SmartCon Live Insight》How fund managers and payment and settlement companies use AI and blockchain to transform finance

Contents of this article

Opening and guest introduction

Ryan Lovell: Well, thank you all for joining this panel, we are actually sitting on the same panel now... This panel is around the modernization of financial corporate actions.

While this in itself is a pretty boring topic, we recently worked with three companies to actually solve corporate action problems in a very interesting way that combines new architectures around the use of artificial intelligence, oracles, and blockchain. , creating a faster, more efficient way to solve corporate action problems. So today's Panel process is going to be divided into three parts, today we're going to talk about the challenges of corporate action; then we're going to talk about solutions, which is the initiative I just described; and finally we're going to talk a little bit about where we want to go in the future. .

So to introduce you to the people I'm sitting with and some of my panelists, so we have Stéphanie Lheureux from Euroclear, she's the head of digital assets at Euroclear; Euroclear is an international securities custody and settlement company.

They hold 40 trillion euros in custody assets, they settle transactions worth 1 trillion euros a year, oh, four trillion dollars, I'm right.

We also have Whikie Liu from Swift, who is the director of capital markets strategy at Swift; Swift is a global messaging system and financial trading system.

They are connected to more than 11,000 banks and financial institutions in more than 200 countries, and they process approximately 50 million messages and transactions every day.

Last but not least, we have Andrew Crawford , who is Vice President of Digital Assets at Franklin Templeton; a global investment asset manager with approximately $1.7 trillion in assets under management, and they have Operating in more than 165 different countries.

Guest list:

Traditional finance enters the corporate action challenge of blockchain technology adoption

Ryan Lovell:

So many guests are here to talk about the problems and challenges encountered in corporate action. Corporate action has been a continuous challenge for financial markets for many years. What makes this issue particularly important to solve this problem now, especially As we move toward tokenized assets?

So why is this really an interesting question at this time? This has been talked about for some time. Maybe I'll start with Stéphanie from the market and I think we'll start with you.

The importance of message unity

Stéphanie: Yeah, hi everyone, first of all I'm not sure, when we talk about corporate actions, does everyone know what we're talking about? A corporate action is anything that happens during the life of a security, should it be a bond or something like that? If you have different types of corporate actions that are very value-added to the company, like dividends or coupons, each of those is valuable, especially if you have some of the more exotic events and very corporate actions.

What's interesting, at least for the notification part, is that this is information, this is public information from the issuer, but this information has to be distributed and travel through the entire value chain until an investor buys a side.

The way we do this is by transmitting this information from one entity to another through messages, when you are a buyer at the end of the value chain, you receive this information from various participants, but sometimes this information is inconsistent.

You have some discrepancies, information that should be the same is different; sometimes you receive it from one side and not the other, so you don't know which one is correct.

So (message inconsistency) is a problem for everyone in this delivery chain, especially for the buyer, as financial market infrastructure, we are concerned about solving industry problems for the users of the infrastructure.

But it’s also a problem for us as CSDs, custodians and our participants.

We are all doing the same thing, which means receiving corporate action events, processing, validating profiles, re-creating and re-sending profiles.

This would not bring value, the value would be having a source of gold and having the same source of gold available to all participants in the capital markets.

Whikie Liu: This is really a topic that we collaborate on in the best possible way. Yeah, if I may, I think as Ryan mentioned, collaboration is not really a new challenge for the industry.

As Stéphanie mentioned, this industry (problem) has existed for many years because there is a lot of unstructured data that needs to flow to the buyer and be processed. So as you can imagine, in the middle, there are a lot of time inconsistency issues and data quality issues.

But we've been talking about it for years, so why are we talking about it now, and why from today on, it becomes even more urgent and important.

I think there are some industry trends that are really driving the need for change, for example, the first one will shorten the second cycle of secure transactions.

The shape of T+1 in the United States, you can also see a lot of announcements in different countries, they want to move the security second cycle from T+2 to T+1. That has shortened collaboration timelines, and the industry is pushing for faster, larger turnarounds for collaboration processing.

Another is that we could also see another regulatory push for transparency and standardization.

For example, in the EU, the SRT2 regulations, they are really positive compliance requirements for shareholder transparency.

We are also seeing a lot of security market structures that are either updating or planning to update their cooperation materials to ISO 20022. So all of this means that the industry really wants to get into automation and standardization.

The last, but certainly not the most important, is the rise of digital assets.

I'm sure most of you (web3) who know how here have a stronger sense than I do, that you can see the heightened development of digital assets, especially tokenized real world assets (RWA).

And then we saw that there was a real need for real-time data flow and better interoperability when it came to collaborative processing.

Ryan Lovell: And then Andrew, what about from an asset manager perspective?

Andrew Crawford: Yes, I look at it from a user perspective. I think BCG released a report in 2022. I think that by 2030, blockchain financial assets will be 16 trillion US dollars in size, two years now. It's past tense that (I believe) at some point we have to see this massive growth. The thing that bothers us the most about this is that we now only offer one primary market fund and invest in U.S. bonds and back deposits, which is the most liquid market and easiest to record on the blockchain, it is immutable of.

If we can't do this, we should go home early.

So I think what this project does, is take the unstructured data from the channel, structure it, and put it into the code in a way that everyone can use it simultaneously, now will allow us to Overcome this limitation of only providing money market funding.

Now will allow us to offer bonds, but that's really more in equities, and that's where we perceive the problem, so I think from our perspective, it allows us to get closer to people, invest in chain diversification investment portfolio.

So far you know, you can buy money market funds and some bonds, but that's about it.

I think people, you know, want to own some stocks, they might want to, you know, there's a lot of talk about real world assets, access to private equity, and I think we just do some basic assets that people are familiar with, you know, this The project is getting closer.

I think that's one of the things about the growth of assets on blockchain systems, it's going to tilt, you know, that's going to allow us to get to a place where it's basically, you know, predictive, so, yeah, it's a big moment .

Ryan Lovell: Yeah, because Benji* is one of the most successful money market funds, I think probably $400 million.

Note: Benji is the first crypto private equity fund launched by Franklin

Andrew Crawford: Most successful because we have the highest yield, - Yeah, that's why you mentioned success, we're getting more money back on our investment. So, I guess that's the measure of success.

Ryan Lovell: Yeah, I think what a lot of people don't realize is when people send you money for Benji, you're taking off the chain to buy traditional Treasury bonds, fixed income instruments.

Andrew Crawford: Correct.

Ryan Lovell: What you just described is that if these assets are on the blockchain, you can't service those assets without corporate action and so on.

Andrew Crawford: They could be on the blockchain. You know, we're looking at having a master fund, which is the fund, and then we can tokenize that online, and we can only do that if we have corporate action.

How to combine traditional finance with AI and blockchain

Ryan Lovell: Maybe off topic a little bit from Andrew, but also very, very exciting, thank you for everything and your thoughts on this, and now we're going to talk about the project that we're all going to successfully complete,

I think internally, we call it Project Calm, credited to Angie Walker, which orchestrates the entire campaign lifecycle management. But from the outside, I think this is just blockchain in oracles and asset services.

So what we really wanted to achieve was just a brief overview of some of the achievements of the initiative, and to give everyone a little different perspective, we tested corporate actions in six different European countries, Euroclear is one of our main sponsors, so We focus on the European region. We integrate three major AI LLM models, Claws AI, OpenAI and Google Gemini.

As Whikie said, we're getting unstructured material like this, and this paper is here, and it's just part of it, but there's critical material on this paper that needs to be structured, and that's for all market participants. meaningful.

You can call on artificial intelligence to ask certain questions about the document, such as, who are the panelists in this group now?

And it can be returned and then compared with other AI LLM models , and we were able to achieve this successfully.

I think some of the asset managers in the pilot are calling this farm-to-table data, where you use automation to take data from human-readable files and put it into machine-readable chunks in the chain.

We've achieved some unified gold records on the blockchain, so we asked our panelists some pointed questions, the first one is for you, Stéphanie.

This concept of a unified golden record is central to the solution, how does this change market infrastructure and the relationship between participants?

Unified golden record standard

Stéphanie: Yes, I'm going to answer that question, but I'm going to answer the question you asked first.

Going back to why we're on this group and why it builds on what you said, when you said, well, we need to collaborate as we build the digital asset project, that's for sure, but for now, this ( The number of collaborations is still limited.

What I love about this project and the approach is that we are actually solving a precise niche in the traditional asset industry, which is a huge opportunity in the market. There was a recent study that mentioned value exchange, which is millions of dollars per institution per year, which means that globally, this is a very large amount and business opportunity, and we are still working hard to improve the efficiency of digital asset adoption.

So it's technology that enables capital markets to become more efficient and solve real problems, which means it has a real opportunity to move into the next stage of adoption, production and scale, which is super, super interesting to us.

And then, one more thing, at the end, we tested a workflow, which was a workflow that used artificial intelligence to build golden records, but that wasn't the first idea of ​​the design.

There were a few iterations between us, but there were also different participants in the project, and at the beginning we were actually saying, maybe we could use fast messages as input, or we could use other types of structured data.

Doing this collaboration with marketing, what's the best use or the best setup and the best workflow that we can design together to create this golden record?

We finally chose the LLM model.

The question then becomes how accurate and trustworthy we can place the information established in this golden record.

If you only use one LLM model, and in fact we have several, we need to make sure they are all consistent and we can build trust in that material.

Because in the end, if we use this information to transfer money, that's a lot of money, real money. So we need to make sure that as an industry we trust the data.

Stéphanie: Next I want to go back to that golden record of Unified, and I like the name Unified because I guess you know, especially in the capital markets, we've been building a lot of DLT platforms, and there's a lot of them, so creating and so on. .

We are now moving towards the concept of a unified ledger , which has been very popular recently.

As a concept of a unified golden record, to me, the fact is that we have established this record, this information, this information is trusted and can be used, and then distributed throughout the industry, on various blockchains .

How we ensure you maintain the integrity of your data throughout its lifecycle.

So it's super interesting for us to have a unified golden record and reorganize the capital market structure, which is the second demonstration today, that it can be restructured and all have access to the same golden record blockchain.

So that's a huge value and it's really going to eliminate inefficiencies, eliminate frequency issues and then allow us to process faster and go to TPC1 or T0 data sets.


Ryan Lovell: Thanks, Stéphanie, I think what we're doing with this pilot is not just getting the material on-chain, it's also distributing it across different blockchain ecosystems.

We actually have three different blockchain partners in this pilot project and broad access has been this key theme as everyone from different jurisdictions, asset managers, market infrastructure have different requirements for the areas they use. Blockchain types have different preferences, so this is another key point of this project.

On this Whikie, I'll turn it over to you and let's talk about standards and SWIFT messaging, right?

Therefore, we can get a unified golden record on the blockchain.

On the other hand , Andrew's team and all asset managers have systems in place that need to capture this information.

So talk to us about connecting blockchain to existing systems.

Existing blockchain wired into the system

Whikie Liu: I think the beauty and dignity of artificial intelligence in this project lay the foundation for everything.

It's not just about solving old challenges in collaboration, like reducing errors or reducing costs, but at the same time, we're starting to think about how to move into a more integrated and digital financial world, like digital assets, so we're putting that together.

So from a SWIFT perspective, it's really about three things, data status, market practice and interoperability .

This is really not a new process for the industry when it comes to collaborative material standards and market practices. We're having a discussion, and we've made a lot of recommendations over the years, but having standards doesn't mean we have structured material, right?

So, for example, if the original announcement is really paper, like a PDF, which is very typically unstructured material, that means you need someone to read that paper and then manually convert it into a structured format, although that This can be agreed in advance by the SNPG or the participant.

So you can see that AI can really have a lot of value, from unstructured to structured, but at the same time, I think the key component to successfully solving AI is that we need to train AI models to be better Understand and fully understand the appearance of standard materials.

This can be consistent and based on established market practices. In this way, I think data standards can bring value to changing AI models and ensuring that they follow the correct logic and ensuring that unstructured data is indeed the industry standard. Looking for something.

Another angle is around blockchain, where we can see more and more institutions getting into tokenized assets or real-world assets. So you could imagine, as Andrew mentioned, it could be a money market fund, maybe a stock or a tokenized security, and every year they would need to announce their co-buying event and how that co-buying event would be distributed as a source of gold Go back to Andrew in your existing system, but release updates to the blockchain at the same time.

I think that's where interoperability comes in, that's not just doing it quickly and right, basically, we need to work with families, dividends, brokers and chains to make that happen.

Market aspect - fund managers’ needs

Ryan Lovell: Thank you, and finally, Andrew, I think you may have hinted at this, which way are you most excited about getting real-time corporate action data on-chain? It sounds like you want to get all the $16 trillion in …

Andrew Crawford: We're hoping to get the transfer done in this chat, but there's a snag. So what's exciting to me is that it means we can start moving more assets.

Because in the past 10 years, we have acquired many asset management companies. So we pretty much span the efficient frontier, with managers in every asset class and strategy. We want to be able to provide these services and I think, what I said this morning, by using Blockchain, we can really reach a wider audience and we think we're good at what we do and we want to make it more accessible People are exposed to this.

I think a blockchain system will allow us to do that, and I think having this unstructured material, it's the same problem that we have with a collection of multiple ledgers, where you have multiple vendors. For example you have Bloomberg publishing there, you might have Reuters, they have different structures, they have different ways of writing, structuring that material and putting it into a unified source and then putting it next to each other so it's Programmable, is a huge bump.

Whether it's to identify investment opportunities, you know, there's a whole bunch of managers who are executing quantitative systems and now they can execute qualitative systems, you know, one of the problems you always have with qualitative systems is you lose sight of the qualitative because there's no structure information, so I think there's a whole world of opportunity.

I think there are other things that can be talked about, you know, because people want to get more from their investments, you know, whether it's around a value, around ESG, that kind of information can be extracted, you know, from this corporate event In, you know, corporate actions, you know, as a public company, you have to provide certain information, good and bad.

Some of that might go to, you know, this company has a lower ESG score than you think... I want to move my company, I might have a criteria, I say I want a 90-year-old, score on ESG is 90 points.

There might be a corporate event that pulls them off and you might say, I don't want them to execute the portfolio.

So I think, you know, this is where you are in the oracle, this can go to a lot of different places, because I think what we're going to see from proponents of this information is that there are a lot of uses. That’s the great thing about public blockchains, someone will come up with a creative idea to use this information to help people make decisions that are more aligned with their goals and values.

Ryan Lovell: Yeah, I think the ESG part is really interesting. Because of the idea of ​​green cleaning, I can issue a security and I can tell my investors, hey, I want to improve my factory.

It's a dirty factory and I want to turn it into a green factory and use the green proceeds to buy a bunch of exotic cars that no one will know about.

But with transparency, oracles, and data, security is just a data container, which is a smart contract.

CO2 emissions are what you can get from the real world.

Imagine a world where investors were given that kind of transparency.

Andrew Crawford: You could balance that need, you might have a source company, an oracle, which comes from a neighbor that's doing, you know, a forest or whatever, and you might have a report saying we're generating 100. The reality is, you have another office that is producing 120, which can create another event.

Ryan Lovell

Yeah, you see, you know, there's, the asset management industry, we did a panel on this earlier, it seems like active management exists until it's passively indexed.

Yes, you are describing a world where investors have ultimate flexibility and customization.

The only way to do that is to have an open platform where the material is accessible so someone can make these changes transparently.

Is this what you see? There is also valid information, right.

Andrew Crawford: Yeah, I mean, you know, most people don't do finance, take financial advice, they ask you financial questions.

They didn't say, what do you think about these issues? Are these issues important to you? You know, because they have no way of claiming, collecting, and monitoring this information. With this type of data, and through oracles, you can ensure that your investments meet criteria beyond your financial goals.

You'll have lifestyle goals or you may have, you know, family goals, so you can make sure that your portfolio meets all of those.

And, you know, you're doing things that can incorporate all of this into criteria to evaluate whether you're actually on track, you know, we just have quantitative targets.

"I want $100,000." You know, I want to get $100,000, but I don't want any of these issues to arise, I want, you know, a reward or whatever, the whole thing becomes interconnected.

Yeah, the core of decentralization, you know, you act as an intermediary for information, you don't have to donate it to someone else to get it.

Yeah, so, yeah, that's great.

Where does the future go?

Ryan Lovell: Last question. You know, we've identified, you know, several key areas for future growth. Maybe I'll go left or right.

In addition to tying into Andrew's goal on the $16 trillion train, where do you see some of the next steps for our initiative that are most critical in the next phase of our initiative?

Stéphanie: Yeah, so from my perspective, what's interesting is the past, how past predictions have been made about this. Honestly, since the report was released, I've gotten a lot of really good feedback from different participants, even from Ishros myself, but if I can bring that kind of value to my investors, I'd be very interested.

Again they implement artificial intelligence which is interesting, but I can also enter the information myself if I'm sure it can be distributed to investors. So we can see that this is really everything, again, the real game plan for the industry today.

We need to continue to work with you and the participants, the capital market participants, to see what the fastest path to production is.

So you're not going to solve all problems for all company types at once, but I believe we can provide value.

Whikie Liu: Yes, I would like to say two things. One is, we started with a single company activity, and maybe we should try some more difficult and complex company activities.

Specifically, for example, if we can bring in reference materials, we can provide additional actionable insights along with the power of artificial intelligence and then make corporate notifications more powerful and actionable.

The other part is really about, this is an idea using artificial intelligence and D&T, how do we turn that into a real business case, which means we need to think about system integration and market adoption.

I think those two really go hand in hand.

For systems integration, I would say, we really need to make sure we're simple and cost-effective and people can leverage existing infrastructure to deliver new services.

I think we can see this as the next step.

Stéphanie: Maybe before we bring you the last sentence, but just to add, sometimes we have mentioned in our work with NPG or market associations that it is good to have a golden record, but we need to make sure that it is the market standard, It will cover everything.

So all this work and discussions with market associations should be an appropriate direction, but if we continue to talk about data types versus market association types, this is a super critical time for collaboration.

Ryan Lovell: Oh, we’re at market zero!

Andrew Crawford: I just felt like, because I came from a math background, the idea of ​​taking unstructured material and putting it into a structured format that was readily available from multiple sources was a big leap.

I guess that's my thing, I don't know where it's going to be used, I mean, I know where we're going to use it , I know there's going to be a lot of fund managers who can do qualitative screening.

The market is actually a bit blind, and it's hard for one person to get through 500 rules (to complete these filters), but for some codes, it is much easier to do so, and I think there is a good open market here.

Ryan Lovell: We appreciate, thank you for coming on the panel, and we're going to be moving this project into phase two, so we're really excited to announce that with some of the folks here and other market participants.

As Stéphanie said, we had a lot of fun and people came up to us at Sibos and said they read the report and they wanted to go to phase two, so we're really excited about that.

Awesome, thank you for coming.