The Bureau of Economic Analysis put third quarter growth at 3.4% on an annual basis this morning, up from 2.2% in the second quarter and ahead of what most forecasters expected. On the face of it that is a strong economy accelerating into the end of the year.

Underneath the headline the composition is unusual enough to change what the number means. Inventories contributed about 1.6 percentage points. Trade subtracted about 3.3. Strip out the inventory swing and growth was closer to 1.7%.

What inventories do to a growth number

Gross domestic product measures production, not sales. A company that builds goods and leaves them in a warehouse has produced them, so they count this quarter whether or not anyone buys them.

That makes inventory accumulation the least durable component in the accounts. It adds to the quarter in which the stock is built and subtracts from the quarter in which the stock is worked down, so a large positive contribution is a loan against future growth rather than evidence of underlying demand. The second quarter had a drag from inventories; the third had a contribution. Neither tells you much on its own, and the pair together tells you mostly about timing.

The import side is the more interesting half. Imports rose at a 19.7% annual rate, led by capital goods other than automotive, which is the category that contains servers, networking equipment and the physical contents of a data centre. Imports are subtracted in the calculation because they are not domestic production, so that surge alone took about 2.6 points off the measured number.

The figure that is doing the real work

Whitmore's Watchlist:
> XLI (Industrial Select Sector SPDR Fund): Domestic manufacturing and capital goods, the part of the economy this quarter's investment number runs through.
> FDX (FedEx Corporation): Freight volumes, which show whether inventory that was built is actually moving.
> IEF (iShares 7-10 Year Treasury Bond ETF): The intermediate curve, which prices the growth and inflation combination this release describes.

Look past both distortions and the cleanest read on domestic demand is real final sales to private domestic purchasers, which strips out inventories, trade and government. That rose 4.9%, against 4.6% in the second quarter.

That is a genuinely strong figure and it is the one the Federal Reserve will weigh. It says private demand is not slowing, which is awkward for anyone expecting the growth side of the mandate to make the case for stopping.

Government added about 0.9 points, with federal spending rising at a 15.8% annual rate. Business and housing investment added about 1.7, and consumer spending about 2.5.

The price indexes went the other way

The inflation measures inside the release were considerably better behaved than the monthly data has been. The personal consumption expenditures price index rose 2.1% in the quarter, down from 5.0% in the second, and the core measure excluding food and energy rose 2.5% against 3.3%.

The monthly figures published alongside tell a less comfortable story. September headline PCE inflation was 3.6% on the year and core was 3.1%, both up from August, with the monthly core reading at 0.3% and August revised up to match.

So the quarterly average improved while the most recent month deteriorated, which is what happens when a price spike that occurred earlier in the year drops out of a quarterly comparison. The Fed held rates at 3.75% to 4% yesterday with two votes for an increase, and described activity as expanding at a strong pace. Nothing in this release argues against that description.

One caution worth keeping in view. This is the advance estimate, built partly on September data that will be revised, and recent revisions have been large in both directions. Second quarter growth went from a first estimate of 1.5% to 2.2%, and the fourth quarter of last year was eventually revised down to a fraction of its first print. The second estimate arrives on November 25.

For a portfolio the question is which of these components your holdings are actually exposed to. An industrial business geared to domestic capital spending reads the 1.7 point investment contribution as demand. A transport or distribution business reads the inventory build as volume that has already been handled and will not repeat. A retailer reads the same build as merchandise it now has to sell.

The inflation half of the release points the other way from the growth half, which is the awkward combination the committee is working with. Strong private demand and a core monthly reading moving up are the two conditions that argue for another increase, and a quarterly price index that looks better than it did is the one that argues for waiting. Both were in the same document.

Whitmore's Take: A 3.4% headline that is 1.6 points of inventory build and a capital goods import surge is a different economy than the number suggests, but the 4.9% in private domestic demand underneath it is real. Worth watching whether the inventory that was built this quarter sells in the next one.

Written by Daniel Whitmore
Millionaire Insiders