GDP. Up until recently those were the best known initials in economics and stand for Gross Domestic Product (today the best known initials are QE).

Still, GDP – calculated by the Bureau of Economic Analysis (BEA) - is the mother of all economic indicators and viewed as the best broad economic barometer.

Needless to say, calculating the price tag of all goods and services made in the U.S. is no easy task to begin with, but the BEA has devised a GDP reporting system that makes the data more confusing than necessary. How so?

GDP is a quarterly data point, yet the BEA releases new GDP estimates every month. Why?

The BEA prepares three vintages of quarterly GDP growth:

• The advance estimate (released about a month after the quarter ends)
• The second estimate (released about 2 months after the quarter ends)
• The third estimate (released about 3 months after the quarter ends)

Even the third estimate is subject to many revisions. Annual revisions are usually done in July. The regular revision period is 13 quarters, but about every 5 years the BEA does comprehensive revisions which impact the entire historical period back to 1929. Revisions can be drastic as the two examples below show:

Q3 2008

Advance estimate (October 2008): 0.3%
Second estimate (November 2008): 0.4%
Third estimate: (December 2008): 0.5%
Revision on July 2009: - 2.7%
Revision in July 2010: - 4.0%
Revision in July 2011: - 3.7%

Q1 2011

Advance estimate (April 2011): 1.8%
Second estimate (May 2011): 1.8%
Third estimate (June 2011): 1.9%
Revision in July 2012: 0.1% 

The chart below graphically illustrates the advance, second, third estimate and the most recent revision. It's somewhat hard to tell on the chart, but out of the 19 quarters from Q1 2002 to Q3 2006 the third estimate was higher or about equal to the advance estimate 17 times (89%).

Since Q3 2006, the third estimate has been lower than the advance or second estimate 12 out of 23 times (52%). The most recent revision is lower (often by a large margin) than the advance estimate 66% of the time. I don't know the reason for this, but the data would support a notion that the advanced estimate (the data that receives most media attention) is purposely inflated.

GDP’s Effect on Stocks

The BEA has a tendency to be more “generous” with its advance estimates while the third estimates are generally disappointments. The most recent revision often deviates so much from any of the first three estimate that one wonders why anyone even pays attention to GDP numbers.

For some reason Wall Street and the media focus more on the advance estimates and sweep bad revisions under the carpet.

Case in point, last Friday’s third estimate came in 26% lower than the second estimate (1.25% instead of 1.7%), but it was hardly reported on. Stocks closed lower, but the damage for the S&P 500 and Dow Jones was less than half a percent.

A Big Fat Flaw

A big fat GDP flaw is embedded in the GDP formula:

GDP = private consumption + gross investment + government spending + (exports – imports)

GDP includes government spending, which has steadily risen since 2008. GDP does not recognize how government spending is financed; therefore all the government spending that’s brought the U.S. to the brink of insolvency is reflected in GDP.

Government entitlements (such as unemployment benefits, etc.) account now for over 35% of personal income. Personal income accounts for about 75% of GDP. This means that government spending makes up over 26% of GDP. What would GDP be without the government’s finger on the scale?