Archive for the 'Telecom Boom, Meltdown and Resurgence' Category

Fueled by the Dot Com boom, billions of dollars were invested in constructing and lighting fiber networks in the late 1990s.   Valuations soared, and those who timed their exits right became multi-millionaires.   Those who didn’t exit by the early 2000s saw their fortunes evaporate.

Like all massive corrections, the telecom meltdown was fueled by truisms.   The Internet would change everything, and it did.  Bandwidth would grow rapidly for as far as the eye could see, and it has.  Fiber would be the workhorse of the Internet, and it is.  These trends are as powerful today as they were speculated to be in the 1990s.  Moreover, they will be prominent a decade from now, and the decade thereafter.   The telecom boom was built on a rock solid foundation.

Only partially true was the notion that the meltdown was caused by the overbuilding of fiber and capacity.   The majority of fiber in use today was constructed after 2002.  Without doubt, > 95% of the capacity in use today was created long after the end of the telecom boom.  Looking forward, more fiber will be built in the next ten years than exists today, and 95% of the capacity in use ten years hence has not yet been created.

So what went wrong?   What caused the meltdown?

First, during the telecom boom, too many fiber companies were created.  In their rush to build network, they focused on the most obvious geographies to build fiber.  They built networks in NYC, in Chicago, in the Washington D.C. market, in Dallas, and in the big California cities.  They connected to the same dozens of buildings in these markets, and they connected up the major markets along similar routes.   In many cases, they traded fiber strands with one another – or leased fiber on a competitor’s network.

Second, these companies rushed to light their networks, creating a glut of finished capacity relative to the embryonic size of the circa 2000 Internet.  Too many competitors combined with too much finished capacity created a blood bath.

Third, over-exuberant investors and inexperienced management teams added fuel to the fire.   Instead of acknowledging the situation, they hid behind story telling, fiber swaps, accounting gimmicks, and acquisition dust clouds.  They looked to buy themselves time, hoping to find an opportunity to exit prior to the boom turning to a bust.   They denied fault and instead blamed their competitors.

For fiber-based providers of bandwidth, the meltdown ended in mid 2005.   Since the mid 2000s, nearly all Bandwidth Infrastructure companies have done very well for their investors.   Revenues grew at double-digit rates.  EBITDA grew even faster, with many producers of bandwidth generating EBITDA margins in the 50-60%+ range.   Most generated free cash flow.

Slowly, the “too many suppliers” problem corrected itself. Today, roughly one of twenty companies remains in business, and only a handful of companies exist with substantial size and scope.  Consolidation will continue–as it does, balance will emerge between those who supply and those who consume bandwidth.

Despite these fact-based observations, a fundamental question lingers.   Is the owning of fiber network, and the production of bandwidth, a path to creating equity value?

The naysayers need only to point to Level 3 Communications.   Level 3 owns the most fiber, across the broadest of geographies, and with the fullest range of service offerings.   It led the consolidation of the industry by purchasing dozens of fiber providers.  Yet, its financial results remain disappointing.   Organic revenue growth is underwhelming and EBITDA margins are in the mid 20%s.  Its stock price remains low.

The dearth of other public companies puts additional weight on Level 3’s results. TW Telecom and Cogent perform well, as did Abovenet when it was a public company.  However, XO Communications (now private) and Global Crossing (acquired by Level 3) performed poorly.   All cite the challenges on pricing pressures, difficulty in growth, and capital intensity.  The development of value added services and the expansion to new geographies are positioned as necessary to preserve revenue and margin, instead of as an opportunity for incremental value creation.  With the exception of Cogent, all are (or were) opaque on their operational financial metrics.

Adding to the murkiness is the impact of traditional telecom companies (Verizon, ATT, Centurylink) and CATV companies (Comcast, Cox, TW Cable).  Will these companies hurt Bandwidth Producers in their quest to develop profitable business models?

Intuitively, powerful trends suggest bandwidth production should be a great way to make money.  Bandwidth demand is growing by 40-50% a year, and this will continue for as far as the eye can see.   Fiber is the critical resource needed to produce bandwidth, and industry consolidation is leading to a favorable balance between those who need and those who produce bandwidth.  New entrant barriers are gigantic, ensuring improving environment will not be disrupted.  Yet, if owning fiber networks and producing bandwidth leads to equity value creation, why aren’t we seeing empirical proof of it? 

The positive results of TW Telecom, Cogent, and Abovenet (when public) are positive signals that bandwidth production might be on a path toward being a lucrative business.   Their business models vary, and their messaging about the state-of-the-industry, raises questions.  Nonetheless, their financial performance in recent years has been strong.  All have seen 8 – 12% revenue growth, achieved EBITDA margins of 35 – 45%, and generated free cash flow.

Perhaps more promising is the valuations credited to private companies that focus primarily on bandwidth production.   Fibertech was purchased by Court Square in late 2010 for 12X LQA.   Abovenet was purchased by Zayo Group, with backing of new investor GTCR, at 9.5X LQA.   Lightower buyout by Berkshire at 12X LQA was announced in early 2013.  Highly respected Private Equity firms facilitated these three transactions.    The healthy multiples suggest a consensus view that owning fiber networks and producing bandwidth will lead to equity value creation.

The ultimate answer to the question seems obvious.  With robust bandwidth demand curve, increasingly healthy industry structure, and high barrier to entry, Bandwidth Production will prove to be lucrative.   However, the path toward this answer is likely to be bouncy.  Not all of the potholes have yet been avoided.

Strategies still need to be sorted out.   Multiple strategies will prove to be effective, but not all strategies.  Some strategies will lead to poor outcomes for investors.

Execution Matters.  Though bandwidth producers have strong tailwinds behind them, they also face execution challenges.  With bandwidth growing at 40-50% annually, the business environment is in a constant state of flux.  Technologies change.  Pricing is fluid.  Integrations are hard.  Industry consolidation creates murkiness.   Processes and systems are complex.

Poor Track Record in Creating Value.  Telecom management teams have a poor track record in creating value for their investors.   Perhaps circumstances dictated this outcome.  Perhaps weaker management teams have been weeded out.   Nonetheless, the fact remains that most telecom management teams do not have a sustained history of running large fiber businesses in a way that creates value for their equity investors.   Until proven otherwise, investors will remain susceptible to unproven management teams.

Effects of Remaining Consolidation.   Most of the consolidation has already occurred.  However, the remaining steps will be the most profound.  The long term industry structure will undoubtedly benefit, but some investors might not fare well along the way.

New Fiber Construction Projects / Entrants will Face Steep Hurdles.   The past five years have proven to be lucrative to many owners of fiber networks.   Some investors will over-react to these successful outcomes and, in hopes of achieving similar results, will make poor investment decisions.   They may overlook the cost of developing new fiber networks, and the difficult inherent in competing against focused competitors with established business models and deep customer relationships.

In the 1950s, Oil Production must have been a lucrative business.   Imagine a conversation that might have taken place between two savvy investors in the space.

“We certainly are fortunate to be in the Oil industry during its heyday,” says one magnate as they sip cognac at their country club.

“Ah Yes.  We are lucky to have been in the right place at the right time,” agrees his colleague, as he puffs his Cuban cigar.

The first ponders out loud: “The question is ‘how long do you think this Oil Boom can last?’.   Should we sell now before this Oil trend subsides?”

Fiber, like oil, is a scarce resource that will be relevant for multiple generations.  Bandwidth production, like oil refinement, will be a lucrative industry for many decades to come.   The telecom boom and meltdown was the first chapter of a long novel, one that will be played out over 100 years.   Though the remaining chapters will be less dynamic, many fortunes will be made and lost in the coming years.


Fathers of Invention of Level 3

Among the most exciting days of my career was April 1st, 1998 (ummm, yes April 1), when USA Today featured a story that unveiled Level 3 Communications.  I was an integral part of the team. The second to the last line made us especially proud: “It’s a dream team with a dream network and a killer business plan.”   I guess we should have focused more on the final line:   “Just as long as they don’t screw it up.”

Caption on the image  that appeared on the USA Today Cover (at right): Billionaires Warren Buffett of Berkshire Hathaway, Bill Gates of Microsoft, and Walter Scott of Peter Kiewit Sons’ crossed paths in 1995 at a manion in Dublin, Ireland, where they discussed the Internet. The result is James Crowe’s Level 3 Communications, a budding force in the Telecom Industry.

OMAHA — Walter Scott, CEO of Peter Kiewit Sons’, calls the group “Our Gang.” They are Warren Buffett’s billionaire and executive friends, including Scott, Microsoft CEO Bill Gates and former Coca-Cola President Don Keough.

Every two years, Buffett invites about 30 of them and their significant others to a multi-day retreat. In the summer of 1995, they gathered at a mansion outside Dublin, Ireland. The topic was the Internet. Gates and his wife, Melinda French, gave a presentation. It jarred Scott.

Over the decades, Scott had so successfully run publicity-shy Kiewit — a builder of roads, dams and other monster projects — that the company often had excess cash to invest. One investment was funding James Crowe in 1989 to build MFS Communications, a phone network that would compete against the big local and long-distance phone companies for mostly corporate business. By the time of the Dublin retreat, MFS was the biggest of the so-called alternate access carriers.

Gates told Our Gang that the Internet was going to be huge and that he was swinging Microsoft to meet it. He said the Net would threaten the traditional phone powers by sucking away data and voice traffic. Scott realized MFS, too, could be hurt by the Internet. “Afterwards, I sat down with Bill and talked with him about it,” Scott says. “My gut feeling was, if you weren’t part of it, you were going to be left behind.”

Scott flew back to Omaha with Buffett, whose office is on a floor of leased space in Kiewit’s headquarters here. Scott sat down with Crowe “and talked to him about where Bill’s head was.” Scott wanted Crowe to look into the Internet.

Crowe says, “He told me that anytime anyone’s told him there’s a risk, there’s always been an opportunity there.”

Crowe launched what MFS called Project Silver to study the the Net and what MFS should do in response. That was the genesis of Level 3 Communications. Today, Level 3 is listed for the first time as a public company on Nasdaq. It begins life with a market value of $10 billion and a loud buzz.

A revolution in the making?

Level 3′s business plan, if correct in its assumptions, could drive down the cost of long-distance phone calls to almost nothing and hasten the decline of the big phone companies.

Level 3 plans to build a global fiber-optic communication network entirely based on Internet Protocol (IP) technology. Crowe is betting that the dominant networks of the world — the ones that carry phone calls using circuit-switching technology based on a 100-year-old design — are dinosaurs.

Not everyone agrees, least of all phone company people. “There’s a bright future for IP, but it’s not a walk-away win,” says Arno Penzias, chief scientist for Lucent Technologies.

And there’s a twist to Level 3′s story that could make things interesting. Level 3 has made a dangerous enemy: Bernie Ebbers, the dynamic and powerful CEO of WorldCom.

Yet for many reasons, the industry is watching Crowe closely. After all, he built MFS into an industry power, then sold it to WorldCom in 1996 for $14.3 billion. And he has an injection of $3 billion in cash and assets from Kiewit for Level 3.

“It’s like watching him trying to make lightning strike twice,” says industry analyst Jeffrey Kagan.
In 1995, the Internet was just taking off and Project Silver team members could find little reliable analysis beyond the fact that the Net was messy, unreliable, interesting and cheap.

Crowe’s epiphany came when he flipped through an industry newsletter and saw a chart from consulting firm North River Ventures titled, Cost to Deliver 42 Page Document. Faxing it from New York to Tokyo using AT&T cost $28.83. E-mailing it over the Internet cost 9.5 cents. “That’s when I realized this was not driven by cool people on the cover of Newsweek,” Crowe says. “It was driven by economics. When we figured it boiled down to bucks, all of us took notice.”

The Project Silver team visited 20 to 30 Internet service providers (ISPs). Crowe went to six. The team concluded that to move fast, MFS would have to buy its way into the Internet. The best ISP to buy was UUNet. The path had come full-circle back to Gates. Microsoft owned 14.7% of UUNet and Microsoft Network was its biggest customer.

Crowe and Scott flew to Microsoft headquarters in Redmond, Wash., to talk to Gates. If Gates had said no, MFS would not have tried to buy UUNet. Gates gave his OK. His only condition was that Crowe had to keep UUNet CEO John Sidgmore. In April 1996, MFS bought UUNet for a stunning $2 billion.

Clash of the titans

Crowe, 48, has big facial features and a booming voice. He prefers to dress casually in a Dockers style. He’s a technology nut who wired all kinds of gadgets together in his home. He eats lunch in the cafeteria on the ground floor of Kiewit headquarters. Level 3′s offices are in one small corner of the stoic 15-story building.
People describe Crowe as incredibly focused, good at communicating a point simply, good at recognizing and hiring talented people and a voracious reader. (One of his favorite books: Snow Crash, a science-fiction novel about cyberspace, by Neal Stephenson.) “He’s very intelligent,” Scott says. “And he gets loyalty and enthusiasm from his people.”

That loyalty is what got Crowe in trouble with WorldCom’s Ebbers. WorldCom, based in Jackson, Miss., has grown by making ever-bigger acquisitions, mostly in the long-distance phone business. Ebbers needed a company like MFS, which has lots of local networks and business customers. UUNet made MFS even more attractive. Within months of MFS buying UUNet, WorldCom made its $14.3 billion offer for MFS.
Crowe did not want to sell MFS, but the offer was too good. He knew the Internet train was coming and that MFS, mostly built on the older circuit-switching technology, would have a rough time making the transition — even with UUNet’s help.

Crowe was to become chairman of the merged MFS and WorldCom, but it was clear he’d be second to Ebbers, who would be CEO. Most of MFS’ top management packed up their Omaha homes and moved to Jackson. Three weeks after the merger closed, Crowe quit and returned to Kiewit.

“I was talking with Jim (Crowe) about what he’d like to do and told him we’d be willing to support him if he was interested in starting a business,” Scott says. Kiewit has two parts. The main one is the construction business. The other one is Kiewit Diversified Group, which holds all Kiewit’s investments in companies. Those companies have ranged from small telecommunications ventures to a power company. Scott asked Crowe if he’d like to take over Kiewit Diversified Group. “I felt Jim would do more with it than I would.” Total value of the group: about $3 billion. It would become Crowe’s seed money.

Reinventing a success

The concept for Level 3 is simple: It’s MFS with a fresh start. Same business plan, but this time based on Internet Protocol (IP) technology. Crowe would use Kiewit’s money to start building a network that could carry information and voice conversations far more efficiently than any network out there, then slowly fill his fiber-optic pipelines by siphoning off business from the telecom giants.

Crowe wasn’t the only one with the idea. Around the time Crowe quit WorldCom, Qwest Communications was starting to get noticed. Based in Denver, Qwest was using money from railroad magnate Philip Anschutz to build its fiber network, which would be partly IP-based and partly circuit-switched. Anschutz put Crowe on Qwest’s board. Qwest also hired fiery AT&T executive Joe Nacchio as CEO. From the outset, Crowe told Nacchio he might start a competing business.

Just before unveiling Level 3, Crowe quit Qwest’s board. Now he’s moving Level 3 to Denver, right in Qwest’s back yard, to take advantage of telecommunications talent in that region. Relations between the two are said to be chilly. Nacchio did not respond to requests to talk about Crowe.

In the meantime, Crowe figured that if he was going to do an MFS all over again, he might as well get the MFS team back. One way or another, he got 18 of his top 20 MFS executives out of WorldCom. Most have taken the exact jobs they had at WorldCom. All were wealthy enough after MFS’ sale to never work again, but they’re in their 30s and 40s and sound thrilled to be back together for another run. Mike Frank, 44, head of Level 3′s human resources, says that at WorldCom: “My usefulness was not appreciated, and I wasn’t fulfilled.”

“I have a lot of loyalty to the guys who got me here,” says Ron Vidal, 37, another member of Crowe’s reconstituted team.

Crowe’s departure from WorldCom irritated Ebbers. But the mass exodus made him boil. The rancor runs deep throughout WorldCom. WorldCom would like nothing better than to bury Level 3 in the marketplace.
That threat notwithstanding, the combination of Kiewit’s backing, the old MFS team and Crowe’s savvy makes many in telecommunications believe Level 3 will succeed. “There’s a lot of faith,” says analyst Kagan.
Investors are expected to snap up the stock. Before the Nasdaq listing, shares were tough to come by. Peter Kiewit Sons’ stock is privately held by employees only. Kiewit Diversified was held mostly by employees and former employees, though shares could be sold publicly. With the Nasdaq listing, an official split from Kiewit and a name change from Kiewit Diversified to Level 3, the stock goes public.

It’s too soon to tell whether Level 3 can challenge some of the regional Bells or top long-distance companies. The telecom giants know about the efficiencies of IP, too, and some are building IP-based networks. They scoff at the notion that circuit-switched networks will become albatrosses. “Nobody’s stupid here,” Lucent’s Penzias says. He holds that communications will wind up being a mixture of circuit-switching and IP. “One protocol to do every job would be a step backward.”

Meanwhile, WorldCom, Qwest and others are building IP fiber networks. Teledesic and Motorola’s Celestri will create high-bandwidth IP networks using satellites by around 2003. Still, demand is exploding. Communications capacity is already in short supply for data traffic, which is growing 150% or more a year. Voice calls over IP networks, a business that barely existed in 1997, will be at least a $1 billion industry by 2002, according to Forrester Research.

Level 3 seems to be a player to watch. “It has an opportunity to be quite disruptive in the industry,” says Mark Bruneau of consulting firm Renaissance Worldwide. “It’s a dream team with a dream network and a killer business plan. Just as long as they don’t screw it up.”