Thursday, February 14, 2008

Real Retail Sales Suggest Slowdown, Not Recession

The chart above shows the 6-month moving average of the series for real retail sales, adjusted for inflation by the St. Louis Federal Reserve using the CPI, through December 2007. Several comments on this recent CD post mentioned that retail sales were not adjusted for inflation, and the graph above is in response to those comments.

1. Note that the CPI for January won't be released for another week, so it won't be possible to adjust January retail sales until then. But the strong January increase in retail sales following a decline in December suggests that real retail sales in January will probably look good, unless inflation comes in much higher than expected.

2. Note also that real retail sales were negative during the last recession. Although real retail sales growth of +1% at the end of 2007 certainly suggests an economic slowdown, it wouldn't signal a recession unless, and until real retail sales growth approaches zero or went negative.


At 2/14/2008 12:32 PM, Anonymous Anonymous said...

U.S. Treasury Secretary Henry Paulson on Thursday said there will be a big effort to get Americans to file their taxes promptly so that they can qualify for rebates under the new U.S. economic stimulus plan

Hurry, get those rrebates and spend, spend spend Oh and economy is strong, and I belive in a strong dollar and it's contained!

At 2/14/2008 1:53 PM, Anonymous Anonymous said...

Why the 6 month moving average? Real retail sales as of December 2007 are negative year over year.

Haver Analytics

St. Louis Fed

January CPI will have to be less than 3.6% year over year to arrive at a positive January real retail sales y/o/y.

At 2/14/2008 2:40 PM, Blogger SoldAtTheTop said...

Glad to see you update your post... perhaps you should update Larry Kudlow... here is my comment from your post of yesterday...

Professor Perry,

Retail sales is reported in current dollars.

Adjusted for inflation (ticking up notable recently) your chart would look substantially more negative.

Click here for a better chart.

Also, overall retail sales includes items like clothing, food and fuel which are essentially non-discretionary.

Non-discretionary purchasing doesn't tell us much about the state of the consumer as cutting back on food and gas is much harder to achieve than say... flat panel TVs and a new sofa.

To get the best outlook on the consumer, par down the retail sales to include only the discretionary items and things look considerably worse and reflect the weakness that has been being reported by retailers.

At 2/14/2008 3:35 PM, Blogger SoldAtTheTop said...

Professor Perry,

I just read your post a little closer...

Please note that real "discretionary" retail sales (the aggregate of all discretionary items) has, in fact, been negative for thirteen consecutive months.

Here's a link to a chart that compares "real" (blue bars) vs "nominal" (red bars) discretionary retail sales.

Also, keep in mind that even nominal discretionary retail sales has been negative for 6 of the last 13 months.

Further, even the complete retail sales series (including all non-discretionary items) was negative for the last two months when adjusted for inflation.

Please correct Kudlow... its doesn't help anyone to get this stuff wrong.

At 2/14/2008 3:38 PM, Blogger spencer said...

Over the last some 15 years the average annual increase in the CPI
was 2.7% while over the same period the deflator for retail sales as calculated by the BEA experienced average increases of 0.7%.

Deflating by the CPI causes one to significantly under estimate real retail sales growth.

In addition I can not find the chart you are using at Haver. I suspect there is an error at your source of the chart.

Haver reports that as of the December data -- and the January data is not yet available -- the Y/Y change in real retail sales was
plus 2.3%.

At 2/14/2008 5:00 PM, Anonymous Anonymous said... it appears that Haver* and the St.Louis Fed are using the BEA Table 5U and deflating by the CPI to arrive at the negative y/y number. But the BEA uses a different deflator in Table 6U to arrive at a positive y/y number.

The Haver chart came from Ritholtz's The Big Picture blog


Post a Comment

<< Home