When Telstra’s Phil Burgess complains that the G9 proposal would effectively give Singapore carte blanche control over Australia’s telecom industry he’s only telling the half of it. A little reported aspect to the G9’s FTTN plan – lodged with the Australian Competition and Consumer Commission last week – would also see a wholesale import of a Singapore policy designed to impose government intervention in network design, costing and ownership.

You see G9 isn’t explicitly asking for the government to accept its specific FTTN proposal. Deep in the fine print of its submissions, the group is calling for a licensing regime – which would grant at least five years exclusivity in return for an open access monopoly – and an open tender to find the best proposal to build a FTTN network. This is the same as a Singapore policy currently underway that is mounting a competitive tender for a gigabit backbone and 100Mbps access network to be provided to all comers on an open access basis.

In the words of the G9 submission, “the Singapore Government has created a competitive environment for the construction of a national broadband network by calling for collaborative interest from the incumbent, and other potential infrastructure providers, for the construction and operation of the network. This project began, in part, from perceived delays by the incumbent in investing in high speed broadband infrastructure and also a desire on the part of the Singapore Government to lead other countries in investing in broadband infrastructure.”

“To achieve this, the Singapore Government has created a competitive tender process and structure for the separation of network construction, operation and management of a national broadband network. This has created a credible threat of entry in the customer access network that will provide very strong incentives for the incumbent to offer reasonable terms of access and to proceed with broadband infrastructure investments.”

This is fascinating on more than one level. Firstly, G9 is clearly led by SingTel Optus, and this, in effect, represents the first public view of what SingTel thinks about this specific Singapore policy (the idea that SingTel could engage in the same anti-government rhetoric in its home market as is deployed by Telstra in Australia is, obviously, preposterous).

Secondly, the actual Singaporean process has inspired a wide variety of bids from an eclectic range of vendor and carrier consortia – some which may well come up with technical and management solutions simply not imagined by the small coterie of Singaporean access seekers. G9 isn’t scared to see its proposal trumped by something else as long as it provides better terms of access for them! Whether the winner of a hand-picked beauty contest designed to fulfill political imperatives could ever be the best economic and technical solution remains to be seen.

Thirdly, there are two broadly competing broadband architectures in Singapore – SingTel’s relatively high quality copper network and Starhub’s newish HFC network. Both are desperate to win the tender as evidenced by SingTel’s willingness to present as a partner in several competing consortia bids and Starhub’s push to deploy a new bleeding edge Vyyo overlay technology that would set world record speeds. In other words, there are high stakes at work here and winning might not be such a pleasurable experience given the terms sought.A cynic might observe that SingTel wants Telstra to cop the same medicine it is about to endure in its own market.

Definitely the actual G9 undertaking – which proposes access pricing ranging from $15 to $50 -  should really be seen primarily as an ambit claim or the first step in a negotiating process designed to ensure that someone builds the network but that their risk as access seekers is minimised.  Because that’s definitely the central theme of the FANOC proposal – minimising risk for the G9. While members would be free to help finance the proposal it’s clear that the collective view of the G9 is that private equity and the like will come along – Macquarie fund-style – and happily accept rates-of-return on capital somewhere south of 10% but effectively hand over all rights of management control and veto to the nine main customers.

SALES FORECASTS: One of the things that’s most notable about the proposal is the governance device the G9 is proposing to control pricing – effectively through mutually agreed sales forecasts between FANOC management and the G9 oversight lord, named Speedreach. Unit pricing would be determined by these sales forecasts, which G9 believes creates incentives for FANOC to pursue efficiencies and maximise the pricing and quality value of its products in order to beat the forecasts. Likewise, it takes the pain if sales forecasts aren’t met.

The mutual agreement bit is a crock of course, how is FANOC expected to reasonably challenge or critique the sales forecasts of members who act as its overlord?

Indeed there’s an assumption here that displays a stunning naivety. It suggests that retail market success is somehow mainly determined by the value proposition of wholesale inputs. It seems a bit rich for G9 members – some of whose sales are currently declining by double digits annually – to impose a discipline on FANOC which they clearly don’t live by themselves. After all these same G9 members are handing in mediocre annual revenue performances in a climate now where regulation is driving wholesale prices down (LSS prices down from $9 to $2.50 in 2 years as but one example) and the costs of most other inputs such as DSLAMs, backhaul bandwidth, mobile termination and so on are also decreasing.Would it not be fairer for G9 members who fail to deliver enough business to FANOC as per their forecasts to themselves pay some kind of penalty? Of course not! The G9 is all about avoiding risk, not taking it.

WRONG MODEL? Maybe the whole open access model itself is the wrong one. Perhaps the onus should be on using policy to encourage alternatives to the copper loop or perhaps even leverage overbuild – as with Broadband Connect in the bush or in the United States -not to standardise everybody on one net.When I explained the topology and access model of the G9 proposed network (and to be fair, the ACCC-negotiated Telstra one) to some of the cognoscenti of the global networking industry last week in China they were resoundingly negative.

The occasion was a global customer and partner meeting for Juniper and the guests of honour included some of the world’s top data networking minds – the people who wrote the MPLS standards and so on. The coffee break consensus view? The proposed Australian FTTN model was retrograde, pushing the service provider’s core too far back into the network, when the technical consensus is that the core should be pushed out to the node. Their arguments were too nuanced to do justice to here, except to say that the feeling was that the Australian FTTN approach would lead to inefficient opex costs, reduced service innovation and a lack of future proofing.

Such network sharing might seem attractive from the point of view of preserving industry structure, but perhaps that structure isn’t worthy of protection from disruption in the first place, opined one. Would not a better solution to their market challenge be for the G9 to formally merge themselves, rather than collectivising their network platform while competing on nothing more than price? At least their mono-services would benefit from scale and a reduction of corporate, sales, marketing and legal overhead.

“Network sharing is a crazy idea”, said another, no less than the VP of strategy at Juniper, Judy Beningson, while yet another – a fellow who serves in the higher echelons of the IETF, which writes Internet standards – looked at me with a pained expression and said “that won’t work”.

Hmmmm.

- Grahame Lynch

add to delicious | add to digg | search technorati