Concept Dollars - oder wie man ein hohes GDP Wachstum "erzeugt" - 500 Beiträge pro Seite

[Ken_meyer]

Auch das US-Statistik Büro hat seinen Teil zu dem
Hype beigetragen.
Ist zwar schon länger bekannt, aber wer will
das schon lesen, wenn es nach ner "kleinen Delle"
sofort wieder steil aufwärts geht


Quelle: http://www.grantspub.com/coins/

Coins of the realm

Grant`s Interest Rate Observer - February 16, 2001

Many are the miracles of the computer age. Though the dot.coms, ASPs
and ISPs etc. may perish, the Commerce Department lives on, and its
economists continue to produce the data to demonstrate the benefits of
faster and more powerful computers. As usual, such numbers featured
prominently in the January 31 advance report on fourth-quarter GDP. Also
according to script, they provoked no searching discussion. Alan
Greenspan accepts them as gospel.
Thus, investment in computers and peripherals in the last three
months of the year totaled 329 billion conceptual dollars at an annual
rate. These are rarefied, mathematical, "chained" dollars. Some dollars
are for spending, others are for measurement. These are the finest, most
expandable measurement dollars ever created. They are called into being
to measure the "real" level of high-tech investment, adjusting it for
improvements in processing speed and hard-drive capacity as well as for
observed changes in patterns of consumption.
What was the volume of fourth-quarter investment in computers and
peripherals in terms of regulation, green, spendable dollars? The answer
is $118.2 billion (also expressed at an annual rate). "Real" investment
was therefore $210.8 billion, or 2.8 times greater than nominal. Then,
again, there can be no statistical integrity in comparing conceptual
money with legal tender.
All of which begs the question of what is really "real" in the
investment portion of the national income and product accounts? Actual
dollars or adjusted ones? Fact and fancy are forever at odds on Wall
Street. Seasoned investors may bridle when a corporate management
invents a new "metric" with which to describe its unique
achievements.Yet almost nobody questions, let alone disputes, the "real"
data produced by the Bureau of Economic Analysis. (We interrupt
ourselves to draw a distinction between our doubts about econometrically
enlarged data and our admiration for the people who think them up and
calculate them.)
Principal blame for the credulousness of world financial markets
toward overbred U.S. economic data, it seems to us, goes to Chairman
Greenspan, who often invokes the supposed statistical proof of
technological progress and productivity growth but never (as far as we
know) cautions against over-reliance on numbers that may not be what
they seem.
In fairness, never before has the econometrics profession been
presented with an indexing challenge like the one that Silicon Valley
has handed it. New computers are incomparably better than the models of
only a few years ago. The price assigned by the government is not the
price in the catalog. It is the price derived by deconstructing the
value of the machine`s characteristics (speed, hard-drive size, graphics
capability etc.). The Commerce Department makes the irrefutable argument
that the price of strawberries is quoted not by the box but rather by
the pound. In like vein, it says, the price of a computer should be
quoted not by the box but by the unit of computational power.
As long ago as the first Reagan administration, the BEA was teaming
up with IBM to devise a rigorous methodology to do just that. Now this methodology--seemingly the very kind of thing that a sensible layman
would cross the street to avoid having to try to understand--has become
a significant factor in the measurement and capitalization of the U.S.
economy. In the boom, the world was content to believe that hundreds of
billions of dollars a year of imputed investment was just as potent as
the other kind. But it wasn`t--and it isn`t. The conclusion of this
spreading work is that the accounting presentation of the macro portion
of the New Economy warrants just as much scrutiny as does the micro. It
would be unfair to draw a close comparison between the BEA`s econometrics and Amazon.com`s holiday-era "Delight-O-Meter," or Cisco`s selective non-GAAP earnings releases. Yet each set of reporting
conventions flatters the economic performance of the nation or
enterprise that disseminates them.
The seed for our undertaking was an article by Lynne E. Browne,
director of research at the Boston Fed, that appeared late last year in
the bank`s "Regional Review." Browne pointed out a red letter occurrence: In 1995, for the first time, the "real" value of
information-technology investment eclipsed the nominal value. The gulf proceeded to widen until, by the fourth quarter of 2000, it reached the aforementioned annualized $210.8 billion. "While nominal investment in computers and peripherals rose roughly 10% per year between 1995 and 1999," writes Browne, "real investment rose 45% a year."
It`s the "real" GDP that passes for bona fide in most discussions.
When, for example, Merrill Lynch recently referred to the "strongest
U.S. capital-spending performance on record," it actually described the
"strongest U.S. capital-spending performance on record as magnified
through econometric adjustment." Similarly, when Alan Greenspan on
January 25 testified before the Senate Budget Committee about "the
improvement in structural productivity," he should have said "the
improvement in econometrically enhanced productivity" (as a separate
matter, the word "structural" bears scrutiny: a decade ago, there was a
"structural budget deficit"; now there`s a "structural budget surplus").
The contribution made by econometric science to the techno-boom of
the 1990s is only rarely acknowledged. Even the econometricians are
diffident. Competition among the purveyors of digital hardware is
notoriously fierce, but only a portion of the improvement in computer
technology is passed along to consumers in the form of lower prices.
Government econometricians, trying to quantify the breakthroughs in
product performance, put an even higher value on them than the
marketplace does. Hence the acceleration in the "real" value of information technology. "As computers get faster," Browne notes, "the
estimated price comes down, even if the price of the physical machine
does not."
From 1995 to 1999, according to Commerce, the average decline in computer prices was about 7.7% a year. Over the same span, the average
imputed decline in computer prices was about 33.3% a year, a difference
of 25.6 percentage points. It will be said that this is not an
unreasonable gap considering the phenomenal advances in product quality.
To which it will be said in reply that the best measure of the expected economic value of a new technology is its actual price.
To determine the value of any product, divide the volume sold by the
price received. Divide a rising volume by a falling price, and you have
the nominal version of the growth in computer and peripheral investment:
Undeniably impressive. Divide a rising volume by a plunging price and
you have the "real" version of the same data: A number so stunning it
may qualify to be cited by Chairman Greenspan in congressional
testimony. The steeper rate of price decline is, of course, the
econometric contribution. The value produced by these adjustments--
constant readers will remember that, for reasons never explained, they
are called "hedonic"--echo throughout the national income accounts. They have contributed to faster measured growth in GDP. And because
productivity is defined as output divided by hours worked, a higher
level of output yields a faster rate of productivity growth than would
otherwise obtain.
Not every effect produced by the hedonic adjustment of digital
hardware prices works to enlarge the GDP. For example, because the U.S.
is a net importer of high-tech hardware, "real" imports run higher than
nominal ones. In the fourth quarter, the difference was an annualized
$66.5 billion, equivalent to 16.1% of the 2000 "real" current account
deficit.
"A review of the data shows that only a small share of the increase
in measured growth in the latter half of the 1990s is associated with
the use of hedonic price indexes." So state the BEA`s J. Steven
Landefeld and Bruce T. Grimm in an article in the December "Survey of
Current Business." Which begs the Clintonian question, what do you mean
by "small"? Except for hedonic adjustment, as Browne points out,
business investment as a percentage of GDP in the late 1990s would have
broken no records. But after econometric improvement, the data do set
records--and contribute to the perception that we Americans are blessed
by a New Economy. Thus, the path of progress of semiconductor
performance has been transformed into a major source of economic growth.
Or, rather--an important distinction--of imputed and latent economic
growth.
It is imputed because it is introduced from outside. It is latent
because a faster computer yields no gains if it is packed away in a
shipping carton. Until it is set to profitable employment, a computer is
a piece of furniture. Like a piano, its utility depends on the
individual at the keyboard: He may play "The Moonlight Sonata" or "Happy
Birthday." The implication of hedonic adjustment is that the computer-
using U.S. workforce studied at Juiliard. To which it may be objected
that the latent musical power of a Steinway is irrelevant as a point of comparison because a piano is not a capital asset.
Computers are manufactured as capital assets, but many serve a
musical function. During the Internet bubble they generated no profits,
but, rather, played the Symphony of the Stock Options. In theory, in a
capitalist economy, the expectation of turning a profit is why an
investment is put in place. Profits may not always be delivered, but
they are usually anticipated. However, in the late 1990s, they were not
even anticipated. Page views, revenue growth and market share sufficed
instead. Net income was considered an affectation.
This is not to say that profitless high-tech investment produced
nothing. Filling out the econometrically amplified GDP data, it
contributed to the making of the greatest bull market. As the price of
capital cheapened, the pace of investment quickened. Now--to judge, for
instance, by the collapse of resale prices of some of Cisco Systems`
like-new networking equipment--there is too much investment on the
ground.
Kurt RichebŠcher, author of The RichebŠcher Letter (published by the
Fleet Street Group, Baltimore), has put his finger on a particularly
glaring omission in the hedonically enhanced GDP. It is missing an
associated stream of income. These hundreds of billions of dollars worth
of statistically conjured "production," he points out, are "nobody`s
expense and also nobody`s revenue." A statistically amplified price
decline is treated as if it were identical to an actual spending boom--
which, of course it isn`t: "The only economic effect that can be reasonably claimed," RichebŠcher goes on, "is an increase in the capital
stock, but that has conventionally zero bearing on GDP."
"Real output figures can. . .give a distorted sense of the health of
the computer industry and the impetus to aggregate demand that
investment in computers provides," as Browne elaborates. "Based on
growth in real spending, one might assume that computer firms have been
doing well, with employment growing vigorously. However, employment in
the manufacture of computers was only 5% higher in 1999 than in 1995.
Over the same period, employment in motor vehicles and parts
manufacturing also rose 5%, while real motor vehicle output grew about
5% per year. Not a bad performance, but only about one-tenth of the real
growth rate in computers."
RichebŠcher, who exhibited an irascible streak even before there was
a PC industry, is none the mellower today for the general obliviousness
of the American investment community to his thoroughly original and
excellent work on the fallacies of U.S. national income accounting for
technology. Early on, RichebŠcher saw the importance of the Commerce
Department`s decision, taken last year, to treat outlays on software as
a capital investment rather than a business expense. The significance of
the change is that, as a capitalized investment, they contribute to the
GDP; as an expense, they didn`t. The numbers are such that not even
Landefeld and Grimm would consider them "small." Thus, in 2000, software
outlays totaled $229 billion. (Incidentally, nominal and "real" software
outlays are virtually identical; Commerce has developed no hedonic software index.)
Combining capitalized software spending with hedonically magnified
computer investment yields numbers large enough to change the face of
the U.S. economy. Last year, "real" investment in "information
processing equipment and software" totaled $676.5 billion. As a point of
comparison, "real" GDP totaled $9,320.4 billion.
Thus, based on RichebŠcher`s work--and, indeed, our own, for which
we once more thank Messrs. Von der Linde, Medoff, Harless and Isla--the
economic substance of the boom of the late 1990s was exaggerated. If
there is a silver lining, it`s that "real" dollars will decline more
steeply than nominal ones during this, the contraction phase of the
cycle. Nominal dollars meet the payroll and pay the interest expense.
According to the logic of hedonic price adjustments, rapid
technological progress is equivalent to rising production: A constant
volume of output divided by a lower price yields a greater value of
output. Yet nonstop innovation is also a destroyer of value. "Nothing
ages more quickly than Internet infrastructure because of the
breakthroughs," Fred Hickey, editor of The High-Tech Strategist, tells
Robert Tracy of the GrantsInvestor.com staff. "The increases in capacity
come so quickly--more quickly than anything else--and that is one of the
reasons why we have such a glut. They come faster than anything we have seen, faster than Moore`s Law." A year ago, Red Herring reported that,
via the magic of DWDM (Dense Wave Division Multiplexing), a single fiber
can carry 6.4 terabits of information per second--"a 3,000-fold increase
over just a few years ago."
Yet, some retort, there is no surplus of high-tech capital goods,
because they depreciate almost as soon as they`re installed. "Last
year," writes Bruce Steinberg, chief economist at Merrill Lynch, "tech
spending accounted for half of capital spending on equipment and
software, up from 40% a decade earlier. Yet the tech capital stock rose
to only 30% of the capital stock of equipment and software last year, up
from 26% a decade earlier. Even very rapid tech spending growth has led
to only a modest increase in tech`s share of the installed capital
stock.
"Obsolescence," Steinberg goes on, "is a major part of the reason
that huge increases in tech spending have had only a small impact on the
relative size of the tech capital stock. Tech depreciates much more
rapidly than most other types of capital spending. So it takes much more
gross investment in tech to grow the capital stock than it does in other
areas."
Rapid rates of obsolescence ought to imply rapid rates of
depreciation. Yet there is no trace of such a step-up in the national
income and product ledgers. Anirvan Banerji, director of research of the
Economic Cycle Research Institute, New York, plotted the ratio of
depreciation to capital stock over about the past 70 years. Depreciation
("consumption of fixed capital") as a percentage of capital ("private
fixed nonresidential equipment and software") rose sharply during World
War II. But that is the only such spurt on record. The ratio of depreciation to capital turned up in the early 1980s, with the advent of
the PC. However, that slight inflection does not begin to suggest the
gale of innovation, i.e., obsolescence, that has swept across the land
in the 1990s. (Nor, incidentally, has depreciation picked up in the
1990s when expressed as a percentage of corporate profits.)
The techno-bulls, innovators in financial perception, seem to have
it both ways: more product obsolescence without a commensurate rise in
depreciation; shorter, and therefore costlier, product cycles without
(until recently) a commensurate reduction in the stock-market valuations
of the companies at risk. In Monday`s New York Times, Floyd Norris
helpfully updated the P/E ratio of the Nasdaq 100 (it is 811 times) and
noted the many ways in which high-tech companies end-run GAAP to
brighten the news they take to Wall Street.
The financial commentator Bill Fleckenstein has poetically said that
"Alan Greenspan invented the Internet." The literal truth is that the
value of capital goods depends on the financial setting in which they
are produced and consumed. Thus, the rate of product obsolescence slows
markedly during a bear market, even though the pace of product
innovation is unflagging. In the post-boom, people no longer change
their computers (and servers and routers and printers) like their socks.
The march of technological progress slowed in the 1930s, but not because
of a failure of science. The problem lay elsewhere.
"The key factor driving the cumulative upward revisions in the
budget picture in recent years has been the extraordinary pickup in the
growth of labor productivity experienced in this country since the mid-
1990s." So Greenspan told the Senate Budget Committee last month.
Extraordinary, perhaps, but hardly unprecedented. There was a wonderful
boom in productivity growth in the 1920s, even without the use of
hedonic enhancements. It just preceded the Great Depression.
It may well be that strawberries shouldn`t be priced by the box. But
the perceived collapse in the price of strawberries doesn`t share
responsibility for igniting a stock-market bubble and confusing even the
chairman of the Federal Reserve Board. The time has come to simplify--
and to think in terms of real money.

[ANDYDERAECHTE]

Great ken-meyer!
Gruß aus BH

ANDY DER AECHTE

[ANDYDERAECHTE]

Great, ken_meyer wie alle Deine Beiträge!

American Style of Life (SOAL). Wohin man auch guckt, beschönigende Kosmetik. Dabei sind die Amis Statistik-Fetischisten. Die Armen. Volk will betrogen sein.... So läßt sich leichter über die Verhältnisse leben und das allein zählt.
Hast Du einmal dran gedacht, Deine wirklich herausragenden Texte mit einer kurzen 3-Zeilen-Zusammenfassung in Deutsch auch einem breiteren Teil unserer User-schaft zugänglich zu machen?

Gruß aus BH
ANDY DER AECHTE