New Questions on How a Key Agency Shared Inflation Data

New Questions on How a Key Agency Shared Inflation Data


Agency emails show that the Bureau of Labor Statistics has shared more information about inflation with Wall Street’s “superusers” than was previously disclosed. The revelation is likely to prompt further scrutiny of how the government shares economic data at a time when such information is of great interest to investors.

An economist at the agency sparked a firestorm in February when he sent an email to a group of data users explaining how a methodological change may have contributed to an unexpected rise in housing costs in the consumer price index the previous month. The email, addressed to “superuser,” quickly spread across Wall Street, where every detail of inflation data can impact the bond market.

At the time, the Bureau of Labor Statistics said the email was an isolated “mistake” and denied that it maintained a list of users who had been given special access to information.

But emails obtained through a Freedom of Information Act request show that the agency — or at least the economist who sent the original email, a long-time but relatively low-level employee — is in regular communication with data users in the financial industry, apparently also with analysts at large hedge funds. And they suspect that there was a list of superusers, contrary to the agency’s denials.

“Would it be possible to be on the superuser email list?” a user asked in mid-February.

“Yes, I can add you to the list,” the employee replied a few minutes later.

Efforts by a reporter to reach the employee, whose identity the office confirmed, were unsuccessful.

Emily Liddel, deputy commissioner at the Bureau of Labor Statistics, said the agency does not keep an official list of superusers and that the employee appeared to have compiled the list himself.

“It is not something that the program office has even put together, maintained or approved,” she said.

In response to the New York Times records request, the Labor Department redacted the names of the email recipients. But their employers are partially visible. Many of the recipients appeared to be in-house economists at major investment banks such as Barclays, Nomura and BNP Paribas.

Others work for private research companies that sell their analyzes to investors. And some recipients appear to have been analysts at major hedge funds such as Millennium Capital Partners, Brevan Howard and Citadel, who trade directly on their research.

Brevan Howard and Citadel declined to comment. Millennium did not immediately comment.

There is no evidence in the emails that the employee provided early access to upcoming statistical releases or directly shared other data that was not available to the public. In several cases, the employee told users that he could not provide the information they requested because it would require disclosing non-public information.

However, the emails show that the employee had extensive personal email exchanges with data users about how the inflation numbers are compiled. Although such details are highly technical, they can be of great interest to forecasters who compete to predict inflation numbers to the hundredths of a percentage point. These estimates, in turn, are used by investors who bet on the vast amounts of securities tied to inflation or interest rates.

Analysts regularly interact with government economists to make sure they understand the data, but “if such access can move markets, the process for that access must be transparent,” said Jeff Hauser, executive director of the Revolving Door Project in Washington. “This stuff is so valuable, and then someone just emails it.”

In at least one case, emails to superusers appeared to contain methodological details that were not yet public. On January 31, the employee sent an email to his superusers describing upcoming changes to how the agency calculates used car prices, which was a crucial issue for inflation watchers at the time. The email included a three-page document with detailed answers to questions about the change and a table showing how it would affect the calculations.

“Thank you everyone for your very difficult, challenging and thoughtful questions,” the email said. “It is your questions that help us flesh out any potential issues.”

The Bureau of Labor Statistics announced the change in a press release in early January, but did not post details on its website until mid-February, two weeks after the employee’s email.

Ms Liddell said it was “not appropriate” to share information that was not public and had not been fully verified.

“When things like this happen, it really undermines our credibility, not just with the public, but with the people who have put their trust in us to provide us with data,” she said.

It’s not clear when the employee began sharing information with superusers or whether he was the only agency economist to do so. Several of his emails were also sent to an internal Bureau of Labor Statistics email alias, suggesting he did not believe his actions were inappropriate.

The superuser problem came to light in February when the group’s employee emailed to say that he had noticed a technical change in a recent data release that explained an unexpected divergence between rental and homeownership costs. “All of you who are looking for the cause of the divergence have found it,” he wrote.

About an hour and a half after this email was sent, a follow-up action asked recipients to ignore it. In a subsequent online presentation, economists from the Bureau of Labor Statistics presented evidence that the change identified in the employee’s email was not, in fact, the cause of the discrepancy.

It wasn’t the first time the employee had provided information that was later found to be unreliable. In an email in mid-February, he told users that the estimates for rent and homeownership were based on separate data sets. A few days later he responded and said his understanding was wrong.

“Because of this misinterpretation, I am now training to become an animal shelter economist,” he wrote. “Hopefully this training will prevent future misinterpretations” of housing cost calculations.

Omair Sharif, founder of Inflation Insights and a recipient of some of the emails, said the practice of notifying superusers via email is relatively new and that it likely evolved along with increased interest in inflation data.

After years of remaining low and stable, inflation began rising in 2021 and has been a major news story ever since. Because it influences Federal Reserve policy, it is a key driver of market trading.

“I just think the amount of questions has increased so much,” Mr. Sharif said. “Staffing didn’t do it. They’re almost certainly overwhelmed.”



Source link

2024-04-05 21:43:41

www.nytimes.com