The Mickey Perspective: What do the ISM Numbers Tell Us? (Part 1)

Earlier this morning, we referenced Mickey North Rizza's (MNR, for short) new blog,, and promised to follow-up with some posts highlighting some of her more informative content to date. We'll start with a post that examines what recent ISM PMI numbers tell us. In this analysis, Mickey shares four specific summary observations we should zero-in on.

First, " The PMI, indicates a contraction in the manufacturing sector for three straight months and is the lowest reading since July 2009."

Second, "The ISM's New Order Index has declined two months in a row and contracted a third time since April 2009."

Third, "The Production Index declined this month for the first time since May 2009."

And fourth, "Inventories have increased for the first time since September 2011."

In short, she argues, the economy, from "the manufacturer's viewpoint, is stalling. When you look at the whole picture of manufacturing contracting, new orders and production declining and inventories increasing you start to paint a bleak picture for US manufacturing businesses."

But is it? It's our belief the PMI is a useful indicator, but there are other leading indicators as well. Moreover, knowing how the PMI is put together each month (including some of the smaller organizations that are part of the sample), the information maintains a qualitative side that most don't bother to fully investigate.

Consider how the PMI is survey-based. It does not leverage actual systems or pricing data directly taken from systems. People fill it in. Which means it's only as good as the homework attention that respondents pay to it. Humans, as we know, are always a weak link when it comes to generating accurate data. Contrast this with the BLS indexes, for example, which are based on transactional or data-driven underlying inputs. And in addition, while the PMI represents a solid cross section of manufacturing, it's only a sample of the broader community.

Mickey also tosses out some other observations that are useful in reading the PMI for what it is (vs. what it's not). To wit, "while the numbers above are just indicators of what has occurred and sets up the question as to why this has occurred, the numbers don't tell you if you need to do anything and what you should do to under the circumstances. The real questions to ask are: What do these numbers mean to our company? What does the company need me to do to respond?"

Indeed. But we'd also add the need to treat the PMI as just a single indicator. As anyone who has done statistical modeling before knows (e.g., for commodity markets), you need to work your way through dozens of data points to identify the collective set of inputs that can lead to a more accurate scenario forecast that holds up over time.

Stay tuned as we offer our own take on the PMI and manufacturing numbers tomorrow.

- Jason Busch

First Voice

  1. Others disagree:

    Calculated Risk disagrees:

Discuss this:

Your email address will not be published. Required fields are marked *