TS Imagine integrates LTX’s pre-trade analytics tool

Spencer Lee and AJ Cass have known each other for some time now.

Lee is a veteran of the buy side, making his bones first at BlackRock, where he led global credit electronic trading and West Coast fixed-income trading. He then co-founded a systematic credit hedge fund called Agilon Capital. 

Cass, on the other hand, hails from the sell side, having called Bear Stearns home for 16 years before joining Goldman Sachs after the 2008 financial crisis. Today, Lee is chief markets officer at trading platform provider TS Imagine, and Cass is head of market structure and liquidity at LTX, the bond trading platform that is a subsidiary of Broadridge.

In many ways, their relationship mirrors how fixed income has evolved through the years and where the bottlenecks still reside. As more electronification and automation of workflows come into the world of fixed income, the chances of making a mistake increase. 

What defines a mistake? Perhaps the biggest one for a buy-sider, says Lee, is if a trader at a hedge fund reaches out to someone on the sell side looking to match a trade, but the trade is not made, and now they’ve leaked that information into the marketplace.

Next week, the two companies will announce an integration that will see LTX connect into TS Imagine’s TradeSmart execution management system (EMS). The aim of this partnership is to provide better pre-trade information to prevent such mistakes. 

“AJ has been the other side of the trade for me for years. He knows his workflows; he knows the importance of a first call,” Lee says, meaning that when he was on the buy side, he knew he could trust Cass. 

As relationships become more electronic, that trust is not always there. “I think that’s kind of what we’re talking about here: AJ and the team at LTX have built this AI tool—this pre-trade synthesis of information—that affords buy-side traders a higher probability of not making a mistake when they’re selecting that recipient of the first call,” Lee says.

While there has been a fair amount of consolidation, at this stage of our evolution to more electronification, there’s more fragmentation than ever

Spencer Lee, TS Imagine

This integration will offer mutual buy-side clients with improved pre-trade transparency, price discovery, and access to aggregated liquidity within their existing workflows. In addition to LTX order staging, TradeSmart users can submit their indications of interest into the LTX Liquidity Cloud. 

According to Cass, Liquidity Cloud consists of roughly $45 billion average daily notional buy- and sell-side IOIs. Through the integration, TradeSmart users will be able to send their IOIs and positions to Liquidity Cloud, and the LTX system will ping back real-time indications of natural contra interest with what the vendor calls Cloud Match Scores—ranging from 0 to 10—that are updated on the TradeSmart screen. 

For example, if a trader wants to buy a bond at $100, and there is a seller at $101, the engine will analyze how often that bond trades at a 1-point bid–offer spread. If it’s, say, 10% of the time, the Cloud Match Score will be 1. Or, if the bid–offer is .1 (10 cents) and it trades 90% of the time, that’s a 9, Cass says.

“[Liquidity Cloud] takes a look at a normalized bid-offer spread on a per-Cusip basis and tells you … how close you are to the other side—both in size and in price—and how often that bond trades inside of where you’re looking to buy and somebody is looking to sell,” Cass says.

He adds that there is also a “policing mechanism” tied to the score where if people are giving LTX their intentions and they don’t act on that consistently, their intentions are now no longer worth a full point. “Think of it as downgrading something—you didn’t do what you told me you were going to do,” he says.

That Cloud Match Score is identified using proprietary analysis in the Liquidity Cloud, which comprises that roughly $45 billion average daily notional volume and which uses machine learning to boil that information down into an easy-to-understand score.

The Cloud Match Scores are also accessible via LTX’s recently launched large language model (LLM), BondGPT+, which answers complex questions in natural language. If a portfolio manager says, “I’m looking to buy BBB-rated utility bonds that trade at a spread wider than +200 and are in a 10-year maturity,” the system will spit back the bonds that fit that description and where the portfolio manager should go to get those bonds. “Ultimately, this application will be embedded into the EMSs as well to assist more in pre-trade,” Cass says.

Lee says buy-side traders are concerned about four things as they build their order book: speed, price, quality, and certainty of execution. 

“We have a lot of options as buy-siders: We’ve got phone, [instant messaging], RFQ, and trading venues with their different protocols. While there has been a fair amount of consolidation, at this stage of our evolution to more electronification, there’s more fragmentation than ever,” Lee says.

As an EMS provider, TS Imagine wants to create a one-stop shop so that traders don’t have to bounce around different systems pulling together information to make a trade. 

He recalls a recent conversation he had with a trader who told him that, on average, it took the trader’s firm between three and five minutes to do a credit trade. For very complex trades, it could take upwards of 30 minutes, Cass says. 

“That’s a lot of time in the middle of a trading day to execute one corporate bond trade.” Lee says.

The EMS serves as a central hub of connectivity for the various corners of the market. The more information you have pre-trade—and not having to scour different systems to find that information—the greater the chances of success. 

According to Lee, adding LTX’s capability to the TradeSmart platform will give a buy-side trader more confidence that they are choosing the right style of trade, with the right dealer, and that there’s the right amount of liquidity on the other side of the trade. And getting this information in the same system they’re trading on makes them faster and more efficient.

“Every trader—whether on the buy side or the sell side—gets extremely frustrated when they see something hit Trace, and they didn’t know about the potential of trading there,” he says. “Now, it could be two people meeting on a phone somewhere, but it could also be two people meeting on a venue somewhere. So if you didn’t have access to all corners of the marketplace with the biggest catcher’s mitt possible, then you’d simply miss these opportunities.”

When asked if Lee has ever had to yell at Cass over a missed opportunity or mistake, Cass laughs, and says that Lee doesn’t so much yell—it’s more like a parent saying they’re not mad at you, just disappointed. The hope with this integration is that there will be less disappointment for buy-siders.

Correction: An earlier version of this story incorrectly stated the number of IOIs brought into the Liquidity Cloud system. Also, the Cloud Match Score does not have BondGPT as one of its components and does not use BondGPT. Rather, it is available within BondGPT. We apologize for the errors.