Tuesday, September 14, 2010

A Bit More on Patterns of Stagnation


A little more on yesterday's discussion, which turns out to be a trickier issue than it looks.  The above shows recent consumption for the major OECD countries (EIA numbers again).  As you can see, much of the non-European decline since this spring is in Japan.  However, it turns out that Japan's consumption is highly seasonal:


So a lot of that decline since January in non-Europe-non-US OECD is due to seasonal factors in Japan.  We don't normally think of global oil production/consumption having a lot of seasonality, but that's because it's a sum of a sufficiently large number of pieces with approximately canceling seasonality.  Clearly, in attributing change, we are going to have to be more careful about accounting for that - more in a future post.

Notwithstanding the seasonality, that's a pretty dramatic decline in Japanese consumption in recent years...

3 comments:

Glenn said...

And eyballing the Japanese graph it looks like the mean would follow the top of a fairly smooth curve. Perhaps the top right of a bell shaped depletion curve...

Glenn

ondrej said...

Very interesting indeed.., it would be interesting to know what is causing such demand reduction.

What about to have look at currency rate between US dollar and Japan Yen in given time frame ?

IMHO we will see a lot of demand destruction in coming times due to buying paritity (not sure if this is correct english term) for given international currency versus oil priced in US dollars

BTW thanks for interesting graphs and insigths:-)

KLR said...

Hmmm - I'm not too sure about those JODI numbers. They are labelled "Refinery Input/Demand", rather than being final demand, and the pattern of them is very different than the EIA numbers.

Au contrarie, mon frere: My chart. And here's my spreadsheet of JODI vs EIA. That took a bit of reformatting just now. Chart didn't include Japan or US to keep scale reasonable but they looked fairly close in shape as well.

Here's the historical JODI data as well, if you're interested. Rembrandt Koppelar sent me that a while back. These are all OO .ods files, can reup in .xls etc if you like. I prefer JODI to EIAIEA for the simplicity of format and wider coverage; never liked this devloped-nations-philia either, of course.

The "refinery input" thing must be why the entries for Singapore are 8 full years of "0." With other nations it's not much of a difference to matter, it seems to me.