Infrastructure based competition is a dead duck
Monday, December 15th, 2008The end state for broadband infrastructure is FttH, perhaps with large sections of the market having their own dedicated fiber ‘tail’ from their home/office out to a central aggregation point. The extremely high capacity of FttH, combined with its shrinking but still significant cost, suggests that a rational broadband policy would not include multiple competing fiber plants in the local loop.
Interestingly, countries who are leading in structural change are often also the countries with overbuilt infrastructures (telco, cable, wireless) such as the Netherlands, Sweden, Singapore and Hong Kong.
The reason for this is that, due to their heavy use of existing infrastructure, these countries have a much better understanding of the need for FttH. They quickly reach the conclusion that when HFC (cable operator plant) needs to be upgraded to FttH, an interesting economic situation will arise, whereby the content providers (cable operators) will need to decide if they want to duplicate telco fiber, or whether they will be able to use (open) networks of the telcos. In Europe some of the HFC networks are already struggling to survive. Recent regulation developments in the Netherlands are aimed at rapidly establishing an attractive mass market for FttH. This will become available for as little as €12 per month.
Obviously there will be densely populated areas that can sustain two FttH cables at current retail rates, but it defies basic logic that there will be all that many of them - especially as access to capital has become expensive (current EU and OECD data indicates that sustainable FttH requires 37%-40% subscription rate penetration).
We also don’t see overbuilt plant in electricity, gas, water, sewers, highways, HFC cable networks, etc. Telecoms engineers are indicating that most HFC networks are under-dimensioned to serve large-scale use of DOCSIS 3, and, even with DOCSIS 4, 5, 6 or 7 HFC, they will eventually be unable to compete economically with fiber.
The transition to competition based on pure broadband capacity is somewhat paradoxical for cable operators. The very high capacity of HFC networks, as compared to legacy telcos’ twisted pair copper, gave cable operators an early lead - one that they maintain - in residential broadband. Yet the basic cable business model entails selecting content for subscribers and delivering it either in packages (standard cable tiers) or individually (pay-per-view).
This business model is at odds with an end-to-end Internet in which users are essentially in sole control of the data they send and receive. So, in addition to ultimately facing the need to upgrade the pure capacity of their networks, cable operators face a challenge to their fundamental business model. If they want to be pure bandwidth providers their networks are no longer state-of-the-art. But if they want to be content providers looking into the future, there is no need for them to have a network at all - if there is adequate capacity available on an open fiber network, whether provided by legacy telcos or otherwise.
This shows that the reason we have two networks - cable and telco - is that in the past entertainment and telephony required vastly different technologies. But that is no longer the case.
Note that when we speak of FttH we are not particularly referring to shared-bandwidth passive optical network (PON) arrangements, which are obvious improvements over telco-based copper DSL solutions, but which can (depending on the particular PON architecture selected) lead to congestion problems like those faced by cable systems if consumers, for instance by telecommuting, radically increase their parallel use of broadband. This is in the national interest but it is not in the individual interest of a network operator trying to minimize capital expenditure and foreclose competitors.
What we are referring to is dedicated or quasi-dedicated fibers to each individual home or business.
In this respect the OECD, for example, is strongly in favor of dark fiber solutions rather than municipally-run networks. They are the closest thing we have to unbundling on copper which has been successful. It does mean, however, that we need point-to-point connections in order to foster competition. A good example here is Citynet in Amsterdam.
Note also that wireless broadband is a very different infrastructure, with different capabilities, and it will develop parallel to FttH. It has a totally different (fundamentally sound) business case, based on the distinctive benefits of mobility rather than raw capacity - so we don’t see this as an ‘overbuild’. Certainly, it is no alternative to, or replacement for, FttH.
That having been said, wireless broadband is going to play a key role in certain regional and rural areas where the cost of laying dozens or hundreds of miles of fiber to serve one or a few customers can be daunting. It also allows for a far more rapid rollout of midsize broadband which after time can be replaced by fiber when the business cases for such deployments are becoming clearer.
Paul Budde
This article is based on international collaboration of a dozen telecoms experts.
See also
- Global - Infrastructure - Strategies for the Digital Economy
- Global - Investing in the Communications Revolution
- Global - Broadband - FttH Overview & Statistics
- Global - Broadband - Internet & Broadband Statistical Tables by Country
- Global - Broadband - Regulating Fibre Access
- Global - Infrastructure - Structural Separation