LME and ARCEP decisions regarding NGA
Since the French senate validated the LME (Loi de Modernisation de l'Economie) dispositions last week - including those related to NGANs - and ARCEP published some additional decisions regarding deployment of fiber networks, I've been wanting to write some commentary, but I failed to find the time. Fortunately, someone way worthier than me wrote an in-depth analysis of the dispositions, namely Yves Blondeel of T-Regs who I had the chance of meeting in Stavänger last April.
T-Regs is an independant regulatory consultancy, and their website, which regularly publishes analysis of regulatory decisions, can be found at http://www.t-regs.com/.
You should first read Yves' analysis here under France: FTTH law adopted, and new ARCEP decision, new access obligations. Here are my own comments on some of the aspects covered.
LME was supposed to be the big boost to FTTx penetration as France finds itself in the crazy situation of having an est. 1 million homes passed by horizontal fiber networks but less than 50k actual subscribers and probably less than 100k eligible connections, simply because operators can't seem to get access into the buildings. The dispositions voted might help, but I hardly think they will make such a big difference. The main hurdles to in-building penetration were administrative (housing associations could refuse to even put an operator's proposal to connect a building to the vote) but also competitive (housing associations were worried that they would be stuck with the operator that first deployed).
While the former issue will be solved by LME, the latter hardly is: beyond a statement of intent, none of the specifics of what will happen in the building are dictated. How does the vertical part need to be deployed, how and where will mutualisation happen, what's a reasonable price for resell of the vertical, etc. None of the questions have been answered. If anything, I'm wondering if this isn't a fake blessing, because it will allow FT's salespeople to state to housing associations that the issues are solved by law when in fact they will be able to stall competitors asking for access for a long-time still because none of the specifics have been decided upon.
As Yves clearly explains, the issue of where unbundling needs to happen has not been sorted either. Issues of interoperability also haven't been addressed. I think this is an area where ARCEP, as it has the power to do, needs to move fast if the deployment is to be harmonious. It's well known that Free and Neuf are working together at making their in-building connection points interoperable, but FT has so far refused to work together with them, and presumably Numéricable is not quite ready to upgrade their in-building wiring to fiber just yet.
Another aspect of the decision that I wanted to react to is the absence of bitstream access obligation. This is, I fear, characteristic of ARCEP's hesitancy between the competitive model it really wants. Clearly an absence of bitstream obligation suggests an infrastructure based model, with a clear risk of a very strong dominance of France Telecom, which ARCEP seems to want to avoid otherwise. Since the unbundling issue at any point in the network higher than the building is also not resolved, the current situation is only viable for alternate operators in areas where overlay makes sense economically, ie. Paris and a few select city centers in Tier 1 cities.
Ubiquitous fiber is not even in the picture yet...
