The gods must be crazy
Duncan Stewart
National Post
I love watching movies with my kids. When somebody onscreen is about to do something really stupid, they yell advice at the characters.
When I was studying history, there was the same tendency: Reading The Iliad, you want to yell at the Trojans, "What are you going to do with a giant wooden horse -- it's a trap you idiots!"
So looking back at the recent investment in long-haul fibre-optic networks, there is a similar sense of inevitable disaster, and a strong desire to ask, "What were you thinking?"
Level 3 Communications is apparently some new sub-basement in stock market valuations, Global Crossing is a financial arm movement that its shareholders should use to ward off evil share prices, and the 360 in 360Networks is the number of degrees in a circle --which is approximately what the company is worth.
Any business plan can be well done or poorly done -- it takes a truly dumb idea to cause everybody that tries it to fail.
The demand for bandwidth is large and new fibre-optic systems will be able to provide needed capacity, but that demand is not infinite and the business model has a fundamental flaw.
The perfect product is one that only you can make and you can charge whatever you want for it, whenever you want to. A product that everybody makes and is perishable is a very pure commodity. Once you make it you must sell it or it becomes worthless.
This is why the grocery business is so tough. All retailing has low margins, but selling food is usually the lowest. Biotech is the opposite. Patents protect you from competition, each drug is unique, products can be quite stable and profit margins are very high.
The business that these long-haul carriers tried to enter was the purest commodity business in history.
There are no differences between one cable and the next. The photons don't know or care which network they are on. The cost of switching from one carrier to another is very low and switching is almost instantaneous. Any company with access to capital can build a network almost anywhere. And the product that these companies sell, bandwidth, is not merely time sensitive -- it is time.
Let us say 10 companies have 10 identical cables, each with a capacity of one terabit per second. Once a second has passed, they can no longer sell that second -- it is gone forever. If each company's cable is using half of its capacity, the pricing pressure becomes enormous. The cost of providing each additional bit of capacity per second is zero. Not close to zero, but actually zero.
Since these companies borrowed billions of dollars to build these networks, they have regular debt payments. In order to finance those, they need to fill as much of their pipe as they can. Since their cost of providing that service is zero, they are willing to charge almost nothing for it. If there were only one cable, or maybe two or three, they could try to manage the market and maintain stable pricing.
But there is no monopoly or cartel on bandwidth providers. So we are seeing the dreadful consequences of a price war in a pure commodity. Prices go toward zero faster than anybody ever imagined. No discipline, no profits, no shareholder capital. |
