[ service_innovation investor ] prev next
Transformation Interrupted - Part II: What Went Wrong?
(Please note: Originally published in July 2002, exhibits from this report can be seen in the downloadable .pdf copy)
II. What Went Wrong?
A Very Large Elephant
In the well-known parable, three blind men touching an elephant each come to entirely different conclusions as to what it really looks like. The telecom implosion is still underway, very large, and, like the elephant, subject to differing interpretations. The two most common views are that the pre-Worldcom meltdown was caused by deeply-flawed market deregulation, and a capital markets bubble which paralleled the dotcom universe.
There is no question that some of the former (bad regulation), and a good deal of the latter (money like water) played large roles in getting us to where we are today. We will discuss each of them – regulation quite briefly – but also introduce what we believe to be a more important causal element: business strategies which made the problem much, much worse. More important because fixing business strategy is well within the control of surviving businesses – it is something they can actually act upon today instead of grimly and passively waiting things out. And more important also, because we believe the future of the telecom industry – restoring growth and innovation – very much depends on whether we learn, and what we learn, about competitive strategy from the current nightmare.
Bon Voyage!
The Telecom Act of 1996 relied on a two-pronged, carrot-and-stick strategy to create a pro-competitive market.
First, Baby Bells (more generally, “Incumbent Local Exchange Carriers”, or ILECs) were required to allow competitors to use parts of their networks (what was referred to as “unbundling”), so challengers could avoid having to duplicate the most capital-intensive of network assets – the famous “local loop”. It would be infeasible, take interminably long, and be just plain wasteful for new competitors to run duplicate sets of wires to homes and businesses. But this regulatory “stick”, the idea that not only could new companies rent out and directly interconnect with incumbents’ facilities, but also pay for their use at “unfairly low” wholesale prices was something that the ILECs never really quite accepted.
Then there was the carrot. ILECs were limited to providing local services within their region. The (then) juicy business of long-distance was reserved for the post-divestiture AT&T, as well as MCI, Sprint, and a few others whose genealogy couldn’t be traced back to the old Bell System. For an ILEC, the upside of complying with the onerous-feeling unbundling requirements was that they might then be allowed to drink from the golden chalice of long-distance revenues. This would require an ILEC to demonstrate not only that it had indeed unbundled its networks in a challenger-friendly way, but also that real, measurable competition had demonstrably resulted as assessed using a fourteen-point checklist known as “Section 271”.
Lawmakers and the FCC assumed that the carrot of entering the long-distance market combined with the stick of enforced facilities sharing would provide [incumbents] powerful incentives…
Lawmakers and the FCC assumed that the carrot of entering the long-distance market combined with the stick of enforced facilities sharing would provide powerful incentives for the ILECs to open up their local service monopoly. In reality, enforcement turned out to be largely non-existent, as ILECs, no slouches at dealing with regulation, stalled, pursued legal challenges to the new regulatory framework, and even when found in contravention of the regulations were handed fines that amounted to no more than wrist slaps. Moreover, the golden chalice turned out to be made of baser metals. The long-distance market dwindled rapidly – prices declined, and margins headed ever-closer to zero. Several of the incumbents appeared to be in no hurry to enter the supposedly-prized LD market.
For new competitors (“Competitive Local Exchange Carriers”, or CLECs), the practical effect of this foot-dragging was myriad obstacles thrown in their paths as they tried to rent out, touch, or otherwise connect with the incumbents’ networks as the ‘96 Act permitted. The true extent of “operational stalling” will probably never be known publicly, but even if incumbents had wanted to make life easy for their competitors, they would have to overcome large operational difficulties. The ILEC’s complex, expensive business processes and information systems had not yet been re-designed to allow the kind of the wholesale trading that the ’96 Act required.
It was unrealistic to assume that the ILECs would not only willingly enable competition in their core businesses, but also invest in speedy overhauls of complex technical barriers to doing so. The cumulative Operations Support Systems (OSS) and billing investments of incumbents totaled billions of dollars and millions of man hours. Pasted together from huge, creaky, non-Y2K compliant programs, these information systems were highly inflexible and resistant to design changes. Changes there could potentially disrupt the core business operations of the ILEC itself.
The ILEC delays have allowed them to prevail, and most CLECs have gone under without any regulatory intervention. But as with the blind men, the story is not as simple as that. While CLECs have always needed some degree of access to ILEC networks, the vast majority always intended to be more than resellers of supply wholesaled by incumbents. CLECs committed vast amounts of capital to transitioning from being (mere) resellers, to true “facilities-based” providers, i.e. carriers with their very own network. And, across the street, backbone data carriers, an almost entirely unregulated business, have failed more spectacularly than their CLEC cousins.
The Capital Made Me Do It
The tipping point, if not the wider contagion, of the telecom collapse is widely blamed on the unsustainability of too many businesses fueled by far too much capital, especially in the form of high-yield debt. The capital markets say capital was the problem. New companies gorged themselves (or were force fed) on nearly-free capital, ultimately defaulting on massive borrowings. Adding insult to injury, a mild recession drove the industry into a cyclical decline – a possibility many investors had forgotten.
The bubble was fueled by four sources of financing, all with historically different aims and risk profiles, yet for the first time all converged on a single exit strategy– early sale of companies at a substantial premium. Venture capital, “mass market” venture capital (i.e. early-stage, small float public offerings), vendor financing, and high-yield debt, were all blended with this single objective in mind.
...the combined results of extreme leverage and overcapacity is … the “stranded assets” problem
While all these forms of financing were used in unprecedented amounts, high-yield debt was the largest source and most dangerous form. Equity can be written off in bankruptcy, and an otherwise reasonably-funded company can continue to operate successfully. The turnaround or unwinding of failed debt-financed businesses is much more difficult. In this case, the combined results of extreme leverage and overcapacity is starting to look like the “stranded assets” problem: infrastructure which, even when its book value is written down to zero, still has a very difficult time earning any returns in a market flooded with low-priced supply. When even large incumbents cannot earn their cost of capital, reincarnated competitors with a cleaned-up balance sheet and a “free” network still have nowhere to go except back to where they started
High-yield debt was an essential component of the telecom bubble, precisely because it required so much money to build all this infrastructure. There was no way through venture capital alone, or supplements from small float public equity offerings, to fully finance these ventures. And time was of the essence. So convertible and high-yield debt investors constructed for themselves a new exit scenario under which they could rationalize the large risks they were undertaking.
Under the exit plan, an entrepreneurial CCC-rated Telco would obtain rights to capture customers, and build out a comparatively small amount of infrastructure. More often than not, this build-out was generously, and dangerously, financed by network equipment vendors whose commitment was often the basis for attracting subsequent financing tranches. The new Telco would later sell itself to a (back then) AAA-rated Telco with its vast infrastructure, ever-expanding need for more customers, investment-grade credit rating, and insatiable hunger for scale2.
Cable Days Are Here Again
It was reminiscent of the early, heady days of the cable industry – stake out a territory, build for EBITDA growth not profitability, or maybe even just revenue growth, and then sell. Except it was much better than the cable days—ready-made, very large buyers already existed! The decades-long (and, until recently) gradual consolidation of the cable industry could be compressed into, say, five short years with PacTel, USWest, Ameritech, Nynex, Bell Atlantic, GTE, BellSouth, et al waiting at the end with grudging admiration and a very large check.
Individually, each deal (fly-specked with covenants) seemed to fit an acceptable risk profile. This attractive exit scenario enticed debt holders to expect both, as the label suggested, high yields, but also some prospect of an investment grade safety net – a no-brainer, a “ten-bagger”, a short ferry ride on an inland waterway, safely disembarking on terra firma with high returns, not the financial equivalent of a winter Atlantic crossing on the Titanic.
The incumbent Telcos did find someone to buy – each other … [they] watched as the challengers went under, occasionally inspecting the debris for cheap assets, but generally keeping the purse strings drawn tight.
But not only was the ferry ride storm-tossed, its very destination evaporated. The incumbent Telcos did find someone to buy – each other. Thus the exit plans rapidly crumbled, not just for a few unlucky sailors but collectively for the entire newly-financed fleet, as the universe of qualified buyers disappeared into each others arms. The incumbents watched as the challengers went under, occasionally inspecting the debris for cheap assets, but generally keeping the purse strings drawn tight.
If the telecom industry absorbed $240 billion of high-yield debt (nearly 48% of high-yield issues from 1996-2000), it is perhaps no surprise that unwinding excesses of that magnitude will cross well over the line of “creative destruction” into just plain old destruction, and for a period potentially as long as the five years or so it took to create the problem in the first place.
Me Too! But Cheaper…
Earlier, we said the entrepreneurial excitement created by deregulation focused on three things: new networks, new services, and new business models. Only the easiest of the three was actually delivered, and in massive quantity: new networks. Not delivered was the industry “transformation”, with tantalizing but frustratingly vague visions of big bundles of customer-friendly, sophisticatedly marketed, convenient, incumbent-killing services.
The principal value created by new telecom service providers and their infrastructure suppliers was lower prices – in the case of data services, dramatically lower. In all other respects, the new competitors were driven by business models largely indistinguishable from those established over decades by powerful incumbents.
Competitive Local Exchange Carriers (CLECs) aimed at the same (midsize business) market, with the same (local and long-distance, data maybe later) offers, in the same (high-density urban/suburban industrial) regions, with the same (15-20% lower) prices, and the same (aggressively commissioned sales force and agents) distribution channels. They clustered in the same “beachfront properties” of telecom competition. In the zone encompassing Boston, New York, Philadelphia, and Washington for example, over a dozen competed to first resell circuits from their incumbent competitors (Nynex and BellAtlantic), and then build the (same) networks to directly serve newly-won customers at lower marginal cost. With price as the value proposition, the CLECs managed to make commodities out of incumbents’ already-unexciting offers before their first switches were even installed.
Businesses got data price wars, and residential data customers got confusion, the increasing sense that there were too many phone companies, and that maybe “doing the broadband thing” might best be left to their local cable company
Customers, not exactly at the center of the industry’s attention, by and large got neither any voice services they couldn’t have gotten a decade or more ago, nor better provisioning times, nor more responsive customer service, nor even an easier to decipher bill most of the time. Businesses got data price wars, and residential data customers got confusion, the increasing sense that there were too many phone companies, and that maybe “doing the broadband thing” might best be left to their local cable company.
In the long run, it isn’t much of a competitive strategy in a scale-based, capital-intensive industry for a smaller business to offer exactly the same thing (or worse) as a much larger competitor, but at a lower price.
For every Southwest Airlines (which does, in fact, offer a differentiated product – convenient point-to-point, on-time service, as well as low prices) there are forty or more failures, depending on how far back in time you care to measure. In fact, the airline industry provides a cautionary example of a deregulated utility decades later. Vast segments of the industry built through cost-driven mergers full of route system synergies have failed to produce a return on capital, the Southwest Airlines, or Virgin Atlantics notwithstanding. As Warren Buffett famously remarked, “despite putting in billions and billions and billions of dollars, the net return to owners from being in the entire airline industry, if you owned it all, and if you put up all this money, is less than zero”.
Analogous to discount airlines, the “me too, but cheaper” CLEC was especially vulnerable to incumbents doing everything in their still-considerable power to protect their markets and shareholders. Most CLECs started first by reselling incumbents’ circuits (whose wholesale prices alone were the subject of a regulatory guerilla war, never mind the inter-company business processes involved), then evolved to only marginally more profitable “on-net” circuits, portions of which were owned by the CLEC as they began deploying their own networks.
CLECs were ultimately dependent on interconnection with Regional Bell Operating Company (RBOC) networks, and their provisioning and billing systems. Very rapidly CLECs began dying the Death of a Thousand Dependencies, skillfully administered by incumbents whose regulatory finesse, “by the book” work rules, balance sheets, and market share enabled them to pull up the drawbridge and wait out the siege.
Freight Cars of Grain
Perhaps greatest optimism and glamour was associated with the new fiber backbone data networks from PSINet, Global Crossing, Williams Communications, 360 Networks, Level(3), Qwest, and many others. Yet they rapidly found themselves squeezed between massive overcapacity and unit price declines more rapid than their wildest worst-case scenarios could have imagined, as shown in, Exhibit 5 and as much as 75% a year in certain city-pair markets.
Simultaneously, their ability to get the very high capacity, fiber-based transport services directly to big-spending large enterprise and internet service provider (ISP) customers was thwarted. The relatively scarce and expensive building interconnections through metropolitan “fiber rings” to their own networks proved hard to come by, and were largely controlled by others.
“customers want sandwiches, but all [the new backbone carriers] can offer is freight cars of grain” – Susan Kalla
The new backbone carriers struggled to take customers away from existing suppliers such as AT&T, Sprint, and particularly Worldcom, who in turn were quick to cut deals which in effect aggregated voice and data services in more attractively-priced bundles than their upstart, data-only competitors could provide. Without direct access to end customers, new broadband carriers were relegated to being wholesalers, or in industry terms, “a carrier’s carrier.” In the words of Susan Kalla, a telecom analyst turned bearish quite early, “customers want sandwiches, but all [the new backbone carriers] can offer is freight cars of grain”.
Silicon Economics: Weapon of Cash Destruction
It was going to be like Intel and microprocessors – silicon economics would be brought to the heart of the telecom industry. Prices per unit of computing capacity fall rapidly and constantly, but customers find new and more uses for ever-cheaper computing power – demand growth more than makes up for unit price declines.
For the new backbone carriers, this intended price elasticity of demand equation began unraveling almost as soon as it was applied. As a crude analogy, imagine that in the early 1980s there are five or even ten brand-new, highly-leveraged Intels, each with the same multi-billion dollar investment in wafer fabs, and identical microprocessor product lines. Further imagine that Microsoft, the MS-DOS operating system and, thus, nearly all the early PC applications (think Lotus 1-2-3) from dozens if not hundreds of vendors, did not exist. The demand curve would have taken a very different and, for the hypothetical Intels, fatal turn – the jurisdiction of Moore’s Law would have been limited to just electrical engineering, instead of silicon economics.
The backbone carriers had planned for Moore’s Law unleashed on their networks to duplicate the silicon economics success story, not just for the telecom industry in general, but for each carrier in particular.
The backbone carriers had planned for Moore’s Law unleashed on their networks to duplicate the silicon economics success story, not just for the telecom industry in general, but for each carrier in particular. It was time to ship the grain. How it would turn into sandwiches would be solved by others – everything would be done on the internet, and it would all grow at 100% or more annually. What the applications were, whether and how to nurture them was immaterial, or at least someone else’s downstream concern “further along the value chain.”
Ironically, the longer-term assumptions about overall data demand weren’t that far off, but carriers found themselves in a three-part squeeze:
- In the short-term, which is what matters with debt-laden balance sheets, they collectively built too many freight cars for the underlying grain demand
- The railways terminated far from the sandwich factories
- As things got worse, they bet on a vague hope that some new class of broadband application from the e-commerce frenzy would arrive just in time to reshape the grain demand curve above its already generous slope, and somehow spike it upwards. Instead of all meals being progressively “super-sized”, the carriers needed a “killer sandwich”
For a brief moment some thought a killer sandwich might be on the horizon. Broadband access penetration was sluggish (Moron’s Law!), business-to-business e-commerce traffic had yet to materialize, but an odd little application called “Napster” appeared. Napster, the online music service ultimately shut down after copyright litigation, briefly promised to be an application that might make the Moore’s Law of networks become immediately self-sustaining.
It would be the consumer-equivalent to data networks of what Lotus 1-2-3 had been to early PCs – the breakout application that catalyzed the industry and drove consumption of underlying infrastructure through the roof, setting PCs safely on their way to corporate desktops.
More important than Napster itself was the hope of its example stimulating a whole new family of data-intensive, made-for-broadband, networked applications just when the telecom industry needed it most. But Napster was a false start whose grass roots appeal had skipped over certain commercial and legal niceties (like who owned the music), only to be squashed in court by “incumbent” music publishers. As a source of broadband demand, Napster’s importance was greatly exaggerated, and betting on a “killer sandwich” to stimulate wholesale grain shipments was a false hope. What was needed was structural changes in aligning wholesale network capacity with retail distribution networks and their demand.
With near-optical speed, backbone carriers became price-based commodity suppliers…
The carriers’ goal of reaching the volumes and sustaining the prices required to achieve profitability receded further over the horizon, dangerously beyond the all-important milestone labeled “fully-funded business plan.” With near-optical speed, backbone carriers became price-based commodity suppliers struggling to bridge last-mile bottlenecks to lucrative customers and reschedule their crushing debt loads. Most have gone bankrupt.
Through prescience or luck, Qwest Communications may have glimpsed the on-coming train early. Qwest won a bruising battle with Global Crossing to acquire USWest’s cash-rich balance sheet and stable RBOC cash flows, thus postponing the day of reckoning, but seeing its debt downgraded to junk and its KPNQwest partnership go bankrupt. Revenue recognition issues surfaced in an SEC investigation, as did suspicions of Worldcom-like accounting “errors” and the spectre of potential bankruptcy.
Worldcom, long the industry darling and a well-established player, found itself simultaneously squeezed by long-distance price erosion in its MCI unit and the same price pressures as its peers in data communications. By the second quarter of 2002, Worldcom had been stricken from the S&P 500 index, confessed to fraud, seen much of its $27 billion in debt downgraded to junk, and faced imminent potential bankruptcy. Level(3) Communications began using scarce cash to make financially-driven, accretive acquisitions of software and systems integration companies as a means to prop up revenues and reduce operating losses
Plenty of Blame to Go Around
CLECs were a legal invention and their demise can certainly be attributed in part to weaknesses in the 1996 Telecommunications Act, and subsequent failures in its enforcement. But backbone carriers were unregulated and most met an identical fate. Excess capital funding an unsustainable number of players set up unsustainable competitive dynamics at the same time that the presumed exits became closed. But sub-sectors that were not highly capitalized – enhanced services, for example – have also withered, though typically without the defining event of bankruptcy.
Regulatory uncertainty and capital markets cannot together explain the suddenness and depth of the telecom crash, nor is it excused by expedient and apparently illegal financial practices by some players. The additional key ingredient was a weakness in business models that relied too much on new technology to provide me-too services using me-too operational processes.
It is ironic that a sector whose product is wholly ‘virtual’ should have drawn so little from the eBusiness transformation under way in many other sectors. Restoring growth will require not only innovations in network technology – we have had plenty of them already – but creative re-thinking of the products and services offered to customers and the manner in which these services are conceived, deployed and distributed. ■
Download this article:: TransformationInterrupted.pdf [4.1mb]
11 July 02
