Global banks appear to be giving top clients private information to win corporate-bond trading business, according to a new study showing the $56 billion-a-day market is stacked in favor of the most active investors with the broadest dealer networks.
The study examined the profitability of trades by insurance companies, which are whales in the credit world. It found that those with the best access to Wall Street’s bond-trading desks often had better outcomes than other investors ahead of market-moving events, such as mergers and acquisitions (M&A) and ratings downgrades.
Critically, the more important an insurer is to a dealer’s business, the more prescient their trades ahead of M&A transactions tend to be—especially if the dealer is part of a bank also involved in the takeover. The pattern holds even when taking into account the pricing benefits offered by large dealer networks, professors at Rice University, Yale University, and the University of Pennsylvania discovered.
Recommended For You
The trio are careful to avoid the term “insider trading” in their analysis of the data, published late last year, noting that dealers could potentially be doling out “private information” without necessarily providing “privileged information.” Still, their findings raise critical questions of fairness in one of the world’s largest and most important financial markets. The results indicate private information is being used as payment for order flow, the authors wrote. That suggests a potential breakdown in the walls that are supposed to prevent information from one side of a bank flowing to another—a mainstay of U.S. financial industry controls for decades.
“The surprising part to us was the results we found around these M&A deals because we know that we have all these ethical barriers inside the banks,” Rice’s Stefan Huber said in an interview. “Picking up at least some trading that looks informed ahead of these deals was a little shocking.”
The authors didn’t link any specific banks to the patterns they observed. JPMorgan Chase & Co. was the most active dealer based on trading volume in the study, followed by Bank of America Corp., Citigroup Inc., Morgan Stanley, and Goldman Sachs Group Inc., according to Huber, who conducted the analysis with Yale’s Edward Watts and Penn’s Christina Zhu.
Representatives for all five banks declined to comment, as did spokespeople for the Securities and Exchange Commission (SEC) and Securities Industry and Financial Markets Association.
Some finance industry professionals are skeptical of the study. William Cunningham has worked on both sides of these trades, having been the head of credit portfolio management, research, and trading for Nationwide Insurance and, before that, a credit strategist at JPMorgan and Merrill Lynch & Co. “People talk and whisper and stuff like that, but it’s not organized and it’s not significant,” Cunningham said.
The researchers chose to zero in on insurance companies because, in addition to being among the biggest players in the U.S. corporate-bond market, that’s where they could source the best data. Every quarter insurers have to report their financial-market transactions to the National Association of Insurance Commissioners, a consortium of state regulators. Those filings contain detailed information on each trade they make, including the dealer on the other side.
By analyzing roughly 2 million trades between 2009 and 2022, the researchers were able to map out the dealer networks of 1,577 insurers. They then overlaid that dealer network data with M&A transaction figures to examine the trading activity of insurance companies in the bonds of takeover targets. What they discovered was that insurers that traded with more dealers showed “improved decision-making in the 30 trading days before these deals are publicly announced.”
And it wasn’t as simple as bigger insurers making better decisions. It turns out that trading with more banks that have merger and acquisition businesses seems to result in better outcomes for insurers.
While the analysis of trading ahead of credit-rating downgrades and defaults was slightly different, the results proved similar. All told, the insurers in the top quintile for dealer-network size got an advantage of about 0.84 percentage point in annual returns compared with all the others. That’s a meaningful difference in a market where U.S. investment-grade corporate debt yields average a little more than 5 percent.
Dealers Seem to Reward Their Best Trading Clients with Information
Still, the outsized gains associated with larger networks don’t necessarily mean that dealers are passing on private information. To test for this, the researchers divided the roughly 370 dealers involved in the underlying trades into two groups: those with investment-bank divisions that advised on at least one M&A transaction in the previous 12 months, and those without.
That data showed that when an investor has a connection with an M&A-affiliated dealer, it sells bonds to a greater extent prior to acquisition announcements that negatively affect debt prices. The more connections investors have to dealers involved in a transaction, the more pronounced the results. What’s more, by taking into account how important an insurer’s trading business is to M&A-affiliated dealers, the researchers determined that the prevalence of informed trades ahead of M&A transactions increases in relation to past order volume between the insurer and the bank.
“These results are consistent with dealers rewarding their best trading clients with private information for order flow,” the researchers wrote, and are potentially “indicative of breakdown of ‘ethical walls’ established by the Glass Steagall Act.”
Kevin McPartland, head of market structure and technology research at Crisil Coalition Greenwich, said he has little doubt “helpful” information is being shared by corporate bond dealers. Still, they may be passing along information that isn’t public, but also isn’t material enough to move the price of a security by itself.
“The question is if the information that’s being traded falls within the regulations of what’s allowed to be shared. And that’s something that we can’t really know based on the study,” he noted. “If you’re a bigger asset manager and you have many traders and portfolio managers who are talking to many dealers and research people, then that’s just more information. And if you have the people and systems in place to distill that information and make decisions from it, ultimately being more informed is going to be beneficial.”
Others question the study’s conclusions entirely. Cunningham, the former portfolio manager at Nationwide, says there isn’t any kind of concerted effort for dealers to hand over key information to bigger investors. “Yes, there’s always leakage, stuff like that, but it’s not systemic and it’s kind of natural,” Cunningham said. “A lot of smaller players—asset buy-side players, be it the insurance companies or asset managers—do complain that the larger shops get the first look in a lot of things.” But it’s basically access to new deals, access to trading flows, he added. “It’s not proprietary.”
Huber, Watts, and Zhu, for their part, say that, if anything, their findings likely undersell the informational value of trading relationships. Insurers are widely considered less active in the market on a day-to-day basis than other participants such as hedge funds, and therefore are less likely to engage in information-based trading. While Huber stressed that the study is not a call to action, the authors do suggest that greater scrutiny by regulators may be warranted in the U.S. corporate bond market.
The evidence “highlights potentially unfair advantages being garnered by informed traders in these markets through investors’ dealer networks,” they concluded.
————————————————————
© 2025 ALM Global, LLC, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to [email protected]. For more information visit Asset & Logo Licensing.
