Wall Street frets about reliability of government data on eve of CPI
This week’s inflation data will be huge for markets, and not just for the numbers. Beneath the Bureau of Labor Statistics’ reports on consumer and producer prices will be simmering questions over the data’s validity. Those concerns have accelerated as budget cutbacks have forced the agency to change the way it collects data. On top of that, President Donald Trump’s decision to fire the BLS commissioner after the July nonfarm payrolls data was released raised worries that the bureau could be politicized. Doubt over the accuracy and integrity of the data is a serious issue considering how much BLS work is used to formulate policy, calculate Social Security payments and inform any number of other political and economic decisions. “I feel like this data that is coming out is getting much less reliable, and this has been building for a long time,” DoubleLine CEO Jeffery Gundlach said last week on CNBC. Staffing and funding is one issue. Trump’s decision earlier this month to fire Erika McEntarfer as the BLS chief is another. The move “has raised questions about whether and by how much the quality of official data produced by official U.S. government agencies could be compromised,” Morgan Stanley economist Michael Gapen said in a note. “Official U.S. data has never been perfect, but has been of high quality, compiled impartially, and useful in the setting of policy.” Questions on multiple fronts Gapen’s sentiments mirror those around Wall Street, where written commentary has broken down the various issues the BLS has already faced in terms of budget cuts that have pushed it to alter the way it collects some data. Also, notable revisions the BLS has applied to its critical monthly payrolls count have raised concerns, including accusations from Trump that the McEntarfer-led BLS was manipulating the data for political purposes. BLS officials weren’t immediately available to respond to CNBC requests for comment. Outside political circles, there are few on Wall Street who place credence in the notion that the BLS is doing anything nefarious with the data. However, the bureau’s often primitive way of collecting jobs data — largely through phone calls and written surveys — has fed into caution about the reliability of pre-revisions data releases. Survey response rates have been steadily sliding, forcing the BLS into bigger revisions. “Significant downward revisions to job growth and increased imputation for CPI have raised questions about the reliability of the official statistics,” Bank of America senior U.S. economist Aditya Bhave wrote. “We argue the data remain reliable, though we would recommend being careful with the initial jobs data.” Looking for inflation clues On the market’s plate this week, though, are the two key inflation readings, first with the consumer price index on Tuesday, then the producer price measure, considered a gauge of costs at the wholesale level, on Thursday. At issue for the BLS is its move to stop collecting CPI data from several cities due to staffing limitations, as well as the growing use of imputed data, or estimating price movements in areas where it can’t get exact information. In those cases, the BLS will try to use data from another source, preferably somewhere near the locale it is surveying, but sometimes it has to use prices from other urban areas as assumptions. When the imputed data comes from a local source it has a higher degree of reliability. But Gapen and other economists worry that if the BLS has to rely on “different cell” imputation, the chances for higher variance increase. Estimates are that some 35% of BLS data for its price reports is affected in some way by imputing. Bank of America estimates that a combination of different-cell imputation and the varied impact of tariffs likely will alter the headline CPI reading by only a basis point or two — 0.01 or 0.02 percentage point. It’s still a consideration, though. “In more normal times, that may be too small to matter, but in today’s environment every basis point counts,” Bhave said. “Still we do not think the BLS’ decision to reduce the CPI sample is enough to warrant alarm over the signal from the inflation data.” Impact on Fed moves Precision will matter as the Federal Reserve closely monitors inflation data and charts its monetary policy course. Economists surveyed by Dow Jones expect the all-items CPI to show a 0.2% increase for July, putting the 12-month inflation rate at 2.8%, up 0.1 percentage point from June. Excluding food and energy, the respective forecasts for core inflation are 0.3% for the month and 3.1% for the year, the latter up 0.2 percentage point from a month ago. The Fed targets inflation at 2%. Beyond those numbers, Wall Street will be poring through the data for readings on tariff-sensitive items. Should those goods and services not point significantly higher, it would encourage the Fed to cut rates in September. However, higher readings could keep policymakers in the wait-and-see posture that has dominated the year so far. “Sequential firming in inflation is one key factor behind our view that the Fed will remain on hold at the September meeting despite recent employment data that point to a sharp slowdown in labor demand,” Morgan Stanley’s Gapen wrote. Traders widely expect the central bank to cut the fed funds in September, then at least once more before the end of the year, but Wall Street economists are in multiple camps. For instance, Morgan Stanley and Bank of America both see no cuts this year, while JPMorgan Chase anticipates three — equal to one at each of the remaining meetings. Those forecasts could change, though, depending on what the data says — and how the BLS data is viewed. “Taken at face value, the new procedures implemented by the BLS should not introduce systematic bias into its estimates of CPI inflation, but it will increase volatility in future CPI prints,” Gapen wrote. “That said, the devil is in the details.”
<
