The business cycle is indeed a cycle. Tech folks would call it a sustained oscillation. They would explain that if a system oscillates, it is usually because there is some delayed response in one part of the system attempting to react to an undesired state in another part of the system. For the bullwhip effect in supply chains we looked at earlier in this series, the undesired state was the size of the order backlog and the reaction took place in the production start rate. But, how about the business cycle?

Quite a few economists and stock market gurus will tell you that the business cycle has very little to do with supply chain dynamics. It's all about the state of the world economy, technology discontinuities, interest rates, productivity gains, exchange rate imbalances and so on. Well, all such factors could undoubtedly be taken into account.

Nevertheless, I strongly believe internal supply chain dynamics and the structures that generate them, not external shocks, are to a large part responsible for business cycles in innovation driven industries. To explain why, let's look at one of my favorite high-tech sectors, the semiconductor industry. This sector had its best-ever year in 2000, when worldwide demand grew some 30%, only to collapse dramatically in 2001. At that time, I was working with one leading integrated circuit ("IC") manufacturer, and their experiences may be fairly representative for the industry as a whole.

In the summer of 2000, one key customer of this company predicted that for the next two years their demand would be greater than what their supplier could manufacture. In September, we started a so-called "collaborative planning" pilot with them, in which supplier and customer shared their internal production and forecast data. It came as a complete surprise that more supplies were in the pipeline than there was demand. It took several weeks and repeated calculations to convince management on both sides that this was not an IT tool error but actual fact. Later, in the first quarter of 2001, the first revenue shortfalls were reported. Nevertheless, several CEO's declared that this was only a temporary dip, and overall prospects were fine. Production forecasts remained optimistic for at least the second quarter of 2001. As it turned out, 2001 was disastrous, and 2002 was only slightly better.

What really happened? Here we have to look at the underlying dynamics. The earliest warning signals came from the 1999 order book. More new orders were booked than deliveries made. The ratio of these two is called the book-to-bill ratio and is a good early warning signal for mood changes in the industry. If more orders are coming in than are going out, order backlog, being the cumulative difference of these two, is rising. And so it happened. In this industry, it is common knowledge that if order backlogs are rising and capacity is fixed, pretty soon customer order fill rate percentages are going to drop and delivery lead times will go up. Everybody knew this at the time, but still nobody dared to say "No," to customers. In fact, those supply chain managers who tried quickly received a phone call from their general manager saying that "No," was not an option for their valued customer.

What about the business cycle? Surely, all this is just bullwhip effects related to material shortages and surpluses, and the business cycle is about discrepancies between capacity and demand, isn't it? Well, yes, precisely. The third accumulation that is evident from company and industry data is the rise in production capacity. As inventory increases started later than order backlog increases, but lasted longer, the same was true of capacity increases. It took some time before company management really believed that the market would be there for expensive new capital investments. Getting new fabrication plants and assembly lines financed, built, tested and in operational state took longer. By the time the industry was in one of its most serious downturns, most of the new capacity was finally coming on line. This belief was based on the situation of the supply chain one to two years earlier: gigantic order books and screaming customers. The book-to-bill ratio had gone below 1.0 as early as the summer of 2000, but that fact was ignored.

The opposite happened in 2002 and 2003, not just with this particular company but also with the IC industry as a whole. It took a long time to accept and implement capacity reductions in line with demand. In 2003, worldwide industry demand increased 16%. But, in the second half of that year, several semiconductor facilities were still being closed down in a delayed response to the production surpluses of 2001. Ironically, in 2004, sales increases will even be higher. Again, there are considerable production shortages and more are expected as demand grows higher and higher. The order books are rising sky-high once more.

What can companies do about this? They could start by believing their own data sooner, but usually this happens too late. It's not a lack of knowledge regarding these supply chain control theory dynamics, certainly not in a company full of highly trained engineers. The problem is that people are collectively trapped in the system. Those that want to row against the current may see their careers endangered by their counter-cyclical behavior. After all, "10,000 lemmings can't be wrong." If you say we should slow down because your data backs you up, but the whole industry is shouting you should go faster, think twice before you raise your concerns too loudly.

Incidentally, there is a separate story to tell about those sectors that supply capital goods to cyclical industries. IC manufacturing equipment is very complex and expensive. For convenience, assume that it lasts ten years. That means you need to replace 10% of your machines a year, on average. Now, say that demand for your IC's grows by 10%. That means you need 10% extra capacity. So, when customer demand for IC increases by 10%, demand for IC production capacity increases by 100%. This is called the investment accelerator, and it is a curse for most machine manufacturers.

Ironically, a little Euro-sclerosis can do wonders here and in cyclical markets in general. If you are in the highly flexible job markets of the Far East or the U.S., you can quickly fire staff once demand drops. Unfortunately, when demand picks up again, you can no longer find the same people. So, the net result is that average production experience in your work force drops. This hurts productivity and production quality. In Europe, CEO's would love to do the same with their work force, but their labor market regulations won't allow it. For a change, that is a government policy that makes good business sense, as most European electronics companies are very happy that they could not fire all the excess staff they had in 2001 and 2002. In 2004, they urgently need all the staff they have...and more.

This article is the fourth in a series of eight articles by Akkermans about Supply Network Dynamics. The 4th Wave will be described in the July/August edition of the Connector. To read the introduction to this series, click here.