WASHINGTON – While dealer groups are pushing the Municipal Securities Rulemaking Board to place more restrictions on its proposal to share trade data with academics, researchers say the ones the MSRB has already floated threaten to render the data hard to use or even useless.
"It's not going to get as much use as we would like it to because of all the legal rules that it looks like are going to be imposed," said Bart Hildreth, a professor in the Andrew Young School of Policy at Georgia State University and former MSRB board member.
The Securities and Exchange Commission and Financial Industry and Regulatory Authority already get the full scope of MSRB trade data, with the identities of dealers, for audit and enforcement purposes.
The academic trade data product, which the MSRB first proposed in July 2015 after academics periodically asked for data for studies, has drawn support from market participants for its potential to increase transparency, but dealer groups like Bond Dealers of America and the Securities Industry and Financial Markets Association are still concerned that the introduction of anonymous identifiers could open their members up to the detrimental effects of reverse engineering. An anonymous identifier would allow the MSRB to show all of the trades of a dealer without identifying the firm.
Under the proposal, the data would be made available only to researchers associated with a higher education institution who subscribe and pay a fee. The data would be that gathered from required reports dealers make to the Real-Time Transaction Reporting System within 15 minutes of the time of trade. The MSRB makes some of that post-trade information available now, but none of it currently contains dealer identifiers.
The dealer groups' concerns led the MSRB to make several changes to the proposed product before submitting it for approval with the SEC earlier this year, including lengthening the wait time before data can be released to three years from two and bolstering the steps the self-regulator said it would take to combat the threat of reverse engineering.
The MSRB also agreed with a dealer suggestion to exclude primary trades from the product's data sets by not including list offering price and takedown transactions. But BDA and SIFMA both asked for further changes in recent comment letters to the SEC.
Leslie Norwood, managing director and associate general counsel with SIFMA, and Sean Davy, SIFMA's capital markets division managing director, said the three-year timeframe before data would be released was still too short and asked for it to be released after four instead of three years.
Mattia Landoni, an assistant professor of finance at Southern Methodist University, said, in response to the proposed longer delay, that it is important that researchers are able to write about topics that people are interested in at the time that the researcher releases his or her findings.
"With a three-year delay, that means I would be able to write [a] paper in the best case, three years later and in the worst case [even] later because [I] will have moved onto something else," he said.
Mike Nicholas, chief executive officer of Bond Dealers of America, said BDA "remains extremely concerned" with the risks associated with the proposal and added it is "simply inappropriate" to give higher education institutions the dealer-specific transaction information that dealers are required to submit to the MSRB.
"Because of the interconnected nature of our markets, it would only take one large dealer working in collaboration with a researcher at an institution of higher education to completely identify the dealer names that match MSRB's 'dealer identifier' and then have full visibility and transparency into the business strategy and transactions of every dealer," Nicholas wrote.
He added that the dealer-specific transaction data that the product would provide could easily be exposed to hacking attempts or a freedom of information act request if the data is being held by an academic at a public university.
Landoni said academics would not be opposed to agreeing to the MSRB rules designed to prevent misuse of the product.
"None of us would have a problem with promising not to reverse engineer individual dealer strategies," he said. "That's just not what we do."
Hildreth said that many universities, especially state schools, are going to have "real difficulty" in agreeing to the liability restrictions the MSRB would tie to reverse engineering that would have to be agreed to if academics wanted to access the product. He also said it is unclear how confidentiality rules tied to the data would transfer if for example a PhD candidate started a dissertation at one school but then moved schools during the several years it took to get the dissertation published.
Both BDA and SIFMA urged the MSRB to group similar dealers together and use the groups instead of the anonymous identifiers. However, SIFMA added that it would like to see the MSRB widen the eventual product's availability to any not-for-profit organization that has a separately identifiable research department and regularly publishes research reports instead of just academics with higher education institutions.
Hildreth and other academics said the dealer identifiers are important.
"Without dealer identifiers [the research process] is going to be less rigorous," he said. "The delay in the data [release] just adds to that."
"It's not going to be as used as the research community would like it to be used out of the gate," Hildreth said. "But then again, I respect MSRB's concern about what the market is telling them."
