Michael Scotti

Hi, welcome to the Market Structure Wave. My name is Michael Scotti. I'm editor here at TabbFORUM, and today we're joined by Jim Kwiatkowski. Jim is CEO of LTX.

Jim, welcome.

Jim Kwiatkowski

Thank you. Thanks for having me, Michael.

Michael Scotti

Jim, you have an interesting background. You've been involved in the order management business, the data business, the sale of data, and the transaction business.

Tell us a little bit about your background and what led you coming to LTX.

Jim Kwiatkowski

Sure. So I guess it all started with my wanting to study engineering and coming out as an engineer. I wound up getting involved in the FinTech business very early, long before anyone called it FinTech.

As you said, I was involved in financial information, financial information distribution systems, but then I got involved in electronic trading, and I started with a small equities electronic trading company.

I did that for several years. That company was sold, and I moved on to foreign exchange, got involved with a leader in the electronic trading of foreign exchange.

That company went through a couple of acquisitions, and I found myself, before LTX, as the global head of transaction sales at the London Stock Exchange Group.

I left there and found an opportunity to join LTX. What's really interesting is that I started out in electronic trading of equities and then went into electronic trading of currencies, and I find myself today running a company that's responsible for the electronic trading of bonds.

Michael Scotti

That's an interesting progression. If you would, tell us a little bit about the niche that LTX is in. I'm sure you've got your own acts in the business. Talk a little bit about that, if you would.

Jim Kwiatkowski

Sure. So LTX is focused on US corporate credit, corporate bonds, and in particular, we're focused on solving a couple of problems that we know are evident for all of our clients.

Number one, most of the clients are trying to get larger trade sizes done. Number two, liquidity, is often challenging for corporate bonds.

So finding the other side of any bond trade can be difficult. So we put some artificial intelligence to work to try to help our clients find the other side of trades, and that's worked out very well for us, giving people pre- trade insights into available liquidity.

Secondly, we have come up with some unique trading protocols that help our clients to execute in larger trade size.

Michael Scotti

Sure. Would you be able to talk a little bit about those protocols?

Jim Kwiatkowski

Sure. So the company really started out with a truly innovative protocol called RFX. RFX was built to be a dealer- centric protocol.

So some buy- side participant can say," I have a trade I'd like to execute." The first thing they're going to do is choose a dealer to trade through. On the other side of that dealer, that dealer may have one or many customers that have opposing interests to the customer that started that protocol.

RFX delivers a couple of things that you don't see in other trading protocols. Firstly, the buy- side can trade through a dealer with their other buy- side colleagues, but you can also trade multiple customers at the same time.

Third is transparency. So each of the participants can see the other participants bids or offers, which allows for some price improvement when multiple participants get involved.

Michael Scotti

Jim, you talked earlier about helping clients to get size done when they're trading. Have there been any recent developments there?

Jim Kwiatkowski

Yeah, there have. Thanks for asking. So in the course of rolling out our product to clients, we've taken as much feedback as we can get, and a lot of our clients said that they'd like us to add more familiar protocols to the RFX protocol that we already had.

So we went about incorporating the RFQ, that's very, very popular and probably the standard for electronic trading and corporate credit today. But we wanted to take on this problem of trading larger sizes and most customers are trading less than $ 1 million in RFQs.

So we came up with a protocol that we refer to as RFQ +. RFQ + is very simple. It's intended to be very familiar. It's an RFQ, except, in the end, when the customer is faced with multiple responses from the multiple dealers that they requested quotes from, they can choose more than one.

In fact, they could choose as many of these as they'd like. We wanted to make this friendly for both the buy side and the sell side. So we gave the buy- side trader the option to choose one or more of the responses, but also to choose the responses that they got from the dealer or to use the concept of the Dutch auction so that all of the responders would print at the same price.

We found that providing the buy- side with these options makes them feel comfortable that they're preserving their relationship with the dealers. That's very important.

Michael Scotti

How long has that been live?

Jim Kwiatkowski

So we announced it just at Phil's in June, and we'll only go live in September.

Michael Scotti

Okay. So that's your next iteration of-

Jim Kwiatkowski

We're very excited about it.

Michael Scotti

protocol. Okay.

Jim Kwiatkowski

Yeah. We spent several months at the beginning of this year getting customer feedback, both buy- side and sell- side. As I say, I'm very excited to roll this out.

Michael Scotti

So recent developments, that was the foundation of LTX, and you've been involved in the news quite a bit with this BondGPT, with the generative AI.

Talk a little bit about that, if you would.

Jim Kwiatkowski

Sure. We're really excited about BondGPT. So we've been speaking with a lot of our clients about data, and clients have been asking us to put some of the unique data that LTX has on the screen, coupled with industry standard data pricing, ratings, just security master- style data.

In talking to our clients about how they'd like to see this delivered, we learned two important lessons. One, there's just way too much data, and the other is, there's not enough screen real estate.

So then, in talking to our clients, we're trying to solve the problem of how do we give everyone what they're looking for and minimize the screen real estate? So I think what BondGPT is doing for our clients right now is giving them a very, very small screen footprint and access to all of this data.

But it goes further than that because while it accesses all of the data that we can make available, both unique data that we have, data that they can get from any market data terminal, and in some cases, data that's unique to the client themselves, what we're able to do is find the insights.

So when clients are looking for data, very often, they don't want that particular data point. That data point is a step along the way. The raw data is not what they're looking for.

They're looking for the combination of a number of pieces of raw data to form an insight, and they want it quickly. So a lot of the comments that we've gotten from our clients is," Wow, I've gotten to the end product, the insight.

This would've taken me minutes, hours. This took me a long time, and you got it back to me in 20 seconds without me having to do any work other than type in a plain natural language question."

Michael Scotti

Wow, that can save a lot of time for a client. The old adage, time is money, and faster to market, obviously.

Jim, what would you say is the next innovation in bond trading? We have AI. That's generative AI. In your case, what do you say?

Jim Kwiatkowski

It's very difficult to say what's going to happen in the next six months. If we go back six, nine months ago, no one was talking about GPT, and today you can't read an article, you can't talk to someone who isn't asking about GPT.

They've either deployed or they have plans to deploy it. I think what's really clear is, the pace of change is accelerating, and everyone is looking for ways to deploy AI.

We're finding that some customers are still cautious about how they deploy AI. We have others that are moving full steam ahead. But I think it's pretty clear, in the next six months, one year, everyone's going to be using AI a lot more than they are today.

The tools that are available today will look antiquated compared to the tools that we'll be talking about a year from now.

Michael Scotti

Isn't that the way technology always works?

Jim Kwiatkowski

Exactly right.

Michael Scotti

Jim, thanks so much for joining us on the Market Structure Wave. It's been great having you. I appreciate your time.

Jim Kwiatkowski

Michael, thank you very much for the opportunity. It's really been a pleasure.

Michael Scotti

Thanks again, Jim. Michael Scotti, with TabbFORUM Market Structure Wave, thanks for joining us.