Industrial Production Increased 0.3% in September, Minimal Impact from Florence

The Fed released Industrial Production and Capacity Utilization numbers for September this morning.

Industrial production increased 0.3 percent in September, about the same rate of change as in the previous two months. Output growth in September was held down slightly by Hurricane Florence, with an estimated effect of less than 0.1 percentage point.

For the third quarter as a whole, total industrial production advanced at an annual rate of 3.3 percent. In September, manufacturing output moved up 0.2 percent for its fourth consecutive monthly increase, while the output of utilities was unchanged. The index for mining increased 0.5 percent and has moved up in each of the past eight months. At 108.5 percent of its 2012 average, total industrial production was 5.1 percent higher in September than it was a year earlier. Capacity utilization for the industrial sector was unchanged at 78.1 percent, a rate that is 1.7 percentage points below its long-run (1972–2017) average.

Capacity Utilization

Econoday Comments

• Focusing on manufacturing which makes up nearly 3/4 of all industrial production, this yearly rate is up 3.5 percent. This looks like a modest rate of growth but this report tracks volumes and when adding in the rate of inflation, growth is closer to 6 percent which is strong.
• Motor vehicles have been a central strength, jumping 1.7 percent in September for a 7.0 percent on-year rate. Selected hi-tech has also been strong, up 0.6 percent in the month for a yearly 6.9 percent gain.
• The manufacturing component of this report, even when adding in inflation, has yet to show the double-digit strength of factory orders and shipments nor even a shadow of the strength of small sample surveys like Empire State and ISM. Yet the gains for vehicles, hi-tech and especially business equipment are pluses that do point to positive momentum. And year-on-year growth for overall industrial production is very strong at 5.1 percent. Watch tomorrow for the Philly Fed’s October update on manufacturing.

Few of these numbers make much sense. Motor vehicle sales certainly are not up 7%. And Econoday is correct that Industrial production has way under-performed the regional Fed reports.

The reason for the latter is a combination of survivor bias, reporting bias, and diffusion bias.

Survivor Bias

Companies that out of business stop responding to surveys.

Reporting Bias

Not all companies bother responding to the surveys. Those who do, appear to be too optimistic.

Diffusion Bias

The region reports and the ISM reports are diffusion indexes. It does not matter if a company adds 1 worker or 500. They are both treated as a gain in employment. The same holds true for shipments. Any tiny gain in shipments counts as much as a large gain in shipment.

And the same holds true in both directions, with results magnified.

Mike “Mish” Shedlock