Wednesday, April 04, 2012

Impressive Sales Gains at Full-Service Restaurants

Sales at full service restaurants increased in January by 11% compared to a year-ago, following a 12.5% gain in December according to the Census Bureau (see chart above, data here). The increasing willingness on behalf of consumers to open up their wallets to spend money at sit-down restaurants suggests that Americans are becoming more confident about the economy, the job market and their future prospects.   

33 Comments:

At 4/04/2012 11:05 AM, Anonymous Anonymous said...

Off-topic, but here's something Mark might be interested in ...

US crude oil production reached the 6 million barrels/day mark last week for the first time since February 2000.

"Petroleum Supply" table, line 1 - 6,049 thousand barrels/day.

 
At 4/04/2012 11:52 AM, Blogger morganovich said...

i am wondering to what extent this may be being driven by differentials in inflation.

food at home has been rising in price far more rapidly that food away from home (according to cpi).

might this have somehting to do with consumers perceiving a relative bargain?

that said, based on personal experience and chats with some restaurant owning friends, i'd guess at least half this gain is from price hikes.

 
At 4/04/2012 1:16 PM, Anonymous Anonymous said...

More on the US oil production.

The EIA has revised upward its oil production figures going back at least a year. January, the last month with full data, averaged almost 6.1 million barrels/day. November and December were both over 5.9 million barrels/day.

U.S. oil output revised up, going back at least a year

 
At 4/04/2012 1:28 PM, Blogger PeakTrader said...

Morganovich says "i am wondering to what extent this may be being driven by differentials in inflation."

Even if there were an inflation differential, it wouldn't be much. It's driven by increased economic activity, higher real income, lower interest rates, the payroll tax cut, pent-up demand, etc..

 
At 4/04/2012 2:04 PM, Blogger morganovich said...

a 3-4% shift in relative prices can be a big deal.

this sure is not being driven by income gains. per capita real income has been going down, not up.

the census data is very clear on that.

check the cps population and per capita money income series.

been down for 5 years in a row in chained dollars.

real DPI dropped in both jan and feb even using the BEA's absurdly low deflators (that are tracking at 1/3 of cpi).

whatever is driving that jump in food, it is not a jump in personal income.

 
At 4/04/2012 3:05 PM, Blogger Jon Murphy said...

Morganovich-

I deflated the numbers using the Food & Beverage CPI. Using this number, it looks like the growth rate is about 6.5%. SO, looks like about 40% (give or take) of this gain is from prices rising.

However, it should still be noted that annual Sales are at the highest level since Dec 2007.

 
At 4/04/2012 4:14 PM, Blogger morganovich said...

jon-

"I deflated the numbers using the Food & Beverage CPI. Using this number, it looks like the growth rate is about 6.5%. SO, looks like about 40% (give or take) of this gain is from prices rising."

yeah, i did that too, but i believe that the CPI number understates inflation very significantly due to piles of bad weighting from substitutions and poor (and mostly one sided) quality adjustments.

the whole geometric weighting idea is totally flawed. it sounds plausible on its face (people swap away from goods that get more expensive and toward those that get less expensive) but the more one looks at it, the more clearly wrong this is.

it assumes that all price moves are supply driven, which is nothing like the case.

the movie sideways pushed pinot noir and belittled merlot. this had very significant effects on demand for the 2 varietals. (this really happened and it was a BIG move) pinot consumption and prices rose, and merlot sunk.

but the BLS would then assume you drank more merlot because of the drop in price, when in fact, the drop in price was caused by people drinking less. thus, their adjustment winds up with the wrong sign.

this is exactly why rent is so badly understated in owner equivalent rents and why food inflation as reported has not caputred actual inflation. they sub in sirloin for porterhouse to keep it from showing price hikes.

i think the actual rate of inflation for food is double digits and that therefore the % of this restaurant gain from pricing is likely 70-80%.

 
At 4/04/2012 4:21 PM, Blogger Jon Murphy said...

I'll agree with you that the inflation is understated. I may not go as high as 70-80%, but I won't dispute it either.

 
At 4/04/2012 4:40 PM, Blogger PeakTrader said...

More imaginary inflation from Jon and Morganovich.

Food and beverage C.P.I. rises 0.1% in February
March 19, 2012

"The Consumer Price Index for food and beverages increased 0.1% in February.

The gain followed increases of 0.2% in both January and December.

The food index was flat in February after rising 0.2% in both January and December.

The only major grocery store food group indexes to rise were cereals and bakery products and other food at home. Other food at home was the big mover, rising 0.4%...while cereals and bakery products advanced 0.2%.

The index for dairy and related products, which rose 0.9% in January, fell 0.5% in February.

Also decreasing during the month was the index for meats, poultry, fish, and eggs, which fell 0.2%.

The index for nonalcoholic beverages, which eased 0.3% in January, was unchanged in February."

 
At 4/04/2012 4:46 PM, Blogger PeakTrader said...

CPI year-over-year - Feb '11 to Feb '12:

All items 2.9%
Food and beverages 3.8%
Food at home 4.5%
Food away from home 3.1%

 
At 4/04/2012 4:59 PM, Blogger PeakTrader said...

Of course, unless core inflation falls, we won't get a QE3:

CPI Feb '11 to Feb '12:

All items less food and energy 2.2%

 
At 4/04/2012 5:54 PM, Blogger PeakTrader said...

Jon and Morganovich, if you want to understand why inflation is not understated, see:

Addressing Misconceptions
About the Consumer Price Index
Monthly Labor Review • August 2008

http://www.bls.gov/opub/mlr/2008/08/art1full.pdf

 
At 4/04/2012 6:20 PM, Blogger Jon Murphy said...

No one is arguing the CPI methodology isn't good theory. But it does have its limitations. Just like any theory, it has a set of assumptions that may not hold in the real world. The CPI is a good indicator of the direction prices are heading, but as a measure of inflation, it quite literally comes up short.

 
At 4/04/2012 10:52 PM, Blogger PeakTrader said...

Jon Murphy, you presented no proof the CPI is bias, one way or the other (and neither has Morganovich).

The BLS methodology gives the best possible unbiased estimate of inflation, although, it likely overstates inflation, since appropriate adjustments have been cautious.

Yet, you assume it understates inflation.

 
At 4/05/2012 8:51 AM, Blogger morganovich said...

peak-

if you want to understand why the current cpi is total BS, read its seminal think piece, the boskin report.

i bet you have never done it.

seriously, do it. it will be an education for you. you will see that the whole system is based on assumption and subjective guesswork. there is no hard data of any kind in it. cpi is no longer a measure of price movement, and that is by design.

you can bury your head in the sand if you want to, but the price of food is up FAR more than 4.5%. the price of breakfast (eggs, bacon, coffee, etc) is up several multiples of that.

if i have to switch to top round from rib eye because rib eye went up in price, that does not mitigate inflation. it offsets costs, but it also has a drop in living standard associated with it.

the boskin nonsense is all based on dropping living standards to hold prices steady. that is NOT low inflation.

i note you never answer the issue about the 70's. was inflation high then? if it was, the it is high now. the underlying prices are moving the same way. will you dent that the 70's had high inflation?

i'm sure you will try your resort to magical thinking that today is so different from then and that products never improved before the 90's and try to compare apples to oranges, but that's just ridiculous.

you have made it clear over and over that you do not really even understand what inflation is.

if you want imaginary, look to the cpi numbers, not to us.

 
At 4/05/2012 9:05 AM, Blogger PeakTrader said...

Morganovich, there've been many rigorous studies on the CPI besides "the boskin report."

Why believe in weak studies or speculation instead?

The link above in the Monthly Labor Review has real answers.

 
At 4/05/2012 9:44 AM, Blogger Jon Murphy said...

The main problem I have with CPI is it's lack of consistiancy. The CPI revises and reverses itself all the time. This makes comparing prices over a time frame very difficult. It also doesn't track quality very well.

In my humble opinion, the ACCRA Cost Of Living Index put out by the C2ER does a much better job at tracking inflation regionally and nationally (they use a national average for prices, but the breakdown is easily obtained and one can deduce an inflation index from that easily). ACCRA uses the same market basket of goods it has used for the past 20 years. It's a highly specific market basket (for example, one item is a 1lb pound of Jimmy Dean brand breakfast sausage). Furthermore, it measures the costs of a middle-class executive household. With the CPI, they send out surveys and ask people to fill them out based upon their buying habits.

No one is saying the CPI isn't useful. But it's just not the best measure for inflation. It has much more value as a general indicator of prices than it does as a deflator.

 
At 4/05/2012 10:08 AM, Blogger PeakTrader said...

Jon Murphy, I doubt there's an easy way to measure changes in the general price level of a most dynamic, large, and diversified economy, particularly in the Information Revolution.

Of course, it's difficult, if not impossible, to quantify quality.

Do you have any proof the ACCRA Cost Of Living Index is more accurate than the CPI?

 
At 4/05/2012 10:21 AM, Blogger Jon Murphy said...

I have evidence, rather than proof. Am I willing to share it? No. I'm writing a paper on this very subject, so if you'll forgive me, I am unwilling to share details before my paper is published.

 
At 4/05/2012 11:28 AM, Blogger morganovich said...

no peak.

there have been exactly ZERO actual empirical studies. they are all just subjective bias piled on top of assumptions.

go back and read the 4.14 post about wine and you will see unequivocally why geometric weighting does not work in the real world.

until you can respond factually to that (as opposed to appeals to imaginary studies) you do not have a leg to stand on. and yes, we have proven cpi understates inflation. it's easy.

i'll do it again.

ribeye goes up in price, so the cpi assumes you eat more flank steak. this substitution lowers reported inflation.

so far, there is nothing here you could possibly argue with. this is simple fact.

no consider: you do not know why prices moved. if they went up because more people wanted rib eye, then the substitution has the wrong sign. if it went up due to supply, then yes, maybe less ribeye is consumed, BUT it was replaced with somehting you wanted less. thus, your standard of living dropped. if you like flank steak as much as ribeye, yiou would have already been eating it.

you are the one who loves to try to impose quality adjustments, but you are being selective with them. flank is lower quality than ribeye, thus, even if the price is lower, you ought to be upping the price due to the quality drop.

but this never happens.

the whole idea of substitution is BS for this reason. it does not use "perfect" substitutes over which consumers are indifferent, but drops in quality, so even if it does have the correct sign (which in many of not most cases it does not) it it directly counter to the hedonic adjustments applied to products that stay the same.

therefore, for any given basket of goods, geometric weighting will ALWAYS underreport inflation.

it's a mathematical certainty.

you can prove this yourself. take a basket of 100 goods of price 1. apply a random price move of +/- 5% to each, but makes sure the sum is zero so the whole basket stays at a total price of 100. save the new values and iterate this 50 times. apply the geometric weighting standards from the bls site.

like we did, you will get large reported deflation every time for a basket of goods that stays the same price.

CPI was deliberately adulterated peak. they made sure it would always read very, very low as a deliberate policy to rein in social security COLA. it was one of the more insidious and perfidious acts of greenspan, that great charlattan and horrifically bad fed chair who gave back the hard won gains of voelcker and then hid the results under lies about the data.

 
At 4/05/2012 11:35 AM, Blogger morganovich said...

jon-

i'd be very interested to see that paper when you are done.

if you would like any help or proof reading, just ask, i'd be happy to render any assistance i can.

the CPI has become a farce. the mechanical weightings get further divorced from reality every year. rent is massively understated, as is healthcare.

the games they play with food are astounding. all the prime cuts of meat have been dropped from the series along with anything else that goes up in price.

the politicization of economic figures has reached a frightening level. it's difficult to even use many of them anymore though i have found that the census numbers tend to be a lot better than anything from the BLS or BEA. has that been your experience as well?

 
At 4/05/2012 1:20 PM, Blogger Jon Murphy said...

I'll be happy to share my paper once it's finished with you, Morganovich. It may be a while, though, as I've just finished my research. I think I have plenty of proofreaders, but another set of eyes never hurts and I'll be happy to forward along the paper to you for comments (this is an open invitation for everyone here). My self-imposed deadline for completion is the Fall, but we'll have to see how my real workload goes.

 
At 4/05/2012 3:36 PM, Blogger PeakTrader said...

Morganovich, you continue to make false assumptions. So, you'll continue to be wrong. Read the link above in the Monthly Labor Review.

 
At 4/05/2012 3:39 PM, Blogger Jon Murphy said...

By the way, I forgot to thank you for that link, Peak. I'm going to use it as a source in my paper. It does a better job explaining the CPI than I could.

 
At 4/05/2012 4:03 PM, Blogger PeakTrader said...

Morganovich says "geometric weighting will ALWAYS underreport inflation."

Another false assumption. Not using the geometric average will overreport inflation over time, because inflation increases exponentially.

You can see this using a compound annual growth rate calculator. Type 140 for ending value, 100 for beginning value, and 4 for number of periods.

The answer is 8.78% (not 10%). The more periods, the lower the rate, e.g. 180, 100, and 8 is 7.62% (not 10%).

 
At 4/05/2012 4:06 PM, Blogger PeakTrader said...

Jon Murphy, sure, you're welcome.

 
At 4/05/2012 4:27 PM, Blogger PeakTrader said...

Morganovich says: "the mechanical weightings get further divorced from reality every year. rent is massively understated, as is healthcare."

Everything has to add up to 100%.

Food and beverages 15.2%
Housing 41.0%
Apparel 3.6%
Transportation 16.9%
Medical care 7.1%
Recreation 6.0%
Education and communication 6.8%
Other goods and services 3.4%

Note, this is the aggregate average, and it my not fit your average spending.

 
At 4/05/2012 4:33 PM, Blogger PeakTrader said...

http://www.bls.gov/news.release/cpi.t01.htm

 
At 4/05/2012 6:06 PM, Blogger Ron H. said...

Jon-

I too would be very interested to see that paper when you are done.

Hopefully you will publish an English translation for us commoners. :)

 
At 4/05/2012 6:59 PM, Blogger Ron H. said...

Peak,

"Not using the geometric average will overreport inflation over time, because inflation increases exponentially.

You can see this using a compound annual growth rate calculator. Type 140 for ending value, 100 for beginning value, and 4 for number of periods.

The answer is 8.78% (not 10%). The more periods, the lower the rate, e.g. 180, 100, and 8 is 7.62% (not 10%).
"

Wait. Are you going to tell me that compounding makes a difference?

Go on! Who would have thought!

How did Morganovich miss something so basic?

[end sarc]

Unless I'm mistaken, Morganovich used a model that produced random price changes to 100 items that each cost $1 to start, but in which no inflation or deflation could occur, as the total basket must always cost $100.

Applying the same geometric weighting the BLS uses caused reported deflation every time. In other words, actual inflation, (0%), was higher than reported.

I can reach no other conclusion, than that the current CPI calculations understates inflation.

I would be happy to hear that his model is flawed, but you haven't addressed it.

As actual inflation is 0 in the model, compounding is no help to you.

 
At 4/05/2012 7:17 PM, Blogger PeakTrader said...

Ron,

1 X 1.05 = 1.05
1.05 X 0.95 = 0.9975

 
At 4/06/2012 2:18 AM, Blogger Ron H. said...

morganovich,

Can you provide more detail on how you ran the basket of 100 goods model as you described below? As Peak correctly points out, a 5% increase followed by a 5% decrease leaves a value of .9975, as does a decrease followed by an increase.

To keep the basket at a constant price of $100 after the first iteration seems to require more price increases than decreases for each iteration.

You wrote previously:

you can prove this yourself. take a basket of 100 goods of price 1. apply a random price move of +/- 5% to each, but makes sure the sum is zero so the whole basket stays at a total price of 100. save the new values and iterate this 50 times. apply the geometric weighting standards from the bls site."

 
At 4/06/2012 2:20 AM, Blogger Ron H. said...

Peak: "1 X 1.05 = 1.05
1.05 X 0.95 = 0.9975
"

That's correct. I need to know more about the model morganovich runs. I will ask.

 

Post a Comment

<< Home