The Consumer Metrics Institute

[neohguy]

I have a low end subscription to The Consumer Metrics institute that I would like to share with the board. I'm having a difficult time posting their charts on our new board. Maybe someone can tell me how to do that?


--------------------------------------------------------------------------------
Consumer Metrics Institute Members News: January 30, 2011 - More Thoughts on the BEA's "Advance Estimate" for 4Q-2010

Lakewood, Colorado


If you have not yet had an opportunity to review the BEA's "Advance Estimate" for 4Q-2010 GDP, we encourage you to at least read our commentary on the report. The report was in several regards remarkable, not the least of which because it could be described as (borrowing Mr. Bernanke's words last summer for the economy) "unusually uncertain." The reported headline number was an annualized growth rate of 3.17%, up about .62% from the third quarter. That number in itself was not surprising -- if anything it was slightly lower than consensus expectations. What's remarkable is how the 3.17% was derived, and how a number of wildly swinging components just happened to magically off-set each other in such a way as to prevent the report from becoming literally unbelievable.

Two components of the GDP had quarter-to-quarter swings in their values that added or subtracted over 5% from the 3.17% headline number. The culprits were imports and inventories. If, for example, the rate of inventory building at factories had just remained constant from the third quarter to the fourth quarter, the headline GDP would have reported that the economy was growing at an annualized rate of nearly 8.5%, probably driving the DOW right through 14,000 on its way to the moon. If, on the other hand, the levels of imported goods and services had simply remained constant from the third quarter to the fourth quarter, the headline GDP would have shown the economy to be contracting with a -1.8% "growth" rate, probably pushing the DOW back down through 10,000.

But instead (by the whim of the Fates) we have 3.17% (+/- 5%), computed to a large extent with two months of real data and one month of guesstimates -- aided and abetted by an overall assumption of a minuscule 0.3% annualized inflation adjustment (even as they oddly report from the other side of their mouth that imported goods were inflating at a 21.8% annualized rate).

It would be charitable to simply call the "Advance Estimate" noisy -- perhaps even too noisy for the extraction of any meaningful data. On the other hand it could actually be scary to take the reported swings at face-value; surely neither factories cutting inventories that sharply nor the skyrocketing cost of oil portend well for employment and the real U.S. standard of living.


--------------------------------------------------------------------------------


Meanwhile, all of our composite measures for on-line consumer demand remain extremely soft. Looking at our Weighted Composite Index over that past year shows only two weeks where the week-long averages showed year-over-year growth -- and both of those weeks are now distant memories:





After bottoming during the first week of August the Index recovered somewhat, but it now seems to be moving laterally in a zone between a 4% and a 6% year-over-year contraction. A longer look at our Weighted Composite (using 48 monthly averages) confirms the shape of the "Great Recession" as it is generally known, except that our data shows that the subsequent "Great Recovery" faded rather more quickly than either the BEA or NBER has reported:





Our favorite way to view the detailed shape of the above chart is contained in what we call our "Contraction Watch," which essentially compares the two dips clearly visible in the above chart by superimposing them, one upon the other. In the "Contraction Watch" chart we actually use our Daily Growth Index (a 91-day trailing average of our Weighted Composite Index) to track the contraction events:



(Click on chart for fuller resolution)

(If a chart is not visible above, please click here to see this commentary as a Web Page.)

In the above chart the day-by-day courses of the 2008 and 2010 contractions in our Daily Growth Index are plotted in a superimposed manner with the plots aligned on the left margin at the first day during each event that our Daily Growth Index went negative. The plots then progress day-by-day to the right, tracing out the changes in the daily rate of contraction in consumer demand for the two events. The 2010 contraction event is now more than a year old, dating back to January 15, 2010. Although the chart clearly bottomed at about 9 months into the contraction (at roughly 270 days), the rise since that bottom has been neither steady nor substantial. In fact, there is no way to forecast when the indicated contraction of on-line consumer demand for discretionary durable goods will end based solely on the recent course of the blue line.

We continue to regard a couple of our sub-sector charts as particularly telling indicators of the comfort that our shoppers have with the prospect of taking on new debt. The first of these is simply new applications for loans for newly acquired residential properties:





Although this chart has arguably turned upward over the past few weeks, it is still showing a greater-than 30% year-over-year decline. Furthermore, that decline is compounding on what was already a year-over-year contraction a year earlier (at the left margin of the chart). With the recent upward trend in mortgage rates we might have expected far more of a rush to lock in the historically low rates. The lack of significant upward movement may simply be signaling that everybody who can qualify for a loan already has one.

The other chart that we think tells us a lot about the condition of our on-line consumers is one that we call the "Personal Finance Sub-Index":





This chart tracks consumer comfort with the state of their personal finances. Among other things it tracks the (inverted) consumer demand for credit counseling services, foreclosure advisors and bankruptcy attorneys. Because we invert the measures of demand for products or services required by financial stress, the line in the above chart goes up as people feel better about their financial situation and down as their circumstances get worse. Clearly the consumers that we track are not feeling great about their finances at the present time.

[scaredshirtless]

I like their "what GDP would have been" stuff for a better feel of what the changes meant. I hope they were accurate.

Good stuff Neoh!!!

[neohguy]

Jan 31, 2011 12:53:50 GMT -5 @scaredshirtless said:

I like their "what GDP would have been" stuff for a better feel of what the changes meant. I hope they were accurate.

Good stuff Neoh!!!

Thanks scaredshirtless. The Consumer Metrics Institute released this yesterday about what they consider flawed methodology in determining gdp:
Consumer Metrics Institute Members News: February 2, 2011 - Answering More Questions about the Q4-2010 GDP Report (Ad Nauseam)

We have received a number of questions about some of the more bizarre aspects of the Bureau of Economic Analysis's (BEA) "Advance Estimate" GDP report for the fourth quarter of 2010 (please also see our earlier commentary on that report). In truth, a fair share of those questions arose because the data in the report itself could be charitably described as noisy, volatile or even peculiar. But many of the questions also indicate a high level of confusion over the methods that the BEA uses to develop the report. We thought that it might be useful to go back over some of the fundamentals:
BEA Methodologies

Most "Main Street" people hear the GDP numbers and think that the BEA simply taps into all of the cash registers around the country and tallies them up to calculate the country's quarterly commerce. That would be nice, and it would certainly be far more accurate than what really happens. The BEA's current methodologies are actually just an extension of a model that they started in 1937, when the best that they could do was to sample activity at a few hundred factories and then attempt to extrapolate that small sample to the entire U.S. economy. Now, 74 years later, they have replaced snail-mailed forms with e-mailed questionnaires, and substituted spreadsheets for the old hand kept ledger books. Sample sizes may have increased to thousands of questionnaires, but the basic concept remains the same: send questionnaires to a relatively minor portion of all U.S. businesses and extrapolate the whole economy from those results.

Instead of tallying up commerce, the BEA runs what is effectively the "mother of all spreadsheets" -- containing complex formulas that try to model the vast number of interactions within the whole economy. Picture in your mind the most elaborate, convoluted and opaque spreadsheet known to man. Increase that by an order of magnitude and have it maintained by hundreds of statistical clerks -- none of whom have purview to the entire thing. Now assume that some of the inputs to this spreadsheet are dollar values from a select sample of transactions, other inputs are sampled counts of goods as they get transported or warehoused, and yet other inputs are from independently derived price surveys. Ultimately all the inputs have to get scaled-up into the aggregate dollar value of all U.S. commerce and inventories by the formulas within the model.
This spreadsheet has thousands of line-item rows representing different parts of the economy (e.g., groceries, automobiles, housing, exported machine tools, imported oil, finished goods inventories and municipal spending). And it now has hundreds of columns, representing each of the quarters since 1947 (and a number of annual columns prior to that). At the end of each January, April, July and October a new column is added to the right hand edge of the spreadsheet containing the numbers for the most recently ended calendar quarter. The numbers are then "rolled up" for publication into spreadsheets with about 60 rows, an example of which can be downloaded directly from the BEA.

Introducing the Adjustments

Further assume that, as economists, the BEA staff is obligated to have multiple versions of this humongous spreadsheet: one with raw numbers, one with the "nominal" raw numbers "seasonally adjusted," and one with the numbers corrected for inflation (the so-called "real" numbers). Seasonal adjustments are meant to smooth out the varying levels of commerce for goods that don't have constant year-long consumer demand (e.g., mistletoe, bikinis or home heating oil), but they are notoriously difficult to calculate and maintain (but arguably easy to fudge). They also introduce spurious results when seasonal patterns get skewed (e.g., by non-seasonal weather patterns or one-time events like 9/11).
Inflationary corrections are handled by a series of "deflaters" that convert the "nominal seasonally adjusted" current dollars into "real" or "chained" 2005 equivalent dollars. Since different types of goods are inflating at different rates at different times, unique "deflaters" are used for many of the line-item rows and every column in the enormous spreadsheet. Statisticians calculating the appropriate "deflaters" aren't necessarily talking to the statisticians extrapolating the "nominal" numbers, nor is it even their job to be concerned with what their "deflaters" are doing to the overall spreadsheet.

The key point to remember is this: the BEA isn't actually publishing measurements of what is happening in the economy -- they are instead publishing what is happening in their humongous, convoluted and opaque spreadsheet as a consequence of keying in the latest results from their questionnaires.

"Deflaters" Gone Wild: Part 1 - Imports
This past quarter serves as a great example of what can happen as a consequence of one of the "deflaters" going crazy. The BEA reported a 3.17% annualized "real" growth rate for the economy. But then they also reported that two of the line items in their spreadsheet had quarter-to-quarter changes that could each individually have changed that roughly 3% number by over 5% -- in one case up and in the other case down. They just happened to offset each other. At face value that observation is nothing short of alarming -- how can we trust a 3.17% overall growth number when two different line items in the detail are swinging so wildly from quarter-to-quarter that the overall number might just as well have been any been anywhere from +8% to -2%?
One of the culprits in this current example is the price of imported goods (e.g., oil) and services, which the BEA says was increasing at a 21.8% annualized rate during the fourth quarter of 2010 after decreasing at a 9.2% rate in the third quarter. This means that the quarter-to-quarter swing in this line-item "deflater" was 31%, enough to cause spurious (and phantom) changes to the "real" economy as purported by the BEA's spreadsheet.

Specifically, imports are generally measured in current transactional dollars. From Q3-2010 to Q4-2010 the "nominal" (in 2010 dollars) annualized value of imports actually increased from $2,399.4 billion to $2,416.0 billion (0.7%). When they "deflated" the latter number using the oil-spiked "deflater" (that was independently derived) they concluded that "real" imports shrank from an annualized "chained" 2005-era $2,184.3 billion to $2,106.1 billion (-3.58%). They then converted those dollars into presumed quantities imported, and reported that 4% fewer physical goods were brought into this country. This was a major surprise to most people tracking consumer consumption of imported goods. And in the convoluted logic of the GDP spreadsheet, it changed the contribution of imported goods and services to the overall economic growth from the third quarter's negative 2.53% to the fourth quarter's positive 2.40%, a whopping swing of +4.93% to the "headline" number reported in the press. What this means is that without the import "deflater" impact, the GDP might have been reported as contracting at a negative 1.76% annualized rate.
"Deflaters" Gone Wild: Part 1 - Inventories
Even more convoluted is how inventories were impacted. Inventory data is largely collected as quantities (e.g., the number of cars or barrels of oil being stored). While one might expect inventories to be valued exclusively using some variation of book-value FIFO accounting logic, they are additionally impacted by an "inventory valuation adjustment" (or "IVA") that utilizes prices from a "Fisher formula" (that according to the BEA's notes "incorporates weights from two adjacent quarters; quarterly indexes are adjusted for consistency to the annual indexes before percent changes are calculated").

This is where a bizarre "deflater" can do its damage, because at the highest level the spreadsheet logic attributes any variations in the quarter-to-quarter rates of change in inventories to changes in production levels. So, even though the BEA reported that "nominal" inventories grew by $5.2 billion, the quarter-to-quarter rate of change for "real" inventories (seasonally adjusted, of course) dropped by $114.2 billion, enough to swing the contribution of inventories on overall GDP growth from the third quarter's positive 1.61% to the fourth quarter's negative 3.70% -- a gigantic swing of minus 5.31% in the overall growth rate. Without this impact the GDP might have been reported to have been growing at an astronomical +8.48% annualized rate!

This (largely phantom) inventory building "reversal" was interpreted by the talking-head pundits on TV as a sign that factories had physically drawn down inventories -- and that ultimately this could only be good news, since those inventories would have to be replenished with geared-up manufacturing sometime during the next few quarters.
Assessing the Damage

How much of the above swings were caused exclusively by "deflaters" run amuck? It's impossible to know for sure from the BEA publications that we have seen, and probably only the statisticians controlling those specific cells in the humongous spreadsheet know for sure. But a good educated guesstimate is that at least several percentage points +/- in the 3.17% are the result of purely statistical anomalies. And an additional percent or so might have been added to the reported growth number by the use of an almost obscenely low "deflater" for the overall economy (0.3%) -- which defies a number of other current annualized inflation readings, including the December year-over-year Consumer Price Index ("CPI") of 1.5%, the December Producer Price Index ("PPI") for finished goods numbers of 4.0%, the BEA's own "deflater" for the prior quarter at 2.1%, and (as mentioned above) the BEA's Imported Goods "deflater" for the fourth quarter of 21.8%.
The conceptual bottom line to all of this is that the BEA is only reporting the results of a huge spreadsheet model of the U.S. economy that collects inputs using techniques first developed in 1937 to target things that mattered to FDR's factory employed constituents. That model samples a small part of the economy and makes enormous extrapolations, seasonal adjustments and price normalizations -- all of which are subject to substantial errors, and some of which are significantly leveraged in the final numbers.
This past quarter provided an opportunity to learn how some aspects of the huge spreadsheet can be taken hostage by statistical quirks. It also painfully reminded us that on occasion the BEA's GDP report is not about the actual economy, it is instead about their model of how they think the economy should be performing. It is small wonder that people on "Main Street" don't really trust the government's data anymore, since it often doesn't agree with what they are seeing with their very own eyes.

[scaredshirtless]

"Two line items coulda changed that 3.17 guess as either 8% or -2%."

So how DO you "trust" a 3.17 print?

Knowing what they put into it kinda scares me...