On Thursday the Atlanta Fed GDPNow Model Estimate of GDP took a big dive.
It's not really a forecast. Rather, it's what the model says would happen with the data we have now if the quarter ended now.
The dive happened following release of the U.S. International Trade in Goods and Services report for July 2021 by the BEA.
Email Exchange With Pat Higgins
Mish: Hi Pat, I noticed the big decline today. What exactly in the Goods and Services Trade report triggered the change?
Higgins: You can see in the TrackingHistory tab of the excel file that the forecast for Q3 real services imports growth increased from 17.1 to 36.7 percent. That was the largest change amongst the trade categories in the international trade report. The raw data is at https://fred.stlouisfed.org/series/BOPSIMP; that series increased more than the model was expecting. Services exports increased by somewhat less than the model was expecting; the third-quarter real growth forecast for that category declined from 8.4 percent to 4.6 percent. The forecasts of goods exports and imports were nearly unrevised, the model had “advance” July data from this Advance Economic Indicators report https://www.census.gov/econ/indicators/index.html released on August 27.
Service Imports and Exports
I created the above chart from the TrackingHistory spreadsheet that Higgins mentioned above.
Service imports and exports went wild, at least according the GDPNow model expectations.
It's not the data that matters to the forecast. Rather, it's what the data does vs. the model expectation that matters.
However, looking at the actual BEA International Trade report I do not see anything unusual at all. Here are the actual raw numbers from the BEA.
Exports of Services
- Exports of services increased $0.1 billion to $64.2 billion in July.
- Other business services increased $0.2 billion.
- Charges for the use of intellectual property increased $0.1 billion.
- Travel decreased $0.2 billion.
Imports of Services
- Imports of services increased $2.4 billion to $46.6 billion in July.
- Travel increased $1.0 billion.
- Charges for the use of intellectual property increased $0.9 billion. July imports included rights to broadcast the 2020 Summer Olympic Games.
- Transport increased $0.4 billion.
- Exports of services were revised up $2.4 billion.
- Imports of services were revised down $0.2 billion.
The revisions appear to be net positive to GDP.
And I struggle to grasp how a $2.4 billion increase in import services to $46.6 billion could have that much of an impact on GDP, but there it is.
Real Final Sales
The key take-away in the GDPNow cast is not the plunge in the GDP cast but rather the Real Final Sales plunge.
Real Final Sales is the true measure of GDP. The rest is a Change in Private Inventories CIPI component which tends to zero over time.
Of the 3.7% GDP cast for Q3, a whopping 3.4 percentage points is due to an inventory build.
Only 0.3% is attributed to Real Final Sales.
This morning I went to update my chart which normally includes Nowcast only to find this message.
The following chart reflects Nowcast as of a week ago.
If the GDPNow estimate proves accurate, third-quarter GDP growth has stalled.
However, there's still plenty of data on the way. The Gross Domestic Product, 3rd Quarter 2021 (Advance Estimate) is not due until October 28.
Meanwhile, the Real Final Sales trend looks ominous.
Finally, and once again, Higgins is very generous with his time. If he chimes in further on the Goods and Services Trade report I will add an addendum.
Like these reports? I hope so, and if you do, please Subscribe to MishTalk Email Alerts.
Subscribers get an email alert of each post as they happen. Read the ones you like and you can unsubscribe at any time.