By Spencer.

I recently ran across an interesting example of how looking at an economic series from different perspectives creates very different conclusions.

The original demand collapse at the start of this recession created an inventory problem across virtually all sectors of the goods economy. We now seen to be in a position where the inventory problem has quit getting worse and one of the big questions in the economic outlook is how long it will take various industries to resolve their inventory problems.

One example of the inventory problem is the information technology (IT) industries like computers, communication equipment and semiconductors. If you look at the first chart it gives the impression that the problem is not all that severe and that the I/S ratio appears to be peaking.

If you look at other data you find that for example the semiconductor industry is operating at 54.7% of capacity, a new all time low at compared to the previous low of 61.9% in May, 1975.
So obviously IT industries are slashing production and imports to deal with their inventory problems.

The apparent peaking of the I/S ratio should be saying that the worse is now behind us and
that industries like semiconductors will soon be able to expand output. Much of the expectations that growth will resume in the second half is based on this type of analysis.

But, if you look at the I/S ratio in a little different perspective it tells a very different story.
The second chart shows that the I/S ratio is over 200% of its long term trend. This suggest that the inventory problem is much more severe than the first chart implies and that it will be some time before IT output will bottom. This calls into question the second half recovery scenario
and the recent rally in IT stocks.

OK, there are a couple of questions about the second chart. The first is a question of distortions from using percent ratios. Assume that inventories are 5 points above shipments. If inventories are 105 and shipments are 100 that generates a ratio of 105%. This does not seem bad. But if inventories are 15 and shipments are 10 — the same 5 point difference — this generates a ratio
of 150%. So the point that the i/s ratio is much lower now than it was 20 years ago causes percent ratios like the second chart to show much more severe problems.

Second is the issue of whether or not the old trend line is still valid. If you look at the I/S ratio one can make the argument that the old trend changed around 2000 and that the new norm is the light solid line on the chart below rather than the dotted line. If this is the case than the percent ratio is about where it was in the 2000 recession — still bad, but not nearly as severe as the chart implies.

I just though I would present this case as an interesting diversion from our usual political discussion and issues.

But even with the questions about the desired I/S ratio, this analysis still calls into question one of the major basic premises behind the assumption that the economy will snap back in the second half.

Hopefully this will generate some interesting discussions.