July 2006 GDP: Will the real growth rate please stand up?

The Bureau of Economic Analysis reported the following on July 28:

Real gross domestic product... increased at an annual rate of 2.5 percent in the second quarter of 2006... In the first quarter, real GDP increased 5.6 percent.

Ouch; sounds like the economy slowed down, doesn’t it?

Don’t believe it just yet.  In my view, real growth didn’t drop, it increased.  I’ll explain below, but for now, here’s how I would have reported the Q2 results:

Based on Q1&2 2006 results, real GDP increased at an annual rate of 4.0%.  That compares with the Q4&1 result of 3.6% reported last month.  The real GDP growth rate increased.

Explanation
To calculate the growth rate, the BEA annualizes the one-quarter change in GDP.  But an alternative way of calculating it is to annualize the two-quarter change.  I’ve been doing that calculation for a year, then posting the result here each month.  Here's the latest result:

[UPDATE: Click here for a brief, explicit summary of the math.]

Why do I calculate it two different ways?  Because the two-quarter method has been a better predictor than the one-quarter method has been.  I’ve back-tested both methods, quarter-by-quarter, from 1970 to now.  Result: The “annual growth rate” from the 2-qtr method is still a better predictor of GDP four quarters into the future.  It has won 58% of the time since 1970, 58% of the time since 1980, and 60% of the time since 1990.  (In an election, 60% would be a “landslide.”)

The 2-qtr method yields a growth rate that changes more smoothly quarter-to-quarter; the 1-qtr method is a lot jerkier.  [Important note: The BEA is not in the forecasting business; they are working hard to report what did happen, not what will happen, and they are doing an excellent job of it.  The 1-qtr method says the economy grew at an annual pace of 2.5% in Q2.  Although it’s also a fact that the 2-qtr method yields 4.0%, the BEA only reports the results of the 1-qtr method.] 

Below is a recap of the two methods, going back several quarters. 

If you’re interested in a more-detailed comparison, click on the thumbnail immediately below.  It’s the simplest way I could think of to show how accurate the two methods are, when used to forecast GDP four quarters into the future.  Email me if you have a question about it.

Click to enlarge:

Bottom line 
Contrary to conventional wisdom, real GDP growth appears (to me) to be accelerating.  If that upsets, or confirms, anyone’s political agenda... well, sorry, but it’s purely an accident.  I do have some private-sector forecasting experience, and I always prefer the method that yields the more reliable prediction, no matter whose pet theory is affected.  In this case, the 2-qtr method has been the more reliable predictor.  Real GDP is growing closer to a 4.0% rate than 2.5%.