Thursday, May 10, 2007

Did You Ever Wonder Your Numbers Don’t Look Like Theirs . . . . .

One of the problems I noticed about much of the reporting in the press, news, or on the internet is the numbers don’t always jive. For example, economic growth (real GDP) discussed in the media and furnished by the US Dept. of Commerce is defined on a quarter over quarter (QOQ) basis. The real GDP for Q4 2006 is 11,513.0 and for Q1 2007 is 11,549.1, therefore, the calculation of economic growth is

(11549.1 – 11513.0)/11513.0 = 0.31355%

Because this is quarterly data to get an annual growth rate you multiply by 4

0.31355% * 4 = 1.25%

The same calculation for a year over year (YOY) estimate of economic growth is

Q1 2006 real GDP = 11,316.4
Q1 2007 real GDP = 11,549.1

(11549.1 – 11316.4)/11316.4 = 2.0563%

Hmmm! Numbers are pretty different 1.25% using the QOQ method versus 2.06% using the YOY method.

This is a common problem, knowing how calculations are made so comparisons can be apples to apples and not apples to oranges.

By convention the US Dept. of Commerce discusses changes in economic growth on a QOQ basis. On the other hand I tend to do both YOY and QOQ, but I tend to emphasize the YOY values. The QOQ method tends to gives results similar to the YOY method, but with greater variability or a higher level of noise (double click on the graph for a larger image).

For those of you that like numbers the average economic growth for the period 2002 Q1 though 2007 Q1 using the QOQ method is 2.93% with a standard deviation of 1.53%. Using the YOY method the same values are 2.87% and 0.93%. Clearly the average levels of economic growth are very similar using both methods, but the variability using the QOQ is almost 65% higher.

A good discussion of the value of the QOQ versus the YOY method can be found in Ahead of the Curve by Joe Ellis

The point of this little exercise is not that one method is better than another, it is just one should know whether the analysis uses QOQ, YOY, or MOM (month over month) and know the positives and negatives of each method.

No comments:

Post a Comment