Blog Post

Are Bandwidth Mergers Enough?

Updated: For nearly two months I had been hearing chatter that WilTel was in play. I had heard that even Comcast was interesting in the company, and there was some talk of Level 3. Unfortunately, I could not pin this one down, and now feel bad about it. The $781 million deal is a good step, but not the final step in cleaning up the bandwidth over hang in the market. My sources tell me that within weeks Level 3 will make a play for Broadwing. Using the same metrics as WilTel, the price for Broadwing could be around $500 million, and the deal could come before thanksgiving.

level 3 wiltel map

Stock markets are rejoicing at Level 3’s attempt to become the consolidator in the bandwidth market. I am not too sure. Ask yourself this question – if Leucadia was willing to sell WilTel, it means they have had enough of this market. They will soon get rid of those 115 million shares of Level 3 as well, soon. Hey anyone know if Level 3 will have to pay Williams Companies for the right-of-way? Or will they de-comission the Wiltel network? From a financial perspective, this will help ease some of the financing issues facing Level 3!

I have written in recent months that the bandwidth prices have stabilized a tad, and demand from new fangled services like iTunes is increasing. I might have been premature in my prognosis, some smart insiders in the wholesale business tell me. Instead, they think it is a calm during the storm. There are still too many bandwidth providers in the market, at a time when large bandwidth consumers like Comcast, Verizon, SBC and Qwest are building their own infrastructure. I see there is room for maybe one or two more wholesale providers in this market.

Lets analyze this market a little bit further. BellSouth is working with Sprint. Verizon has started aggressively to move its traffic onto UUNet/MCI. SBC has AT&T assets to play with, Qwest has its own pipes. Google is building its own backbone. Comcast has bought dark fiber and has basically started to build its own private Internet. Time Warner is doing the same. Level 3 faces a future where a lot of its IP Transit business is going to vanish, because most of these big players are now peering directly with each other. Yahoo is one of Level 3’s biggest customers, but they are striking a pose with the Bells these days. Will that become a problem for Level 3? Possibly?

Level 3 also has some other issues it might have to contend with: how to replace the dial-up business that is drying up? Where is the future growth – not in bandwidth wholesale because even MCI and AT&T have started to build their own networks again. The glimmer of hope for Level 3 is to become a player in the content distribution business. Wiltel has Vyvx, which could give some expertise to Level 3, and possibly make it a competitor to Akamai. Nevertheless, what happens in the future, only time will tell. But I do applaud Level 3 management for making the move.

9 Responses to “Are Bandwidth Mergers Enough?”

  1. What about this blimp technology I’ve heard about is there really going to be old style airships delivering serious banwith to rule areas like mine in upstate new York and if so how long do you think it will take ?

  2. Re: Bandwidth Mergers

    There may be another way to look at this. Google’s dark fibre
    acquisition gives a hint.

    Specifically, I see some separate networks, some physically separate
    like Google, some technically separate by being encapsulated in the
    current nets (note plural).

    Your comment about the deficiencies of Internet 1.0 are dead on, but
    certainly not the fault of the originators in 1968 and 197x. Our use is
    outgrowing the design of the original net, with somewhat predictable
    results. QoS, priorities, filtering, etc. are all adaptive attempts to
    remedy these limitations.

    The high speed Internet2 used by science and government is slightly
    different than 1.0, but more specifically, limited in who connects for
    what currently. It will probably remain that way.

    What will be the result? We will eventually realize that not only does
    one size not fit all, it *cannot* fit all. We are in the process of
    building multiple nets that will interconnect at well defined points,
    not everywhere, with well defined protocols and limits. It must be that
    way to serve the separate interests.

    Yet it will not result in Balkanization, but speciation. Nets will be
    optimized for classes of traffic and classes of applications. They will
    be invisible but there if you need them, efficient for specific

    Overall, a network of nets will be more robust and more efficient than
    one master net that has to either warp itself or warp traffic to work.
    One net cannot work well with all these new demands, therefore we are
    seeing sprouts, seedlings of a new growth that will deliver the MultiNet

    Bill Nicholls

  3. Jesse Kopelman

    Jacob, Ed Whitacre is not so much generically crazy as meglomaniacal. The Internet is the worst possible thing he can imagine, because it involves cooperation for the common good. Ed wants control and restriction, with his company being the gatekeeper. His concern is about squeezing the maximum profit from things he can envision and control, not facilitating new services and having to live with ancilary revenue.

  4. The real issue is missing last-mile diversity. As long as a broad band connection is as expensive and as slow as it is now, there will be no traffic to feed this backbone. Right now, it doesn’t look like the US will ever catch up to more modern countries like China and Korea.