our current model of transactions and assets. It’s obsolete.
So where’s a better alternative model? Well, let’s have a look at the entities and objects that we need in this alternative model. And this is getting on to, `What is the digital issuance optimum model?’ Well, first of all, we have our two key parties: the capital issuer and the asset owner – issuer, recipient. And then we need some smart tokens, a smart token. And then we need the things that the smart token is going [to do]. That’s it. That’s the whole thing that we need.
So that’s the issuance model. If we look at the operating model, how does that manifest itself in a transaction? There is our issuer and recipient again. The issuer does what its name suggests. It issues pledges. It issues – mints – smart tokens, which it then edits a pledge on to. And it then transfers that to a recipient who is willing to receive it. (They might be making their own pledge or transferring some tokens in consideration, but that doesn’t really matter – that’s just a pledge the other way.) So the smart token is transferred to the recipient. And then, critically, the issuer makes the pledged tokens visible and available on its node. And that is how trust is established. When the smart token wakes up, its trigger happens. That makes it live. And it transfers those tokens to the recipient. So it’s done what it’s defined to do. Having done its job, it can then be transferred back to the issuer. And the issuer can then do whatever it likes with it. It can burn it. It can stick it on a shelf. It can use it as an indication of interest. It can use it for another pledge to another party. So it can do what it likes. That’s the whole thing.
So we have one issuance model, and we have one operating model. Because we can represent all of those asset classes that I showed you, all of those processes that I showed you, all of the things the other side of the balance sheet that I showed you, then that one operating model will work for all of those.
So when you have a new asset, you define a new set of tokens. You don’t have to build any new technology – it’s already there. It works because the underlying operating model works. So that’s it. That’s the one operating model. Two entities, two objects, five processes. And that means that regulation can go from being a very dark cloud to being a very small one.
So what about trading? Trading smart tokens? Well, we can trade them, as I said, individually or in clusters or in fractions – whatever you want. So if you received a smart token, there is a guaranteed secondary market. You can trade however you like. Clusters simulate conventional assets. But if you create a bond, it’s a cluster of tokens. But once you’ve traded it, the recipients have just got a bunch of tokens, so they can trade those individually, in fractions, whatever they want. So the asset doesn’t persist. Only the tokens persist. Unissued tokens can be kept visible and searchable on the node of the of the issuer and then they can be searched by willing recipients. The node, if you don’t want to be searched, or if you don’t want to make yourself public, you can create the smart markets or you can create a node that matches anonymous pledges, so you can contract as a kind of, sort of central, sort of market.
And orders follow the same operating models, so you don’t need to build an order management system or an execution management [system]. The orders follow the same operating model, because they’re just back-to-back pledges. So they are identical to other tokens.
So screaming, shouty man says, `Hang on a minute, this is all a very radically new way of looking at this. This is digital issuance, seeking to be optimal, and seeking to have a single operating model. I’m sure it’s happening all the time anyway.’
And it is, but I’d point out two flaws. When we have lots of projects, lots of people in this room are really involved in [inaudible] which are doing, usually, title token issuance. Well, [inaudible] will often get better than that. And we’re doing it for sub-classes of assets, often sub-sets of asset classes. And we are doing it, making it all up for the individual project. So we’ve got a proliferation of methods, a proliferation of approaches. And this is in the future going to make it very difficult to inter-operate. It’s going to make inter-operation complex and costly. And it’s going to limit what we can achieve and what benefits we can deliver.
The second thing that we’re doing wrong is that a lot of digital issuance – almost all of it at the moment – is at the asset level. And as soon as you receive at the asset level, you concretise that asset class, you concretise the asset, and its lifecycle, its operating model and its terms and conditions. And that will lead to [inaudible] benefit.
So what do we need to do? What are the rules if we’re going to optimise? What are the rules for optimum digital issuance? The first is obvious: issue the tokens at the flow level. This is all about future flows, not at the asset level. Make the token smart because the tokens can then operate themselves and they can give us a very simple operating model. To do that, they have to be self-actuating, self-executing and self-limiting. And we need them to be individually traded. They’ve got to be liquid. They’ve got to be something we can trade. And then measure [inaudible] risk of the tokens and the flows and not of the assets.
So what does this leave for business systems? Well, we’ve transferred so much onto the token, even though the tokens themselves are pretty simple. Business systems shrink, and they become standardised. So we still need business systems to operate the node that represents the business location. But what they become … Well, they’re a receptacle for tokens, they’re just a node holding tokens, and that’s where the tokens … Tokens don’t issue themselves and edit themselves. They can’t do that, so the business system does that. And it sends the tokens to its initial recipient.
But the business has to have visibility and understanding of the tokens it has in issue. So it has to know what it has got inbound, and what it has got outbound – so what it’s pledged to others [and] what is pledged to it. And it has to look at those in terms of risk, valuation and, critically, asset and liability management. So the inbound and the outbound have to be matched up. And you have to know what your asset and liability position is over time. We’ve talked about searching for matching pledges issued as indications of interest. And you have to maintain permissions for who can look at the indications on your node. That’s about it.
So when you’re creating your asset class for a new product, you don’t need to change any of that. So you don’t need to create a new asset class. You just create a new asset class out of new cluster of tokens. And you don’t have to change the security master. You don’t have to have new maintenance facilities. You don’t have to have new transaction processes. Any of it. You can just do it immediately. Your new product? You can do it immediately.
So that gives you a flavour of [the] benefits. All the benefits of this model spring from the fact that there’s one digital representation that is optimised or seeks to be optimised. And there is one operating model.
Benefits for different parties? The asset owner/investor gets much more product choice because it’s so much easier to create products. Asset sourcing is simplified because you can get … you can trade whatever you want at whatever level you – the individual, the fraction, the cluster, the individual token or the asset representation. So asset sourcing is simplified. And because everything has an asset side and a liability side, you see the whole asset and liability picture in one place.
From the asset manager’s perspective, creating new product, creating personalisation, is straightforward. And moving into new asset types – which we all know if we’ve been involved in derivatives and new products and new structures, this has been a nightmare, it’s taken forever, it’s been expensive and high risk. So from an asset manager’s perspective, change becomes cheap.
For the issuer, the capital issuer, matching funding requirements [is a benefit]. Instead of approximating your asset-liability to match your pension scheme or insurance scheme, you can create it precisely without needing to go into kind of complex swap contracts to get rid of the approximations. So you can match your funding without needing to go into complex and expensive transactions.
From a service provider’s point of view – I don’t know whether this is a good thing or a bad thing – but from a service provider’s point of view, there is no income processing. It does it itself. There is no settlement management. There is no order management. There is no security data maintenance, because there’s no security master. The smart tokens that are out there are the security. Where they are is the security master. There’s no registry maintenance. It does it itself. There’s no entitlement calculations. So what we depend on in asset servicing pretty well goes away, pretty well evaporates.
And, for everyone, we get radically simplified regulation because the number of entities, processes and interactions is so low. You get no boundaries between asset classes, so the straitjacket of dealing with conventional asset classes, which we all just live with and think is normal, it’s not normal at all. It’s not right. Asset classes shouldn’t be fixed. Assets should do whatever you want them to do. And we should have a secure operating model that works for all of them. And that’s what’s on offer.
IT architecture becomes standardised. So the business systems will all be the same. It doesn’t matter who you are. You can be an issuer, a recipient, an asset manager but whoever you are your IT architecture is the same.
So, do we want to do this or stay with what we’ve got? Well, if we do stay with what we’ve got, well, then, the blizzard of entities, processes, regulations, controls, we stay with that. We keep that. Our costs stay where they are or go higher, because over time, it’s getting difficult to make further savings. Change is slow and expensive. And as other markets, other jurisdictions, get a bit smarter our competitiveness will decline. [Inaudible] forms of digital issuance, and the proliferation of complexity will continue. And that means that it will get more and more hard to inter-operate and harder and harder to create a universal view of assets across ledgers. We’ll be stuck with the current definitions of asset classes and their concretised life cycles and operating models.
And overall, we’ve got a generational opportunity here to do something different, to create a new model, which optimises our deployment of digitalisation and deployment of digital ledgers. And if we don’t take that, then our clients may start to do more and more of this themselves. Decentralised finance gives them the platform to do that. So if we don’t do this, then people may run out of patience. I think there is evidence that people are [inaudible] with organised finance and are going to decentralised finance. That’s a threat – an existential threat – to all of us. So that doesn’t look very good.
So I’ll leave the last word with Arthur C. Clarke, who said every radical idea provokes three sequential reactions. The first of those is, you’re a clown, you’ve got to be joking, this’ll not happen. This second is, `Okay, it may work but no business case; it’s not worth it; carry on doing what we’re doing.’ And the third is to say `It was my idea.’
Thank you. I’ve hoped, this afternoon, I’ve persuaded you that trying to optimise issuance, and creating a single operating model for digital issuance is at least the start of a really good idea.