Using GDP shares in 2007Q4 as weights, the contributions of the various components are:

Consumption = -2.0%
Government spending = 0.1%
Business/residential Investment = -7.3%
Inventory adjustment = -1.8%
Net Exports = 3.5%

Clearly, the collapse in business/residential investment is the dominant contributor to the decline. The minus 2 percentage point contribution from consumption also strikes me as large, with the contraction apparently gathering pace given what we know about retail sales. The strong positive contribution from net exports is striking, as Alan Matthews had anticipated. Finally, the large negative contribution from inventory investment is unlikely to continue as stocks are run down.

A 15% fall in GDP would put is mid table on the largest GDP falls in the last 100 years.

http://www.economist.com/finance/displaystory.cfm?story_id=12852043

http://media.economist.com/images/20090103/CFN620.gif

The big fall on the expenditure side is in capital formation — construction of course but not only that; a big fall-off in computer and transport machinery evident from the trade stats published today.

Not much evidence here that cross-border shopping or consumer risk aversion were big drivers in the substantial fall in the domestic expenditure contribution to overall activity.

The important point that you keep making, correctly, is that this is a high-frequency macroeconomic correction - real GNP seems to be falling about 1% per month - and the yoy crowd are just lazy. The monthly and quarterly data are really important, and time aggregation is a big trap just now.

Dreaded Estate wants to know the peak-to-trough decline. So de we all. To date (Table 4, expendture table sa) says about 7% to Q4 ‘08, but it’s not over. My guess is (quarterly sa real GNP) that double figs is likely at this stage, low teens probable and mid-teens entirely possible. These words to be eaten in due course.

Three papers have been published to my knowledge on the Irish quarterly national accounts, one on volatility (by me in ESRI QEC), one by me and John Lawlor on seasonality (also in ESRI QEC), both a few years back, and one recently by Patrick Quill of the CSO on revisions (also ESRI QEC). All are downloadable free from ESRI webiste.

The main points made are, briefly and in order: (i) the Irish qtly numbers are very volatile by international standards, not CSO’s fault, just the way it is, I think an important point. (ii) there are substantial seasonals, and you can argue the toss about whether CSO’s sa adjustment (X-12 on the aggregate data) is the best - I think not, but a small point, and (iii) Quill thinks the scale/frequency of revisions are not out of line and I agree.

If hacks continue to quote yoy % changes for monthly and qtly figs, that’s their problem, but Karl is 100% right to insist on proper attention to the CSO’s high-frequency numbers, the best we have and a huge improvement on where we were.

Our low corporation tax rate would seem to favour the latter route. Maybe others can explain the kinds of situation (tax or otherwise) that typically motivate the former route for some MNCs.

On carry-over, the sa fig for GDP is -5.0 and for GNP is -3.2. There is no chance of these figs being achieved, since a Q1 09 deterioration is pretty certain, as Karl argues. If GNP drops 3% in Q1 and stays there (a fairly rosy scenario?) the fy decline is 6.2%.