I read this fellow as presenting a very mechanistic version of a Marxian 
over-investment cycle, in which over-accumulation leads to crisis, which in 
turn leads to the purgation of imbalances and then recovery of 
accumulation. (This shouldn't surprise us, since the "Austrians" and 
Schumpeter owe a lot to Marx, as S. acknowledged.) He only looks at the 
excess capacity dimension of the cycle, ignoring the debt overload and the 
problem of pessimistic expectations. Unlike the Roach article that preceded 
it, he ignores international dimensions. Further, he misses the downside of 
the purgation cycle, i.e., that the purgation may be _too_ painful, leading 
to what I've called the "underconsumption trap" (wages being pushed down to 
save profits, but hurting consumption demand and profits) and what Fisher 
called  "debt deflation." It also ignores any social disorder encouraged by 
recession, which can hurt capitalist expectations.

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

United States: Purging Excess -- The Capacity Overhang

Richard Berner (New York)

That there is a capacity overhang, especially in high-tech industries and 
investment, has long been a cornerstone of our U.S. recession call (for 
example, see "The IT Crash: How Big? How Long?" Global Economic Forum, 
February 26, 2001). We've highlighted the grim implications: First, 
overinvestment crushed returns on invested capital, massively squeezing 
profit margins. Second, the glut killed pricing power, worsening that 
squeeze. And in response, Corporate America would have to slash capital 
spending to purge the excess, triggering recession. The eternal question: 
How long would it take to end the glut -- months or years? Believe it not, 
now there's light at the end of the tunnel: The capital spending bust is 
purging the excess, and with the exception of communications equipment, the 
overhang could be gone in a year. That will help set the stage for 
recovery, one that likely will surprise by its vigor.

A year! That timing may come as a shock to those who agree with my business 
cycle logic, which harks back to the Austrian theorists like Schumpeter and 
Haberler, but who believe that the excess is so massive that it will take 
years to shed. Indeed, my colleague Barton Biggs eloquently summarizes the 
Austrian point of view to question the common view that recovery is around 
the corner: "For the economy to truly recover, ...bad investments must be 
liquidated. This is the 'creative destruction' that leads to eventual 
recovery." (see "What am I Missing?" Investment Perspectives, November 6, 
2001).Barton is right. Recovery is not around the corner. But recession is 
telescoping the restructuring process, and it is not too soon to anticipate 
the turning point in the economy that will result (see "Restructuring 
Corporate America," Global Economic Forum, October 29, 2001).

There's no mistaking the excess in capital spending. Measured by the growth 
of capacity relative to output, the glut is the biggest on record. 
Different metrics tell the same story. Over the past seven years, the 
growth in business capacity (measured by the real stock of equipment and 
structures) averaged 7.1% -- double the pace of GDP. That capital 
deepening, so essential for productivity improvement, simply went over the 
top. The more dynamic equipment and software component (including capacity 
resulting from both high and low-tech outlays) soared at nearly a 24% 
annual rate. And while the Federal Reserve's measure of overall industry 
capacity rose at a hearty and unsustainable 5% annual pace, that in 
high-tech industry grew at a staggering 50% annual clip. This glut, coupled 
with the economic slump, brought operating rates down by nearly 10% over 
the past year.

Correcting such a record glut surely must be a Herculean task, but the 
process is well under way. Although it is only a year old, the capital 
spending bust in a still-growing economy has so far nearly equaled cyclical 
norms for recession. Business capital spending plunged by 8.3% over the 
past year, while in the typical postwar recession, the peak to trough 
decline averaged about 10%. (Certainly there were some, as in the 1950s, 
that were far deeper.) Getting this head start is good news for 
restructuring, and judging by the recent plunge in nondefense capital goods 
orders excluding aircraft -- down 26.5% year-on-year -- the downdraft is 
accelerating. But will even this freefall be enough to correct the overhang 
relatively quickly?

Three very different sets of calculations all come to the same conclusion: 
They suggest that the glut -- or at least the bulk of it -- could be gone 
in another year. One way of making that judgement is to ask what would it 
take to stop the growth in capacity. Note that capacity is still growing 
despite the plunge in investment outlays, because new spending net of 
depreciation is still positive. We calculate that another 15-20% decline in 
real business outlays would halt that net increment to capacity, allowing 
operating rates first to stabilize and then rise even with meager growth. 
That's not wildly different from our current projected decline of 13%.

Statistical relationships designed to predict investment outlays provide a 
second set of calculations for measuring the glut and how long it will take 
to eliminate it. These rely on a notion of the long-term optimal 
capital-output ratio, given the cost of capital. Such a notion is 
especially important for gauging excess, because I believe that whatever 
the other incentives to invest heavily in IT and other outlays, the era of 
low-cost capital encouraged excess. Sure enough, such relationships 
consistently underpredicted IT spending over the past five years, even 
allowing for falling IT prices and a favorable financing climate, hinting 
at excess. Cumulating the errors from such relationships yields a crude 
measure of the overhang, with the excess for computers and software at its 
peak amounting to about $30 billion in real terms, or about 3% of the 
outstanding stock. Using this crude metric suggests that the bust has 
already trimmed the overhang significantly, and that it will be essentially 
gone by early 2002. While that conclusion is too optimistic, in my view, 
this calculation gives a starting point.

A third and final computation comes from the Federal Reserve's own capacity 
growth projections that are used as inputs to their measures of industry 
capacity utilization. Fed staff estimate that the growth of industrial 
capacity has already slowed to about 1.5% annualized in the current 
quarter, compared with 4.6% a year ago. This slump in capacity growth is 
the most dramatic in the history of these data, which began in 1967. 
Extending that decelerating trend would arrest the growth in capacity by 
mid-2002, corroborating the timing of the other two calculations.

That three sets of calculations add up to similar results is powerful 
support for the notion that purging the capacity glut won't take years, but 
months. The bad news, of course, is that further staggering declines in 
capital spending are likely -- and needed --in the immediate future. And 
working off the glut in some key industries, such as communications 
services and equipment, will take much longer; there, the overhang is 
bigger than macro data suggest. Communications analyst Simon Flannery 
expects that U.S. telecom capex will plunge by 20% next year, with possibly 
more to come. But such purging likely will correct at least the bulk of the 
much-feared overhang of capital. For an economy handicapped by falling 
profit margins and sinking returns on invested capital, that is good news. 
It will lay the groundwork for an end to the bust in capital spending. And 
it will set the stage for recovery in operating rates and in profit margins.

Reply via email to