ECB Paper Finds Early Trading Before U.S. Economic Releases

  • Study says leaks probably helped traders make $160 million
  • Strict release procedures made suspicious trading less likely

A European Central Bank-published study found suspicious trading ahead of seven regular U.S. economic releases, saying information leaks probably helped traders make more than $160 million in profits in two markets over six years.

Between January 2008 and March 2014, prices in Treasury futures and equity index markets drifted in the “correct” direction before the release time for 11 of 21 indicators, with seven showing a major move about 30 minutes ahead, according to the paper posted Monday on the ECB’s website. Three of those seven had release procedures that the researchers said were the least secure of the 21 indicators; some have since seen security beefed up to make leaks tougher.

The findings may add to concerns that data providers, which can include central banks and governments, are failing to adequately protect market-moving communications such as economic indicators and monetary-policy decisions. Last month, the Federal Reserve’s internal watchdog urged beefing up safeguards for giving journalists some information under embargo, while the Reserve Bank of New Zealand announced an overhaul in the release of decisions after a journalist was able to leak a surprise interest-rate cut.

Strict Procedures

“We find that an implementation of strict release procedures makes pre-release drift less likely,” according to the paper, written by a four-person team that includes one ECB researcher and three from U.S. and U.K. universities. The study, which has been circulating in earlier drafts for almost two years, doesn’t analyze other nations’ indicators.

The paper finds “strong evidence” of pre-announcement moves before the Conference Board’s consumer confidence index; the National Association of Realtors’ existing home sales and pending home sales reports; the Commerce Department’s preliminary GDP report; the Fed’s industrial production; and the Institute for Supply Management manufacturing and non-manufacturing indexes.

Jobs Report

The U.S. nonfarm payrolls report, which markets are particularly attuned to, wasn’t associated with pre-announcement movement, according to the study.

The researchers estimate that the “price drift” in the seven flagged indicators probably resulted in a total profit of about $119 million in the E-mini S&P futures during the roughly six-year period. Profit in 10-year Treasury note futures probably amounted to about $46 million.

“It’s sizable,” said Alexander Kurov of West Virginia University, one of the study’s authors, noting that the two-market analysis provides an incomplete picture of how much money is made on informed trading. “It was a way for us to see if the results were economically significant.”

While it’s difficult to pinpoint whether the early trading happens because some market participants are better at forecasting or whether it stems from leaked data, evidence suggests that it’s some combination of the two, the researchers wrote.

It’s “conceivable that several factors combine to cause the drift,” according to the paper. “The source of informed trading merits more research in view of the public interest in the safeguarding of macroeconomic data.”

Four of the data points associated with notable moves are released differently today than they were during the study’s time period.

The Conference Board, whose report is listed as among the least secure, changed in 2013 from e-mailing embargoed material to journalists to publishing live.

Media Lockups

The Fed’s industrial production data, also underlined as particularly insecure, was distributed under embargo during the study period, including via a secure website; it’s now released in a media lockup at the Fed where all communication with the outside world is severed until release time.

The ISM reports, previously posted live, are now sent to journalists under embargo.

By contrast, the realtors’ association hasn’t changed its procedure for releasing pending and existing home sales since the studied period.

The NAR has strict rules that prohibit the early communication of data, spokesman Adam DeSanctis said. While Internet access isn’t actually shut down when journalists receive the existing home sales report in a room at the NAR office in Washington, reporters agree they won’t communicate the information outside of the room until the release time, and a press officer stands by watching.

The pending home-sales releases are e-mailed to journalists under embargo, and media organizations that release the information prematurely face penalties. No one else besides research and communications staff see the data in advance, DeSanctis said.

ISM chief executive officer Thomas Derry said in a statement that it has made appropriate changes in its release procedures over time, some of which were proprietary, and adheres to “policies and protocols to protect the security and integrity of its data.”

The Commerce Department didn’t have an immediate comment on its GDP release, also unchanged. GDP data are disseminated at a government lockup where Internet communication with the outside world is closed entirely before the embargo lifts.

The authors of the paper, in addition to Kurov, are Alessio Sancetta of the University of London, Georg Strasser of the ECB and Marketa Halova Wolfe of Skidmore College in Saratoga Springs, New York.

Before it's here, it's on the Bloomberg Terminal.