What does the April ISM really tell us?
There are several different ways to look at The April ISM. We learn something from each way of looking at it.
It was a surprise.
It was out of kilter with other MFG diffusion surveys for the same period of time.
It shows trends different than those in durable goods/factory orders.
The Fed offers up five MFG reports each month. And the ISM in April is stronger than those reports on average. However, the Philadelphia Index was quite close in standing to that of the ISM while the Richmond report was even stronger than the ISM. So while the ISM was a surprise there were some regional reports that now lead us to look at the ISM as though it is NOT simply an outlier.
These monthly diffusion reports are of two varieties so we must transform then to compare them. One set of reports is up minus down or 'net' diffusion, the other is the ISM style which takes all the up observations and half the unchanged as a percent of the total. The ISM gauge spans a range of zero to 100 the net indices span a range from -100 to +100.
To bridge this gap in methodology I take the reports over a single time horizon, currently that is from September 2005 to date. For each report metric we compare the headline index, employment and outlook (no outlook for the ISM). To do this we simply rank each observation in its queue of values from September 2005 to date. In that way we rank each index relative to its own historic standards. In March, the ISM stands in the 66th percentile of its ordered queue while the five MFG indices from the Fed surveys stood in the 49th percentile of their queue, after averaging. The ISM was much stronger. But the Philly regional index by itself stood in the 64th percentile of its queue and the Richmond regional index was in the 88th percentile of its queue. So the ISM while not 'predicted' by the Fed surveys is not inconsistent with 'them' either. The Dallas current index was negative and stood in the 38th percentile of its range for the period while the KC index stood at the 21st percentile of its index. The NY index while expanding was as low in its queue as the Dallas index. These three regions dragged down the overall assessment made by the Fed surveys.
The Fed indices taken at their average seem to paint a picture that is now in doubt: the faltering current reading coupled with a lower outlook reading and ongoing strength in employment; that is a classical mature cycle progression. After a recovery the industrial sector cools, the outlook is less ebullient and jobs often continue to mount as jobs are a lagging variable.
But with the new ISM reading we have something different. We have strong current conditions, there is no outlook metric per se, but orders are also strong and jobs are strong as well. While the Fed surveys seem to paint a picture of a mid-cycle slowdown, the ISM itself paints a 'Go-Go' profile. No wonder markets liked it.
The ISM is different from the durable goods report, too, where data lag a bit. Even so in the durables report orders are weak and order backlogs are strong. In the ISM orders are strong and order backlogs are weak- exactly the opposite. That's a bit of mystery. But diffusion is a report on breadth while durable goods/factory orders measures strength; it may be that the orders strength is more widespread and that order backlogs are simply more clustered or concentrated. That could produce such a difference between the ISM and durables report.
We can also look further into the rankings of the sector components of the ISM. The ISM index stands in the 68.8th percentile of its queue. Thus is has been higher only about 31% of the time. Current production stands in the 81.8th percentile and is super-strong. The import gauge at 53.5 is higher only 37% of the time. Since imports are a function of domestic demand strength the ISM is implying that MFG output right now is stronger than domestic demand, if we gauge demand by the strength in imports.
One place that the Fed surveys and the ISM agreed was on job growth. The ISM jobs metric is stronger than it current level only 8% of the time. For the shorter period we to compare with the Fed indices the ISM rank was 82% compared to the average Fed reading of 78%, both quite high.
When we use different measurements and get the same results it encourages us to believe the findings. We are encouraged therefore, that jobs strength in MFG is for real. The weakest region for jobs was Dallas and it was in the 68th percentile of its queue. MFG jobs look to be on a tear.
We also observed export orders as strong; at a raw reading of 59 they stand in the top eight percent of their queue. This is a huge reading and surprising given the state of the world economy. Meanwhile current production is stronger only 20% of the time yet there are no delivery lags. Delivery lags are slower than this about 88% of the time. So US factories are operating at a high level with no bottle necks.
One caution is that order backlogs and employment go together. When firms have backlogs of orders, they hire. But we are now in a period where jobs are stronger relative to order back logs than what is customary. And having backlogs of orders and jobs moving in opposite directions, as they are now, is usually a short-lived experience. So while the ISM is encouraging it also offers a warning.