It would be all too easy to say that after six short years, the submarine cable bubble is back.

After all, with much existing capacity still unlit in both the North and South Pacific, what is one to make of actual and proposed builds from the likes of Verizon, Tata Communications, Reliance, Pacnet, the Telekom Malaysia-helmed Asia America Gateway, Pipe Networks and Telstra? Plus plans for massive upgrades on existing cables such as Australia-Japan Cable and Pacific Crossing?

Milling around the conference halls at last week’s Pacific Telecommunications Council meeting in Honolulu it’s fair to say that no one was in denial about the fact the 2001 bubble was based on similar causative factors to those that exist now, nor the fact that there was trepidation about the industry marching into another one. One common sentiment was that it is unlikely that every single announced upgrade or build will go ahead. Another was that things were different now, and financing was considerably harder to get, acting as a brake on inferior plans.

That certainly seems to apply in the case of perhaps the most left-field of all the new builds: Pipe’s Australia-Guam cable. Pipe’s cable has undergone a very public and prolonged gestation period as the company sought to write enough business to cover the cable cost.

With five announced customers and several more significant unnamed ones, Pipe didn’t confirm the cable until it could point to enough business. Somewhat amazingly, brokers have now upgraded their targets for Pipe’s ASX listed stock, apparently endowing the planned cable with zero risk and almost unlimited upside. Pipe passed its test as it were.

Another factor promoting the economics of new builds has gone less remarked ... the fact that vendors such as NEC, Huawei, Alcatel-Lucent and Xtera have dramatically improved the cost efficiencies and productivities of their technologies. I helped moderate a session at PTC that featured representatives from several of these vendors and it’s clear from their testimonies that the price-productivity trends that have so infused and fuelled other parts of the telecom business such as HSPA and DSLAM ports are also now manifesting themselves in wet long haul.

This in turn has promoted the viability of builds on second and third-tier routes: for example, Hawaii to French Polynesia, Singapore-Batam-Jakarta and Sydney-Noumea. It’s quite interesting that these days you can build a multi-thousand km cable with change from $200m or even $100m, when not so long ago you were talking change from $1 billion.

HOSTILE REACTION: Of course that’s not to say that everything is rosy and you can just lay it, as it were, and the business will come. I was a little surprised at the hostile reaction I received in a panel session when I suggested that some of the builds on thinner routes were vanity builds and were based more on national prestige and political largesse than actual business cases.  A fair percentage of the industry still seems to attach evangelical goodness to the very notion of undersea builds—forgetting the fact that you can’t fill those pipes if you don’t have favourable regulatory and investment reforms in the access network at the other end as well as basic pre-conditions such as mass PC literacy and affordable services.

From a political point of view, broadband is a fashion—a problem if you don’t embrace it, and a opportunity if you do, and nothing can take away from the fact that this year’s pet project can become next year’s white elephant.

But, of course, in other cases, new builds on thin routes will actually liberate and unleash repressed market forces. I was lucky or unlucky enough to share a corridor in my Honolulu hotel with Matrix Networks—the Indonesian equivalent of Level 3 or Pipe Networks—who are building a new cable connecting various points of Indonesia to the regional hub of Singapore. Judging by the constant flow of industry types to their suite at all hours last week there is incredible demand for alternative ways to land traffic in that country. Ditto for the fellows from Pipe.

However if there was one topic that was taboo at PTC it was the real “elephant in the room”: Google. As CommsDay revealed last September, Google is developing its Unity project which aims to partner with new builds in the Pacific so it can access connectivity at build prices and peer with rapidly rising Asian and Australian markets behind their own international gateways.

Google is incredibly secretive and coy about this—and demands no less of its discussion partners. One executive from a potential partner told me he was threatened with the sack for mentioning the Google plan in public! Hey, do no evil, Google!

But despite the lack of discussion, the move of Google and others of its ilk—the Yahoos, the MSNs and the Skypes of this world—into the connectivity ownership game is at least as significant a trend as what might be a mere momentary capacity glut. By some estimates, Google—mainly as a result of YouTube—now generates about quarter of the world’s Internet traffic. Yahoo and MSN must also account for meaningful shares, as must other significant servers of video content such as the big US entertainment corporations like Fox. Theoretically they don’t need to buy capacity at all—they can let the world’s ISPs come to them, but in practice they need to reach out and improve the technical and cost equations for their sources of eyeballs as much as possible in order to achieve growth rates that back their heady valuations. This means peering with as many ISPs at the best point of connection for those ISPs and minimising barriers to consumption of their offerings.

If one content player such as Google gets into this game and secures a cost break, the rest will be compelled to catch up. This could have a significant effect on industry margins, if, in effect, the wholesale and retail margins from the biggest generators of traffic disappear as they become owners and operators of capacity in their own right. It could mean leaner times for telcos and perhaps a perpetual boom-bust cycle for the undersea cable sector—very much in the interests of the content servers who want to reach as many eyeballs for as cheap as possible and who, after all, monetise their businesses in completely different ways.

It’d be great to have an industry discussion on this but it won’t happen soon given that nearly everybody is in courting mode and wants Google to buy a wavelength or a fibre pair on their particular project. One thing I can confirm: Google executives were at PTC! It’s just that every time I got close to them in the bar they managed to get away!

by Grahame Lynch 

add to delicious | add to digg | search technorati