By Karl Smith
I want to mention a couple of things about the BEA Release.
There are at least three sections you may wish to concern yourself with
- Percent Change From Previous Period
- Contribution to GDP Growth
- Level and Change
Percent Change From Previous Period
The first section tells you the annualized rate of growth of different sectors of the economy. If you have an intuitive sense for the structure and dynamics of the economy you may find this section the most useful. It tells you at a glance how all the different parts are moving.
An important thing to remember is that for our purposes the long run average growth rate of each sector must be equal to the long run average growth rate of the economy.
So suppose that as in this release the growth rate of a sector was 0.2%. This is far lower than either the current growth rate of GDP or its long run average.
However, because the long run averages must be equal, either this sector is going to be pulling down on GDP, GDP is going to be pulling up on this sector or some combination of the two.
We can try to make out from thinking about the sector at work.
In this case the sector was Personal Consumption Expenditure on Services. Now, over the long run if PCE-S pulls down on GDP that means that GDP is shifting away from services. Does this seem likely? Probably not. This tells us that this sector is likely to rebound, mechanically.
That is, to say you are probably looking at a deviation away from the underlying trend. Once, the deviation moves away growth will look fast though its just because the deviation is now gone.
Contribution to GDP Growth
This section is the most fun but it can be dangerous. Mathematically what this is, is the change in spending in this sector divided by the size of the entire economy. What that means is that all the numbers have the same denominator and so you can add them. If you add up all the sectors then you will have all of the change divided by the size of the economy and so you will have GDP growth.
Another way to think about it is that it normalizes each sector by how big it is.
This section also allows you to do little thought experiments like – if it wasn’t for X then GDP would be Y. That is what is fun.
What is dangerous is that these thought experiments often don’t make sense.
For example, I hear a lot of people saying final demand only contributed about 0.8% with nearly 2.0% of growth coming from inventories. That portends bad things.
Yeah, accept that story doesn’t really jive with the rest of the chart. You don’t inventory services or government production. You inventory goods. Consumption, Net Export and Fixed Investment in goods added about 1.65 this quarter and about 1.85 last quarter. While inventories subtracted 1.35 last quarter.
So, even if you just go back one more quarter the story switches to one of quite solid growth in goods production and mild inventory accumulation.
In general I think its best to be wary of these games unless you have a solid macro story to go behind it.
Level and Change
Level and change is helpful if you know something about the underlying industry.
For example, this time non-residential structures fell by 6.2 Billion. The major mover here is in Oil and Gas instillations. Construction in that area rose about 2 Billion in October and November. so there is a fair chance this is just wrong. On the other hand December could have been brutal or there was an unexpected shake out somewhere else in the commercial construction world.
In any case, without further information we would be hesitant thinking that this tells us a lot about the underlying economy.
We also see another negative next to software. Its not clear what’s going on here. Maybe lots of folks are migrating to Goolge Apps, but in any case this is not likely to last. Software goes up.
You can also see both imports and exports are well below their peak levels. Unless we think the dollar or some other factor is altering the baseline composition of import and exports this means they will likely drag on growth as growing imports will eat up more of expanding consumption than exports will support expanding production.