Copyright © 2000, CIMI Corporation, Voorhees, NJ, USA. All rights reserved. This issue may be printed or stored in Web format for personal, non-commercial, use, provided that the entire issue including this copyright notice is reproduced and included. Portions of this issue may be quoted, printed, or stored providing that the subject portion is annotated with the issue identification above, and is attributed to the copyrighted material of CIMI Corporation. Other publication, reproduction, electronic storage or retrieval of this material, in whole or in part, without the express written consent of CIMI Corporation, is prohibited.
This section, offered four to five times per year, provides an overview of the opportunities, issues, and development of key telecommunications service and product market sectors. We're sorry, but it's never available online.
It’s hard to find anyone who hasn’t been following the recent slump in the stock market, and in technology issues in particular. Many are wondering just what this might mean to the future of our industry, but many are also interested in the near-term impacts. Will capital dry up for the high-tech startups? If so, what will happen to the golden age of teen-aged entrepreneurs?
Well, on the one hand, we can’t claim to have all the answers. On the other hand, we have been indicating in our Annual Technology Fore-casts for the past two Decembers that the valuations on high-tech firms were becoming dangerously high. A crash of some proportion was inevitable, and we’re having it (or have had it, one could hope). Now what?
To answer that question, we have to start off with the “why” of all this.
Stock prices are set not by absolute value metrics, but by the laws of supply and demand. When a stock goes up, it doesn’t mean that the underlying company is good, but that there are more people buying its stock than are willing to sell. That, in turn, could be an indication of public perception of value, but “public perception of value” isn’t always a good indicator of real value.
Historically, buyers have been somewhat constrained in bidding up stock prices by concerns over the relationship between price and earn-ings—the PE multiple. In the good old days, companies were usually priced at about 15 times the earnings per share.
High-growth companies like technology companies quickly put a crimp in the old PE principles, though. If a company’s earnings were to be growing at a very fast pace, it would be logical that investors would anticipate that growth—buying the stock before future earnings pushed it up and thus gaining more value. This resulted in a growing tendency to “price future earnings”, and created a spread between the “growth” stocks (that were priced at perhaps 50 times earnings) and the more traditional ones.
In the late 1990s, as a long economic boom created an appetite for high returns, PE multiples in general began to rise as more people en-tered the stock market through indirect channels like the purchase of mutual funds. This influx of new money also increased the spread between the high-flying sectors (particularly the tech stocks) and the rest of the market. PE multiples soared to 100, 200, or more.
Unfortunately, technology isn’t something that the general public or the investment community understands in detail. Any claims, news, advertising, articles, or manipulations that were credible to the masses could be used to drive up expectations on future earnings, be-cause…well…they were in the future and couldn’t be disproved quickly. We described this situation as “defining success as the absence of provable failure”.
The premium in PE assigned to tech stocks—particularly to the networking sector—grew to the point where it created irresistible desires in the hearts of VCs. Long accustomed to demanding a short-term “exit strategy” for companies who might not make it in the long term, VCs found the market climate was favorable to creating an exit strategy as the only strategy. Companies were funded to be bought—the VC term is “flipped”. The term “burgers” became part of the Wall Street lexicon. The result of this was a series of market-hype-driven stock waves like the so-called “dot-coms”, sectors that left investors overall holding the bag.
But hype-driven sectors die off because even the future eventually becomes the present. As some of these “never-real” junk issues began to falter, it made Wall Street more suspicious of tech-sector stuff in general. After all, one technology looked about as good as another to these guys, and our industry ballyhooed good and bad stuff indiscriminately. As a result, we started to see expectations of some realization of earnings potential develop. It started in Wall Street demands for real earnings prior to an Initial Public Offering—something the dot-coms rarely had.
Ever clever, the VCs solved the earnings problem by financing buyers as well as sellers. The CLECs were a gold mine concept; buy equip-ment from the VC-funded equipment players and validate them, and have the purchase demonstrate the aggressive deployment by the CLECs at the same time! In effect, we were using one pyramid scheme to fund and validate another.
This unhappy trend of “laundering” venture money to make it appear like earnings is another of the millstones around the market’s neck, but there’s more. Other larger equipment vendors, trying to counter the insider buying trends that seemed to favor startups (because they did), began to hold the financing on equipment sales to the CLECs. Debt of this sort approached a half-billion dollars by the third quarter of this year.
Many measures were taken to accelerate revenues, keeping in mind that “revenue” was only what was owed to a company under the pre-vailing rules of accrual accounting. Thus, many firms extended credit to companies unlikely to be able to pay, taking the risk that they’d either collect or replace the loss with greater earnings in the future. Many offered buyers founder shares, so that “buyers” were gambling that the stock appreciation their equipment vendor might enjoy after going public would more than pay for the equipment they “bought”.
Justifying all of this was a truly forward-looking piece of nonsense, the concept of capitalizing customer value. DSL customers were said to be worth fifteen, or twenty, or forty thousand dollars. How that “worth” was justified was another matter, but nobody was watchdogging this process and nobody cared. Those who did try to look inside were treated to the same story that the dot-coms gave—advertising revenue was the golden key to the future success of everybody.
But now the problem arises. The dot-com failures began to make people think about the wisdom of capitalizing customer advertising value. That, in March, began to put pressure on the CLECs’ credibility in the capital markets. Raising new money became tough.
Capital, unlike earnings, has finite limits, and CLECs spent theirs through the summer. As the capital reserves of the CLECs fell, their credibility fell further, and the sector fell into chaos. Compounding the problem was the tightening of collection assumptions through the industry, with auditors telling many companies to write off their uncollectable debt—with an accompanying hit on their bottom line. The recent reports by relative CLEC giant Covad that it has a serious accounts receivable problem with many of its ISP partners is only the lat-est round of this problem.
With the CLEC problems and the prior dot-com flop, the market seemed to finally reject the capital-value-of-customer concept of judging a company’s assets. It’s now extended to requiring that earnings reports in the near term validate the heady assumptions of future earnings growth. That’s what is pulling down the market in the near term, because those assumptions were and are patently stupid and unattain-able, so near-term progress toward them is hardly credible.
The bad news is that things will probably be a bit testy for a while here. The tech wreck is threatening to undermine the economic expan-sion because of the wealth loss, and that shows up in declining prices of non-tech stocks. Technology companies themselves have been big spenders in some sectors, and those are starting to feel some pinch as well. Overall, we will probably test the 10,000 level on the Dow and the 3,000 level on the NASDAQ.
Long-term recovery of stock values isn’t a given, either. Much of our problem here is that stocks are over-valued by traditional metrics. That’s what has caused this whole mess. The question, then, is how the market will deal with this loss of valuation. There are three clear possibilities:
- The market could accept the loss of market cap overall, meaning that the total value of stocks could take a long-term dip. That would mean that investors would have to pull money out of stocks and accept the returns and risks of another sector, like real estate or CDs. We don’t think this is a high-probability event.
- The market could accept a new paradigm on PE valuation, that would allow for greater price growth without accompanying growth in earnings. That would have the effect of reducing pressures on companies to produce earnings in the near term, re-moving the stimulus for market instability when they don’t produce.
- The marketplace could respond to the demand for high-quality investment opportunity by producing something other than ba-nal, stupid, companies to invest in. There are many good ideas today that could justify megabucks of investment, but they’re not easy, short-term good ideas. VCs and investors need to get real for this to happen.
As this issue goes to press, the markets are still in limbo. Threats are emerging in the optical area, fueled by a revenue shortfall by Nortel. We’ve always said the optical sector was exaggerated, and it may well be that this sector will now take a hit, though we expect it will re-main overvalued well into next year.
In the long term, it looks like a bake-off between the second and third choices, here. We think that what will finally happen is most likely to be a combination of the second two possibilities, with the deciding factor being the nature of a particular company’s business.
Decent technology companies like Lucent, Dell, Intel, Nortel, and so forth will recover some of their stock prices fairly quickly as investors adjust themselves to accepting a higher static PE multiple on their stocks. In effect, the market will be saying that high PE is bad only if a lot of people start thinking it’s bad, and as more people accept higher multiples, the risk of a stock having such a multiple getting dumped on will fall.
The general rule in this sector will be simple. A tech stock battered below a PE multiple of 40 or 50 is almost certain to recover to roughly that range. One with a multiple between 50 and 100 may experience some upside. Those with multiples above 100 will have to show the Street something.
The PE tolerance Wall Street will afford these market giants won’t solve the problem of where to put investors’ money, though. What will be needed are some other high-growth-opportunity new players. In the short term, we will probably see a lot of this cash flee to other sec-tors, including biotech. In the longer term, we’ll see a new round of companies funded to absorb the excess influx of funds.
That’s where the future of this industry will be decided. We’ve done the worst possible thing—substituted nonsense for legitimate value when we had the value for the asking. Where was the legitimate value? New-age fiber remotes, which will account for over $30 billion in spending. No startups have stepped forward with a correct product here. Service points of presence, which will generate over $100 billion in spending. Here, there are some players emerging. Content networking, which is a market worth over $200 billion in services, and for which CPE is only now appearing, and service management, the glue that would bind all of this into a reasonable network.
The problem is that revenues from these new areas won’t just fall into the laps of companies, and VCs have lost interest in funding some-thing that has to be deferred in terms of bail-out or realization. It wasn’t always like that, though, and so it clearly can be the old and good way again—if we as an industry want that badly enough.
If we can now find the strength to put forward real companies with real value, the investing community will support us. If not….
Well, there’s always time to get a new degree in bioscience.
Our general disdain for VoIP is well-known to our readers, and we want to assure all of you that we haven’t caught IP religion here. What we do want to do is explore a very different interpretation of a so-called “VoIP” service, and ask some questions about its viability and im-pact.
Traditional VoIP is drawn from the same well of imagination as convergence. The theory is that voice networks would be re-engineered to run over something like the Internet. Given the fact that the Internet earns about a hundredth of the revenue of voice, the economic impact of such a move would be obvious, technology considerations aside.
Suppose we change the rules, though? We can’t redefine economics, but we could sort of redefine the concept of VoIP—and its mission. If we do that j…u…s…t right, we might create something very interesting. Different, not convergence, but interesting.
Voice over IP has meant calling using phones like those of today, but over IP network infrastructure. “Voice over IP” is what VoIP has al-ways meant.
There’s another interpretation of the acronym, though. What we might call the “voice of IP”, on the other hand, is voice that originates on an IP appliance—a PC. In most cases, this kind of voice would be related to multimedia collaborative exchanges that might—but wouldn’t necessarily—involve video.
Suppose this kind of voice, which is virtually nonexistent today, were to become at least a minor market? How would we support the voice of IP off the site or LAN where it originates?
One answer would be via POTS. Obviously, we could buy a gadget that would link a data network containing some voice sources with a PSTN line. This would give us free interoperability with traditional voice users, and would also leverage the current infrastructure. It doesn’t require capitalization, and given the current small market, that might be a commanding benefit. In fact, everything seems in favor of exactly this approach.
Well, not quite everything.
Suppose you’re an incumbent LEC. You are required by the Telecom Act to wholesale network elements and services related to your voice business because you were a major player in that market during the reference period defined in the Telecom Act. You can’t very well argue that all your voice revenues from the early ‘90s were not really derived from the PSTN services at all, but rather were the gift of a grateful public to support the good public works you do. In fact, that’s about the only argument the ILECs didn’t try in court.
But could you perhaps argue that you were not an incumbent in the “voice of IP”? There was no computer voice service to speak of this year. There was less in 1995 and earlier. How could an RBOC be an “incumbent” in a non-offered service? One could perhaps—just perhaps—make the argument that they could not have been. That would mean that the voice of IP service would not be explicitly regulated under the Telecom Act. An RBOC would be free to offer that service from an unregulated subsidiary without any concern about wholesal-ing it or its elements.
So what, you may ask? Well, keep in mind that the single-minded fixation of the RBOCs since the Act was signed has been to minimize wholesale access asset risk and service element wholesale risk. It’s very possible that the RBOCs would see a regulatory guarantee of im-munity from wholesaling in our voice of IP service as the scales-tipper. They might offer such a service.
One good way of dodging wholesale risk is offering something nobody would possibly want to buy. Such a decision is immune even to the regulatory uncertainties of Washington. Dodge opportunity and you dodge competition. Is that perhaps what’s going to happen should “voice of IP” service become a reality?
If we’re going to have a voice of IP market, we have to give voice to IP in the first place, or more accurately give voice to the LAN. If there is to be any regulatory claims of non-incumbency, we have to clearly separate our voice of IP from legacy black-phone voice. In short, we have to develop a market in which computers are used as primary voice instruments instead of phones.
Harkening back to one of our prior issues, we’ve noted that the market opportunity space for LAN-based telephony is about 800,000 sites in total, with somewhere between a third and a seventh of that becoming accessible in a given year, depending on economic conditions and various market timing issues. These sites represent about 2 million workers, and generate on the order of $6 billion per year in access and service revenues for voice. That, in itself, isn’t a particularly big pie.
What may make this group more interesting is its potential future. That population has nearly ten times the potential for developing video revenue as the general business population. It’s potential data revenues are five times those of the average business of the same size. Thus, this particular market segment might well represent over $40 billion per year in incremental revenue opportunity—properly devel-oped.
The “properly developed” part may prove a bit difficult, though. One seemingly critical problem is that the LAN voice segment has been interested in a voice instrument that’s a composite of the computer and the desktop phone. That marriage may be made in heaven from the perspective of the user, but it’s regulatory hell to the RBOC because the use of legacy instruments would significantly weaken (if not com-pletely kill) any claim that this was a market apart from PSTN voice.
Then there’s the fact that this space has a history of rejection. LAN-based voice products have been available for a decade, and their pene-tration into the target market has been negligible. In fact, the target market size is actually shrinking slightly at the moment, reflecting a general trend among high-tech LAN users to reinvest in standard key telephone or PBX technology.
Another factor limiting the potential of this market is the loss of account control of the players most likely (in the past, at least) to influence these buyers toward a LAN voice decision. LAN switches have largely commoditized, and the result is that most of these 800,000 sites don’t have real sales contact with anyone offering LAN equipment, or promoting the use of the LAN for anything other than data.
In fact, we could say that the biggest problem that LAN voice has faced is the lack of credible sponsorship. When we’ve surveyed buyers in this sector, we’ve learned that they have discounted LAN voice strategies out of hand, often rejecting sales efforts in that space with dis-dain. Why? Because the sellers were perceived as sleazy techno-hacks—as they often were.
Could anything turn this around? Yes, said our research. A LAN voice product that was targeted at working in the desktop computer for as much call/message management as possible, that was software upgradeable to add new features, and that had an open architecture that supported third-party development.
Hey, we can get that now. Clearly there’s more to it. The “more” is that all of this would have to be presented by a credible source and in a credible package of product, installation, and support. According to sellers in the space, this is hard to put together because the selling cy-cle tends to be long, the conditions on the LAN vary significantly and this variation impacts installation procedures and cost, and the buyer tends to kick any LAN problem encountered after installation back to the LAN voice player, whether voice is involved or not.
Could an RBOC solve these problems? We think that may be possible.
The most credible player in the voice marketplace, according to buyers, is the local exchange carrier. The LEC rarely slams customers, re-sponds (in most cases) to problems promptly, has relatively few problems to respond to, and is perceived as a sound, stable, well-run busi-ness. The LEC, through Centrex, is also a long-term provider of what could be called “virtual premises voice” services.
Because the LEC has a continued relationship with the customer that can earn ongoing revenue, it can justify a continued account relation-ship more easily. LECs would also have multiple revenue sources from our voice of IP application—the premises voice revenue and any on-going voice service revenue as well.
RBOCs also have an opportunity to use premises-based voice of IP applications as the basis for both a new generation of Centrex service (one that, bypassing CO switching, would be immune from wholesaling if the regulatory exemption were to be admitted) and a unified mes-saging service.
On the downside, the whole thing hinges on two basic questions; is there legitimate argument that the voice of IP is a new voice application, and can the RBOCs address that application given the need to thread carefully through the regulatory minefield?
We’re of two minds on the question of whether the voice of IP application is really a different kind of voice. On the one hand, it’s tempting to argue that human voice is the foundation of both applications and thus a regulatory-enabling link between them. On the other hand, it’s true that cellular/PCS voice is also based on human voice, and is not “legacy’ voice in a regulatory sense.
The whole thing seems to hinge on the execute. We think that the RBOCs would probably have to establish a legitimate claim that the voice of IP application existed on the LAN separate from the wire-line execute of the WAN connection, and separate from the instruments currently used to support PSTN voice. To do that, we’d have to socialize the concept of making calls using a computer alone, which isn’t as easy to do as it sounds.
The net-net of all of this may be that the success of the RBOCs here will depend on just who they go to for the solution to the problem of equipment and software technology. Voice-of-IP solutions require prem hardware, prem software, a CO architecture, and a service archi-tecture overall. The RBOCs, never great innovators, are likely to run out and try to find a TAPI/TSAPI VAR, developer, or other player of more-or-less traditional bend, and this may involve them in things like IP phones and legacy-to-LAN conversions. That would start to look to regulators like collusion to translate regulated activity into an unregulated space, which is a sure way to end up being regulated through and through.
On the other hand, the problem or opportunity of the voice of IP application is one of the many casualties of the convergence and voice-over-IP hysteria. We didn’t launch anybody with the specific objective of solving the problems of this space. Can we expect that somebody solved them accidentally, or saw through the nonsense enough to bravely go right when everyone else was turning left (no political references in-tended, Internet communists)?
The interesting thing is that we’ve found persistent evidence that the RBOCs are looking at this very issue, and in general the RBOCs have been extraordinarily careful students of regulatory mindset. We believe that the FCC tea leaves are showing a predisposition to accept at least some arguments that the voice of IP is a real, and really separate, market.
Like all new market crazes, this one (if it emerges as we expect it to) will be hyped beyond recognition. Look for heady forecasts of tens of billions of dollars per year in “voice of IP” service revenues. It’ll be nonsense, of course. The real market for this type of service is probably less than $10 million in the near term, and probably won’t reach a billion per year until around 2005. But truth doesn’t sell stock in equip-ment companies (see the first section of this issue) and we can expect relatively little of it will emerge here.
It’s also far from clear whether this voice of IP market will represent a long-haul IP service market at all. Certainly there will be an oppor-tunity to add features to traditional voice. Certainly there will be some regulatory pressure on RBOCs to use something other than conven-tional voice technology—new stuff would strengthen their claim of regulatory immunity. But will this be based on IP? We don’t think so.
But there’s another regulatory truth yet to be weighed in. RBOCs were clearly not incumbents in the long-distance market. A separate RBOC regional or national toll subsidiary would be exempt from wholesaling as well. So far, there’s no indication that the RBOCs would build IP voice infrastructure for these applications (Verizon, for example, has announced it would use ATM), but it is possible that the toll voice strategy (ATM?) and the voice-of-IP application might develop based on common technology. Yet another mystery added to the equa-tion.
One thing’s for sure; RBOC profits are good in a market where service providers in general are struggling big-time. Historically, the RBOCs (and in particular SBC) have been uncannily savvy in timing their entry into new market spaces to obtain benefits from the new without sacrificing the old. New-gen voice infrastructure, for such cautious players, is most likely to be credible when new revenue oppor-tunities for data justify build-out of more data-centric transport and service resources. With RBOC data services enjoying rapid growth, that could come later next year. With it might come the first visible signs of whether the RBOCs will attempt to develop our “voice of IP” application as an independent business.
If the road to hell is paved with good intentions, the road to future networks may be paved with content. Content networks—which we’ve called “service extranets” in our material—offer a large potential revenue stream in the near term. They dodge traditional incumbencies in the service provider space, so they represent at least a potential market for new players. They are traffic hogs so optical players love them. They are enabled by residential DSL so the FCC loves them. Gee, how can you lose?
Easily, of course. The challenge of content networking is how to approach the market, and there’s a good chance that the typical content network players won’t do it right.
A content network is a network (generally, but not always an extranet) that supports the delivery of an information product rather than the communications experience. Consumers of content buy the content, and the content provider will normally pay the fee to have it delivered via a network.
From this basic definition, it should be clear that content is the key element to content networking. In fact, it’s the launch point of defin-ing content networking strategies as well.
Sorry, Internet readers; the rest of this section is for our subscribers only!
Netwatcher is a copyrighted information product of CIMI Corporation. Subscription is on a controlled circulation basis only. Subscription information may be obtained by clicking here.