Rashe,
That was a very good reply, indeed. You brought a few things to the table that I omitted in my message. Certainly, as you state, capacity swapping has been going on for a very long time.
But many of the deals you're referring to, in all fairness to my argument <smile>, have to be regarded in the context of the times they were done, right up to the current day. Most of them were and continue to be wholesale-priced leasing arrangements. Or, they are/were sometimes even-up actual facilities swaps.
Some entailed raw fiber, as they still do even recently, but the majority of the deals were about the wholesale leasing of electrified, or SONETized, capacity, and not full fiber spectra by competing carriers, although there were, and continue to be, exceptions.
There is a difference, as I am certain you already know, between an IXC leasing an OC-48 from a CLEC (or even an OC-192 pipe), on the one hand, and on the other hand the IRUing of one of the newer grades of fiber by the entire strand (not to mention 24 strands).
An OC48 is but a drop in the bucket compared to the total potential capacity of a DWDM'd strand. And, from what I've seen, they are probably comparably priced if you look at one-to-three-year costs, depending on distances, after the OC-48s have been provisioned with SONET and other transport element gear by the selling entity. In other words, for a very low cost, you can purchase the strands that would yield many times the capacity as a comparably priced OC-48 (for the first three years) and then you have a gravy train for the remaining life of the fiber (~~ the next 10 to 15 years).
To keep this discussion honest, though, I also acknowledge that DWDM gear and SONET elements are not exactly inexpensive, either, nor are the staff and other corporate infrastructure elements to properly operate it, so one has to do the actual trade off arithmetic and engineering in order to stay above tide level.
But I think that one of the major points I was making is getting lost here, and that is the difference in the business models between an electrified carrier, and one that sells dark fiber for a living.
If you look at MFNX, they install 432 strands "per inner-duct tube" in every major route they cover. Sometimes they stuff all three or four tubes per conduit this way. That's like gold to them, since they are primarily interested in selling silica (for the moment, anyway), and not the upper layers of transport that must be maintained for years on end with sometimes dubious paybacks, and uncertainties about the next technology coming out of the factory.
In fact, it would be counterintuitive for them (MFNX) at this point to start selling SONETized and upper layer services, since they are in the business of selling "glass," for to do so, it would undoubtedly eat away at the early potentials of the glass, at least along those routes which are capacity-starved at the present time. And let's not forget, MFNX is not simply a carrier's carrier. They also sell to large enterprises and other forms of service companies.
Take a look at this LVLT announcement snippet from one of LVLT's press releases toady:
==begin snip
"Level 3's agreements for construction of a network along the BNSF and Union Pacific rights of way follow an earlier data network agreement with Frontier Corporation [NYSE:FRO - news] to lease capacity on Frontier's 13,000 mile SONET fiber optic network. The lease arrangement gives Level 3 an IP-capable network over which it expects to begin providing service to business customers in several cities during the third quarter of 1998, while Level 3 builds its own IP-based network -- the first national communications network to use Internet technology end-to-end."
====end snip
Wow, FRO really gets around, eh? They buy fiber capacity from QWST, and then they lease back SONET capacity to LVLT. Hmmm... But my point can be made here. In the first instance (where FRO is wholesaling SONET to LVLT), it is expensive, so in the second instance, LVLT will build its own. A momentary shot in the arm for FRO, but what happens when LVLT's routes are in place?
I recall some of the early fiber swap deals of the early Eighties between CLECs (there was only one then) and RBOCs. The maximum transport rates of the era was ~280 Mbps, and then ~560 or 565 Mbps (the equivalent of the payloads of a SONET OC-12, but this was before SONET.) These deals later turned out in some cases to be recriminatory and contentious types of situations, when higher transport levels were achievable. Fiber miles were tallied or granted in a "credit" scheme, in other words, and over time, a twelve mile stretch became a lot more valuable than what it was worth at the time of the deal. There was even talk about invoking a passive form of bit rate limitation (through the use of a cunningly contrived "bit rate limiter" or "dispersion inducer" made by Corning) at the time.
I suppose that the things I was questioning in my first post, and even here, were whether there is any real wisdom in selling off key competitive assets to the competition in the first place... or whether this was just a sign of being overly eager to raise capital beyond that which was achievable through securities... or a sign of hedging... or a sign of fearing a distressed state at some point, due to uncertainties about the future levels of potentially more sophisticated forms of competition?
Any thoughts along those lines? Rashe? Anyone?
Frank C. |
