Who Are You Going to Believe -- the Commerce Dept. or ISM, Autodata & the BLS?

This past July 29, the Commerce Department surprised the economic cognoscenti by reporting that its advance estimate of second quarter real GDP annualized growth was a paltry 1.2%. The consensus estimate of Street economists was north of 2%. Oh my, with growth this anemic, the Fed certainly would not entertain raising its policy interest rates at its upcoming September 20-21 meeting, would it? Yes, it might. And here’s why.
Firstly, the advance estimate of GDP is called that because it is made in advance of the Commerce Department having all of the data that goes into the GDP calculation. For example, Commerce does not have complete inventory, trade and construction data. So Commerce makes educated guesses as to what the missing data might turn out to be. Moreover, some of the “hard” data Commerce has is actually “soft” and will be revised in the coming months. As previously incomplete data are available and revisions to previously available data are made, Commerce releases a revised estimate of GDP a month after the release of its advance estimate. And then a “final” estimate of GDP is released by Commerce a month following the release of the revised estimate. But the final estimate is not really final inasmuch as Commerce keeps revising a given quarter’s GDP years after the release of the not-so-final estimate. To illustrate this revision process, I pulled out of the air (thin air? drink) estimates of the annualized change in real GDP for the first quarter of 2014. The advance estimate was 0.1%. Two months later, the “final” estimate was
minus 2.9%. As of July 29, 2016, the estimate of the annualized change in Q1:2014 real GDP was minus 1.2%. So you can see that there is many slip twixt the cup and the lip when it comes to advance estimates of GDP and later estimates.
Secondly, even if we made the heroic assumption that the advance estimate of Q2:2016 real GDP growth at 1.2% were close to being accurate, it understates underlying demand for U.S. goods and services. If the guessed-at inventories component of GDP are stripped out, real final demand for U.S. produced goods and services grew at an annualized 2.4%. This compares with annualized growth in real final demand of 1.2% and 1.3% in Q4:2015 and Q1:2016, respectively. So, if the Fed were inclined to believe the advance national income and product data for Q2:2016, then it might be impressed by the acceleration in real final demand growth.
To get a fix on the current performance of the U.S. economy, I prefer to look at some indicators that have been reliable in the past, are more timely than Commerce Department GDP estimates and do not get revised nearly as much as GDP. Those indicators are the ISM new orders index, monthly new car/truck sales and monthly number of unemployment insurance recipients. The last 12-month behavior of these indicators is shown in Charts 1, 2 and 3.
Chart 1 shows the monthly levels of the weighted-average index of new orders for manufacturing and nonmanufacturing businesses surveyed by the Institute for Supply Managers. The weights are derived from the value-added data of manufacturing and nonmanufacturing firms contained in the national income and product accounts. After dipping in May, business new orders picked up in June and July. Chart 2 shows that car and truck sales accelerated to an annualized pace of 17.9 million units in July after braking in June. Chart 3 shows that the number of people receiving unemployment insurance has been trending lower since blipping up in May. Except for new seasonal adjustment factors and/or changes in the relative value added of manufacturing and nonmanufacturing firms, the ISM new orders data will not be revised. Except for new seasonal adjustment factors, the car and truck sales data will not be revised. The unemployment insurance data will be lightly revised in the next four weeks, but that’s it.
These three indicators show that the U.S. economy was doing just fine last month. They could reverse course between now and September 20-21. But given strong growth in bank credit, I doubt they will. Don’t count out a policy interest rate increase at the upcoming September FOMC meeting.
© The Econtrarian

Read more commentaries by The Econtrarian