Demand was robust through the 90s as businesses expanded capacity; that is they purchased capital goods in anticipation of stronger demand, especially going into the Y2K stampede. But of course the bottom fell out -- a calamity that capital goods producers haven't forgotten. The peak-to-trough percentage decline from September 2000 to February 2004 is 24 percent. Shipments are now beginning to approach the 2000 peak, but these comparisons are in nominal dollars, not real dollars that reflect the erosion of inflation. In real terms, shipments have even a little further to go.

Dictionary please
Capital goods are the equipment and materials needed to produce other goods, whether consumer or other business capital goods. Capital goods used to make defense goods may be related to the production of nondefense goods indirectly as they help keep business conditions secure, but they are not physically used to make these goods. Hence they are excluded from the above category. Aircraft, which are limited in number and very expensive of course, are excluded for a different reason. They are usually omitted because of their month-to-month volatility. But including or excluding aircraft, or even defense, the shape of the capital-goods curve remains the same.

Capital goods aren't immediately consumed or placed on our mantel to catch dust. Rather they are used to make other goods and often goods in great quantity. That's what makes them special and something to watch as an indication of confidence in the economy. Companies aren't often in a hurry to buy such equipment given the high cost. And when they do, it's typically at the end of the economic growth cycle, a repeating cycle of capital-goods boom and bust that was most dramatic at Y2K and may be repeating itself right now.

Ahead of the times
Right now demand for consumer goods isn't that great, in contrast to the strong demand for capital goods. And this is the rub. Below is a graph comparing the output of business equipment (dark line) against consumer goods output as measured by the Federal Reserve. The graph shows the swings in the capital goods sector. Businesses were trying to catch up to demand through the mid-90s, but they ended up over doing it which led to the painful contraction after Y2K. And now the trend is reappearing as the slope of business equipment output is once again steeper than the slope of consumer goods output. But that has been a classic pattern for capital goods.

The below graph offers another example as measured by the Commerce Department's GDP accounts, showing the uneven pace of investment in industrial equipment (dark line), exaggerated by its smaller total, relative to household furniture and equipment.

Businesses got burned badly by the Y2K capital expansion, a memory that has made them wary not to repeat the mistake. Cautious or not, four years of strong demand has forced businesses to increase investment.

Lead time squeeze
Adding to the unevenness of capital goods production are long lead times. Once a machine is in place, it can stamp out a great number of goods in a limited time. But the production of that machine, typically much more sophisticated than the goods it's stamping out, takes a great deal more time, especially if demand for the machinery is growing sharply and testing the available capacity of machinery manufacturers and their suppliers.

The above graph shows the rising trend in capital lead times as measured by the Institute for Supply Management. The trend is in line with the rising overall trend in manufacturing capacity. The graph below tracks machinery, showing the sharp acceleration underway for new orders and unfilled orders.

In contrast, new orders and unfilled orders for household appliances in the graph below, unlike those for machinery in the graph above, haven't been much better than flat over the past five years. And also unlike machinery, new orders for household appliances are 2-1/2 times greater than unfilled orders, evidence of the relative ease in which consumer products can be stamped out and also the relative lack of manufacturing constraints on the consumer-products side. Unfilled orders for machinery are the opposite, 2-1/2 times larger than new orders.

Machinery now and then
Before turning to the stock market, let's take a quick look at labor, specifically in machinery which is a useful category when looking at capital goods. Virtually all titles within the machinery group (beginning 333 on the North American Industry Classification System) fall in the capital goods category, and is specifically tracked across a wide variety of data sets, including the Labor Department's establishment survey.

The above graph compares recent machinery employment with retail trade employment, offering a contrast between current strength in the capitals goods sector and not-such-great strength in the consumer sector. The graph below begins in 1999 and offers a reminder of the human cost that a capital-goods collapse can have. Machinery jobs contracted sharply following the Y2K bust and, unlike output, have never recovered.

The Nasdaq danger
The Nasdaq has more than its share of capital goods producers, specifically makers of electronics equipment and electronics machinery. The graphs below compare shipments of nondefense capital goods excluding aircraft against the Nasdaq and against the Dow industrials, where consumer products makers, belying the name industrials, make up about half the index. Start-up electronics firms may not be the best investment in the unlikely event the capital goods sector begins to turn south.