Optical Delusions and Illusions

April 4th, 2012 by · 22 Comments

In the Wall Street Journal yesterday there appeared an article entitled "Optical Delusion? Fiber Booms Again Despite Bust" which requires a response. While noting the importance of where the fiber is in determining its value, the author continues to speak from the doubtful perspective that there is some pile of fiber out there that we have to use up before we can justify putting in a route to someplace that doesn't have any.

Nobody is out there building gobs of fiber and hoping they will come, at least in the US.  Those days really are over. The industry is building out fiber based on customer demand, meaning actual revenues and a return on investment. The business case for building that fiber does not come from traffic growth guesswork for the next decade, it comes from sites that need the bandwidth now and can't get it from copper and will sign a contract to get it.  Even the limited speculative buildouts out there that don't have customers attached ahead of time are to places with known demand and little competition.

The article quotes 19 million miles of fiber as having been built out in the US in 2011, which I have no cause to doubt. But a key point here is that the vast majority of these miles were not intercity routes like those specifically mentioned in the WJS article, but rather metro and regional routes hooking up stuff that was never hooked up before because it wasn't viable then and it is now. They didn't add to a glut because there wasn't any fiber at all there until now.

The problem here is that too many look at fiber buildouts and see the total theoretical capacity put in place as the key number. Of course there's always too much of it that way, but that isn't the right way to look at it and wasn't why the bubble burst last time.  The problem back then was that there were too many trying to sell the same exact pipes to customers who didn't really need it yet.

If an operator builds out a 432-count metro loop to virgin territory, he has just put in a massive amount of theoretical capacity that will probably never be exhausted.  Has he created his own glut?  No, because he's the only one selling it.  If two dozen other companies match him on the same route, or if he sells too much of it to competitors at prices that didn't make back his construction costs then yes things might get dysfunctional.  But that's not the fault of the fiber.

The 'fiber glut' was never about a glut of actual fiber capacity or mileage -- that was always an optical illusion.  The fiber glut was about too many companies targeting the same opportunity by spending billions up front on a business model they didn't understand. It didn't matter that traffic growth was slower than expected back in 2000, it could have been faster than expected and the system would still have collapsed.

So if the market wants to keep an eye on the risk of the 'next fiber glut', it should stop throwing around raw total capacity numbers and fiber mileage counts and focus on whether there are too many participants in the market for it to be healthy. Last I checked, the number of fiber operators was still dropping on all fronts in the US, and more and more of them are either profitable or knocking on the door.

The key question one must ask to determine the health of the fiber sector is quite simply this:  Are pricing trends stable enough that fiber builds are generally able to make back their construction costs within a period they can reasonably project?  Because once a fiber build passes that threshold, it's home free.

Categories: Internet Backbones · Metro fiber

Join the Discussion!

22 Comments So Far

  • B says:

    Well put and thought out. I guess I will try and temper my anger and disgust for Crowe & Co. Just a little longer and assume their capex and fiber builds has or will be justified someday.

    • Rob Powell says:

      Heh, I was careful not to specifically endorse anyone’s efforts there, although the WSJ article quoted Crowe as being interested in more speculative buildouts. One could argue though that all he is seeking is something Zayo, AboveNet, and tw telecom have each had over the past few years – surplus cash to invest in new directions. He wasn’t talking about anything similar to pre-2000 buildouts at all.

  • Jeff Storey says:

    I nailed it, you mean !

  • Dan Caruso says:

    Anton Trioanovski was the Wall Street Journal writer. I obviously have a different view on the theme (“Optical Delusion?”) of his article. Time will tell, but plenty of recent evidence supports the view that this is a legit resurgence.

    In my opinion, Anton displayed an extraordinary lack of understanding of the topic of his article. He failed to mention any of the significant activity of the past two years. He made no mention of (in alphabetical order) 360networks, Abovenet, Abry, AFS, AGL, Alpheus, Bank Street, Cavalier, Columbia Capital, Court Squared, ENA, Extanet, Fibertech, Fiberlight, FPL Fibernet, Lightower, M/C Ventures, NextG, Sidera, Sunyses, and Zayo. He didn’t talk about the extensive fiber construction underway for Fiber-to-the-Tower, DAS, school districts, data centers, Internet companies, and hospitals.

    How can the WSJ publish an article about the potential resurgence of bandwidth and fail to mention any of the players who are aggressively buying, selling, investing, and building network, nor the key drivers of today’s fiber construction? With all due respect for Spread and Allied Fiber, their role in the resurgence is a minor part of the story. I understand why Level 3 was a focal point of the article; perhaps the contrast in Level 3’s recent construction investment and views to those of focused bandwidth infrastructure providers would have been an interesting angle for Anton to take.

    If this was a random blog post, I would chuckle. However, I expect more from a reporter who gets a by-line on the cover of the Wall Street Journal.

  • Finally a solid conversation on the matter. As a PR professional in this industry and as an individual who has spent hours of my time with Anton before this article was written trying (in vain) to educate him on our industry, including introducing him to the “right players” to talk to, as Dan mentioned above, I could not be more disappointed with his coverage. I understand, as a former journalist, the need to try to cover all sides of the story and include skeptics as well as fiber builders, but I felt the entire story lacked the metro lit fiber perspective. Moreso, he missed the entire point: there needs to be new fiber and in key locations– so that America can compete in a global marketplace.

  • Anonymous says:

    I’m not sure I understand the backlash to that WSJ article at all. I thought the article was pretty clear that most of the recent builds are to places that need fiber either because there isn’t any now or because the demand seems clear.

    At worst, it fired a warning shot that cavalier attitudes about builds, demand, etc.. are what got the industry into trouble the first time.

    • Anonymous says:

      I should add “arrogance” to the list of qualities that got the industry in trouble the first time.

    • Anonymous says:


      I completely agree with you. In the late 90s up to the tech wreck, investors had blinders on about investing. Whether it was investment banker greed or the one-eyed thinking executives, over-investment and build-it-and-they-will-come strategies abounded and many investors were hit hard.

      For the last few years our industry’s buzzwords have been success-based builds. Naively, I ask, where is the line between a success-based build and a speculative build?

      Must a success-based build pay for itself 100% from the first customer?

      What about the second customer? Third?

      When does the line move from first customer to second to third? Should that be an internal red flag if it does move?

      If the first customer doesn’t cover 100% of the build how much should they cover to qualify as success-based?

      • Dan Caruso says:

        The point about “success-based” is very perceptive and worthy of introspection. It is easy to label an expenditure as success-based if the term success-based is ill-defined.

        At Zayo, we publish an earning supplement each quarter with ample operational metrics (on our web site). In it, we break down our growth capital into carefully defined buckets. One bucket is “network capacity”. This capital is expended to support new revenue but it is clearly not success based because it is not associated with a specific customer contract.

        Another category is “speculative”, which captures network expansions that are not funded by in-hand customer contracts. At times, speculative has revenue associated with it. However, we classify a project as speculative when the in-hand customer contract doesn’t produce a positive IRR. At times, 100% of the capital is speculative, as no customer contract is in hand when the expansion capital is authorized. If a customer contract is at hand, we classify the portion of the capital that is justified by the contract as SBC, and the remainder as speculative.

        Speculative doesn’t imply a decision is a bad one–it just means there is a bigger bet on yet-to-be-signed contracts. For example, most colo expansions are speculative. However, a lesson learned from the meltdown is that operational finance metrics must be tracked rigorously and with objectivity.

        • Anonymous says:

          Dan, thanks!!! Very instructive and certainly makes sense. I suspect your summary above is not necessarily universal among your peers?

  • Thank you Anonymous– I hope that others read the article as such– making a clear distinction about “where” fiber is needed– particularly to the towers, remote data centers, carrier hotels, inter-city, governments, educational buildings etc. That these necessary fiber builds are what will strengthen America and allow us to compete internationally.

    I also want to mention that Anton did do a lot of legwork on the article, interviewing many C-levels from key fiber players, and asking follow-up questions, such as for maps, client testimonials, financial figures etc. I believe he did his best to understand our industry and write a thorough article, probably twice the size as printed, and that his editor, without the same level of investment of time and research into our industry, cut the story down, removing these critical pieces. Perhaps we can inspire Anton and/or other journalists out there to write a follow-up piece on why the “right type” of fiber– in the “right locations”–is critical to the growth of our country.

  • Dan Caruso says:

    I agree with anonymous that there should be reason to be cautious. The industry struggled in the early 2000s, so the burden of proof is on the operators. My point was that Anton showed almost no understanding of the industry–a complete lack of awareness of the significant events and activities of the past 2 – 3 years. As a WSJ reporter, he should have done a better job showing command of the facts.

    • Anonymous says:

      What did he miss other than not mention every player in the industry and what they’ve been doing?

      He did draw the distinction between what happened in the late 90s and what’s going on now, but in a way that made you wonder if they were going to screw up again by making the same mistakes.

      • Another Anonymous says:

        The author didn’t identify why the builds are happening. I think that Dan’s point is that the original boom was speculative (bandwidth is growing>we should build!) vs. success-based (I have a specific customer who wants service at this unserved location>we should build).

      • Anonymous says:

        He would have been about 12 -16 years old in the late ’90’s, I doubt he would have much memory of what went on then (not that he couldn’t have researched it by talking with others). The Journal ran a “fiber” article around Jan ’11 talking about the market heating up. They may be more lagging vs leading indicator. Have to agree with Mr. Caruso’s original point.

  • Crossy says:

    One important issue behind the surface – of the WSJ article and of the discussion here – might be found in a basic strategy conundrum: whether this industry is inherently extremely cyclical, which would call for something like a “rule of four” players (courtesy BCG, Boston Consulting Group) in each market or whether the avalanche of new players created temporary overinvestment a mistake that future players might be able to avoid by using the proper toolset and forecasting methods. If the latter assumption holds true, the next question would ask to identify the industry’s coordination mechanism(s) to better align supply and (latent) demand and how to improve these.

    The conundrum isn’t new at all by the way. For example Bob Atkinson of CITI (Columbia University), Andrew Odlyzko and others have devoted a lot of time and energy into researching these issues, while arriving at rather different conclusions.


    • Anonymous says:

      The fundamental question I still have is what is a success-based build? It can’t be as simple as we have a customer with a b-end at location z. (If it were that simple, wouldn’t that potentially leave stranded investment at location z if that customer didn’t cover 100% of the build-out to location z?) Therefore, what percentage of the build-out cost must be covered by that first customer to prove in the build?

      Do fiber-based companies have, at a minimum, a heuristic model and at the other end of the spectrum a purely quantitative model, that says we build if conditions a, b, & c are met?

      a=customer (order);
      b=x% of costs covered by a; and
      c=any number of other possible variables (to recover the remaining costs) including other potential customers located at location z or some expected percentage of future growth from customer a over the next 3 years or ______ (fill_in_the_blank).

      Am I completely off-base here?

  • CarlK says:

    Long ago and far away, there was an Indian mathematician who turned his skills to the telecom industry in predicting the Mother of all capital spends, one that would certainly leave the players involved along with their industry in turmoil for what continues to appear as eternity.

    By contrast, today, we have industry mavens who have endured the corresponding bust attempting to teach us about LLL concepts for fiber as it relates to real estate, in addition to “success based” capex methodologies designed to capture economies of scale if they’re deciding to build it before they come according to important economic regions.

    Dan The Man, why don’t we confer with this GURU for his take on conditions today, versus yesterday, when he presciently predicted a bubble in “fiber” on the long haul side well in advance of breaking up the monopoly bottlenecks inside the last mile, which is still a necessary component for creating internet nirvana today?


    Which one of you Anonymous Dummies doesn’t understand me heading into this religious weekend?

    Got fiber? Get a new religion! 🙂

  • Anonymous Dummy says:

    “Better to remain silent and be thought a fool than to speak out and remove all doubt.” – Abe Lincoln

  • The WSJ doesn’t have to understand our industry, it was about fiber and speculation, something the WSJ readers enjoy.

    To be fair, I’m not certain most people that work in telecom understand it either. I can’t fathom the amount of debt that this Industry keeps accumulating.

    I don’t get why Jaymie is upset – Allied Fiber once again got quoted. Hunter gets more quotes in the media than he does for fiber, I’ll bet.

    Bet Corning stock bumps up today.

Leave a Comment

You may Log In to post a comment, or fill in the form to post anonymously.

  • Ramblings’ Jobs

    Post a Job - Just $99/30days
  • Event Calendar