Post of the Day
July 6, 1998
Qwest Communications Board
Subject: Re: Depressing article
Warning: Long reply, which may not address everything about that artical that depresses you.
General Comments: I like Qwest. I hold some of its stock. This does not mean I believe Qwest is a risk-free investment. It is possible that Qwest's business plan will fail. I'm just not concerned enough about that to unload my Qwest now at about $36 when I bought in at a merger-adjusted $25.22. Greedy me.
The above-cited article tries to discuss the business of five "bandwidth barons": Frontier, IXC, Level 3, Qwest, and Williams. Lumping these particular five companies together seems a bit odd, but let's consider the article as it applies to Qwest. I see 12 basic arguments in the above-cited article that I can comment on.
1. Capacity is at a premium on the Big 3 (AT&T, MCI, Sprint). Surprise, surprise, surprise! If it weren't, Qwest wouldn't have formed its business plan. To a certain extent, this is a poker game. We don't know whether the Big 3 are really crunched for bandwidth. They might have made policy decisions not to sell bandwidth cheap above the T3 level because that would be subsidizing the competition.
Historical note: In the late 80's, AT&T was ordered to lease bandwidth to its competitors while they built their networks. It was basically in the business of providing wholesale service to its competitor for the purpose of allowing those same competitors to win retail market share. (Ain't regulation great!) Fast forward to now. AT&T is no longer considered a dominant carrier; MCI and Sprint were never considered dominant. None of them are under any legal or regulatory obligation to subsidize new competitors.
2. Qwest shows all the signs of being built to be sold. The authors may be right on this. I'm not sure whether, as an investor, I particularly care. Either Qwest is trying to grow the business (good news for shareholders), or Qwest is trying to be taken over at a premium (good news for shareholders). The article has a negative attitude towards a "build to be taken over" strategy because it is written from the viewpoint of a customer of telecommunications services. The customer at times has different interests than the shareholder. Just ask anyone whose bank has been taken over by another bank.
3. New fiber vs. old fiber, nondispersion-shifted vs. dispersion-shifted. An interesting discussion of a technical point that should be next to meaningless to customers. Totally devoid of information that is useful to investors because none of the Big 3 would tell the author what they have. Basically, a buch of guesses. See point 4.
4. Bandwidth: The article is down on new carriers for lack of bandwidth. It quotes figures that don't seem to tie to what I know about Qwest and Frontier networks. This doesn't really matter, though. Total network bandwidth is a hype number. You try to engineer more bandwidth than you think you're going to sell, but you don't engineer massively more because there are costs involved. When the fiber is in the ground, bandwidth can be added as it is needed.
5. Sale of wholesale bandwidth and dark fiber, vs. Big 3 having committees doling out bandwidth higher than a T3. I view Qwest as being in the wholesale bandwidth business, primarily at bandwidths higher than T3. This business is viable for a startup because the Big 3 aren't in it. I don't know whether they aren't in it because of network constraints or because of internal business decisions. See point 1.
6. Discussion of switch locations, allegedly low number of switches. This is similar to bandwidth. You don't put in more switches than you need, because there are costs involved. Qwests admits that haulback is part of its business. Haulback is part of the business for the Big 3, too; the author just didn't tell us that. AT&T does not have a switch at every POP, and it hasn't seemed to hurt AT&T network quality.
7. Distance as a measure of network size. I don't think distance is a particularly useful measure, even though Qwest cites it in every press release. Number of locations (and where they are) is far more critical. By locations, I mean nodes on the network, not necessarily switches or routers. If bandwidth is cheap enough, haulback can be used to reduce the number of switches you have to deploy. Let's deploy switches because we need switching capacity, not because we want to serve more locations.
8. Linear SONET vs. 4 fiber biderectional line-switched ring (BLSR). The article claims that only IXC has BLSR. Excuse me, Qwest announced what it claimed was the world's first OC-192 SONET BLSR. Kind of makes me doubt the quality of the author's homework.
Granted, Qwest's BLSR is not universally deployed right now. Hey, we all now the network is under construction! But to a certain extent, the linear SONET vs. BLSR concern is overstated. Standard engineering practice would be to build a linear SONET ring (a.k.a. "collapsed ring") when a fiber path is available, then expand it to BLSR when a diverse path is available. [I suspect that the marketing value of being able to say you have SONET rings with route diversity exceeds the actual engineering value, but if it helps Qwest sell stuff that's okay by me.]
9. The IP network, IP over ATM, IP over SONET. The IP network is mostly hype. What makes a network packet-switched rather than circuit-switched is the switches or routers, not the circuits. Between two switches, you will have dedicated circuits of defined bandwidth. It's how the switches used these circuits that make them circuit-switched, packet-switched, ATM, etc. Sometimes a network is engineered one way rather than another because of marketing hype; the "native IP network" is a marketing hype driving engineering. If the network works, and Qwest can honestly claim the buzzwords customers want, I'm content.
10. IP telephony. This is going to be a temporary phenomenon. It works because of a loophole in regulations. Traditional interexchange carriers (IXCs) pay incumbent local exchange carriers (ILECs) per-minute access charges to connect to the ILEC networks. Internet Service Providers (ISPs) don't. Sending telephony over the internet has a cost advantage solely because of this regulatory anomaly. Unfortunately for IP telephony, this anomaly is going to be temporary. While all ILEC attempts to get regulators to make ISPs pay access charges have failed, IXCs have been very successful in getting the FCC to force per-minute access rates down. I expect that by 2002, per-minute rates for all major ILECs will be under a half penny a minute. Longer term, they will go to zero. The cost advantage of IP telephony will then be eliminated.
11. Standard Guarantees. Give me a break. If customers want them, carriers will implement them.
12. Price. The one positive comment the article had about the bandwidth barons. The article is reaonably accurate in its forward-looking statements here; who knows what market conditions will be like in two years?
On further thought, grouping these companies together makes sense from the perspective of the article, which was to discuss communications services from the standpoint of a large business customer. It seemed odd to me to group them together as investments, but that wasn't really the point of the article.
Frontier: Roots as an ILEC. More than a decade of business as an interexchange carrier. Now getting into the CLEC business. Has mobile interests. Building a facilites-based network by adding its own electronics to long-term lease of dark fiber from Qwest.
IXC: A history in microwave transmission, now getting into construction of a fiber-based network. I don't know much about them.
Level 3: Background in construction and mining. (Reminiscent of MFS, which had its origins as a child of Peter Kiewit, a construction firm.) Has leased capacity from Frontier, which leases the underlying dark fiber from Qwest. Has announced the acquisition of rights-of-way, presumably to build a large network. Probably a couple of years behind Qwest in this endeavor.
Qwest: A startup that billed itself as constructing networks. Started out with rights-of-way from Southern Pacific Railroad. (An earlier Souther Pacific Railroad network, sold years ago, goes by the name "Sprint".) Got Frontier and GTE to sign on as major clients for its signature fiber network. Regards "construction revenue" as being worth breaking out for public release of financials. After becoming established as a supplier of fiber, starts to get into retail services. This is accelerated by the purchase of LCI.
Williams: A pipeline company that built Wiltel, at one time one of the top 10 interexchange carriers in the country. Sold Wiltel to WorldCom, which is where a big chunk of WorldCom's facilities-based network came from. Now building another network, and heavily promting itself as a wholesale carrier.
So while all of these companies might provide T1, T3, and higher bandwidth private line services, where these services fit into the companies' respective business plans varies quite a bit. IXC and Level 3 might end up trying to do most of the same things as Qwest; I don't think Williams will, and Frontier's ILEC business will keep it from looking like Qwest for the forseeable future.
Let's be honest here. This whole segment isn't very Foolish, as explained in the Fool books. Their products don't have any particular consumer visibility. They require huge startup investments. And while the investments are being made, profits aren't happening. This sounds a lot like what the Fools tell us (in "You Have More than You Think") that they don't like to invest in.
Throw in a property the Fools don't discuss: regulation. A lot of money in the telecommunications industry shifts around because Congress and the FCC make policy decisions that affect the costs (and in extreme cases, the viability) of firms in this industry. I bought LCI because I thought they were the best in the industry at adapting their business to the insane changes promulgated by the FCC. I stumbled into another pretty sharp operation, Qwest, because I was holding LCI. One of my biggest worries is that the merger consolidation eliminates the competency that attracted me to LCI in the first place.
I guess this is a roundabout way of saying that trying to invest in a group of companies building fiber networks doesn't make sense to me. The details matter. It's easy to lose a pile of money in this industry by not knowing what you're doing. I prefer to invest in companies that I think have good plans and are avoiding stupidities. If I didn't see any company here that I liked as a specific company, I wouldn't be investing in this segment.