« "It's Go Time Beardie" | Main | The Exchange Rate-Consumer Price Puzzle »

Sep 16, 2006

Retail Sales

There's discussion about how to interpret the latest retail sales figures - William Polley has the links, as does James Hamilton. The main point is one we've heard many times, be careful about interpreting monthly data as it is very noisy. To see this graphically, here's something I put together the last time the retail sales data came out to make a similar point, but never got around to posting (this is updated with the latest data):

Retail191606

As you can see, the average over five months is fairly stable, but there is a lot of noise in the monthly series, too much to make much of a single months movement (William Polley looks at the large standard errors). Here are typical movements by month, along with historical data for July and August:

Retail291606
Retail391606
Retail491606

There isn't much information to be gleaned from monthly retail sales data even though, as William Polley documents, that doesn't stop people from doing so. For a bit more, this post links quarterly retail sales to nominal GDP.

    Posted by Mark Thoma on Saturday, September 16, 2006 at 11:17 AM in Economics | Permalink | TrackBack (0) | Comments (3)



    TrackBack

    TrackBack URL for this entry:
    http://www.typepad.com/services/trackback/6a00d83451b33869e200d834b4159b53ef

    Listed below are links to weblogs that reference Retail Sales:


    Comments

    Feed You can follow this conversation by subscribing to the comment feed for this post.


    spencer says...

    The sample for retail sales is only one-third of the firms so it takes three months to get a full sample.
    this is one of the reasons it is so volatile. I often say the most significant information in the retail sales report is the direction of the revisions.

    I use the BEA unpublished retail sales data because they also publish real retail sales data. Last month this series showed that total real retail sales growth had slowed to 2%. But excluding autos & parts the growth rate was still 8.7%-- what may be a record divengence between the two series.

    Moreover, in the recent data if you exclude autos and fuel the increase was 0.4%. I do not have this exact deflator but since the deflator for GAFO sales is running at a minus 2% to minus 3% rate it probably is negative. That mean that real sales excluding autos and fuel was probably something on the order of 0.5%-
    0.6% or still at a 6% to 8% rate.

    Posted by: spencer | Link to comment | Sep 16, 2006 at 12:10 PM

    Movie Guy says...

    Ha. I was going to post that Spencer was the person to ask about retail sales.

    And here he is.

    Good.

    Posted by: Movie Guy | Link to comment | Sep 16, 2006 at 08:16 PM

    says...

    Hey, MG, you need to be quicker. So much to tell us, but make it snappy or we'll think you are just trying to look good after the fact.

    Posted by: | Link to comment | Sep 17, 2006 at 10:55 AM



    Post a comment

    If you have a TypeKey or TypePad account, please Sign In