This week on The Defiant Podcast we speak with Tyrone Lobban, the head of blockchain at JPMorgan’s Onyx, a business unit focused on digital assets.
In the craziness of the past couple weeks, it’s easy to overlook or forget good news and positive developments. JP Morgan recently made history by becoming the first major bank to execute a live DeFi trade. The trade involved tokenized forex and government bonds and it all happened on the Polygon blockchain.
In our conversation, we dive into the details of the trade and discuss the implications of major banks like JP Morgan and the rest of TradFi using DeFi, and start by getting a better understanding of Onyx.
JP Morgan’s first DeFi transaction took place on 2 November 2022 and was facilitated by Project Guardian, a pilot initiative of the Monetary Authority of Singapore. The trade had a lot of moving pieces and was rather complex. Tyrone takes us behind the scenes and walks us through the steps taken to ensure adherence to KYC/AML standards.
Due to regulatory and privacy concerns, JPMorgan is currently running its DeFi experiments on a permissioned sidechain. We discuss the likelihood of a shift from private infrastructure to a public blockchain.
🎙Listen to the interview in this week’s podcast episode here:
đź“ş Watch the video here:
🙏 Thanking our podcast sponsors:
- This episode is brought to you by Circle, the sole issuer of USDC, and a leader in crypto that’s held to a higher standard. Learn more at https://www.circle.com/en/usdc
- Aave Grants is a community-led grants program, focused on growing a thriving ecosystem of contributors within Aave by funding ideas, projects and events that benefit the protocol or surrounding ecosystem. Head to aavegrants.org to learn more and apply for a grant.
- Now for the first time, you can keep your crypto and NFTs safe with tech the “big guys” have used for years, thanks to ZenGo. Download the ZenGo app from the App Store or Google Play and use code defiant to get $20 back on your first purchase of $200 or more. Terms and conditions apply. See site for details.
- Built by the team behind Avalanche, Core is an all-in-one command center for Web3 supporting Avalanche, Ethereum, and Bitcoin, a rich ecosystem of dApps, Bitcoin & Ethereum bridges, NFTs, Subnets, and more. Visit core.app to download now.
What follows is a summary of the podcast episode.
So you are Head of Blockchain at JP Morgan’s Onyx business unit. What exactly does Onyx do?
Onyx is JP Morgan’s blockchain-focused business unit. What that really means is we have a group of people, somewhere around 250 people, that are exclusively focused on building new infrastructure and new applications, new solutions, that leverage blockchain technology.
The JP Morgan blockchain program started in late 2015. Since then, we’ve been trying to figure out where blockchain makes sense for a large, regulated financial institution and how we can use this technology not just for ourselves but also for our clients.
The things that we’ve really been focused on are how do we create better-operating models, better processes, but I think most interestingly, new products. New revenue streams for our clients and for our own internal businesses.
Since we actually began the blockchain program at JP Morgan, we identified three key areas that basically had enough confidence to build this business around.
The three key areas are:
- Being able to exchange information related to payments in a better way and deliver new insights off the back of that, once you have a shared ledger of payments information.
- Transfer of value. Specifically, money.
- The idea of tokenization of traditional assets. Bringing them into blockchain, and creating new products.
I think that all of this is wrapped up in a way that is obviously very focused on blockchain. And especially in the last couple of years, we’ve been thinking a lot about Web3. How do we create better solutions around identity? How do we interact with public blockchains? What are the DeFi opportunities? This is a very important way for us to look at these three different opportunities that we’re really focusing on.
Maybe earlier on, JP Morgan was using more of a private blockchain solution and that’s been the direction that most institutions and traditional financial services take when looking at blockchain. It’s really notable that you have shifted towards a more public blockchain approach. When did this happen? Was this just a deliberate decision?
I would say we’re shifting. All of the products we have now are focused on these three opportunities: information, value, and tokenization.
We actually have three private, almost actual EVM side chains really, that we run for each of those opportunities. We use the Quorum Protocol which is now owned by Consensus. This blockchain protocol is actually something that we created in 2016 primarily to solve some of the challenges that we as a financial institution have to meet when we’re thinking about working on blockchain.
If Ethereum was first created with built-in privacy and better scalability at the time, there would be very little reason outside of things like AML and KYC as to why we couldn’t just use a public blockchain. At the time it was the privacy aspects and the scalability aspects that drove us to have to explore private blockchains which we’ve now stood up as EVM side chains.
Private networks are still important for how we are really allowed to operate or how we can operate, because we can add privacy to them and have more control in terms of the congestion of the network.
We’re not at the whims of some spike in activity on the chain, that doesn’t necessarily have anything to do with what we are doing, the business that we’re doing. They’re important, but we’ve also, for a long time, viewed the public blockchain infrastructure and ecosystem as ultimately the destination and the direction of travel.
We’ve been working over the past few years to actually get to that point. We’ve done a lot of thinking about how we can establish our networks to be fully Ethereum-compliant, for example. So all of our code is obviously written in Solidity. We utilize the normal ERC standards for the tokens that we create. We utilize all of the toolings that exists in the public ecosystem. We want to use our existing networks as a springboard to the public infrastructure when we can.
I think this is happening now because of how the Monetary Authority of Singapore is running things. They basically came out and said, “Hey, we want to look into this idea of institutional DeFi,” which is the idea of putting traditional assets on the public blockchain ecosystem in a way that regulated institutions like us can use.
That’s where we get into things like having to solve for identity so that we can meet our AML and KYC requirements. But there are still open questions around privacy and around the ability to have a business transaction not necessarily be impacted by other activity on the network too negatively.
Do you see a clearer path to actually using public blockchains while preserving privacy and having it be sustainably scalable right now versus what you were seeing before?
I think we’ve definitely made notable progress on the identity aspect. Going into the project, we knew that we needed to have a way of proving that, JP Morgan and SBI, for example, could actually participate in these polls. But over and above that, we needed a way to ensure that our traders were correctly permissioned and authorized to actually transact.
We didn’t want anyone within JP Morgan to just be able to execute the trade. So these were open questions around how we were going to achieve that. We had a few key design principles that we wanted to hold to as we were looking to solve this identity sort of challenge.
One of them was we didn’t want to reinvent any existing processes around KYC or being able to identify our own employees. For example, we wanted to reuse the existing identification processes. Secondly, and for me, actually almost most importantly was that we didn’t want to have any of the DeFi protocols make any changes on their end in order for us to interact with them in a more compliant way.
We wanted to abstract this identity, this identification, and the identity verification from the DeFi protocol itself. Over the past month or so, this conversation is really important as people have been talking about regulating DeFi front ends and having KYC checks on the DeFi protocol. We think that there’s actually a cleaner way to do that, and that’s what we really proved through the project.
We wanted it to be really easy and frictionless for our traders for them to utilize a DeFi protocol. We didn’t want things to have to go through a whole lot of steps to be able to do that. It needed to be a sort of easy trader experience for them. So, we put that into an institutional wallet that we made for the project, but it was really based on the idea that we should have a stronger identity before we interact with these DeFi tools.
So summarizing the key three requirements for this trade were:
- To have an identity system in place for those executing the trade
- To have DeFi protocols maintain their current structure.
- Make it easy for your traders to use.
I would say on the compliant access component of the project, those are the three things. In addition to that, prove that we could use tokenized traditional assets, real-world assets like deposits.
We also wanted to clearly utilize the public blockchain infrastructure to the extent that we could. And the reason why I say to the extent that we could is because from a JP Morgan perspective, we currently are not able to hold crypto. We actually had to make all of our contracts meta-transaction compliant so that we could enable gasless transacting. We had a gasless provider that sort of stood in the middle of us and Polygon. That enabled us to actually transact whilst not necessarily having to hold crypto ourselves.
Compliance around identity, traditional assets on a public blockchain, and then utilizing public blockchain protocols and DeFi pools.
Why was it important to you to use a public blockchain to do this?
Partially because of the fact that we do see this as the trajectory. Even now in the turmoil that we’re currently going through what you have with a public blockchain infrastructure is a persistent settlement rail. Something that if everything works correctly, which for the most part does in ecosystems like Ethereum, you have this infrastructure that is fully persistent, always on, highly available, and redundant. That’s a really good place to be if you’re looking to settle financial transactions.
Then you overlay the additional important aspects around transparency and ensure everyone is actually viewing the same source of truth as it were. Because those types of things start to break down a lot of the complexity that we have around post-trade processing. Once a trade has actually happened, there’s a lot that happens in the background to settle that trade, update, and exchange ownership of the specific asset.
If you can do that in a programmatic way using smart contracts, then you can really simplify that. And then you overlay on top of that the composable nature of things like DeFi tooling and DeFi protocols. You can start to actually build new products much more easily. The innovation happens much more quickly. I think that just accelerates the types of products that we could offer or that others can actually offer. For example, we can start to provide liquidity into.
There are lots of reasons why public blockchain makes sense. And one that I didn’t mention is just the fact that when we look at what’s happening today in the sort of permissioned blockchain space, we are seeing quite a lot of fragmentation. That’s just gonna grow over time. You’re gonna have liquidity fragmentation. And I think if you can get to this point where we can utilize public blockchains, that potentially could be reduced. Obviously, there is still a decision to be made in terms of which L2 or which side chain, or even which L1 is going to be used. But I think it becomes an easier problem to solve.
Now let’s get into the trade itself. I’d love to just, if you could walk us through, like just step by step how it went down.
We started with two assets that we wanted to exchange already.
On the one hand, we had Singapore dollars and on the other hand, we had Japanese Yen. We actually tokenized Singapore dollar deposits. This is a term that most people won’t necessarily be familiar with, but a deposit, a tokenized deposit is really a liability of the bank. It’s a liability of JP Morgan.
When you open up a bank account somewhere there’s a liability that you have that gives you the right to go and withdraw the amount of money that you deposited. This is exactly what we established with through Project Guardian tokenized deposits.
This is a different construct from a stablecoin. A stablecoin is fully reserved. There are a lot of merits around doing that because people feel very confident that this asset that I hold on-chain is fully backed by some real-world asset that exists off-chain and at any point in time I can go and redeem which is great. That has solved a very clear problem in the crypto space because of the volatility of cryptocurrencies.
The challenge with stablecoins is when we actually think about processing and payments at scale. So to give you an example, JP Morgan currently processes $10 trillion worth of payments every single day. To put that into perspective, that’s the entire crypto market as of today, every two hours. So every two hours we’re processing the entire value of the crypto market and actually, it further translates to processing the entire value of USDC. Every 10 minutes. Or every six minutes.
The scale that we have to operate at when we think about traditional finance is vastly different from the scale that the crypto ecosystem is at today. When we think about the stablecoin model, if you’re gonna have to lock up literally trillions of dollars of assets, to fully reserve them, you’re actually sucking a lot of liquidity out of the system.
It becomes very difficult to create money, new money supply, and credit extensions for loans, mortgages, car loans, etc. So we think that actually having a tokenized deposit model where you still have all of the guarantees of getting your $1 out for every dollar that you hold in token form, this can scale in a much better way.
Now, the obvious thing that’s important here is you are reliant on the credit risk of the bank. You need to make sure that this bank is stable and not going to not be able to return your assets, but that’s where we have important things like regulation that ensures that in the US for example, your FDIC insured up to some amount and typically you will be able to get your value out.
An important point here was a tokenized deposit as opposed to a stablecoin.
So the JP Morgan branch in Singapore received deposits from clients in Singapore and then in exchange for those deposits issued a token that gave those users the right to exchange those tokens back to Singapore dollars at a one-to-one rate whenever they wanted to get their money back.
That is essentially correct. There is like a few nuances in terms of how the tokens are actually created in the first place. The right way to think about them actually is bear instruments. The token is the deposit in actual fact. But yes, what you describe is correct. You as a holder of that Singapore dollar token have the right to go and exchange that for actual Singapore dollars at JP Morgan.
And the reason why this is different from a stablecoins is because JP Morgan is not required to be actually holding Singapore dollars in its account at all times. It can go and use those funds in other ways.
That’s correct. So the world today works on a fractional reserve banking system. As a bank we will hold some amount of money at a central bank and then for every deposit that we take, we can potentially utilize some of those funds to extend credit elsewhere, like car loans or mortgages, etc. So this concept obviously is quite jarring in the crypto space because there is an inherent trust that you have in the bank that you’re gonna actually get your deposits back.
But it is important to note that this is how the world works today. As I said, it’s important to be aware of the creditworthiness of your bank. We think that JP Morgan is a very reputable institution in the world, clearly, and that your deposits are extremely safe. We have trillions of dollars of assets on our balance sheet.
There is actually essentially zero risk in this model. As I said, the reason why it’s actually important is because we’re thinking about this at scale. We’re not thinking about the current crypto environment. We’re thinking about what it looks like if we need to be processing 10 trillion, 10 trillion worth of payments every single night.
That’s just on the payments front. That’s not even looking at the size of our markets business, for example. So we need to consider the scale and the tools that we have available to do that.
How do you think this fractional banking can work in crypto? Because it just hasn’t so far.
All of these things actually are being addressed in various regulations, specifically for tokenized deposits, but are already addressed in today’s banking world. So you as a consumer, we as consumers, we have assurances through regulation that we can actually rely on these institutions.
Now again, I fully recognize that this is a little bit “anti” to where the crypto industry is trying to go, and that’s totally fine because I think the point here is that there should be optionality. If you feel comfortable with using USDC, and that is the model that you want to support, that’s your prerogative.
The entire sort of ethos of crypto is inclusion and being inclusive and having optionality and being open. For a large bank like ourselves, that model unfortunately has limitations. And so we need to be able to offer our clients a different model that they can still trust us on.
Our clients trust us. As I said we have $10 trillion worth of payments that we’re supporting every single night. Those clients are trusting us to make those payments every day reliably. So, I think the dial of trust changes a little bit depending on what the use case is.
The other important thing to call out here as well is part of the challenge with like FTX and Terra Luna is this idea of having a token that sort of enables that ecosystem. That token, the collateral it’s social collateral in the sense that it’s what people believe the price of it to be. It’s not a hard asset that’s backed by anything necessarily. I think we were seeing quite clearly that those models cannot work. Obviously, USDC does not follow that model. USDT does not follow that model, but I think we’re seeing these token enabled stable coins, or even collateral uses, as not being something that’s going to, really work even in today’s ecosystem going forward.
I agree. Okay, so let’s move on. Tokenized deposits. Where did those funds come from?
We have a trading business internally so they have access to funds. So we tokenized funds that they typically have access to. We did a very small trade.
How large was it?
A $100,000 Singapore dollars, which was roughly 10 million Japanese Yen. Part of the reason why we kept it small was because we wanted to make sure that everyone was comfortable with what we were doing from a controls risk compliance, audit, legal, and regulatory perspective.
When you start to reduce the value of a transaction, people just naturally get more comfortable. If something goes wrong, it’s not the end of the world. So we tokenized $100,000 dollars of Singapore dollars. SBI who is the other party to the live trade tokenized the equivalent amount in Yen. So about 10 million Yen, and this is where we actually get into the transaction itself.
In normal life this looks like is just an FX transaction, like a spot transaction. You are swapping Singapore dollars for Yen. People ask you why you think you need to do that with a public blockchain. It’s true that the FX markets work pretty well. But again, what we were trying to actually understand here is whether could we use DeFi to simplify not only the trading aspect but also the settlement and the post-trade processing.
As we know, when you are actually transacting through a DeFi pool, the trade is the settlement. There’s no sort of post-trade reconciliation and trying to move assets around. It all happens automatically. And so that’s some of the things that we wanted to actually explore. Another sort of weird thing I guess here that we ended up doing was we used Aave for this transaction.
Aave is obviously not an FX protocol like Uniswap but we used Aave because it already has the ability to provide some permissioning through the Aave concept that they have. So this was already made out of the box deployment that we could use to provide some level of permissioning that we could ensure that only JP Morgan and only SBI could actually participate in this transaction.
Then what we actually did for the transaction was JP Morgan deposited the $100,000 Singapore dollars into the Aave pool. We obviously had to set up the pool in the first place. SBI deposited the 10 million Yen. And then JP Morgan borrowed, basically the full amount, the 10 million Yen and SBI borrowed $100,000 Singapore dollars.
The net effect of all of that is that we started with Singapore dollars and we ended with Yen and SBI started with Yen and ended with Singapore dollars. And that’s essentially an FX transaction. And then we reversed that.
Why did you have to borrow? Why wasn’t just a clean swap of Singapore dollars for Japanese?
Really simply because Aave doesn’t have that capability, it’s not an FX protocol like Uniswap. Uniswap doesn’t currently have, at least, the sort of permissioning aspect.
It would’ve been a little bit more difficult for us to actually build in some of these compliance checks that we actually needed. But this is also an important part because I spoke a little bit about the fact that we have this identity aspect. The ability to identify who was actually participating in the pools, but also identify the traders themselves.
We used something called verifiable credentials for this. Verifiable credentials are an emerging standard. We’ve actually been working with them for probably about three years now. The standard itself has been evolving over that time, and so we’ve been keeping up with it. But we use verifiable credentials to provide fine-grain access control to these DeFi pools.
The Aave system essentially allows you to list a set of addresses so only these addresses can actually interact with this DeFi pool. But that doesn’t quite solve the problem for us from an institutional perspective because what do you end up doing is allowing listing like all of JP Morgan or this specific business within JP Morgan. There’s like some address that represents an entity or a business.
We needed to get much more granular. We actually needed to prove that this specific trader was allowed to participate in this trade. I shouldn’t be able to, I’m not on the trading desk, I shouldn’t be able to deposit JP Morgan’s funds into this Aave pool and actually, trade with JP Morgan money.
This is where the verifiable credentials piece comes in. In my mind, some of the most exciting work that we did.
We built an on-chain verification system. Basically a series of smart contracts that can verify or validate verifiable credentials and validate the accuracy of them. The specific example here is our traders were issued a verifiable credential that says they are an authorized trader. JP Morgan, for example, was issued a verifiable credential that said we are an authorized entity for this particular pool. Then when the trader actually came to do the trade, depositing the liquidity or, actually doing the borrower end, they had to attach their credentials to the trade instruction. So you’ve now got the trade itself and the identity information all traveling at once in, one atomic unit on the blockchain and being verified prior to the transaction actually hitting the Aave pool in the first instance.
Now you’re a verified trader. You attach your credential, you have this on-chain verification system that says, “Yes you’re good to go”. At which point the trade actually happens.
The flip side, of course, is if you don’t have that credential, you’re unable to actually interact with that, in this case, the Aave protocol. But going back to what I mentioned at the start or around the principles, here Aave doesn’t have to know about verifiable credentials at all.
We provided all of this ability to have fine-grained KYC without Aave having to make any changes at all in the front. I think is some of the most exciting stuff and really can be used as a blueprint for not only how institutions can use DeFi pools but if there is more of a requirement for consumers to also KYC at some point, this is potentially a way that it can be done without these DeFi pools having to worry about the KYC aspect.
That’s super exciting. As you said, it’s been such an important topic lately, with Tornado Cash, OFAC sanctions, and DeFi protocols scrambling to keep up with these sanctions so they don’t break US laws.
That has meant that DeFi front end has had to be censored, which goes against the very idea of DeFi. So, if you can come up with a KYC and identity process that lets you keep these permission protocols, I think it’s a huge development. It’s a really big step.
I think so and the last thing I just wanna add here is, because the verification happens on-chain through smart contracts, that means that those acts of verifying some identity credential can become fully reusable and pluggable. So in this project, we issued this authorized trader credential, and that credential is used to access a DeFi pool.
We have the ability to verify that credential on-chain now. So why can’t another process, another Dapp, utilize the same verification and now know how to verify an authorized trader? If we were doing a bilateral trade, for example, with SBI, instead of going through this DeFi pool, they can actually just use exactly the same verifier to ensure that they’re actually trading with someone who’s compliant without actually any of that trader information being stored or submitted onto the blockchain.
We wanted to have identity without the loss of privacy. I should not be submitting on-chain and storing forever my name and my address or whatever other information about me is needed. That should all be mine. That should all be private to me. And that’s what verifiable credentials I think, really shine.
It’s this plugability and this reusability component of what we designed that I think is almost the most exciting thing because you start to get to programmable identity. We speak about programmable money. Now we have programmable identity as well.
How does it work? Who issues the credential? Was it JP Morgan who had to give the trader his badge to do this trade? How would this work for a regular user that wants to KYC? How do you see it scaling?
In any identity model that can be trusted ultimately you need a set of trust anchors/identity authorities. This is how our world works today. Your passport is issued to you by your government. Your driver’s license is issued to you by your motor vehicle licensing agency.
Your employee status. My employee status is issued by JP Morgan. They are the authority and the identity authority of that specific piece of information.
In the project, specifically, exactly how you described it, JP Morgan is the one that issues this credential to our traders to say, “You are authorized to trade on behalf of JP Morgan”.
We actually embed within that credential the fact that JP Morgan itself is an authorized entity. So we’re not saying any random entity in the world can now go and create these credentials and you just have a free for all. There has to be some level of trusted authority that says, “Yes, you are allowed to participate”.
I think how this works at scale is there will need to be over time a set of standard identity authorities that can issue these credentials to people. Maybe over time, it is your government that issues you a verifiable credential that says you’re a citizen of whatever country. In fact, Europe is already making a lot of progress on this front.
There are regulations coming in the next year or so that essentially will ensure that any citizen within Europe has access to a digital wallet so that they can have digital identity credentials. Now, this is not something that necessarily will give them some sort of surveillance control over you. That’s not how Europe, at least, is setting that system up.
Some of the guarantees around that are the fact that they’re actually supporting this idea of verifiable credentials, which are fully off-chain, privately held credentials that you as the person or the entity have the rights to share with whoever you want, only the information that you want to share. It really is user-controlled.
I think the scaling aspect does need these identity authorities to actually come into play. And they’re already, lots of efforts around this, and some notable authorities globally that are actually coming into the space to do this today.
If you had already given these verifiable credentials to the different parties, why did you need one to begin with?
We wanted to do a trade through a DeFi pool, so we didn’t want this to be just a bilateral transaction.
But why? You mentioned you couldn’t use Uniswap.
So one was just the permissioning of the entity itself. If you think about a DeFi pool that exists in the world on public Polygon or Ethereum right now there’s actually nothing stopping anyone from depositing liquidity into that pool and trading against that pool.
What the Aave Arc model actually allowed for was essentially putting up guard rails around that to say, “Not anyone can actually interact with this pool”. The reason why this is important is because what we have tried to solve in this project through Project Guardian was the idea that we may not necessarily know who we’re trading against, but we know that we’re trading against someone that has been KYC to some standard, is a known actor that is not on a sanctions list, for example.
So we can keep some level of anonymity perhaps, but with the assurances that we are doing business with a well-known actor and you need that sort of base layer guard rails around the DeFi protocol in the first place to get to that.
So essentially there were two levels of permissioning.
First, there were the guard rails around the liquidity pool on Aave that ensured that anyone interacting with that pool was either from JP Morgan or from SBI or the institutions that you wanted to be transacting in that pool and nobody else.
The second level was within JP Morgan, there were specific individuals who were allowed to do this trade, and that was achieved through this verifiable credential system.
Exactly right. That’s exactly right. You can imagine over time that you can also have verifiable credentials that say, “You’re allowed to trade up to this limit and only on these assets.”
You can get very specific and granular around the types of access that people have. This is important from an institutional perspective. We need to ensure that people are not able to just trade whatever they want up to whatever amount they want. There are controls in place that we have to adhere to.
This is, potentially, a way for us to be able to actually do that.
What you did there was you had this $100,000 dollars worth of Singapore dollars. You traded that for the equivalent of Japanese with SBI through this Aave liquidity pool and then you did the reverse. Everyone got their money back.
Yes. And if you think about it, Aave truly out of the box, there’s an interest rate, right? So interest will accrue, as you’re borrowing money or, accrue to you if you’re lending money.
There are also liquidations that occur if you fall below your collateral value. So one of the constraints and controls that we put in place for this trade was that we wanted it to be a zero P&L trade. We didn’t want to get less than $100,000 thousand Singapore dollars back at the end of the trade.
The modifications that we made to Aave system were:
- Actually essentially setting interest rates to zero, as close to zero as you could get.
- Deactivating liquidations.
- Fixing the interest rates.
That allowed us not only to ensure that we were able to trade at market, it’s really important that, if the FX rate in the market today is one Singapore dollar equals a hundred Japanese Yen legally, and from a compliance perspective, we can’t trade away from that price.
We have to ensure that we have to trade at market. So we set the FX rate through the protocol as well. We did have the zero P&L set up. It would have been nice to be able to see interest accruing over time through the Aave pool as we were borrowing and lending. But we’ll obviously look to do that in a future state, I would say.
With interest rate rates at zero, and no liquidations, what were the fees associated with this trade? Was it just like Polygon gas fees? How would that compare with what you would have had to pay in fees in a traditional effects trade?
So yes, all the fees were in MATIC so we paid in MATIC using the gases provider. We used Biconomi in this case. I can’t remember exactly how much it was. I think it was something like $18 for this trade. That’s probably a little bit more than what you would actually end up paying for a larger trade on the FX market, but that’s okay as well.
There are a bunch of things that are embedded really within this trade that are not necessarily embedded within a standard FX trade. A standard FX trade is, struck at some fee amount, but then you have all of this post-trade processing to settle the transaction and move money around, and there’s reconciliation. All of that’s encapsulated directly in the trade here through the DeFi pool.
We haven’t done a full assessment. That wasn’t really the goal of the project in terms of how much cheaper, if at all, this is. That is certainly something that we would start to look at. But quite honestly, that’s probably not gonna be the driver to, why we would or would not do this at scale.
What would the driver be?
- Whether or not we can start to rarely rethink how our current systems work and do so in a way that can scale. I talked about how much money we handle every night. The FX market is much bigger. Could we actually get there? And if we can, are there long trade asset pairs that we couldn’t normally trade that are more expensive to trade, that we could now possibly trade through this pool? Are there currency pairs that are a little bit more difficult to trade? Are there things that we can do outside of standard trading hours?
- There are lots of currency pairs in the world that have very narrow windows that they can actually trade and settle in. So can we do that, real-time, 24/7, or some given currency pair because we’re now utilizing these blockchain protocols. Especially if you think about AMM, where the price is actually, at any point in time you can get a price, then, that potentially is interesting.
- New products. So composability of DeFi pools and different DeFi protocols. Can you actually create something new that we can’t do today? Obviously, there is a commercial aspect. Can we make money? Are we disrupting our existing businesses? Can we save money? I think those things are a little bit harder to quantify at this stage, but certainly will be a factor going into any sort of project that we take forward.
What are the new things you can do besides just improving efficiencies? How do you imagine traditional banks can start to interact with DeFi in a way that they can offer new products?
I’ll give you one example that we actually have live today that is not using a DeFi pool, but it is using smart contracts and blockchain infrastructure that could quite easily actually be converted into a sort of DeFi.
One of the networks that we have, one of the EVM side chains that we have is called Onyx Digital Assets. Onyx Digital Assets is this network that is really centered around tokenizing traditional assets. We have an application on that network. It essentially makes it easy for banks, broker-dealers, and our clients to borrow from JP Morgan on an intraday basis and on a secured basis.
So what that actually means in real words is that we have clients who need to get access to some sort of funding during the day. Let’s say they wanna borrow $500 million from JP Morgan. Right now, the way that they do that is by drawing down on a credit line, which can be costly especially when they have to hold assets on their balance sheets.
In times of stress, they need to prove that they can actually repay JP Morgan. So what we actually set out to do was to create an intraday borrowing facility that allows our clients to put up collateral, and by putting up collateral, they can actually borrow at a cheaper rate.
And the way that we did this was we used what is called a repurchase agreement, which is a very standard financial instrument. It’s one of the largest markets globally. The repo market. It is basically core to the sort of global financial system. But the challenge with a standard repo is that it only settles either overnight or term, like two weeks later or some other point in the future.
It’s not a financial instrument that actually settles and matures on the same day. We developed a system and a set of smart contracts. That means you can tokenize some assets like US treasuries, for example. You can put up those assets as collateral and then you can borrow tokenized money off the back of that.
Now, that sounds very similar to what Aave does. That’s basically the same construct. We haven’t used a DeFi pool. This is all happening on a bilateral basis. But that is an entirely new product that did not exist in the world and could not exist in the world because of the fact that our systems and processes today don’t actually allow for you to borrow money from me in the morning and return the money to me on the same day and for you to get your collateral back at the end of the day.
We’re doing a billion dollars a day through this product. We’re just coming up on 500 billion of traded notional through this application. That is real money that’s moving every single day using blockchain and smart contracts. These are ERC20 tokens that, potentially over time, could be even further streamlined through something like a DeFi pool.
So interesting. What kind of clients are using this?
We have banks and broker-dealers that are using this. We have a number of asset managers that are looking to onboard onto the platform as well and they’re using it for their intra-day financing and funding needs.
That’s in the US?
US right now, but actually just expanding to Europe and Asia.
We announced Goldman Sachs is on the platform. BMP is on the platform as well. There are a number of other banks and other financial institutions that are in the pipeline to come on board as well.
I think the really important thing here is that this is one product that they have access to, but because they’re now on this shared blockchain infrastructure, they also get access to other products that we create as well. You start to really get these network effects and be able to use tokens that are being created in one Dapp over in this other Dapp that is being established and created for some other purpose.
So you have Goldman Sachs tokenizing assets to use as collateral and draw credit lines over. Do they have to pay them back on the same day?
Exactly that. The construct is an intraday trade. So Goldman will tokenize US treasuries, for example, in the morning. Let’s say they tokenize a billion dollars of US treasuries. They will then post those tokenized collateral assets as collateral.
Then JP Morgan will post in this instance JP Morgan coin, which is tokenized. And basically, it’s actually a blockchain-based bank account. So they have access to these funds on-chain. Goldman will then use those funds for whatever they need to do during the day. Then once the trade matures, say three or five hours after the settlement the swap will happen in the other way. They will get their collateral back and we will get our cash back.
Would they need to exchange JP Morgan coins for dollars? How would that work?
Right now they would because they’re looking to actually make payments outside of the system. They would effectively redeem that amount of JP Morgan coins for actual dollars. Those dollars would be available in their actual bank account at JP Morgan, and then they can go and make a regular way payment.
You can imagine over time as more institutions come on, that you don’t even need to redeem out of JP Morgan coin or out of whatever the tokenized deposit is. You could potentially just make those payments directly on blockchain.
And right now I imagine that’s happening on the private chains.
That’s correct. Some of the reasons for that are really around regulatory challenges and also privacy.
So ensuring that we have exact definitions of what these tokens are, what happens in the events of default, how those tokens are actually treated and handled and processed, that we can ensure that you have title over the actual underlying assets if you hold the tokens. A lot of that actually requires quite a constrained boundary in which to operate so that, legally at any point in time, to what these assets actually equate to.
And so for now, there are reasonably strong reasons why those types of trades need to happen on a commissioned side chain. But this is part of the reason why we’re taking this step with Project Guardian is because we want to explore how this could actually look on a public blockchain.
How likely do you think it is that these transactions that are happening on the private blockchain will start to happen on the public chain?
It’s gonna take time. It will take time. We still have not solved it.
I mentioned privacy a number of times, but we still do not have that fully solved. There are obviously lots of solutions that are in the works. One of the reasons why we decided to do this on Polygon is because they have their nightfall implementation in the works.
And so we thought that would be an interesting privacy protocol for us to explore in time. It’s not something that we actually used for Project Guardian. But, we need privacy in place. We do need to get to a point where there’s obviously a lot of conversation around scalability utilizing Rollups and L2s.
It’s actually really important because as I mentioned, we can’t have the situation where we are trying to do this FX trade, or if you think about these intraday trades that need to settle at a specific time, we can’t have congestion on the network actually preventing that supplement from happening.
Then you have considerations around MEV, what does that actually look like and how do you go about ensuring that you’re not somehow being handicapped in any way. So there are certain things that we still need to work through and that’s not even talking about the regulatory side of things.
But I think what we’ve shown in this project is almost a blueprint for how it could be done and something to work from. Something to say to our regulators. And we’ve obviously had to have lots of conversations with all of our regulators around us at least relevant to the transaction.
Getting them comfortable with, “Hey, we’re using Polygon and this thing called Aave, and here’s how gasless transactions work. Ensuring that we’re not holding cryptocurrencies. Here’s how we’re ensuring that this thing is gonna settle and that there’s not gonna be a failure”.
All of these things are little things that move the ball forward, and that ultimately, hopefully, can get our regulators comfortable with this type of work.
And, I’ll come back to the identity piece. That’s a big one, right? There is clear concern around enabling money laundering and falling foul of some sanctions lists, inadvertently, even if we can get to a model where we have compliant access in a privacy-preserving way with guarantees that you’re not doing something that you shouldn’t be doing. That is a big step forward.
Assuming that all of these problems, like privacy, scalability, MEV, and regulators getting used to crypto and blockchain, can be solved in some way, what do you think crypto’s role will be in the financial system in, say, 10 to 20 years? What do you think the future of money will look like?
This has been a tough week. We’ve had the FTX blow up. I think that has shaken a lot of people, very frankly that has not installed confidence in the areas of the financial ecosystem that you want to install confidence in.
For the right reasons. There are questions being asked and unfortunately/fortunately, depending on where you stand on the spectrum, that is going to result in more regulation.
My personal view is we want people to feel safe. We don’t want people to put up some assets that might be their life savings. It might be something that they’re looking to do something really meaningful in their life with, and then all of a sudden they don’t have access to that anymore. I don’t think anyone wants that world.
So I think if we can through projects like this educate the benefits and also how some of the risks can be mitigated, create solutions to solve some of the challenges that people have, then I think we can start to make progress.
I firmly believe I’ve been doing this for seven years. It feels a little bit like a kamikaze mission some days where we’re hell-bent on being able to make these changes because we believe in it. I do firmly believe that we will get there. I think the future of finance is going to have a very strong blockchain element to it for settlements of traditional assets, for payment rails, for businesses running their infrastructure on a public blockchain.
I’m not super clear on exactly what happens to the crypto aspects, meaning the crypto markets. I think that, and also, personally, that’s not the most interesting thing for me. I’m very interested in these new use cases that we can develop. And I’m less interested in the price of any given asset.
I think that crypto as an asset will have someplace, but I think the jury’s out in terms of how big that can be. I don’t really have a doubt, right now at least, that blockchain and the infrastructure that the public ecosystem is creating, is gonna be an important and a big part of finance in the future.
Such an interesting take. It’s become a cliche for some people to say blockchain, not crypto. And it’s kind of an old criticism of crypto. To me, it’s just a very kind narrow-minded way of looking at the space.
You’ve given it a new spin. Traditionally when people say blockchain or crypto they’re referring just to private blockchains or stripping all of the tokens and digital assets from the equation.
In your case, what you’re saying is exciting. Blockchains in a much kind of broader sense. Public blockchains and maybe not cryptocurrencies like Bitcoin, Ether, but digital assets that are linked to real-world assets.
I would add one slight nuance there, which is maybe something like blockchain enabled by crypto. On public blockchains you absolutely have to have a cryptocurrency for the incentives to be aligned. They just do not work in the public sphere without that. As long as there is a blockchain in the public ecosystem, there will be cryptocurrencies to enable that. And that’s some of the greatest inventions that we’ve seen are exactly that. Being able to incentivize behavior economically through these cryptocurrencies.
I think where the story of blockchain gets a little bit lost is in the token projects that fly by night trying to create rich quick schemes per governance around centralized actors. That is really tainting the space in a way that is not helpful, not actually necessary, and not going to be part of what the future looks like.
I do see public blockchains enabled by cryptocurrencies supporting really strong use cases, like some of the ones that we’re looking at now.