The Fed Is on the Right Side of Its 'Transitory' Bet
The Federal Reserve receives a lot of criticism for the way it conducts monetary policy, but it shouldn’t be faulted for delivering a hawkish message at last week’s policy meeting in the face of data showing a marked slowdown in first-quarter growth. The May meeting came off largely as expected, with policy makers leaving interest rates unchanged and the post-meeting statement containing a clear message that policy makers remained set on a June rate hike.
From the statement: The Committee views the slowing in growth during the first quarter as likely to be transitory…
The Fed is betting that residual seasonal adjustment issues negatively impacted the first-quarter gross domestic product and that the data does not reflect the true pace of economic growth. If they are correct, then the pace will pick up over the next two months and they will likely hike rates in June. But if second-quarter data shapes up like the first, they may decide they have overreached in their policy expectations once again. And so far it looks like the Fed is on the right side of their bet. The employment report for April reveals solid job growth and low unemployment as the economy charged into the second quarter.
The hypothesis underlying the Fed’s decision is that the process used by the Bureau of Economic Analysis -- the agency responsible for the GDP report -- for computing the data leaves behind some seasonal effects that were supposed to be removed statistically. This causes the reported data to be low in the first quarter relative to the actual performance of the economy. This makes an already possibly volatile number even more difficult to interpret.
Is there a problem with residual seasonality in the GDP data to justify the Fed’s position? If the seasonal effects are removed from the data, performing a second seasonal adjustment should not have a meaningful impact on the data. If I perform this experiment -- essentially double seasonally adjusting the level of real GDP since 1984 and converting to growth rates -- I get this picture:
This results in a 1.8 percent growth rate for the first quarter after removing the residual seasonal impact, more than double the published 0.7 percent rate of growth. Still, on first glance, it is not obvious that one can identify a persistent residual seasonal impact. But if I look at the difference in growth rates for each quarter, I get this picture:
This suggests a fairly large residual seasonal impact in the first quarter that has been growing over time. Also, the offset -- if the seasonal adjustment is wrong in one quarter if must be wrong in the opposite direct in a different quarter -- has historically been concentrated in the second and third quarters but is currently concentrated in the third quarter. Also note that given the persistence of these effects (which I think is an artifact of the statistical technique used to seasonally adjust data), we can be reasonably confident that they will remain consistent this year and only slowly change over time.
There are a number of obvious takeaways from this analysis. First is that the Fed is probably correct to discount the first-quarter GDP data. There looks to be a high likelihood that the report understates the strength of the economy. Second is that the second- and third-quarter published data are more likely than not to accurately report the health of the economy. Third is that the third-quarter data likely overstates the strength of the economy. So basing policy on the first- or third-quarter GDP data is a dicey proposition.
This won’t be the last word on seasonal adjustment issues in the GDP data. The BEA is working on addressing these issues and may hopefully be able to resolve the situation in the near future. But for now, one should be cautious in interpreting quarterly movements in the data, especially in the first and third quarters. Market participants should probably use averages of recent growth rates to track the underlying economy (this is reasonable given the non-seasonal volatility in the data alone). The Fed is making the right choice in weighting the second-quarter data more heavily than the first when deciding on their next policy move. If the second-quarter data disappoints, they will likely need to take that number at face value and adjust the policy direction accordingly. Early indications, however, are that the data will be there to justify a June move.
Bloomberg Prophets Professionals offering actionable insights on markets, the economy and monetary policy. Contributors may have a stake in the areas they write about.
To contact the author of this story:
Tim Duy at firstname.lastname@example.org
To contact the editor responsible for this story:
Robert Burgess at email@example.com