Will Capacity Utilization Tell Us the Recession Is Over? 5 comments
-
Font Size:
-
Print
- TweetThis
By Dirk van Dijk
Next week, we will get an important clue if the recession is ending or not in the form or the Industrial Production and Capacity Utilization reports. The recent decline in the four-week moving average of initial claims for unemployment has suggested that the recession might be coming to an end. It is due out again tomorrow morning, and will look for confirmation in that indicator after it ticked back higher last week.
Certainly the credit market indicators -- like the TED spread and the spreads of junk bonds over treasuries -- have improved dramatically from the awful levels seen during the winter. Commodities, particularly industrially sensitive ones like oil and copper, have rebounded strongly from their lows, even if they remain far below their highs of a year ago.
All of these things suggest that it is possible that the recession might be coming to an end, at least in terms of when the NBER will eventually date the end of this downturn. However, the end of the recession just means that the economy has stopped going down, not that it is going up, and based on most economic numbers it will be a very long time before they exceed their previous peaks.
The capacity utilization report will be very interesting. In particular, pay attention to the manufacturing utilization number, as the total utilization number includes utilities and that can be affected by the weather as much as by economic activity.
The graph below (from here) shows capacity utilization in Manufacturing going back to 1948, and also capacity utilization for the Auto industry. Overall manufacturing utilization is volatile and cyclical, but the auto industry is much more so. The thing to note is that manufacturing utilization almost always bottoms at the very end of a recession and does not produce a lot of false signals during a recession.
Auto industry utilization, while more prone to false signals, has a tendency to turn up just a little bit before overall utilization does, which is to say it gives a bit of a leading indication that a recession might be over. This has been particularly true in the last four recessions. During this downturn, both the auto and the manufacturing utilization levels hit record lows (at least since the data started in 1948).
The auto industry, despite having two of its biggest players in bankruptcy, has actually seen a slight rise in utilization since it hit 38% in January. In April it got all the way up to 42%.
In April, manufacturing utilization fell to 65.7%. If the May number comes in above 66%, then I think we can say that the initial claims data has been confirmed and we are near the end of the recession. If utilization continues to decline, then I would have to say that the initial claims data is giving a false read. If flat-to-up but still below 66%, then the jury would still be out and I would want to see if the trend continued with the June numbers before making a call.
However, even if that happens the level of utilization will still be extremely low. This means that manufacturers have no incentive to buy capital goods from the likes of Ingersoll Rand (IR), Parker Hannifin (PH) or on the tech side companies like Applied Materials (AMAT). I doubt that the order books of the capital goods firms will start to fill up until the manufacturing utilization level starts to get closer to 75% than 65%.
[click to enlarge]
Related Articles
|























This article has 5 comments:
As wrixon points out both capacity and the number of employed are dropping making the same numbers worse than previous. 500k people losing their jobs is much worse now that there are already about 5 million jobs lost. Likewise 40% utilization is much worse if 10% of all factories are already retired.
I am not convinced we are seeing the bottom yet. Perhapse in Q4. And even then, there is no guarantee things won't slip back down after the holiday season and with higher inflation on the horizon. Thus my thinking moves at earliest to 2010 not 2009 for a solid verifyable bottom.