Broadridge has developed an artificial intelligence (AI) engine that can work with the company’s post-trade processing platform for U.S. fixed income to create an electronic market.
Working with Jim Toffey, founder of Tradeweb Markets and CEO of LiquidX, Broadridge’s LTX moves beyond the usual 1:1 sales process and creates an electronic
market that aggregates liquidity by engaging multiple participants simultaneously.
Although the U.S. corporate bond market has grown to $10 trillion, only about 25% is traded electronically, according to Broadridge, which says it has about two-thirds of the secondary market.
“Broadridge sits in the middle of a whole bunch of fixed income trade flows,” said Vijay Mayadas, president, global fixed income and analytics at Broadridge. “We capture a unique set of data which is the foundation behind LTX.
“LTX’s RFX protocol allows dealers to unlock more value from their data and customer network by smartly identifying and efficiently aggregating liquidity across multiple buyers bidding for their desired amount, which in turn helps improve BestEx for their customers,” the company said in its announcement.
Mayadas said Broadridge surveyed users and found a pain point in the marketplace. It was hard to trade larger illiquid bonds, hard for dealers to hold positions and hard to aggregate liquidity across the buyside. The company created LTX at the end of 2017 and has a team of 60 experts in fixed income building the platform that can support multiple participants, unlike today’s way of working by phone.
“You ask four guys for a quote and they have to bid on the amount you are trying to sell, so at the end of the day you are getting liquidity from one person at one price. It’s very restraining, you can’t take liquidity from four or five providers, aggregate and get the best trade.”
The Broadridge team set out to create RFX Next Generation where users can see their price and other prices and can improve their price, Toffey said. A large bank has thousands of buyside clients, but no way to identify, at scale, the natural buyers and natural sellers of a particular CUSIP among the millions of bonds out there. A rich database and AI could provide the tools to help salespeople and traders work smarter.
“Broadridge, as the back office for banks, has permission to use that data and make that data smarter. It was a natural business for Broadridge to use its position in the back office to help the front office, by using AI data and a neural network to help identify natural buyers and sellers for any trade,” Toffey said.
Instead of making a call, or series of calls, that lead to a first come, first served sale with no real market activity, a trader can invite 40 or 50 participants and then rank the bids in terms of price/size with priority to larger blocks. Bidders can see where they stack up in the market and improve their price.
“Price improvement and transparency are what we have over RFQ,” said Toffey. “Plus, the ability to go to the customer base with one click is very powerful,” added Toffey who expects most sessions to run about 10 minutes. Unlike selling by phone calls, the LTX platform captures the trading information and displays it on a screen which shows the trade evolve so a user from BlackRock or Fidelity can understand the liquidity of the bond.
When trading activity indicates a bigger market demand than expected, a trader may go looking for other potential sellers. “When we show buyside customers the view, so they get to watch the dealer build their order in real time, they almost can’t believe it actually can happen.”
The asset manager doesn’t just see the final prices, but he gets a the sense of the liquidity.
“If they went to 20 guys, did 20 show up? How much price improvement, how many buyers, was there unfilled liquidity? I think over time this information and this view will become a prerequisite. We live in a world where we want to see more. The dealer is the middle of every trade, doing more, just with better tools. We are allowing the dealer to serve his customer base smarter and more efficiently than they do today.”
This could help portfolio managers with trapped assets, illiquid bonds that are very hard to sell with current tools. But with ILX, sellers can aggregate liquidity. It may take five or six buyers, but because they will be able to see each other’s prices, they will have some assurance they are paying a reasonable amount. Toffey thinks that an illiquid bond session may run longer, 15 to 20 minutes, to attract liquidity.