Leveraging Smart Contracts for Complex Financial Products

as you may know fluidity is the company behind the air swap project which of course is one of the leading decentralized exchanges and when air swap launched their product it really set the groundwork for how we could represent financial transactions and execute transactions in an optimal way and what that allowed fluidity to do is take a step back and understand how we could build on that technology in order to affect a lot of transactions in the financial space so for the past several months we've taken a look at the capital market space and we've looked at the types of transactions that happen in that in that industry and tried to figure out how we could develop on the technology that has already been built so that we could have an impact in changing the way financial transactions are done within capital markets and really when we've we've gone through that kind of investigation we've started to develop views on how a financial transaction can be represented through smart contract technology so what this discussion really is is about how we can apply a framework to capital markets transactions that are complex in a way that leverages the benefits of smart contract and blockchain technology but before going into really understanding how we can leverage the technology I think it's good to level set on exactly what a financial product is in a very rudimentary way so what a financial product is is really a representation of a contractual right or ownership to some sort of asset classically there are two ways that financial products are really described and stocks and bonds and a stock generally is described as a fractional ownership of an entity and that ownership really gives you rights to the income of that entity so if you owned stock in Amazon you would be granted ownership potentially to a dividend if Amazon was going to distribute that dividend periodically alone on the other hand or I Bond on the other other hand is really alone and what that represents is ownership to a stream of cash flow that is really required for the repayment of that loan so generally that repayment is broken up into two different components principal and interest payment but when you hold a bond you are owed principal and interest payments because you granted a loan to an entity based on your choice so this is a very simple way of understanding what a financial product is as you move into capital markets financial products become a lot more complex and our task is really to understand how we can model that complexity using smart contract technology so to understand that really need to step back and think what drives complexity within capital markets transactions so within capital markets transactions I'm going to use nomenclature here and abstraction nomenclature around vertical structure and horizontal structure for a capital markets transaction what I mean by vertical structure is the characteristic of a capital markets transaction in that it is often broken up into different classes and you'll hear the word tranche thrown out but I was a large financial product is typically broken up into different components and those components represent different tiers of risk sometimes they're known through the terminology of senior through junior but if you own a senior portion of a Capital Markets product you are owning a less risky portion of that product and what that really means is in the form of payments and losses when payments are made against a capital markets product and you own a senior portion of that capital more markets product you're entitled to be paid first and in the case of a law you are basically suffered loss in in the last case because you are holding the least risk conversely if you're owning a junior portion of that product you will face losses first in the case of a loss scenario but you will also be paid last and the trade-off to holding a junior portion of this type of product is that you typically are paid our rate of return for taking that risk horizontal structure is the characteristic as I describe it here of Capital Markets transaction as having value that is based on a and a base of underlying assets so you can think of a pool of different assets which generate cash flow which then flow to the product level that is offered by a Capital Markets institution one of one very prominent example of this is a commercial mortgage-backed security where underlying assets are in effect loans and when you're investing as an institution or institutional participate in a captain markets transaction that includes a commercial mortgage-backed security you're really entitled to the payments that are based on the underlying loan set flowing through that product at at a higher level so the two aspects which drive complexity are vertical structure which is tranching of the specific product into into different portions of risk and horizontal structure which is the amalgamation of different assets which produce cash flow that then flowed through the larger financial product so given that there is complexity within Capital Markets products the question becomes how can we use blockchain and smart contract technology to model that complexity so that we can impact transactions using blockchain technology and so the first place I'll start is around vertical structure remember vertical structure is the represent the characteristic of a capital market in that it is broken up into different tranches which respect which will reflect different risk so the need is for a smart contract to be able to reflect the different tiers of risk that are associated with this larger financial product the way that that can be done is through a design of a of a smart contract that has a parameter around class and that class really identifies the the portion of risk that a token holder that is buying or holding a token within that smart contract would wouldn't would have so in this example this particular product you can think of as a loan that's broken up into four different tranches there's an a note of you know a senior mezzanine component and a junior mezzanine component and it goes from most senior to most junior the least risky at the top to the most risky at the bottom the way that we would model this in a smart contract is we would build a smart contract that would have a parameter around class and then we would be able to create for er C 20 tokens each of those representing a specific tranche within that financial product at fluidity we've taken a stance or we've taken the position that it's very advantageous for you to represent the entire capital stack within the smart contract programming that has been that is used to develop to support a deal even if a specific deal only requires the issuance of one part of the stack and the reason for that and there are many reasons but the one that we'll focus on is valuation the process of valuing a given token really requires mechanics around understanding the cash flows associated with holding that token and the cash flows were due because you own a specific token and the only way that you can really perform that analysis is if you understand how each part of the capital stack is performing for example if you were a token holder of the B know the amount of cash flow that you would potentially be owed is really determined by the performance of the note because if a note isn't performing well you're less likely to be paid what you were owed because you're more junior in the capital stack and as I explained more the more junior portions are paid last so for us to really have a process to be that is impactful in the capital market space we need to have mechanisms where we can value transactions accurately and that really requires us representing the entirety of the capital stack within a smart contract so a lot of the smart contract development that we've done even if there's an issuance of a specific portion we will program the entire stack within our smart contracts an example of this is some work that we did on a property that was not too far from here under the deal called factor 805 so this particular deal required a loan of 32 million dollars to a developer to demand to develop a condominium here and we developed the smart contracts to represent this all on the watching specifically when we were doing this for the particular issuer in this example we were really only focused on the Jr portion of a class the eight point seven million dollar portion of the deal however we developed a contract I was able to produce the RC 20 tokens across the entirety of the capital stack even though there was no tokens distributed for the senior class and that was a design decision that we took honored the premise that you can really only come to a inaccurate valuation of it for specific token in this case the jr. portion if you have the full capital stack modeled in and within the smart contract that's really how we think about representation of vertical structure within smart contracts the next aspect of complexity is around horizontal structure and as I mentioned this is really around the representation of underlying assets that are often associated with capital markets transactions so for any given capital markets deal it can rely on thousands of underlying assets that are producing cash flows that need to be in some way monitored in order for us to provide a drug exhaustive coverage at the deal level so in this particular example we have we have a sequence of different types of collateral which underlie a capital markets product at the top end the way that we are able to accomplish accommodation of the horizontal complexity is really by creating a smart contract that produces that defines non fungible tokens and each non fungible token is made to represent each piece of collateral that needs to be defined under this larger deal so if there were a thousand pieces of collateral we would define a thousand non fungible tokens and each of those would allow us to get a picture of the entire entire product for each non fungible token we would take specific attributes of the underlying asset so if it was a property we might take in specific attributes of that underlying property which would allow us to at an aggregate level understand characteristics of the entire product but the base way that we accommodate for this design is by producing a contract that allows us to model these as as non fungible tokens so an example of accommodation of horizontal structure through this design is for example the modeling of mortgages which is a very core financial product in in our society so a lot of a lot of capital markets transactions and products rely on underlying collateral in the form of a mortgage so we need to think about how we can accommodate the modeling of a mortgage within a non-fun jabal token so that the deal level were able to take in information on the underlying data so for a mortgage we may think about what sort of data we would hold within our smart contracts that would be useful for the process of having a analysis done at a deal of a commercial mortgage-backed security deal level for example and an example of the type of parameters we would capture are listed here so we would actually take in data around the loan but we would also take in data around the property so we would understand where it's located how big it is if it's being appraised we were taking that information and then finally we would look at the servicing data our pains payments being made on these properties and that would help us understand how the top-level security is performing based on the underlying loan performance so given that we're able to model complexity both through understanding complexity around the different classes of risk and then accommodating for a whole pool of assets were able to really model very complex capital markets transactions so here's an example of what a securitized product would look like in that an example of that is a commercial mortgage-backed security the process of securitization involves taking a pool of assets and grouping them together and then building a financial product on top of it which is then transfer different risk levels because we can through our ERC 20 tokens represent different risk tokens and this at the collateral level I represent different types of collateral and different pieces of collateral within our non fungible token we can create a representation of an entire deal which really takes in the full complexity of a securitization now this is quite significant if you think about problems that occurred during the financial crisis a decade ago a lot of that had to do I had to a lot of that was a result of us not having full transparency on the quality of the underlying assets and when we build in design around smart contracts that require us to take in information of the underlying collateral we have more transparency on how underlying assets are performing so we can understand if a commercial mortgage-backed security should be valued the way it is so the power of this design is that it accommodates for a lot of transparency and information that then can flow up to a capital markets investor I've described how we can accommodate for complex teeth in really defining capital market products through smart contracts one of the benefits of defining complexity is that it sort of changes the process that a capital markets institution has to go through when they put a deal into market so I've put in up here an example of what the current process is and why it's somewhat problematic when a deal goes to market there's a process of doing due diligence that's often done on spreadsheets very manually and then from that diligence a term sheet is provided between the borrower the borrower and lender to review and once that's finalized we produce loan documents which represent the final agreement between a borrower and lender once the deal is signed through the life of the deal a lot of the servicing is still handled on spreadsheet form this is very problematic not only because it's quite manual a lot of the constructs that are used to model deals or somewhat archaic and we're using spreadsheets and documents but it's also problematic in that it's very error-prone and often times in capital market institutions you'll see that the cash flows that are monitored during servicing which should have been predicted by the initial diligence documents are not the same when you modeled these through smart contracts and you're putting information on the blockchain and then we're cat we're capturing information relative to all the assets also on chain we're able to automate a lot of the information gathering and produce these documents and produce a lot of the calculations and payment reports in a way that's much more automated and less likely to have errors within the process so a benefit of really applying applying smart contract technology to define complexity is that we remove somewhat problematic process with one that is a lot more efficient and a lot more less less likely to errors so I think that brings us to to where do we move beyond where we are currently in order to really impact the capital market space we understand that we have the ability to model complexity and smart contracts we can model the different risk tranches we can model collateral but really what's going what is going to be required for us to scale this in a massive way in capital markets and I think a lot of the process that's involved in producing smart contracts that are comprehensive and covering a capital markets deal is not so much in the actual modeling of the contracts but it's in understanding the information that is required to go into those contracts and right now there's not really a good process by which different deal terms can be taken from the legacy documents and then used to program a smart contract so what we think potentially is a good next step for the industry is to develop sort of a uniform framework that capital markets institutions across the industry can align on on how they would transfer information relative to it to their deals to some type of format then that can be used to produce smart contracts right now the process involves sifting through a lot of loan documents sifting through a lot of terms that are sometimes bespoke and varied across institutions we need a way where we can really centralize that process in a way that the information we're taking in is somewhat standardized so that it's more efficient in translating the information into our smart contract design so I think before we get to the point where we're really able to impact the industry through modeling of complex transactions we need to have a uniform adaptable framework where many institutions can transfer information about the nature of their deals in a way that can be consumed by technology companies to produce those contracts and with that I'll leave you with a link here to our token white waterfall paper which really describes this lot in a lot more detail on how we can model complexity we've worked with many different institutions we haven't come up with a single standard now and I think it's probably going to be driven by a larger understanding of the universe of transactions that are out there before we propose something that probably is going to be adoptable by a lot of institutions so I think we're a little bit away from what we need to be there yeah I think there I think this presentation here is kind of agnostic to the type of watching that that would be used to model complex transactions this process that I described of understanding complexity in different risk tranches and complexity in the base collateral is really something that needs to be solved across different types of block chains so yeah I'm not implying that we need to necessarily use a public blockchain I'm not necessarily sure in terms of like processing time of transactions but I think a lot of the efficiency is in is gained in kind of the servicing aspect of taking in information around payments and producing reports that need to be done in a periodic basis that we haven't really quantified but there seems to be a lot of manual processes that occur in under that part of the process that we could maybe automate yes so for for the capital markets transactions I'm not necessarily sure how we manifest himself in terms of rates but yeah definitely we think that there's going to be cost savings across the entire lifecycle and I don't know if that results in rates at the commercial mortgage-backed security level changing but I think ultimately yeah we probably see changes in rates that and at the end consumer level it's joe flushing though if you see this is kind of a tiny mecca technology to be able to help automate everything is going advancing pretty quickly but the same systems that have been in place for a long time are still there yes are you seeing something abnormal left at the technology it's going to be so mature by the time that these lagging yeah I think yeah I think maybe the evolution will require development of new applications that's built upon a new data layer which is a blockchain data layer so you won't necessarily be using those legacy systems that are currently being used out I think it's a problem with how easily accessible the data is and how standardized that data is for someone to do diligence on and the power of doing this through this technology is that it forces you to standardize the way that data is is added onto the onto the chain and it also allows for accessibility of that data because depending on the blockchain solution you use some of it may be public I think a lot of the problems before the financial crisis had to do that were driven by the opacity around that data people didn't really understand the quality of the underlying assets here where everything is modeled and required to be modeled in order for you to have the transaction in the first place it removes that kind of barrier yeah when I use that to turn modeling it's not necessarily us determining what the levels are for the tranching I think that analysis is still done by capital markets professional but it's really understanding how we can represent the decisions made there within the contracts yeah other questions [Applause]