August 13, 2009
The headline U.S. retail sales numbers are at first glance disappointing but if one probes a little more deeply into the data, the way the numbers are arrived at and in particular the seasonal adjustment factors that are used to adjust the raw data it is wise to be cautious in leaping to conclusions.
First of all, these are advance estimates from a sample survey of 5000 employer firms. There is therefore a margin of sampling error involved which has to be factored in .
Secondly the headline number is derived from the seasonally adjusted data which the Bureau of the Census explains is derived from a collection of adjustment factors which ”is an approximation based on current and past experience. Therefore, adjustment could become less precise if current competitive pressures, changes in consumer buying patterns during holiday periods and other elements introduce significant changes in seasonal trading and holiday trading patterns.” (Bureau of the Census, Advance monthly retail trade and food services adjustment factors for seasonal and other variations of monthly estimates) These are not trivial amounts. The adjustment factor for July 09 in the auto sector, for example is 1.092. Hence the raw unadjusted numbers obtained from the sample survey is divided by 1.092 to get the adjusted value.
In the case of automobiles and dealers their unadjusted sales in July were 58,404 million dollars but the adjusted number was 52,240 million dollars.
The June 2009 numbers were 55,166 unadjusted and 50,797 million seasonally adjusted. It is possible that no distortions have been introduced by the deep slump in seasonal behaviour but frankly I doubt it. In addition the cash for clunkers program is an unusual event which ought not to be smoothed away by seasonal factors.
Automobiles and motor vehicles and parts incidentally represent almost 3 % of the American GDP. for example in 2007 prior to the recession the U.S.GDP was 14,077 billion and the motor vehicle sector was 400.3 billion.
No comments:
Post a Comment