To: Raymond Duray who wrote (550 ) 8/3/2000 3:00:17 AM From: Raymond Duray Respond to of 46821 Economics 1000000000000001: Wanna own a network? How about a working one? Interesting and suggestive article in TotalTele: totaltele.com <snip> Bottom Line: How much will your network really cost? By David Molony 17 July 2000 The economics of communications seems to have turned madcap. That's not just because dot.coms are falling out of the sky. But the sums of money being raised by telecoms companies - not the Internet start-ups - are so huge that burnrates of $100 million a year are made trivial. Almost monthly, telecoms sets a new record for the biggest bond issue the capital markets have ever seen. Last month Deutsche Telekom raised EUR15.3 billion in debt to add to funds for acquisitions and international expansion. The good news is that the markets have confidence in the industry's prospects and find the money easily enough. The bad news is that the cost of communications development is going up much faster than most of us thought. The networks are going to cost much, much more to build than anyone believed. Five years ago industry analysts were estimating that U.K. cable telephony companies would have to raise 12 billion between them to build a national broadband network over seven years or so, and eyebrows were raised then. Deutsche Telekom has raised 2.5 times as much in the last six months. Mobile operators in the United Kingdom have to pay nearly twice as much just for the licenses to build broadband mobile networks. According to analysis by Kagan World Media of the business models of each of the five winners of the U.K. spectrum auction, the operator at the top end of the network cost table must find a way to collect $99 a month from every one of 5.1 million subscribers it will need by 2007 just to break even. Economists can reasonably argue that commercial investors will not invest if they do not think it is realistic to do so. But it is far from clear that anyone knows what is realistic any more, precisely because the sums are changing so fast. Ask the representatives of Arthur Andersen, UBS Warburg and J.P. Morgan, who were among 150 or so telecoms executives and consultants at the first quarterly meeting in London last month of the new Telecommunications Executive Network. Their questions all amounted to the same one appeal: Tell us how much your businesses are really going to cost to operate and what revenues you will get, because we are going crazy trying to figure out how to value them. Some tried to work out comparisons with the payload economics of the airlines industry. Others worried that network overbuild might produce cyclical revenue flows, as tends to happen in commercial property markets. A panel of chief executives of companies building pan-European fiber systems did their best to reassure the bankers that networks would be filled and that even voice minutes would always have a selling price. They included Charles MacGregor of Fibernet, Rod Matthews of Global Crossing, Jack McMaster of KPNQwest and James Dodd of European Telecoms & Technology. At their end of the business, the economics ought to be simpler. But there are always more costs than you think. Jack McMaster reckoned the cost of lighting a fiber - covering the electronics and software need to open the windows into the network at every point of interconnection - is one-third of the cost of a live network, and adds $500 million to the cost of a $1 billion ring. That's why, he says, he has lit only 1.5 fibers from KPNQwest's European network. According to one industry estimate, Europe's pan-Continental network builders have so far lit just 12 fibers overall. McMaster says the successful telecoms operator would combine Web hosting, peering and value-added IP services; rather like KPNQwest, in fact. But McMaster was frank enough to admit that network service providers were in danger of overlooking the overriding cost of lighting fiber. <end snip> Es la luna, redux? Comments welcomed.