Buy side demands better data aggregation for primary corporate bonds

“Data is the new oil.”

The title of British mathematician Clive Humby’s talk at a National Association of Advertisers conference in 2006 has been repeated over the years, even as some say data and oil can’t be compared because data is anything but scarce.

But Humby wasn’t talking about scarcity; rather, his assertion referred to refinement. In the 2006 talk, he posited that like oil, data had to be refined in order to have value. Data-driven companies can’t sell raw data just as oil companies can’t sell oil without first refining it.

In the financial markets, in addition to the refinement and cleaning of data, aggregation and enrichment can also be key to making data actionable. Take the primary market for corporate bonds, just one piece of the overall fixed-income market. Data in this market can come from multiple sources such as Ipreo, DirectBooks or Bloomberg. But one piece of data from one provider doesn’t guarantee a full piece of the picture on a new-issue deal. And with time constraints—only a few hours, or at most a day to execute—for buy-side firms, having the full picture or the “golden source” is a necessity. Achieving it has been a challenge.

There are a lot of inefficiencies across the board but from the start, you’re a little behind just from the perspective that you have less robust information

Timothy Kurpis, AllianceBernstein

The primary market is where securities originate, while in the secondary market, previously issued securities and bonds float between buyers and sellers. The secondary corporate bond market has well-established platforms like Tradeweb and MarketAxess for market liquidity, but the primary market doesn't have the same options. “The primary market still looks a lot like it always has,” says Timothy Kurpis, senior vice president and credit portfolio manager at AllianceBernstein.

Kurpis sees challenges for the buy side in three areas: information gathering, quick turnaround times, and communicating participation to dealers.

“In the investment-grade market, there are often three to eight dealers, and you have to let them each know your order size and manage that through the process,” he says. “There are a lot of inefficiencies across the board but from the start, you’re a little behind just from the perspective that you have less robust information.”

From a risk perspective, the primary market remains the most efficient way to get risk into a portfolio due to buying in bigger sizes and at more attractive levels. But from a liquidity perspective, those outlying inefficiencies remain a barrier.

‘Mind sharing’ for solutions

Where there have been inefficiencies, there have been efforts to improve the process.

Kurpis points to “mind sharing” between the sell side and buy side around pain points and what solutions should look like in the primary market as a reason why there has been some progress made in recent years. That communication has helped DirectBooks, one of the three main data platforms in the sector, in its efforts to streamline processes. “It’s been a big effort from the dealer community to solicit feedback from the buy side around the information that’s needed and how to better communicate from the start to remove some of the potential frictions that develop,” Kurpis says.

Some larger efforts to improve information gathering are also underway. Genesis Global, a low-code technology provider for financial markets, has built an offering that specifically targets data normalization and aggregation for the buy side. The offering was originally built for a global asset manager with more than $500 billion in assets under management.

“Everyone has their own data format, and everyone has access to a specific market at a certain point in time,” says Rahul Kambli, senior product owner and manager at Genesis. As an example, Ipreo may deliver data faster than other providers, while Bloomberg may deliver more data due to having more market coverage. Once data is delivered, the clock starts to tick. And multiple things—data asymmetry, reading data from multiple platforms in their own formats, referring to internal analytics information, and communicating with colleagues internally as to whether there’s a deal—quickly eat away at the small span of time firms have to act.

“For that decision-making, you need the data in the best shape and format, and one that the user understands. No one really had that,” Kambli says. People could find themselves monitoring siloed applications for new data, miss new issues, and not have enough information to make a decision. This was the problem the asset manager took to Genesis.

As a solution, Genesis built what can be described as a one-stop shop for viewing data in a normalized format, with additional analytics (like ESG scores), a chat function that can include either Symphony or Microsoft Teams, and integration with an order management system (OMS)—currently Bloomberg AIM, with a Charles River option in the works. The offering uses natural language processing to take any important data from emails and documents as well.

On the detection screen, a user can see a list of every new issue that has been detected in the market and has passed the threshold of required information for an investment manager to make decisions. The system also alerts users to new data that has been added since the last time the detection screen was accessed, with updates timestamped. A breakdown of what information is coming from what provider is also available.

Efforts to consolidate and aggregate data for the buy side aren’t new. Software-as-a-service (SaaS) providers target the buy side to unify data across business lines. And vendors like Snowflake have come to market to break down data silos as firms ingest data from multiple third parties.

In fixed income, specifically, efforts around sharing and aggregating golden source data have ramped up in the last year. This past December, data platform Versana launched for the syndicated loan market with the aim to provide the buy side and the sell side with a centralized golden source of data that would normally be kept in siloed systems. APIs are a key component of the platforms offered by both Versana and Genesis. Kambli says efforts in the last few years by data providers to improve API access and move away from data files allows Genesis’ primary markets solution to collect the new issue data.

AllianceBernstein, for its part, has worked to solve some of these issues internally. Kurpis says the firm has seen a fair amount of overlap in deal information the data providers have, so one of the frictions it has looked to solve internally is creating a security. “In the new-issue market, you have a lot of attributes, but the security doesn’t actually exist,” he says. “If you were trading a secondary bond, you would have an Isin, a Cusip, all the reference data that comes with it.” In the primary, all of that is fluid.

Additionally, AllianceBernstein looked to automate the process around portfolio and sizing. “Mondays are a very big new-issue day, and we know very quickly what deals we’re likely to act on so we can respond a lot faster,” Kurpis says. The automation has paid off: New issue participation rates have grown from 14% to 25%.