market intelligence
Undersea bandwidth prices submerge
BY STEPHEN HARDYBack in the old days—say, five years ago—the undersea bandwidth capacity market was the land of the giants. Big national carriers combined into consortia to spend big bucks to build big pipelines for their big bandwidth requirements. If there happened to be capacity above and beyond their needs, they might dole out scraps to smaller carriers at a hefty premium.
With the number of new carriers rising almost as quickly as Internet-driven bandwidth demand, submarine cable capacity has suddenly become an entrepreneurial hot bed. It seems that a month doesn't go by without the announcement of a new cable system planned to hit the water in the next two years. Capacity will soon become more readily available than ever before, which will lead to new options for a wider variety of customers and new pricing strategies for capacity providers.
New outlook
Changes in the submarine market landscape have resulted in a change in capacity pricing. "It used to be kind of straightforward under the consortium system," explains Tom Soja, president of consultancy T. Soja & Associates (Newton, MA). "You got your cost, you got the number of circuits, just divide, and that's the price."Routinely, large carriers and their cable-laying partners were the companies performing such mathematics, says Stacey Yates, an analyst at KMI Corp. (Newport, RI). The carriers would reserve the capacity they required, then sell any excess to other carriers, usually in fairly large allotments. These indefeasible rights of use (IRUs) could cost the second-tier carriers as much as twice what the initial investors paid, estimates Michael Ruddy, an analyst at Pioneer Consulting (Cambridge, MA). That was assuming capacity was available at all.
Several factors, however, have changed the pricing equation. Foremost among these is the advent of dense wavelength-division multiplexing (DWDM), which has significantly boosted the amount of capacity individual cables can carry without a proportionate rise in the cost of the entire undersea system. The price per bit therefore decreases dramatically, particularly as systems approach total capacities in the terabit range. The new economics of capacity cost have changed the pricing equation completely.
"With the newer systems that have so much capacity, the capacity becomes too cheap. So they have to set a lower limit on what they call the 'notional capacity' of the system, so that the investment for funding the system is based on the lower amount of capacity," says Soja. Thus, if a system were capable of providing a total capacity of 100 Gbits/sec, it might be priced initially at what have become market rates as if it could only provide 75 Gbits/sec. For example, the Japan-U.S. cable system has a total capacity of 640 Gbits/sec, according to Soja. When it initially goes into service next year, however, it will only carry 80 Gbits/sec of traffic, with an assurance to the initial investors that a year after its cutover date, the capacity will rise to 400 Gbits/sec. That leaves 240 Gbits/sec of unused capacity, which the owners of the system will sell off separately.
In fact, the consortium has reopened the construction and maintenance agreement (CMA) to new customers, offering them a second chance to own a piece of the cable, as opposed to scrambling for IRUs later on. Ruddy reports that the price of capacity under this new deal may be only 10% higher than that paid by the initial investors.
The arrival of DWDM has affected prices dramatically. "If you look over the last 10 years on a compounded basis, prices have decreased about 25% per year in the Atlantic market, for instance," says Soja. Yates reveals that KMI had forecasted a 39% drop in prices six months ago for capacity this year.
That said, the bottom hasn't fallen out of the undersea bandwidth market. All three analysts report that while pricing has decreased, the deal makers behind the consortium cables are still doing quite nicely, thank you. "At one point in time, when cable capacity was relatively low...the price that each carrier paid was, in effect, the market price. There was no independent market price for the capacity. With the new higher-capacity systems, the underlying costs of each circuit has come down tremendously. But the market price for circuits has not come down in a similar way," says Soja. "The market prices are still pretty high."
"The trend is just keeping in line with the erosion of costs," agrees Ruddy. "The prices that you see on these cable systems going into service in 1999 and 2000 still represent a fairly healthy margin over their actual costs."
New kid in town
The arrival of competition, first in the Atlantic and now in the Pacific, has also upset the pricing apple cart. First the Gemini cable system between New York and London, then Atlantic Crossing 1 from Global Crossing offered emerging carriers an avenue to undersea bandwidth that didn't run through the high rolling consortia.Interestingly, companies such as Global Crossing have not focused on price as the key differentiator in establishing their market niche. "Traditionally, the Global Crossing prices have been a little bit higher, because they, first of all, allow you to purchase smaller increments of capacity. You can just go to them and buy a single STM-1, whereas in the consortium systems, the larger carriers were purchasing a lot more capacity," explains Ruddy. "But Global Crossing really marketed itself as a city-to-city operator, so that you would be buying capacity directly from 60 Hudson Street—the international switch in New York—all the way to the international switch in London, TeleHouse. And you wouldn't have to worry about paying for the backhaul."
Global Crossing gave the new emerging carriers services they required, but that they couldn't get from the consortium cables. "With consortium systems, there are very few— as a matter of fact, I can't think of any consortium systems—that actually include backhaul. Because the consortium systems are constructed pretty much for the initial parties, the really big guys, the AT&Ts, the Sprints, the MCI WorldComs, British Telecoms, and these are guys who already have interconnection and backhaul agreements in place. So traditionally, the consortia have not offered any backhaul or city-to-city packages," Ruddy says.
Thus, independent carrier's carriers such as Global Crossing can actually charge a price premium for additional services. As the number of emerging carriers looking for undersea capacity increases, the consortium cable groups have taken notice. "If Global Crossing hadn't come along and offered a better alternative for the smaller carriers, then the consortia probably would have gone their merry way," says Ruddy. "I think the new packages that you're seeing from the consortium systems and the fact that they're reopening the CMA on Japan-U.S. and the fact that they're offering IRUs on TAT-14—this is in direct response to Global Crossing, and the consortia are trying to capture the market that Global Crossing is going after."
Another influence of the advent of Global Crossing is the notion of varied prices for a menu of options. In particular, new cable systems will offer different prices tied to when the customer requires the capacity. The longer the carrier can wait, the lower the price it will pay. As an example, the table shows the current options available from the Japan-U.S. consortium.
"So it's not just strictly a one-for-one cost divided by capacity or even cost divided by notional capacity. The consortium systems have come around to this notion of more market-based pricing," Soja explains. "So that those that need the greater capacity sooner can bulk up, and they pay a different price than if people are willing to commit to the investment now but not get the actual circuit until one or two years from now."
Wait your turn
Of course, one of the reasons that the new cable systems offer discounts for carriers willing to wait for their capacity is that the majority of these networks won't be finished until the middle of next year, at the earliest. Carriers with an immediate need for bandwidth will find the reception a bit less welcoming."Right this very day, there are some routes where you'd have a tough time getting cheap capacity," Ruddy comments. "Transpacific is probably one of them; North America-South America is one of them."
"Prices for circuits that are available today by and large from what I've seen have not come down," agrees Soja.
The capacity crunch is more acute along the Pacific route than its Atlantic counterpart, the analysts say. "Depending on what your need is, if you need them sooner than later, you may have to pay a dear price for them," Soja warns.
Capacity across the Pacific has traditionally proven more expensive and harder to acquire anyway. New technology such as DWDM generally finds its way into the Atlantic runs sooner than in those of the Pacific, Soja says. The cable runs tend to be longer and more expensive to implement, Yates adds. Thus, Soja reports, the recent China-U.S. cable system can offer STM-1 circuits for $17.3 million; while this amount represents a bargain compared to past prices for Pacific bandwidth, a similar circuit across the Atlantic might cost as low as $5 million to $7 million.
The arrival of Japan-U.S. and PC-1 promise to bring Atlantic economics to the Pacific. In keeping with the new market dynamics, prices on the Southern Cross cable between North America and Australia have shrunk significantly. For example, a 15-year IRU for a 155-Mbit/sec connection between the United States and New Zealand dropped from $37.8 million to $12.9 million. A similar connection between Hawaii and the U.S. mainland fell to $1.4 million from $6.2 million, while the same connection between Australia and New Zealand slipped from $9.9 million to $4.1 million.
Water seeks its own level
As new systems come online in South America and Africa over the next few years, pricing around the world will become increasingly similar, some predict."Increasingly, what we're seeing is a very strong price convergence, where the price of capacity in each of the transoceanic regions is falling in the same line," says Ruddy. "I think increasingly what you'll see as more and more private cable systems are built is that each of those routes—North America-Australia, North America-Asia, North America-Europe, and North America-South America—will quickly fall into the same price range."
This doesn't mean prices will be similar in the short term, however. For example, capacity in and out of South America will remain precious, Soja believes. "I think you won't see prices fall quite as fast as you did in the Atlantic and the Pacific, simply because the traffic streams are somewhat smaller," he says. "You have to have the volume for the prices to come down, no matter what the technology is. So you'll see that trail a little bit; I think prices will be a little bit higher for South America."
Overall, the distance traveled by the circuit will also become less critical to the price, Ruddy believes. "What we're moving toward now is almost the model that was proposed by Project Oxygen, which is pricing irrespective of distance," he says. "We haven't quite gotten there, but we're increasingly moving toward a pricing model [like that]."
Meanwhile, other carrier's carrier cable systems, such as FLAG-Atlantic 1 and Hybernia, and perhaps some of the consortia cable will begin to offer city-to-city links and even financing to attract emerging carriers. "The emerging carriers account for a fairly considerable percentage of the market now," offers Ruddy. "And in order to capture that market share, you really have to operate as a full-service carrier's carrier, because a lot of the emerging carriers are modeled on their marketing skills, going out and selling services, and they may not be as adept in the network operations side of things. And that's really a good market opportunity for companies like Global Crossing, to offer full network operations and even financing services to emerging carriers."
Soja cautions, however, that the impact of competition from what appears to be a plethora of cable systems, particularly across the Atlantic, remains somewhat uncertain until the cables are in the water and the owners reveal their intentions. "There's no doubt that we're going to continue to see prices come down, particularly as competition heats up. Now how soon the competition comes on and whether the nature of the competition is really to truly serve the open, competitive market for circuits or whether it's people building cables to serve their own interests remains to be seen," he says.
The private cable companies now pushing their first systems into the water will be best served by thinking big, Ruddy believes. "Once these private operators get individual cables into the water, the way to really capitalize off that is to build multiple connecting systems and build a whole network," he states. This would increase the number of circuits they can offer quadratically and hasten the market toward the global "anywhere for the same price" model first touted by Project Oxygen.
The bottom line, says Soja, is that the market will soon reach a point where price is not the deciding factor when the customer chooses where to place his order. Instead, the cable companies will focus more on meeting the customer's needs. "There's a lot of healthy innovation going on in the industry, which is really focused on having a customer orientation. What does the customer want? It's not just the owners trying to innovate just so that they can get an edge on their competition; really, it's really a response directly to the customers," he says.
This focus represents a significant shift in perspective at the supply side. "Much more is being done on a business basis, rather than the clubby type of atmosphere that you used to see before with the consortia," Soja concludes. "And I think that's good for the business as a whole."