Version non confidentielle 1 M. Dominique Vosters Président

Transcription

Version non confidentielle
M. Dominique Vosters
Président
Conseil Supérieur de l’Audiovisuel
Boulevard de l’Impératrice 13
1000 Bruxelles
Monsieur le Président,
RÉPONSE DE TELENET À LA CONSULTATION DU CSA CONCERNANT LA RÉVISION DU
RETAIL MINUS
Telenet remercie le CSA pour la consultation publique portant sur le projet de décision concernant la
révision des tarifs de gros pour les services d’accès aux réseaux câblés. Vous trouverez ci-joint la réponse
détaillée de Telenet en anglais, que nous avons résumé ci-après en français.
Telenet demande respectueusement au CSA d’adapter le projet de décision sur un certain nombre de
points. La régulation de prix proposée dans ce projet serait arbitraire, imprévisible et manifestement
disproportionnée, et ne s’accorde pas avec les objectifs du cadre réglementaire. Le projet contient
également un certain nombre d’éléments factuels qui sont incorrects ou incomplets.
Les principaux points qu’il conviendrait d’adapter sont les suivants :
1.
La méthodologie pour déterminer la valeur des services additionnels qui sont inclus gratuitement
dans les offres groupées ne tient pas compte de l’utilisation effective de ces services et doit
prendre en compte les coûts de production de l’opérateur PSM ;
2.
Le « benchmarking » est infondé et doit être revu, étant donné que la plupart des services isolés
inclus dans ce « benchmarking » ne sont pas comparables aux services additionnels de Telenet ;
3.
La formule de calcul pour les tarifs de gros des profils large bande spécifiques aux bénéficiaires
n’est pas adéquate et se base sur des données factuelles erronées ;
4.
Les modalités de la régulation des tarifs de gros manquent de prévisibilité et de sécurité juridique ;
5.
En surévaluant des services non régulés – par exemple la téléphonie fixe, le modem et les services
additionnels – et prévoyant une période transitoire pendant laquelle les coûts de promotion sont
incorrectement alloués aux seuls clients nouveaux, la régulation de prix proposée crée (ou
renforce) des distorsions de concurrence en facilitant une entrée inefficace au détriment des
investissements dans les réseaux et d’une concurrence par les infrastructures durable.
1.
Manquements méthodologiques de la régulation proposée
Le projet de décision considère que la valeur des services additionnels inclus dans les offres groupées avec
télévision et internet peut être déterminée sur la base des prix de détail des services vendus isolément. Le
projet de décision se base soit sur le tarif de détail de Telenet pour le service « standalone » ou, à défaut,
sur le tarif de détail d’autres opérateurs (« benchmarking »). L’utilisation effective des services ou les
1
coûts de production y afférents ne sont pas pris en compte. A aucun moment, le projet de décision
n’essaie de valider ou réconcilier la valeur retenue avec les coûts sous-jacents du service.
Cette méthodologie ne tient pas compte du fait que le prix de détail des services additionnels, lorsqu’ils
sont vendus isolément, est fixé en fonction de plusieurs considérations, dont les coûts, l’élasticité de prix,
et d’autres facteurs relatifs à l’offre et à la demande. Le CSA le reconnaît d’ailleurs dans le projet de
décision, mais ne semble se préoccuper que de l’hypothèse où le prix des services « standalone » serait
« anormalement bas », sans prendre en considération l’hypothèse inverse d’un tarif « anormalement
élevé »… Cette incohérence dans la motivation trahit un manque d’objectivité dans l’analyse.
Le CSA applique une méthodologie qui réduit artificiellement la valeur de détail des offres groupées (et
par conséquent aussi le tarif d’accès de gros qui en découle) en déduisant de celle-ci des montants qui sont
très largement supérieurs aux coûts afférents aux services additionnels concernés. Les graphiques cidessous comparent (i) la valorisation des services sur la base de la méthodologie du CSA avec (ii) une
valorisation prenant en compte l’usage effectif de ces services et (iii) les coûts de production de ces
services. Il en ressort entre autres que la méthodologie du régulateur résulte en une valorisation différente
des services additionnels dans les offres Whop et Whoppa, ce qui démontre déjà en-soi que l’approche
suivie conduit à des résultats arbitraires.
[confidentiel : comparaison entre les valeurs de détail retenues par le régulateur et l’usage réel des
services additionnels]
[confidentiel : comparaison entre les valeurs retenues par le régulateur et les coûts de production
des services additionnels - Whop]
[confidentiel : comparaison entre les valeurs retenues par le régulateur et les coûts de production
des services additionnels – Whoppa]
La différence entre les valeurs que le régulateur a attribuées aux services additionnels et leurs coûts
correspond en fait à la marge artificielle que le CSA accorde aux bénéficiaires de la régulation, alors que
ceux-ci, en tant qu’opérateurs efficaces, supportent en principe (ou devraient supporter) les mêmes coûts
pour la fourniture de ces services. Telenet n’a pas été désigné comme opérateur PSM pour ces services
additionnels, et il est donc tout à fait injustifié d’octroyer aux bénéficiaires de la régulation un avantage lié
à la fourniture de ces services. Il est d’ailleurs remarquable que le régulateur accorde un tel poids à ces
services, alors que l’IBPT a récemment exclu ceux-ci de son comparateur de prix meilleurtarif.be,
considérant que leur inclusion était « fort peu utile ».
L’effet pervers de la méthodologie est qu’elle décourage Telenet à enrichir ses offres d’autres services et
d’entrer ainsi en concurrence pour ces autres services. L’exemple des adresses de courrier électronique
est parlant. En ajoutant des adresses de courrier électronique à ses offres groupées, Telenet entre en
concurrence avec des fournisseurs tels que Google (Gmail) et Microsoft (Hotmail) qui mettent ce service à
disposition soi-disant « gratuitement » (alors que l’utilisateur paie en fait en « data currency » 1 ). La
méthodologie préconisée a pour effet que Telenet, parce qu’elle inclut des adresses de courrier
électronique dans son offre, doit accorder aux bénéficiaires de l’accès une ristourne sur les tarifs de gros
1
C’est à dire les opportunités commerciales qui résultent de la possibilité d’accèder à la boite mail du
consommateur.
2
qui dépasse très largement les coûts de provision du service. Autre exemple : le service Wi-Fi (« WiFree ») est un service de convergence qui permet à Telenet d’entrer en concurrence avec les services
mobiles à large bande, qui sont nettement plus chers en Belgique en comparaison avec les autres pays
européens. Pourquoi Telenet devrait-elle accorder une ristourne de € 1 aux bénéficiaires parce qu’elle
inclut ce service dans ses offres, alors que cela est à l’avantage du consommateur et que les bénéficiaires
sont parfaitement en mesure d’offrir un service similaire avec leurs propres modems ?
Une telle régulation a des effets pervers et va à rebours des objectifs du cadre réglementaire. Elle
constitue un excès de pouvoir manifeste dans le chef du régulateur. La réponse détaillée en annexe
démontre d’ailleurs que dans la pratique réglementaire en matière de ciseaux tarifaires, les composantes
d’offres groupées sont valorisées sur la base des coûts afin de permettre aux nouveaux entrants efficaces
de réaliser une marge suffisante en aval.
2.
Erreurs factuelles
Dans la réponse détaillée en annexe, Telenet a exposé pourquoi l’exercice de « benchmarking » que le
CSA a effectue est factuellement infondé. En résumé :
• Le CSA compare des services qui sont intrinsèquement différents et qui ne sont donc pas
comparables ;
• Le CSA effectue une comparaison qui est incomplète, sans motiver pourquoi certaines offres sont
prises en compte et d’autres pas ;
• Dans certains cas, le CSA déduit certains éléments deux fois.
3.
Calcul des prix pour les profils large bande spécifiques
A part le fait que les points de mesure utilisés dans la feuille de calcul comportent plusieurs erreurs,
Telenet fait remarquer que ceux-ci sont tout simplement trop peu nombreux pour donner des résultats
solides. La moindre modification des points de mesure a un impact important sur le résultat du calcul.
Telenet convient qu’il est nécessaire de trouver une méthode pour déterminer un prix adéquat pour les
profils large bande spécifiques que les bénéficiaires peuvent demander, mais s’oppose à l’utilisation d’une
formule de calcul à ce point volatile et incohérente. Telenet propose dès lors de suivre une procédure où
les opérateurs régulés élaborent eux-mêmes une proposition tarifaire qui serait ensuite soumise au
régulateur pour validation.
4.
Les modalités de la régulation proposée sont contraires au principe de sécurité juridique
La méthodologie du CSA a pour conséquence que les tarifs de gros dépendront d’éventuels changements
des tarifs de détail d’autres opérateurs, y compris des concurrents de Telenet. De surcroît, le projet de
décision prévoit que d’autres services additionnels, qui ne sont actuellement pas encore pris en compte,
pourront dans le futur également être intégrés dans le calcul, réduisant encore davantage les tarifs de gros.
Ceci constitue en-soi un facteur d’incertitude. Les tarifs de gros de Telenet dépendront ainsi du
comportement de tiers, qui auront même la possibilité d’aggraver l’impact de la régulation sur Telenet en
modifiant stratégiquement leurs propres prix de détail.
3
Par ailleurs, le projet de décision laisse au régulateur la possibilité de réduire les tarifs de gros
rétroactivement. Une telle modification rétroactive est contraire aux principes généraux de bonne
administration et est donc en tout cas à proscrire.
5.
La régulation crée des distorsions de concurrence et va à rebours des objectifs du cadre
réglementaire
La régulation imposée par la décision-cadre de 2011 ne devrait en principe que concerner les services pour
lesquels les câblo-opérateurs ont été désignés comme disposant de pouvoir de marché et d’avantages
éventuels en comparaison avec leurs concurrents. Les bénéficiaires ne doivent toutefois pas rester à l’abri
de pressions concurrentielles qui les incitent à déployer leurs activités de manière efficace. Une assistance
à l’entrée excessive pour promouvoir la concurrence par les services n’apporte que des avantages
éphémères aux consommateurs, mais a un impact négatif sur le bien-être des consommateurs à plus long
terme.
En l’occurrence, la méthodologie retenue par le régulateur pour le calcul de la valeur de détail résulte en
une assistance à l’entrée excessive (voir point 1 ci-dessus). En outre, le régulateur part de la prémisse que
les bénéficiaires souffrent de l’absence d’économies d’échelle en ce qui concerne les promotions, et leur
accorde une réduction supplémentaire sur le prix de gros (via un minus plus élevé). Pourtant, les deux
seuls opérateurs qui ont jusqu’ici formulé une demande d’accès, Proximus et Mobistar, disposent
d’avantages d’échelle importants. A certains égards, ces opérateurs ont même une position plus forte que
Telenet. L’accumulation d’adaptations méthodologiques pour avantager directement ou indirectement les
bénéficiaires, conduit selon les estimations de Telenet à une assistance à l’entrée pour Mobistar de l’ordre
de € [confidentiel] millions, alors que Mobistar a déjà bénéficié pendant 17 ans – donc encore longtemps
après son entrée sur le marché belge en 1995 – d’un régime régulatoire favorable au développement de ses
activités…
Ci-dessus, il a aussi été démontré que Telenet est grevé d’un handicap concurrentiel en ce qui concerne
des services pour lesquels elle n’est pas en position dominante, mais doit au contraire entrer en
concurrence avec des acteurs de marché beaucoup plus puissants. Ceci crée des distorsions de
concurrence qui nuisent en dernière analyse également aux intérêts des consommateurs.
En espérant vous rencontrer dans un proche avenir pour pouvoir vous fournir une explication orale de la
note détaillée ci-jointe, nous vous prions d’agréer, Monsieur le Président, l’expression de nos sentiments
respectueux.
4
Non-confidential version
15 JULY 2015
TELENET REPLY TO THE REGULATORS’ (BIPT, CSA, VRM) CONSULTATION
DOCUMENT ON THE REVISION OF RETAIL MINUS
Telenet thanks BIPT, CSA and VRM (hereafter, “the Regulator”) for the public
consultation on their respective Draft Decisions proposing to review the retail minus
methodology (the “Draft Decision”) and has set out its comments below. Based on these
comments, Telenet requests the Regulator to withdraw its proposed methodology or at least
to substantially reassess it as it is an unprecedented and isolated form of regulatory overreach
without legal justification. Consumers will not benefit from the proposed revision but will,
on the contrary, be worse-off as product innovation, network investments and structurally
viable competition will be stifled.
Telenet acknowledges that the validity of the regulation of broadcasting services
foreseen in the CRC decision of 1 July 2011 (the “Framework Decision”) has been upheld by
the Court of Appeal of Brussels and that this regulation remains in force until the revision of
the market analysis (which we understand is underway) will be concluded. Telenet’s
intention is not to undermine the effectiveness of this regulation but to ensure that its
implementation complies with the fundamental principles of access regulation and of the
retail minus methodology. Our contribution shows that the objective of market entry can be
achieved with a less intrusive wholesale pricing regulation based on efficient market
performance. The current regulation artificially facilitates market entry of one or more
specific third parties at the cost of undermining the entire retail strategy of the cable
operators. It also shows that the methodological principles of pricing regulation retained in
the Commission’s 2013 Recommendation on Cost Methodologies and Non-Discrimination
would allow the Regulator to achieve the objectives of the regulation and align itself with the
harmonized approach which the Commission is seeking to secure to achieve the objectives of
the Digital Agenda for Europe, which extend far beyond the users of electronic
communications services.
For the sake of completeness, Telenet also refers to its previous submissions
regarding the wholesale pricing regulation which is currently being appealed before the Court
of Appeal of Brussels. In an effort of concision, Telenet has not reiterated these
considerations in this submission and focuses only on the changes foreseen in the Draft
Decision.
This submission is structured as follows:
 Section I sets out certain procedural considerations.
 Section II examines the substantive legal principles which the proposed tariff
regulation is violating.
 Section III examines the factual errors and arbitrary factual assessments contained
in the Draft Decision.
 Section IV sets out the reasons why the tariff regulation will undermine the
objectives of the regulatory framework.
1
EUI-1200160743v1
15 JULY 2015
I.
PROCEDURAL CONSIDERATIONS
The European and national regulatory framework require NRAs to review the market
analyses every three years. 1 This obligation ensures that the SMP regulation reflects the
prevailing market conditions and is, more generally, proportionate to address the potential
market failures. In 2014, the Commission has adopted a revised list of markets which can be
subject to ex ante regulation (the 2014 Recommendation2) following which NRAs have to
proceed as soon as possible with a review of their market analysis.3
The Regulator’s decision to proceed with a substantial overhaul of the tariff
regulation for wholesale access on cable networks is difficult to reconcile with these
principles and the priority which should be given to a reassessment of competitive conditions.
The market analysis set out in the Framework Decision of 2011 is based on data that is four
to five years old and ignores a number of important developments which are changing the
prevailing market dynamics. The development of OTT-based services and the further
decline of analog TV affects the analysis set out in the 2011 Framework Decision. Telenet
invites the Regulator to proceed with the periodic market assessment and considers that it is
problematic to justify the steep decline in wholesale rates foreseen in the Draft Decision (as
well as its methodological underpinnings) on the basis of a Framework Decision whose
review is overdue.
Telenet further recalls that, irrespective of its objections against the initial Framework
Decision, it complies with the requirements contained therein. It implemented operational
access to the cable for Mobistar, in line with its obligations pursuant to the Framework
Decision, and within the strict time limits imposed by that Decision. In addition, Telenet has
made some amendments to its original Reference Offer following some suggestions from
BIPT and/or Mobistar, in order to further improve render the wholesale access conditions.
Telenet would therefore like to emphasize that effective access regulation should not only be
approached from the angle of ensuring low wholesale prices, but rather as a combination of
reasonable wholesale prices with an effective and efficient implementation of the operational
access.
The Draft Decision indicates that the final decision is due to enter into force 1 month
from its publication. This implementation period is far too short. For Telenet, 4 months is the
minimum time required to implement the pricing changes in its IT systems, as this is
dependent on IT release schedules. This should also be considered for any subsequent
pricing change (and retroactive reductions should be excluded as set out in more detail in
Section II.B below).
1
Directive 2009/140/EC of the European Parliament and of the Council of 25 November 2009, Article
16 § 6, a. Article 40/11 of the Law of 30 March 1995 on Broadcasting Services for the Brussels Capital
Region.
2
Commission Recommendation of 9 October 2014 on relevant product and service markets within the
electronic communications sector susceptible to ex ante regulation in accordance with Directive
2002/21/EC of the European Parliament and of the Council on a common regulatory framework for
electronic communications networks and services (“2014 Recommendation”).
3
Directive 2002/21/EC of the European Parliament and of the Council of 7 march 2002 on a Common
Regulatory Framework for electronic communications networks and services (“Framework Directive”).
2
EUI-1200160743v1
15 JULY 2015
In the context of the current consultation process, Telenet has requested an oral
hearing and full access to the retail minus model. This oral hearing has taken place on 15
June and an extension for the reply to the consultation was granted until 15 July 2015.
Telenet thanks the Regulator for its positive response to this request and hopes that the
Regulator will take due account of Telenet’s comments below.
Finally, Telenet refers to the pending legal proceedings (i.e. the pending appeal
against the previous tariff regulation decision) and the arguments that are made in these
proceedings which may also affect the Draft Decision. These arguments are, however, not
reiterated here.
 The Framework Decision dates back to 2011 and is based on market data of 2010
and older. It is problematic to justify a steep reduction of the wholesale prices on the
basis of a Framework Decision that is overdue for review.
 Telenet has complied with its regulatory obligations and implemented operational
access within the strict time limits imposed by the Framework Decision.
 The implementation period of 1 month is too short.
II.
VIOLATION OF TARIFF REGULATION PRINCIPLES
The methodology set out in the Draft Decision violates the following fundamental
principles of tariff regulation, (i) the principles of proportionality, objectivity, duty of care
and motivation and (ii) the principles of non-retroactivity and legal certainty.
A.
Violation of principles of proportionality, objectivity, duty of care and
motivation
The principle of proportionality is an essential principle of the regulatory framework.
Article 8(1) of the Framework Directive provides that: “Member States shall ensure that in
carrying out the regulatory tasks specified in this Directive and the Specific Directives, the
national regulatory authorities take all reasonable measures which are aimed at achieving
the objectives set out in paragraphs 2, 3 and 4. Such measures shall be proportionate to
those objectives.” As regards the imposition of remedies in particular, Article 8(4) of the
Access Directive provides that “[o]bligations imposed in accordance with this Article shall
be based on the nature of the problem identified, proportionate and justified in the light of
the objectives laid down in Article 8 of Directive 2002/21/EC (Framework Directive). Such
obligations shall only be imposed following consultation in accordance with Articles 6 and 7
of that Directive.” In the Food Supplements case, the Court of Justice confirmed that this
principle is the boundary to the principle of freedom of trade and the right to individual
property, which are not unlimited but can be restricted through regulatory interventions,
provided that the proportionality principle is complied with:
“[I]t is clear from settled case-law that the right to property, […] and likewise
the freedom to pursue an economic activity, form part of the general principles
of Community law. However, those principles are not absolute but must be
viewed in relation to their social function. Consequently, the exercise of the
right to property and the freedom to pursue an economic activity may be
3
EUI-1200160743v1
15 JULY 2015
restricted, provided that any restrictions in fact correspond to objectives of
general interest pursued by the Community and do not constitute in relation to
the aim pursued a disproportionate and intolerable interference, impairing the
very substance of the rights guaranteed […].”4 (emphasis added)
The principle of objectivity has been recognized in established case law of the Court
of Appeal and directly follows from the obligation to duly motivate decisions and ensure that
the regulatory obligations are proportionate. In an appeal regarding cost modeling of
Proximus’ wholesale tariffs, the Court of Appeal of Brussels recognized that even if the
Regulator may regulate tariffs on the basis of a cost model of a hypothetical efficient
operator, it had to ensure that this model remains as close as possible to the reality of
Proximus’ operations and costs. In its judgment of 29 June 2011, the Court of Appeal stated
that: “[…] il appartient à [l’IBPT] de construire son modèle de la façon la plus réaliste
possible, en tenant dûment compte du contexte et des faits et circonstances pertinents
reflétant la réalité dans laquelle un opérateur hypothétique efficace devrait fonctionner pour
éviter des écarts trop importants avec la situation de l’opérateur historique. […] Un tel
modèle doit donc être le plus possible fondé sur des principes et des paramètres reflétant
cette réalité de façon fiable.”5 The Court of Appeal of Brussels further insisted on the fact
that the model would have to reflect effective demand for the service in order to ensure that it
would be a viable and realistic model: “[l]’un des éléments essentiels permettant de garantir
le caractère réaliste et fiable d’un modèle ascendant est que le réseau modélisé à partir
duquel les coûts sont calculés doit être dimensionné en fonction de la demande comme le
confirme l’IBPT lui-même dans la Décision Attaquée (page 8).”6
In the Netherlands, the College van Beroep voor het Bedrijfsleven has recognized the
requirement for NRAs adequately motivate their decisions and ensure they have taken all
necessary measures of investigation and fact-finding to assess that the tariff regulation
imposed is proportionate and does not overreach by imposing an excessive burden. Tariff
regulation is recognized as a particularly intrusive form of intervention which should only be
imposed if it is necessary to achieve the regulatory objectives. There is a particular burden
on the NRA to scrutinize the proportionality of any measure which would imply that the
SMP operator is burdened with additional cost to assist competitors.7 The assessment of the
4
Joined Cases C-154/04 and C-155/04, Alliance for Natural Health e.a. v. Secretary of State for Health,
12 July 2005, § 126 (emphasis added).
5
R.G. N°:2010/AR/2695, 29 June 2011, at § 28 al. 3.
6
Ibidem, at § 29.
7
“9.2.2 De grieven 2.2 en volgende van KPN klagen er kort samengevat over dat OPTA bij het opleggen
van de verplichting tot proportionele doorrekening van wholesalespecifieke kosten verder is gegaan
dan noodzakelijk en niet heeft voldaan aan de op haar rustende onderzoeks- en
motiveringsverplichtingen. […] Het College overweegt hieromtrent het volgende. De in artikel 6a.7
Tw genoemde verplichtingen met betrekking tot kosten en kostenoriëntatie hebben een ingrijpend
karakter. Prijsregulering is een vergaande vorm van regulering van de markt. Dit geldt in het
bijzonder voor de hier aan de orde zijnde proportionele toerekening, die er op neerkomt dat de
gereguleerde onderneming een – aanzienlijk – deel van de kosten die moeten worden gemaakt om
anderen met haar te kunnen laten concurreren, zelf moet dragen. OPTA dient bij het opleggen van een
dergelijke verplichting te onderzoeken of deze geschikt is voor het bereiken van het doel en of deze niet
verder gaat dan voor het bereiken van dit doel noodzakelijk is. Gelet op de in artikel 1.3, eerste lid,
onder c, Tw, genoemde doelstelling van het bevorderen van belangen van eindgebruikers wat betreft
keuze, prijs en kwaliteit, dient OPTA hierbij mede in ogenschouw te nemen in hoeverre de
proportioneel toegerekende wholesalespecifieke kosten door de gereguleerde partij aan
4
EUI-1200160743v1
15 JULY 2015
possibility for competitors to effectively compete on the market with a lighter form of
intervention therefore has to be examined by the NRA.8
As regards the remedies set out in the Framework Decision, the Court of Appeal has
confirmed that the objective of the wholesale access regulation is to reduce barriers to entry
and enable potential entrants and operators in the relevant retail market to provide an offer
that is fully competitive with the cable operators, allowing them to build a lasting business
relationship with the end customer. This access regulation should allow beneficiaries to
invest in their own infrastructure and offer competitive multiplay offers by using their own
infrastructure.9
In conclusion, the regulation needs to respect the principle of “minimum
intervention”, which implies that the imposed regulation should be limited to what is
appropriate and necessary to attain its objective. In accordance with the Commission
Guidelines, a proposed measure is compatible with the principle of proportionality if it
pursues a legitimate aim and if the means employed to achieve that aim are both necessary
and the least burdensome (i.e. the minimum necessary to achieve the aim).10 In order to do
so, the Regulator must take into account the relation between the costs and benefits of the
contested regulation to justify why a certain regulatory measure is necessary and effective,
indicating how the cumbersome effect of the reasonable measures was impossible to avoid.11
The proposed tariff regulation is violating the above principles for the following
reasons which are set out below:
 The regulator does not retain the bundle which is the closest to the regulated
market and regulated wholesale access.
(continued…)
eindgebruikers worden doorberekend. Bovendien dient OPTA de aan haar beslissing ten grondslag
liggende overwegingen – inclusief berekeningen – zoveel als mogelijk inzichtelijk te maken. Aan deze
verplichtingen heeft OPTA in het Tariefbesluit niet voldaan.” College van Beroep voor het
Bedrijfsleven, AWB 07/36, 07/68 en 07/69, 12 September 2007.
8
“Indien het een bestuursorgaan is toegestaan om een ingrijpende maatregel op te leggen, zal zij
immers in de regel ook de bevoegdheid hebben om een minder ingrijpende maatregel op te leggen en
op basis van overwegingen van proportionaliteit hiertoe zelfs gehouden zijn indien ook deze minder
ingrijpende maatregel effectief is. Uit het bovenstaande volgt dat beroepsgrond 3 van KPN gegrond is.
OPTA heeft het zorgvuldigheidsbeginsel van artikel 3:2 Awb en het motiveringsbeginsel van artikel
3:46 Awb geschonden. OPTA dient eigen onderzoek te verrichten teneinde na te gaan welke mate van
proportionele toerekening van WSK is vereist om voor WLR-afnemers een business case te bieden voor
het bedienen van hun klanten via WLR HC.” College van Beroep voor het Bedrijfsleven, AWB 10/85
en 10/86, 9 November 2012.
9
Brussels, 12 November 2014, R.G. 2011/AR/2289 et al., § 123-125, p. 69-70.
10
Commission guidelines on market analysis and the assessment of significant market power under the
Community regulatory framework for electronic communications networks and services, 2002/C
165/03, 11 July 2002, point 118, p. 22.
11
Brussels, 12 November 2014, R.G. 2011/AR/2289 et al., § 123-125, p. 69-70.
5
EUI-1200160743v1
15 JULY 2015
 The methodological approach values additional services (which are independent
from the regulated input) on the basis of the retail reference price (instead of
avoidable cost) and does not take account of effective usage.
 The methodology does not exclude legacy offerings.
 The assessment of promotions is arbitrary and contrary to accounting principles.
 The methodology grants an excessive and generic market entry assistance.
 The accompanying transparency measures imposed in the Draft Decision are
unclear and create an excessive burden.
 There is no impact assessment.
1.
The Regulator should retain the reference prices of the bundles which are
closest to the regulated wholesale access
At § 140 of the Draft Decision, the Regulator decides that when the methodology
yields a lower price for a 3P bundle (compared to a 2P bundle), it will consider the price for
the 3P bundle as reference price. Although it recognizes that the pricing strategy which
consists of increasing the attractiveness of a 3P bundle is legitimate, it refers to this practice
as a form of internal discrimination pursuant to which the cable operator would be applying a
lower access price for a 3P offering compared to a 2P offering.
This approach contains a fundamental methodological error. According to the
principle of proportionality, the methodology should give preference to the retail service
offer which is closest to the wholesale service that it wants to insulate (i.e. prefer double-play
over triple play where both are available). By using the triple-play offer for its calculation,
the Regulator penalizes SMP operators for broadening their bundles to include voice
telephony (or potentially other components such as mobile) which are generally procompetitive as they concern services on which Proximus generally has the incumbent leading
market position and for which cable operators will seek to aggressively profile themselves.
By applying such methodology, the relevant weight of the regulated products in the bundle is
reduced even below a certain acceptable minimum value for these services. The regulated
television and broadband components combined should represent at least 75% of the total
value of the products in the relevant bundle. If this is not the case, it should be concluded
that the television and broadband component in the bundle cannot be considered a relevant
point of reference within the scope of the Framework Decision.
If the cable operators would launch quadruple play offers, this logic would lead the
Regulator to use quadruple play bundles as a basis for its wholesale price calculations, i.e. to
take the retail pricing of mobile telephony into account to determine the wholesale price of
television and broadband. Telenet notes in this respect that the Regulator has decided not to
include the Numericable 4P offering in its assessment and therefore assumes that this implies
that the Regulator recognizes that it does not consider it appropriate to consider the mobile
component in a 4P offering as an “additional service” but leave such offers unaffected.
Telenet would appreciate it if the Regulator would make this explicit in its final decision.
Finally, in selecting a reference product, the Regulator should take into account the
relevance of the product. Products with a limited number of subscribers (e.g., less than 10%
6
EUI-1200160743v1
15 JULY 2015
of the total installed customer base) cannot be considered as a relevant reference product and
should be excluded.
2.
“Avoidable costs” and usage cannot be ignored to assess the economic value of
the services
In its methodology, the Regulator considers “costs” for additional services only to be
the “third alternative” method for assessing value without providing any justification to this
effect (Draft Decision, § 4912).
In the retail minus methodology, additional services should in principle be considered
as avoidable costs which have to assessed on a cost-based methodology as this is the less
intrusive (and therefore more proportionate) methodology. This approach implies that the
non-regulated service elements are assessed on the basis of the SMP operators’ own costs.
The table below compares the impact of the proposed retail benchmark method compared to
a cost-based approach for each additional service. The costs were assessed on the basis of
the incremental costs associated to the provision of the service which have been allocated to
the entire relevant subscriber base (i.e., all subscribers that are entitled to use the service even
if they do not actually use it). Given the short consultation period, the cost analysis is
provisional and only provided for illustrative purposes. Telenet has already provided the
Regulator with cost information that can serve as a basis for a cost-based methodology.
Another fundamental methodological failure in the approach is that the Regulator is
not consistent in taking into account the actual usage of the offers upon which it relies for
benchmarking purposes: actual usage is taken into account for content (Sporting Telenet), but
not for the other “additional services”. Telenet does not have access to all relevant
information (particularly insofar as it concerns offerings of other operators) but has compared
in the chart below the benchmark values used by the Regulator with these values adjusted for
the effective usage of the additional services concerned. The case law referred to above
regarding cost modeling principles confirms that the price regulation should stay as close as
possible to the underlying reality of the regulated operator. By relying on offerings which
are of marginal commercial relevance, the Regulator is violating this requirement. To the
extent the retail values are being considered, the methodology should ensure that effective
usage is taken into account when assessing the relevance of the service as it otherwise adopts
a methodology which is completely theoretical and grants excessive economic weight to
these services. Telenet is willing to provide monthly updates to the Regulator of effective
usage figures of the relevant services.
The charts below confirm that the method retained in the Draft Decision is far more
intrusive and implies that SMP operators suffer from a more rigorous wholesale price
regulation as they expand their bundled offerings with the inclusion of additional services.
12
“Zonder gepaste referentieprijs (bijvoorbeeld omdat de dienst in kwestie nooit los van andere diensten
wordt verkocht en het niet mogelijk is geweest om bij andere operatoren referentieprijzen te vinden)
vindt het BIPT dat het beste alternatief erin bestaat de productiekosten per eenheid te gebruiken (d.w.z.
de kosten per gebruiker) als benadering van de marktwaarde. De productiekosten van de diensten
kunnen ook als nuttige referentie dienen in de gevallen waarin de effectieve prijs nul bedraagt of erg
laag is: blijkt de voor een dienst gehanteerde prijs lager te zijn dan de productiekosten, dan vormt deze
laatste een betere indicator van de marktwaarde.” Draft Decision, § 49.
7
EUI-1200160743v1
15 JULY 2015
[confidential chart: benchmark values vs. actual usage]
[confidential chart: valuation of additional services - Whop]
[confidential chart: valuation of additional services - Whoppa]
The regulatory approach which consists in allocating a “nominal value” in excess of
production costs is an approach which is disproportionate insofar as the retail pricing for
these other services is determined by competitive dynamics which can be different from costs.
Stating that nominal prices are the best estimate of the value (Draft Decision, § 46) is wrong
insofar as products are generally more expensive when sold separately (as opposed to a
bundle). Separate sales create additional costs and change common costs and margin
allocation.
Furthermore, cable operators have no SMP for the provision of these additional
services and the beneficiaries of the regulation have the ability to add the same or similar
services to the bundle. There is no indication whatsoever that the beneficiaries of the
regulation could not compete with the SMP operator if the avoidable costs associated to the
provision of these additional services would be deducted in the “minus”.
In its Draft Decision, the Regulator refers itself to potential distortions which may be
observed in retail price setting which may stem from cross-subsidization between various
retail services. The Regulator recognizes that certain stand-alone services could be priced at
abnormally low tariffs as a result of which the retail price may not be the adequate basis for
assessing their value. 13 In reality, the opposite concern often applies, i.e., standalone
offerings are offered at premium prices in order to increase the attractiveness of bundles.
The table above has demonstrated that the retail pricing for the single play products is often
significantly higher than the corresponding production costs.
The Regulator further states in the Draft Decision (§ 49) that the production cost of
the additional services would only be a useful reference in cases where the price charged is
zero (or very low). According to the Regulator, if the applicable price for a service turns out
to be less than the cost of production, the production cost is a better indicator of the market
value. As production costs are the only objective point of reference available, they should, as
indicated above, always be the preferred basis for assessing the value of the services. In
order to be consistent, the Regulator must therefore also rely on the cost of production where
the applicable price for an additional service exceeds the production cost.
The inconsistency in the approach is also confirmed in the Draft Decision itself as
regards the proposed calculation for the decoder. The Regulator emphasizes at § 87 the
disadvantage of making the wholesale price dependent on a service (in this case the decoder)
which is not part of the wholesale offer.14 The Regulator points out that the beneficiaries of
13
Draft Decision, § 46.
14
“87. Deze aanpak heeft als nadeel dat ze de wholesaleprijs laat afhangen van de huurprijs van de
decoder op de retailmarkt, terwijl de decoder geen deel uitmaakt van het wholesaleaanbod (de
begunstigde van het wholesaleaanbod levert zijn eigen decoder en koopt dus niet die van de
kabeloperator). Wanneer een kabeloperator dus beslist om de huurprijs van zijn decoder te verhogen,
en al de rest blijft gelijk, dan zal de wholesaleprijs stijgen als gevolg van een verhoging van de
retailprijs (terwijl de wholesaleverrichting niet wordt gewijzigd).”
8
EUI-1200160743v1
15 JULY 2015
the wholesale offer will provide their own decoders, which they do not buy from the cable
operators. This reasoning undermines the entire methodology which is relied upon for
assessing the value of the additional services which is also made dependent on retail pricing
of the SMP operator (or other operators retained in the benchmark).
The most proportionate and objective methodology to determine the value of these
additional services is to consider their incremental costs or, if the retail value is considered,
an adequate account of the usage of the service is considered. This ensures that cable
operators do not suffer from an artificial competitive handicap by including these services in
their bundles as opposed to the provision of these services on a standalone basis. The CRC
has itself recognized the need to avoid a regulatory overreach regarding costs which the
beneficiary of the access has to control and which concern services for which the cable
operators do not have SMP.15
A methodology which deducts additional services included in a bundle as an
avoidable cost is also consistent with the price squeeze and replicability analysis which has
been relied upon in ex ante and ex post enforcement cases. These methods are particularly
relevant given that the Commission has conceived such a replicability test in its 2013
Recommendation on Costs Methods and Non-Discrimination, of which the Regulator is
required to take the “utmost account”.16 It confirms that a “cost-based” approach for the
additional services allows the development of efficient market entry.
In the 2013 Recommendation on Cost Methods and Non-Discrimination, the
Commission has indicated that “(64) NRAs should ensure that the margin between the retail
price of the SMP operator and the price of the NGA wholesale input covers the incremental
downstream costs and a reasonable percentage of common costs. […] The use of the EEO
standard enables NRAs to support the SMP operators’ investments in NGA networks and
provides incentives for innovation in NGA-based services”. The Recommendation further
specifies at § 67 that “the economic replicability test set out by the NRA in advance should be
adequately detailed and should include as a minimum a set of relevant parameters in order
to ensure predictability and the necessary transparency for operators. NRAs should apply a
LRIC+ model while taking into account the SMP operator’s audited downstream costs and
assess the margin earned between the most relevant retail products including broadband
services (flagship products) and the regulated NGA access input most used, or identified,
under a forward-looking approach, as the most relevant for delivering the retail products for
the market review period in question. The design of the test applying to the SMP operator’s
audited downstream costs and only for flagship products, aims to ensure that NGA
investments and the effect of the recommended pricing flexibility are not hindered by this
safeguard.” (emphasis added)
15
“Alors si, effectivement la CRC doit veiller, par le prix qu’elle fixe, à faciliter l’entrée d’acteurs
efficaces sur le marché de la télédistribution et ainsi promouvoir le développement d’une concurrence
effective, il ne lui revient pas de faire peser sur les opérateurs régulés le coût des éléments qui sont
sous controle de Mobistar ni celui d’éléments qui ne relèvent pas de la saisine de la CRC en l’espèce
[…]”. Trial pleadings of the CRC of 11 May 2015 before the Court of Appeal in Case 2014/AR/328
e.a., § 53.
16
Commission staff working document impact assessment accompanying the document Commission
Recommendation on consistent non-discrimination obligations and costing methodologies to promote
competition and enhance the broadband investment environment, 11 September 2013, p. 6.
9
EUI-1200160743v1
15 JULY 2015
Annex 2 of the Recommendation on Cost Methods and Non-Discrimination also
confirms that relevant downstream costs (which correspond to the avoidable costs to be
deducted from the minus) include “the SMP operator’s own downstream businesses (EEO
test). NRAs should use the SMP operator’s audited downstream costs, provided they are
sufficiently disaggregated”.
A concrete application of this methodology can be found in the Commission’s Article
7 comments letter regarding Ofcom’s proposed margin squeeze analysis for BT’s bundle
including broadband and premium sport content. In the proposed methodology, Ofcom did
not go as far as the Regulator in retaining the retail pricing (for the standalone content
offering) to reduce the relevant retail revenues, but proposed a cost-accounting methodology
for the associated costs which would be more favorable to market entrants. Following its
review, the Commission warned Ofcom that the
“proposed approach lacks the necessary flexibility in particular with regards to the
treatment of costs for BT Sports. The Commission considers that the proposed static
approach unduly limits BT’s commercial activity with regards to a market in which it
does not have SMP. Indeed, Ofcom’s purpose in designing the test is to address a
competitive concern in relation to pricing of a regulated product, i.e. VULA, for
which it does not propose a charge control. However, the design of the test proposed
by Ofcom may result in BT choosing to address a failed test by changing its
behaviour on a market other than the regulated market, for example by increasing
retail prices or reducing costs for BT Sports. Given the magnitude of the costs
involved, and the uncertainty of future costs and revenues of BT Sports as new rights
auctions approach, there is a risk that Ofcom’s regulatory intervention would have a
significant impact on non-regulated markets, without necessarily affecting the price
of the VULA input. Ofcom will therefore have to remain vigilant that the application
of the test does not have unintended consequences in markets where the application
of ex post competition law would be sufficient, or where BT’s SMP in the WLA market
does not necessarily play a role. […]”17 (emphasis added).
The table above comparing the impact of a cost-based and a retail benchmark
approach shows that the methodology foreseen in the Draft Decision is even much more
distortive than Ofcom’s proposed methodology in affecting the SMP operator’s incentives to
compete more aggressively on the retail market by expanding its bundle to non-regulated
additional services for which it may be confronted with strong competitors.
The BEREC Guidance of 5 December 2014 on the Regulatory Accounting Approach
to the Economic Replicability Test (i.e. ex-ante/sector specific margin squeeze tests)
provides further guidance on the margin squeeze assessment.18
It recognizes the elements that should be part of a margin squeeze test and states that
non-regulated products are assessed on the basis of their input costs (including own network
costs):
17
Commission Decision concerning Case UK/2015/1692: Wholesale local access at a fixed location in
the United Kingdom – Remedies Comments pursuant to Article 7(3) of Directive 2002/21/EC, 13
February 2014, p. 6-7.
18
BoR (14) 190.
10
EUI-1200160743v1
15 JULY 2015
“Regarding the general formula for the margin squeeze test nearly all NRAs take into
account the following items:
- Retail price of the SMP’s downstream service,
- Regulated wholesale costs needed to provide the downstream service charged by
the SMP’s upstream division,
- Non-regulated input costs (incl. own network costs),
- Retail costs.”
This Report also explains the cost/revenue elements that need to be taken into
consideration when applying an ex ante margin squeeze test to IPTV service.
“Bundles including IPTV service
The following revenues and costs are typically relevant for margin squeeze testing:
Revenues:
o TV package revenues;
o TV service set-up fees;
o All TV related revenues such as subscriptions to VOD services (film, sports events),
or pay per use charges;
Costs:
o TV platform costs;
o Program content costs (live TV, Video on demand etc.);
o Set Top box (STB) costs;
o Installation cost;
o Transaction costs (ordering, jumpering, completion costs, etc.).”
BEREC further recommends that, even in a REO (“reasonable efficient operator”)
approach, the calculation should be based on costs when available, rather than retail prices:
“In the context of a REO/adjusted EEO approach, where information is available on
the network costs of the alternative operator, the calculation could be based on these
costs. Alternatively and in cases where these costs are not readily available, the
prices commercially agreed on the carrier market could be used as a first proxy for
cost.”
Calculating the value of additional services on the basis of the standalone retail price
is contrary to the approach followed in ex-post price squeeze cases. In Telefonica, the
Commission favored an aggregated retail service/cost analysis approach based on the
11
EUI-1200160743v1
15 JULY 2015
objective that margin squeeze should allow competitors to replicate the product pattern of the
dominant undertaking.
“In the case at hand, the margin squeeze test has been conducted on the basis of an
aggregated approach, i.e. on the basis of the mix of services marketed by Telefónica
on the relevant retail market. This approach (referred as to the “aggregated
approach”) is based on the principle that competitors must at least be able to
profitably replicate Telefónica’s product pattern. This is the approach most
favourable to Telefónica, since it gives it maximal flexibility to spread the costs which
are common to its retail products (provided that the margin squeeze test yields a
positive result with the aggregated approach). The aggregated approach is consistent
with a new entrant’s internal decision making process in that it assesses the
profitability of its investment in a network by considering the complete range of
products that it is able to offer in the relevant downstream market”.19
In TeliaSonera, the Court of Justice also held that pricing practices should be assessed
on the basis of costs of the dominant undertaking.20
In conclusion, pursuant to the proportionality and objectivity principle, the Regulator
should rely on the costs associated to the provision of the additional services. It cannot
consider this as the third best alternative without any motivation to that effect. The costbased approach is a more reliable and less intrusive form of assessing the value of the
additional services and should therefore be the preferred approach. Telenet has provided
initial cost information to that effect.
If the Regulator retains the retail tariffs, it should, as it did for the content component,
at least apply a correction factor to take account of the effective usage of the service.
3.
Legacy offerings should be excluded
The Regulator’s approach could result in excessively low wholesale tariffs when
applied to some of Telenet’s legacy offers (i.e. retail offers that are no longer actively offered
but still have subscribers and are covered by the resale remedy). As mentioned above, when
selecting a reference product, it is important to take into account the relevance of the product,
19
Decision of the Commission, Telefonica, Case COMP/38.784, § 388.
20
Case C-52/09, TeliaSonera Sverige ECR, 2011 , p. I-00527:
“In order to assess the lawfulness of the pricing policy applied by a dominant undertaking, reference
should be made, as a general rule, to pricing criteria based on the costs incurred by the dominant
undertaking itself and on its strategy (see, to that effect, Case C ‑62/86 AKZO v Commission [1991]
ECR I‑3359, § 74, and France Télécom v Commission, § 108).
42 In particular, as regards a pricing practice which causes margin squeeze, the use of such analytical
criteria can establish whether that undertaking would have been sufficiently efficient to offer its retail
services to end users otherwise than at a loss if it had first been obliged to pay its own wholesale
prices for the intermediary services (see, to that effect, Deutsche Telekom v Commission, § 201).
43 If that undertaking would have been unable to offer its retail services otherwise than at a loss, that
would mean that competitors who might be excluded by the application of the pricing practice in
question could not be considered to be less efficient than the dominant undertaking and, consequently,
that the risk of their exclusion was due to distorted competition. Such competition would not be based
solely on the respective merits of the undertakings concerned.”
12
EUI-1200160743v1
15 JULY 2015
i.e. a product with a limited number of subscribers (e.g., less than 10% of the total installed
customer base) cannot be considered as a relevant reference product.
The Regulator should thus exclude all offerings which are not relevant or no longer
being commercialized (at the very least those which were launched prior to the
implementation of the retail minus methodology) and which are therefore not relevant for the
prevailing competitive dynamics. The Regulator should also provide for a or an absolute
price “floor” to ensure that the regulated wholesale prices do reach pricing levels which
would be below costs.
4.
The treatment of promotions
The Regulator is treating “promotions” as avoidable costs instead of considering them
as a reduction of the ARPU (Draft Decision, § 109). This is inconsistent with the CRC’s
framework decision and decisional practice on price squeeze.
Furthermore, Telenet has provided the Regulator with its promotion costs
demonstrating that all (or almost all) these costs are related to the promotion of triple-play
offers including broadband, television and fixed telephony. However, the Regulator
allocated those costs only to the broadband and television services instead of applying a pro
rata allocation to all three services.
In the CRC Decision of 11 December 2013, promotions were taken into account via
the ARPU, i.e. on the income side of the calculation of the minus (the denominator of the
fraction). In the Draft Decision, promotions are added to the avoided costs, on the cost side
of the calculation of the minus (the numerator of the fraction). In a standard cost minus
methodology, where the avoided costs are deducted as an absolute amount from the retail
revenues, it is neutral whether promotions are added to the avoided costs or deducted from
the revenues. Because of the Regulator’s methodological choice to apply the minus as a
percentage, this is no longer the case and adding promotions to the avoided costs or
deducting them from the revenues has a different impact.
The only motivation provided in the Draft Decision for moving the promotions from
the denominator to the numerator of the fraction is to correct for the “dilution effect” due to
Telenet’s large existing customer base. However, Telenet fails to see the logical connection
between this operation and the so-called “dilution effect”. Moreover, this “dilution effect” is
already being spreading promotion costs on gross adds only (instead of the entire customer
base) for a transitional period. Telenet fails to see, and the Regulator fails to explain, why an
additional correction by moving the promotions from the denominator to the numerator is
necessary. To the extent that there is any logical connection between this change and the
“dilution effect”, it can only be concluded that the Regulator corrects for this effect twice,
without any justification or assessment of the envisaged impact (and efficient scale retained).
Moreover, the Regulator does not make any distinction between promotions in the
form of one-off advantages for new customers (e.g., free installation and activation, free
gift…) and discounts on the subscription price. Even if we accept, for the sake of argument,
that it is “necessary” to make an additional adjustment to the minus calculation to correct for
the “dilution effect”, it is simply wrong to treat discounts on the subscription fees as “avoided
costs”. So while there is some logic to treating one-off advantages such as free installation
and activation as costs, discounts on the subscription fees should be treated as “revenues
foregone” and should therefore be deducted from the ARPU. As further explained in the
13
EUI-1200160743v1
15 JULY 2015
attached paper by RBB, it is contrary to basic accounting rules to book the discounted
subscriptions as if the customer paid the undiscounted price (i.e. to book a fictitious income)
and then to amortize the discount as a cost.
5.
Excessive market entry assistance
The Draft Decision claims that wholesale tariffs should temporarily facilitate market
entry and foresees a calculation method for promotions which limits the allocation of these
costs to a smaller customer base (i.e. the “gross adds” of the SMP operator, instead of taking
the entire SMP operator customer base). The Regulator is claiming that the beneficiaries do
not have an installed client base and suffer from the fact that the current methodology
determines the minus on the basis of the totality of the cable operators’ client base (Draft
Decision, § 111).
Telenet contests the proposed methodology on the basis of the following grounds:
 Inconsistency with the allocation methodology of the wholesale costs;
 No individualized analysis of the beneficiaries of the regulation;
 Promotion of inefficient and excessive market entry.
a)
Inconsistency with allocation method of wholesale cost
Granting temporary market entry assistance by calculating the impact of promotions
on a smaller subscriber base is methodologically inconsistent with the fact that wholesale
costs were allocated over the total subscriber base of the cable operators. This inconsistency
shows that the regulatory approach is not objective but geared towards promoting an artificial
and inefficient market entry. In the appeal procedure regarding the first tariff regulation, the
CRC itself rejected Mobistar’s claim to calculate the promotion impact on a smaller
subscriber base (in order to favor its market “entry”), emphasizing the need to ensure a
methodological consistency with the wholesale cost allocation.21
b)
No individualized analysis
So far, only two operators have expressed interest in the regulated wholesale access:
Proximus and Mobistar. Both operators have the necessary scale, customer base and
financial means, which they have established with the protection of exclusive (or special)
rights and years of high mobile pricing. Both have extensive distribution networks.22 Their
ability to bundle various products gives them unique leveraging opportunities. Neither
operator can reasonably claim to be sub-scale or suffer from entry disadvantages.
21
“La CRC observera que quand il s’est agi de déterminer les coûts d’investissements non récurrents à
charge des bénéficiaires, Mobistar s’est réjouie de la prise en compte par la CRC des clients existants
des câblo-opérateurs. Cette approche a en effet permis de diluer, c’est-à-dire diminuer sensiblement,
les coûts à charge de Mobistar”, cf. § 112.
22
For instance, Mobistar has over 154 Mobistar Centers all over the country, usually in prime retail
locations, as well as a well-developed online distribution channel.
14
EUI-1200160743v1
15 JULY 2015
As far as Proximus is concerned, it suffices to note that the Regulator considered that
Proximus should not be entitled to benefit from the wholesale access all together. 23 The
proposed tariff revision should in any event also foresee a carve-out for Proximus insofar as
most components of the bundle are offered by Proximus in its own bundle. Proximus is even
market leader for most of these components.
As far as Mobistar is concerned, this operator has almost 3,8 million mobile
subscribers. CRC and Mobistar have recognized the relevance of this subscriber base
indicating that the wholesale cable access would allow it to engage in cross-selling activities
and reduce its churn rate for its mobile subscribers by migrating them to multiplay offers. 24
Mobistar has also benefited from unparalleled market assistance (which was
significantly financed by Telenet). As second mobile entrant it enjoyed special rights and
advantages in the inherently oligopolistic mobile telecommunications market. Moreover, due
to a favorable regulatory regime, Mobistar was able to charge termination rates (MTR) that
were well above the rates of a competitive market 25 and allowed it to subsidize the
commercialization of its services.26 Subsequently, with the installment of a glide path by the
Regulator in 2006, Mobistar continued to enjoy a special regime 27 despite the European
Commission opposition. 28 The Commission considered that Mobistar’s MTRs had to be
brought down to reflect the cost of an efficient operator as soon as possible. 29 Overall,
Mobistar benefitted from positive discrimination for almost 17 years, while the European
Commission in 2009 recommended that termination rate asymmetry benefiting new entrants
should be limited to a transition period not exceeding four years after market entry.30
Mobistar has also been able to deploy fixed services and roll-out a network on the
basis of regulated access products since the full liberalization of the sector in 1998. It can
therefore no longer be considered as a “new entrant” under any circumstances. Its failure to
23
Telenet refers to the access dispute which is currently ongoing following Belgacom’s request for
access in this respect.
24
Cf. Mobistar “2013 Analyst & Investor Presentation”, p. 42 which is also referred to by the CRC in its
trial pleadings of 11 May 2015 before the Court of Appeal in Case 2014/AR/328 e.a., where CRC
states “[c]onstrairement à ce qu’elle présente dans sa requête, il convient en effet dans une telle
analyse, de tenir compte de l’impact des services TV et Internet sur la rentabilité future de tous les
autres services que Mobistar commercialise sur son réseau mobile. Au moyen de l’accès au câble,
Mobistar pourra ainsi s’engager dans des stratégies de cross-selling, limiter le taux d’attrition (churn)
de sa clientèle mobile et participer pleinement à l’essor des offres multiplays”, cf. p. 23.
25
BIPT Decision of 23 September 2003 regarding termination costs of Mobistar, § 5.8; Decision of the
BIPT Council of 11 August 2006 regarding market definitions, analysis of competitive conditions,
identification of SMP operators and determining appropriate obligations for market 16: mobile
termination on individual mobile networks, § 2.2.6.
26
BIPT Decision of 23 September 2003 regarding termination costs of Mobistar, § 5.8.
27
BIPT Decision of 11 August 2006 regarding market definitions, analysis of competitive conditions,
identification of SMP operators and determining appropriate obligations for market 16: mobile
termination on individual mobile networks, § 2.2.6.
28
Letter from the European Commission, Case BE/2006/0433 Voice Call Termination on Individual
Mobile Networks in Belgium, Comments pursuant to Article 7(3) of Directive 2002/21/EC, p. 5-6.
29
Ibidem.
30
European Commission Recommendation of 7 May 2009, Recital 17, p. 6.
15
EUI-1200160743v1
15 JULY 2015
achieve a significant size and network for fixed services is largely attributable to France
Télécom’s disinterest in investing in fixed networks notwithstanding Mobistar’s record
financial performances over the years. Its incoherent commercial strategy, recognized by the
CRC, also certainly played a role but is no ground for assistance.31 This is particularly as, its
fixed activities have already benefitted from assistance. Mobistar’s network was acquired by
KPN, which had, in turn, purchased the Tele2/Versatel network which was (along with
Telenet) the only fixed alternative operator to have benefited from entry assistance through
asymmetric FTRs. Back in 2006, however, the Regulator already indicated that this regime
had to be phased out and that the mark-up on Versatel’s FTRs in comparison with Proximus’
(Belgacom at the time) cost-oriented FTRs had to be capped at 15%. 32 In 2012, BIPT
imposed symmetric FTRs on all operators.33
Telenet estimates that the market entry assistance it would be forced to provide to
Mobistar during the three-year transition period would amount to over € 62 million,
including € 13,6 million in respect of promotions, € 17 million in respect of implementation
costs, € 12,6 million in respect of additional services and € 19 million in respect of the
deduction of fixed telephony. Nothing justifies Telenet being forced to provide such a large
subsidy to a well-established competitor.
If the Regulator is considering a regulation which aims at offering an extra entry
assistance, it should ensure that this regulation is sufficiently individualized and effectively
granted to a real hypothetical “new entrant”. The Court of Appeal has recognized that
remedies should be based on such an individualized analysis, insisting on the fact that:
“Le régulateur doit rechercher, au cas par cas, les remèdes appropriés pour
rencontrer de la façon la plus adéquate tous ses objectifs qu’il doit tenter de concilier
entre eux. Cette appréciation faite in concreto, est susceptible de varier dans le
temps.”34
“L’IBPT doit donc apprécier, sous sa seule responsabilité, les remèdes qu’il décide
d’imposer dans le cadre de ses compétences territoriales et il ne doit s’aligner sur
une position commune des ARN – si elle existe – ou sur une ‘jurisprudence d’avis’
suffisamment claire, précise et stable de la Commission européenne – si elle existe –
que dans la mesure et pour autant qu’il puisse raisonnablement considérer, au terme
d’une analyse sérieuse menée in concreto, qu’il n’y a pas lieu de s’en écarter au vu
des circonstances nationales qu’il est le plus apte à apprécier.”35 (emphasis added)
The in concreto/individualized approach is confirmed by the case law of the
Constitutional Court36 and the Court of Justice of the EU.37
31
[confidential].
32
BIPT decision of 11 August 2006 concerning the cluster “fixed telephony”, pp. 225-227.
33
BIPT decision of 2 March 2012 concerning the analysis of the market for call termination at a fixed
location, pp. 117-123.
34
Brussels, 30 June 2009, R.G. 2006/AR/2332 et al., p. 48-49.
35
Brussels, 30 June 2009, R.G. 2006/AR/2332 et al., p. 48.
36
“En ce que le législateur belge traite en l’espèce tous les opérateurs de manière identique, alors que
les articles 12 et 13 de la Directive sur le service universel exigent un traitement différencié de tous les
16
EUI-1200160743v1
15 JULY 2015
Such an individualized approach has also been emphasized in the context of the
regulation of MTRs in Belgium. In the regulation of MTR of 2006, the Regulator has
applied an analysis that takes into consideration the situation of each operator on the market,
by allowing different termination rates for BASE, Mobistar and Proximus. The Court of
Appeal of Brussels specifically criticized the Regulator for not explaining further the
different situations of Mobistar and Proximus:
“La cour ne peut que s’étonner de la justification avancée selon laquelle Mobistar
doit bénéficier de l’asymétrie parce que Base en bénéficie. En effet, la Décision
n’indique pas comment ni pourquoi Mobistar et Base se trouvaient dans des
situations objectivement comparables au moment de l’analyse ; bien plus l’IBPT
indique dans ses conclusions qu’ils ne l’étaient pas!”38
Similarly, in setting glide paths, the Commission has insisted on the need to assess the
excessive gain of “each” operator.
“TKK did not provide any specific justification for the necessity to allow MNOs to
achieve the cost-oriented level only after more than six years. In particular, TKK did
not give particular reasons showing the necessity to allow the MNOs substantial
gains in order to obtain a more competitive market in the long term, or showing that
a faster reduction would lead to negative effects in competition in the given
circumstances. Given that TKK did not provide any reasons showing that achieving
cost orientation within a substantially shorter time frame would be disproportionate
the proposed glidepath of more than six years appears to be excessively long and
does not seem to be the appropriate measure to tackle the competitive problem
identified. […] the Commission services are of the view that an estimation of the
excessive gains of each operator is useful in order to decide about the
appropriateness of the length of the glide path. In any case, at least establishing the
information about the evolution of overall excessive gains could be a valid tool in
order to appreciate the length of the glide path envisaged.”39 (emphasis added)
It follows from the above that, with regard to the current Draft Decision, the impact
of the regulatory regime should be assessed depending on the beneficiary of the regulation
and take into account the principle of forward-looking economic efficiency. Proximus nor
Mobistar cannot claim entry assistance, as they does not suffer from entry disadvantages.
(continued…)
opérateurs in concreto, il a également violé le principe d’égalité et de non-discrimination garanti par
les articles 10 et 11 de la Constitution”. C. Const., Arrêt n°7/2011 du 27 janvier 2011.
37
“The unfair burden which must be found to exist by the national regulatory authority before any
compensation is paid [for universal service provision] is a burden which, for each undertaking
concerned, is excessive in view of the undertaking’s ability to bear it, account being taken of all the
undertaking’s own characteristics, in particular the quality of its equipment, its economic and
financial situation and its market share”. Case C-222/08, Commission v. Belgium, ECR, 2010, p. I09017, §49.
38
Brussels, 30 June 2009, R.G. 2006/AR/2332 et al., p. 43, § 100.
39
Case AT/2005/0256, 4 November 2005, SG-Greffe (2005 D/206043).
17
EUI-1200160743v1
15 JULY 2015
c)
Promotion of inefficient market entry
Even if the Regulator were to consider a hypothetical new entrant (i.e. not Proximus
or Mobistar), it should not promote inefficient market entry and should verify if there are any
economies of scale associated to the avoidable costs which efficient new entrants cannot
replicate. Here again, reference can be made to the Court of Appeal’s judgment of 7 May
2009:
“[…] Cette absence quasi-totale de données actuelles sur l’état de la concurrence
potentielle au moment de l’adoption de la décision en 2008 n’est pas justifiée par
l’IBPT […]. Il résulte d’un défaut manifeste d’investigation et d’analyse et s’avère
inacceptable au regard du comportement que l’on pouvait raisonnablement attendre
de la part de l’IBPT en sa qualité d’autorité administrative […]. […] [L]’IBPT
devait rechercher, analyser et décrire dans sa décision, les capacités et les plans des
acteurs lui permettant d’établir des prévisions réalistes relatives à leur entrée sur le
marché et d’évaluer la réalisation progressive de leurs plans.
Les motifs de la décision sont donc incomplets et de surcroît, contradictoires sur la
question de l’analyse de la concurrence potentielle.”40 (emphasis added)
Such a thorough investigation of the existence of economies of scale is all the more
necessary as all precedents in which entry assistance was provided concerned cases with rollout obligations or at least extensive network roll-outs for which there were significant
economies of scale.
The Regulator considers that market entry assistance would be justified by Recital 20
of the Access Directive which states that “the method of cost recovery should be appropriate
to the circumstances taking account of the need to promote efficiency and sustainable
competition and maximize consumer benefits”. This is, however, not a valid basis for a
regulatory approach seeking to promote extensive market entry assistance. Market entry
assistance is not mentioned as an objective of the EU regulatory framework. This reference
also ignores that the regulatory framework has been modified in 2009 and that such
modification made clear what needs to be understood by promoting efficiency and
competition. Article 8 (5) of the Framework Directive, which sets out the regulatory
principles that must be pursued by NRAs, states that the regulatory framework seeks to
achieve infrastructure-based competition: “safeguarding competition to the benefit of
consumers and promoting, where appropriate, infrastructure-based competition”.
The Regulator also erroneously refers to the European Commission’s 2009
Recommendation on the Regulatory Treatment of Fixed and Mobile Termination Rates in the
EU to justify its decision to grant the beneficiaries of the wholesale obligations a transitional
period of three years to facilitate entry. The Regulator cites a passage from the
Recommendation where the European Commission considers that “it can be expected to take
three to four years after entry to reach a market share of between 15 and 20%, thereby
approaching the level of minimum efficient scale”. The Regulator omits the following,
crucial, sentence: “This is distinct to the situation for new entrants in fixed markets which
have the opportunity to achieve low unit costs by focusing their networks on high-density
40
Brussels, 7 May 2009, R.G. 2008/AR/787, § 246.
18
EUI-1200160743v1
15 JULY 2015
routes in particular geographic areas and/or by renting relevant network inputs from the
incumbents”. 41 By selectively quoting the Recommendation, the Regulator thus fails to
recognize that the mobile and fixed markets are not comparable in this respect, as stressed by
the Commission. It has in any event, also been demonstrated above that Mobistar has already
benefited from the entry assistance referred to in the recommendation and BIPT itself has
considered that such assistance was no longer justified.
Regulators have also generally opposed favoring inefficient entry as a policy
objective. In its Guidelines on price squeeze, the Regulator itself recognized that a price
squeeze test should be based on the costs of the SMP operator because it would “avoid to
stimulate entry of less efficient operators”. The Regulator’s position regarding Telenet’s
entry as full MVNO in the mobile market is also relevant in this context as the Regulator
refused to allow Telenet to apply asymmetric termination rates. In a Draft Decision
concerning the mobile termination rate of Telenet, the Regulator clearly distinguished an
MVNO from the network operators, stating that:
“[D]e economische efficiëntie van de investeringen wordt geoptimaliseerd door een
virtuele mobielnetwerkoperator aan te sporen om het bewijs te leveren van een
efficiëntie die ten minste vergelijkbaar is met die van zijn host-MNO-operator. Het
toestaan van een stelsel van asymmetrische regulering aan MVNO-operatoren ten
opzichte van de tarieven van de host-operator zou dergelijke operatoren kunnen
afraden om hun eventuele inefficiënties weg te werken, terwijl niets zou kunnen
rechtvaardigen dat een dergelijke speler, die voor het grootste deel van zijn mobiele
netwerk, een beroep doet op de middelen van een reeds goed gevestigde
MNOoperator, economisch minder efficiënt kan zijn dan zijn host.
86 Zoals voordien is uitgelegd, zet een virtuele mobielnetwerkoperator of MVNO
zoals Telenet zelf slechts enkele specifieke elementen van het kernnetwerk in
(schakelaars en databanken) en steunt hij voor de grote meerderheid van de
middelen die nodig zijn voor de levering van mobiele diensten, op de
wholesaleverrichtingen die eraan worden verstrekt door een MNO-operator van een
echt mobiel netwerk (in casu Mobistar in het geval van Telenet). De verrichtingen
die aldus worden verstrekt aan de MVNO door zijn host-operator omvatten het
grootste deel van de kosten die inherent zijn aan de mobiele diensten (in het
bijzonder het radiotoegangsnetwerk, waaronder ook de kosten van de vergunningen
voor de radiofrequenties, alsook de talrijke bijbehorende transmissiemiddelen).
87 In het specifieke geval van Telenet zijn, buiten de vergoeding van zijn hostoperator voor het gebruik van het radionetwerk, de kosten die dit bedrijf oploopt om
mobiele diensten in een kader van full MVNO te verstrekken des te meer beperkt,
aangezien deze operator reeds actief is op het stuk van diensten voor vaste telefonie
en omdat er dus synergieën en besparingen mogelijk zijn, rekening houdende met de
middelen waarover deze operator reeds beschikt voor de levering van deze andere
diensten.
41
Commission Recommendation on the Regulatory Treatment of Fixed and Mobile Termination Rates
in the EU [2009] OJ L 124/67, para. 17.
19
EUI-1200160743v1
15 JULY 2015
88 Het BULRIC-kostenmodel van het bottom-up type dat in 2009 door het BIPT
ontwikkeld is met de assistentie van een externe consultant met het oog op de
regulering van de MTR-tarieven voor mobielgespreksafgifte in België en dat de
grondslag vormt voor het price cap-mechanisme dat vastgesteld is in het
basisbesluit, heeft het mogelijk gemaakt om het relatief beperkte aandeel (ongeveer
15%) naar voren te brengen van de elementen van het kernnetwerk in het totaal van
de kosten die verband houden met een mobiel netwerk.
89 Vanuit economisch standpunt zou daarom niets kunnen rechtvaardigen dat een
virtuele mobielnetwerkoperator een MTR-afgiftetarief mag toepassen dat hoger zou
zijn dan datgene wat door het BIPT reeds wordt opgelegd aan zijn host-operator.
Het MTR-afgiftetarief van Telenet moet dus worden geplafonneerd op het MTRafgiftetarief van Mobistar, zoals dit wordt gereguleerd door het basisbesluit.”
(emphasis added)
The Commission has confirmed in its Article 7 practice that no entry assistance is
needed when the infrastructure used is composed of inputs from established players:
“The Commission recognizes that new entrants to the mobile market may initially
incur higher per-unit costs before having reached the minimum efficient scale,
however, in case of MVNOs due account should be given to the fact that such
operators may have the opportunity to achieve low unit costs irrespective of their
actual market shares by renting relevant network inputs from more established MNOs
thereby benefiting from their economies of scale and/or scope”.42
The methodology further fails to take account of the fact that the wholesale regulation
already integrates a number of measures which are aimed at favoring entry. The distribution
of wholesale costs and the determination of the value of the “retail” revenues considered for
the methodology are all measures which favor the beneficiaries.43 No assessment whatsoever
is made on the effective proportionality of the combination of all these measures seeking to
promote market entry.
RBB Economics has examined the entry assistance considered in the Draft Decision
in more detail and confirms the excessive and unjustified nature of the methodology. This
report is attached as an annex. The main findings can be summarized as follows:
 Adjusting the retail minus methodology would only be acceptable insofar as the
entrant would suffer from diseconomies of scale.
42
Decision of the Commission in Case DK/2009/1013.
43
CRC's previous decision of 11 December 2013 already contained several measures to facilitate new
entry: (1) deviating from the principle of cost causality, which would have commanded that the
wholesale implementation costs be borne entirely by the beneficiaries of the wholesale services, CRC
distributed those implementation costs over all active lines ("aansluitingen"), including those of the
regulated operators, as a result of which the regulated operators end up financing the vast majority of
those costs (CRC Decision of 11 December 2013, section 7.3.2); (2) CRC already ensured that the
beneficiaries of the wholesale services could achieve a minimum level of profitability, by including a
return on sales (RoS) percentage in the calculation of the minus (CRC Decision of 11 December 2013,
section 7.5). In combination with the entry assistance measures contained in the new Draft Decision,
these measures lead to excessive entry assistance which could facilitate inefficient entry.
20
EUI-1200160743v1
15 JULY 2015
 The regulatory regime should only facilitate efficient new entrants which have
similar avoidable costs – a (temporary) additional discount to the wholesale price
due to diseconomies of scale runs contrary to the Regulator’s retail minus
regulation.
 The most probable entrant, Mobistar, is an established mobile operator, which
already has scale advantages and would not suffer from entry disadvantages.
 Giving entrants a significant advantage should depend upon the specific
characteristics of the entrant.
The Regulator does not demonstrate that (i) an additional discount is necessary to
facilitate entrance, or (ii) the absence of such a discount would constitute a real obstacle to
use Telenet’s wholesale offer.
d)
Duration of the transition period
The transition period for market entry is currently not clearly defined, as the Draft
Decision only indicates that “a period of two to three years is appropriate” 44. The Draft
Decision announces that the exact duration will be set by the final decision, which implies
that stakeholders will not have been consulted on this important aspect of the draft measure.
Telenet understands that the transition period will only start as of the date of entry
into force of the final decision. However, the Regulator should take into account that Telenet
has had to comply with an implementation period of six months for the previous tariff
decision of 11 December 2013 so as to ensure that access seekers could already get started
while the Regulator worked out the final details of the regulation. This means that six
following the access request of Mobistar, everything was in place to allow access seekers
(and in particular Mobistar) to launch a commercial offer. The fact that Mobistar has not yet
proceeded to a commercial launch, whereas Telenet has complied with all its obligations
under the regulation also entails an important opportunity cost for Telenet and Mobistar
should not be encouraged for its own delays which cast doubts on its commitment to
effectively launch a retail offer on the basis of cable access. Mobistar’s claim to rely on
cable may only have been a strategic behavior to position itself in the ongoing consolidation
wave. 45 The date of implementation should thus logically be the starting date of the
transitional period.
The duration of any entry assistance should be set based on an adequate motivation
and also have a clearly defined cut-off moment. The duration of the entry assistance should
not be longer than reasonably necessary to allow new entrants to achieve minimum efficient
scale. In the past, when the Regulator imposed a decrease of the mobile termination rate
(MTR), the Court of Appeal has also insisted on the need to ensure that if the Regulator
grants an “inefficient” market entry assistance, it should be duly motivated and restricted in
time so as to reach as soon as possible the efficient tariff level. 46 This reasoning should
44
See § 112 of the Draft Decision.
45
Currently this opportunity cost is estimated to amount to €[confidential] million.
46
Court of Appeal, Brussels, 2006/AR/2332, 2006/AR/2628,
2008/AR/425, 2008/AR/427, 30 June 2009, §95-98.
21
EUI-1200160743v1
2006/AR/2629,
227/AR/3394,
15 JULY 2015
equally be observed here as the regulation is granting an extra promotion margin for new
entrants (at the expense of the cable operator) in order to allow these operators to build-up a
client base.
Also, Telenet would like to remind the Regulator that retail promotions are only
granted to clients during acquisition phase and for a limited time (i.e. for one year
maximum). Even assuming that the application of reduced wholesale tariffs reflecting such
retail promotions would be justified, this should in any case be limited in time, i.e. maximum
one year on a client per client basis.
6.
Transparency measures imposed on SMP-operators
Because of the links which the methodology makes between the retail and wholesale
tariffs, the Draft Decision also foresees an obligation for the SMP-operator to communicate
the future evolutions of its commercial conditions regarding the bundles including TV. 47
This communication is a cumbersome process which affects the flexibility and swiftness of
Telenet’s ability to respond on a commercial basis. The scope of the prior notification
obligation also lacks precision [confidential].
It should also be noted that because of the methodological links between retail and
wholesale pricing, there is a risk that such notifications (which are of course highly sensitive
from a competition law point of view) may in some form also be communicated to the
beneficiaries of the regulation which would raise competitive concerns and also affects the
ability for Telenet to innovate.
7.
No impact assessment
The Regulator does not make an overall assessment of the impact of the further price
reduction weighing “benefits” against “costs”. Although Telenet realizes that this analysis
cannot be done in a mathematical fashion, it has been demonstrated above that the regulation
will have far-reaching effects as it drastically further reduces the level of the wholesale rates
and also relies on a number of assumptions which are solely targeted at promoting entry.
The tables below shows how significant the wholesale reduction is compared the prevailing
reference retail tariffs for WHOP and WHOPPA. Imposing such wholesale conditions can
lead to a complete overhaul of the market and should therefore, in any event, be introduced
with the necessary accompanying measures to ensure the stability of the SMP-operators.
47
Draft Decision, § 135 (“135. Daartoe moeten Brutélé, Numericable en Telenet aan het BIPT hun
voornemen meedelen om extra diensten of voordelen op te nemen in hun retailaanbiedingen. Deze
kennisgeving zal:
- een beschrijving van de dienst, een voorstel voor waardebepaling, de retailtarieven van de
aanbiedingen waarin deze moet worden opgenomen, bevatten, alsook de wholesaletarieven die daaruit
voortvloeien;
- moeten vergezeld gaan van alle nuttige informatie om de marktwaarde van de betrokken dienst of het
betrokken voordeel te beoordelen (in het bijzonder, verwijzingen naar de tarieven die worden
gehanteerd door andere operatoren en de productiekosten);
- moeten plaatsvinden binnen een termijn van twee maanden voordat de diensten of voordelen in
kwestie beschikbaar zijn voor de operatoren.”).
22
EUI-1200160743v1
15 JULY 2015
It has also been demonstrated above that the way in which the regulation is conceived
creates links between retail and wholesale pricing which are going to inherently affect the
commercial practice of cable operators in their marketing and innovation strategy, including
in a number of service segments for which they have no dominant position whatsoever.
Regulators should assess whether the extreme scenario which results from the
methodology is necessary for the short-term objective of market entry and consumer choice
23
EUI-1200160743v1
15 JULY 2015
which the regulation seeks to achieve, or whether a smaller reduction and a more
refined/segmented approach (e.g., carving out Proximus and Mobistar) could not achieve said
objectives.
 The reference price should be based on the retail offers that most closely correspond
to the wholesale offer (i.e., in principle, dual play bundles where available).
 The additional services should be included in the minus as avoided costs. This is not
the “third-best” approach but the approach that is most consistent with the retail
minus methodology and the objectives of the regulatory framework. If retail values
are retained, a correction based on usage should be applied.
 The application of the Regulator’s methodology to legacy offers could result in a
disproportionate impact or even wholesale prices below cost.
 The methodological adjustment with respect to promotions is not duly motivated,
there is no logical link with the “dilution effect” (which is already being corrected by
taking into account only gross adds during a transition period).
 The Draft Decision provides excessive entry assistance to the beneficiaries of the
wholesale access obligations, without taking into account the fact that the only
operators that have requested access so far are already established operators. At the
very least, entry assistance measures should be individualized, taking into account
the actual scale disadvantage (if any) of each beneficiary.
 The obligation of prior notification stifles price competition and innovation at the
retail level.
 The Regulator fails to make a proper assessment of the cumulative impact of the
adjustments contained in the Draft Decision.
24
EUI-1200160743v1
15 JULY 2015
B.
Violation of the principle of legal certainty and non-retroactivity
The principles of non-retroactivity and legal certainty are well-established principles
of administrative law. The Regulator’s methodology runs counter to the principle of legal
certainty which implies that the regulation should allow legal subjects to reasonably foresee
the effects of their actions in advance and to confide in the reliability of governance.48
The Court of Appeal of Brussels recognized these principles in its case law.49 In its
judgment of 7 May 2009, the Court of Appeal referred to the restrictive case law of the
Council of State stating that:
“L’autorité administrative, dont une décision individuelle a été annulée peut, dans
certains cas, remédier au vide créé par le caractère rétroactif de l’annulation
lorsqu’elle adopte une nouvelle décision à laquelle elle peut légalement conférer un
effet rétroactif. Cependant, la jurisprudence du Conseil d’Etat n’admet qu’il soit
dérogé au principe général de la non-rétroactivité des actes administratifs que dans
certains cas, qui sont fonction notamment de la nature ou de l’objet de l’acte à
refaire ou de la compétence mise en œuvre.”50
The European Commission has also criticized retroactive measures proposed by
NRAs. In Case PT/2010/1077, the Commission stated for example that:
“Retroactive effect of the proposed price levels
Whilst the Commission notes that provisions of the Portuguese Telecommunications
Law explicitly allow for the imposition of retroactive remedies, it invites ANACOM to
consider whether an imposition of prices with retroactive effect might lead to legal
uncertainty for market players. Against this background, ANACOM is invited to
ensure that the proposed prices, applicable with retroactive effect, do not impinge on
legal certainty for operators currently providing services on the basis of previously
imposed regulation.”51
In the Draft Decision, the Regulator is foreseeing a tariff regulation which would be
contrary to these general principles insofar the SMP operators’ wholesale pricing could
potentially be reviewed with retroactive effect52 depending on:
48
M. Van Damme, A. Wirtgen, “Het rechtszekerheids- en vertrouwensbeginsel” in I. Opdebeek a.o.,
Beginselen van Behoorlijk Bestuur, Bruges, die Keure, 2006, 315; RvS 22 February 2010, nr. 201.138;
RvS 21 February 2011, nr. 211.392.
49
See for example the reference made to the principle of non-retroactivity in the appeal regarding VAS
numbers. “Dès lors que l’IBPT n’a pas fait rétroagir les adaptations requises par lui, à l’offre
tarifaire de Belgacom par la décision attaquée en ce qui concerne les numéros VAS, Belgacom ne
pourrait se plaindre d’une atteinte au principe de non-rétroactivité et/ou d’une atteinte à la sécurité
juridique qui serait la conséquence de son intervention en dehors du délai prévu par l’article 16 du
même arrêté royal.” Court of Appeal of Brussels, R.G. 2004/AR/1777, 16 June 2006, § 33, al. 4.
50
Brussels, 7 May 2009, 2008/AR/787.
51
Decision of 25 April 2010, Case PT/2010/1077.
52
Draft Decision, § 135-137.
25
EUI-1200160743v1
15 JULY 2015
(i) the potential future inclusion of other additional services included in the bundle
(currently excluded from the calculation);
(ii) the potential evolution of changes in pricing (including pricing changes made by
third parties) which have been retained in the calculation.
When retail prices of other operators are used to determine the market value of the
additional services, this creates a situation whereby Telenet’s wholesale prices become
dependent upon retail price decisions of other operators, which may even be potential buyers
of the wholesale service (e.g., Proximus). The situation is further exacerbated by the fact that
the Regulator reserves the right to retrospectively adjust prices when additional services are
added (Draft Decision, § 136) or currently excluded services would all of a sudden be
considered “relevant”.
Such an uncertainty about future wholesale prices and expected wholesale revenues
creates an environment in which informed commercial and investment decisions are rendered
impossible. This therefore creates a risk which hampers efficient investment and innovation,
disregarding the dynamics of the digital economy, ultimately leading to stagnation, and runs
counter to the objectives of the regulatory framework.53
Moreover, the retroactive corrections provided in the Draft Decision (§ 135-136) are
constructed in a way that negatively impacts the SMP operators. The retroactive effect on
Telenet’s wholesale prices only seems to benefit the beneficiaries of the regulation. After a
freeze of one year (Draft Decision, § 133), the prices can only be adjust downward, not
upward which constitutes a discriminatory assessment given that the Draft Decision
acknowledges the need for pricing stability a concern which applies both the beneficiaries
and the SMP-operators.
Telenet further refers to its previous observation according to which the proposed
pricing modification could only be implemented within four months because of IT
requirements and that any future pricing modifications should also need to take account of
potential IT constraints.
It will also be demonstrated in Section III.D below that the formula relied upon for
the calculation of the wholesale prices for the broadband profiles is unreliable and is
excessively sensitive to changes making unfit for purpose and contrary to the above
principles of legal certainty and predictability.
During the Q&A session of 15 June 2015, BIPT confirmed that the “normal
principles” of a retail minus methodology would continue to apply, in the sense that every
change to the retail tariffs would trigger an adaptation of the corresponding wholesale tariffs.
With the retained methodology and the potential retroactive adaptations this basically means
that each attempt of Telenet change the retail price of its bundled offers will (retroactively)
benefit Mobistar in the form of a wholesale tariff reduction, even if the discounted pricing
concerns an unregulated component of the bundle. This constitutes a major competitive
handicap for the regulated operators. It also means that the more competitive the market gets,
53
Framework Directive, Article 8 (5), d).
26
EUI-1200160743v1
15 JULY 2015
the more the wholesale tariff will decrease, creating further disincentives for Mobistar to roll
out its own network.
 By reserving the right to retroactively adjust wholesale prices when additional
prices are added or currently excluded services are brought into scope, the
Regulator creates an environment of legal uncertainty, hampering efficient
investment and innovation.
 The retroactivity is a one-way street, since wholesale prices can only be adjusted
downward, which is discriminatory.
III.
FACTUAL ARGUMENTS
The Draft Decision’s benchmarking methodology (set out in § 37-81) is subjective
and arbitrary.
The value of additional services included in a bundle are (i) the standalone prices of
Telenet’s services, or (ii) the standalone prices of services of other domestic and foreign
operators. As indicated above, production costs are only considered a third best alternative.
The Regulator’s benchmarking methodology looks at operators which, in exceptional
circumstances and for commercial reasons which do not accord with those of Telenet, offer
an additional service on a standalone basis.
The Regulators also ignores that the additional services considered have a relatively
low penetration rate compared to the core bundle services and in particular when offered on a
standalone basis. However, the same high tariffs are now being applied to Telenet’s services,
under the assumption that 100% of Telenet’s installed customer base would be willing to pay
such tariffs, whereas Telenet can provide concrete figures showing that the additional
services have a very low adoption rate even when offered for free as part of a bundle (e.g.,
[confidential]% adoption rate for webhosting services). These usage rates are of course
relevant when setting the pricing for bundles.
The detailed critique set out below will demonstrate that the Regulator’s analysis
factually fails for the following reasons:
 The methodology contains elements of double-counting as it deducts certain
elements twice leading to an artificially low wholesale value;
 The benchmarking analysis for the individual additional services is unfounded;
 The methodology relied upon to deduct modems is unfounded;
 The methodology relied upon to assess the wholesale pricing for broadband
profiles is unfounded.
27
EUI-1200160743v1
15 JULY 2015
A.
Double-counting
Telenet has identified three problems of double counting, i.e. errors whereby certain
costs are being deducted twice from the retail tariff to calculate the relevant wholesale tariff.
1.
Double counting of avoidable costs
The avoidable costs (i.e. the costs which the cable operator avoids by providing a
wholesale service instead of a retail service and should be deducted from the retail price to
determine the minus) are not necessarily specific to each considered retail service. The
standalone price for the additional services will generally also cover these avoidable costs.
Conversely, when additional services are offered as part of a bundle, common retail costs are
only incurred once.
The proposed methodology ignores this which leads to a double
deduction of avoidable costs: (i) once as a part of the standalone price of the additional
services which is deducted from the reference price; and (ii) subsequently as part of the
“avoidable costs” included in the minus.
The bundle discount which is taken into account as a correction factor to assess the
relevant value of the service in the bundle does not fully or necessarily address this double
counting issue as the bundle discount may also be driven by other considerations than costsavings.
2.
Double-counting of copyrights
It is not clear from the Draft Decision (§ 60-64) whether the additional services used
by the Regulator in the benchmark include copyright costs. For instance, the benchmarks
used to determine the value of the 2nd screen service appear to also include copyright costs.
By deducting from the reference retail price values that are based on benchmark prices which
include copyright costs, the Regulator deducts copyright costs twice (as it also subtracts
copyright costs separately from the reference price).
3.
Double-counting of the modem
The Regulator makes a double-counting error by deducting from the hypothetical
reference retail price both a value for the Wi-Free service and a value for the modem, while
Wi-Free is in fact a functionality of the modem.
B.
Erroneous assessment of individual services
The Regulator establishes a reference price for the following services:
 Wi-Fi access (homespots and hotspots);
 2nd screen;
 E-mail;
 Webhosting for personal web pages;
 Content (Sporting Telenet);
28
EUI-1200160743v1
15 JULY 2015
 Fixed telephony.
As a general preliminary remark, Telenet notes that the methodology followed by the
Regulator gives a lot of weight to additional services such as e-mail and webhosting in the
determination of the wholesale prices, even though BIPT has removed these services from
the retail price comparator tool bestetarief.be / meilleurtarif.be because they were deemed to
lack relevance.54 BIPT is thus on record acknowledging that these additional services are of
little importance for end users. This confirms that these services are overvalued in the
calculation of the wholesale tariffs.
1.
Wi-Fi access
The benchmark estimates the value of the Wi-Fi access offered via homespots (“WiFree”) and hotspots at € 2. This is based on the combination of the Telenet retail tariff for the
hotspot service (€ 1/month) and a benchmark for the homespot service offered by third
parties (€ 1/month).
Wi-Free uses the Telenet users’ modem network to create a service that is available
for all Telenet broadband customers. Wi-Free is therefore already included in the current
model as part of the avoidable costs related to the modem (in the old methodology) or as an
additional value for the modem to be deducted from the retail base (under the new
methodology). Adding Wi-Free to the additional services implies a double-counting, as the
same element is subtracted twice. As Wi-Free is a functionality of the modem, it is not
possible to detach it and value it as a standalone service.
It should be stressed that Wi-Free is not really a “free” service, since the related data
usage is counted and imputed to the customers’ monthly download allowance. The Wi-Free
service is in fact a form of network sharing, whereby Telenet’s broadband customers can
make use of the wireless modem of other Telenet customers on a reciprocal basis. As the
network only exists due to participation of individual broadband customers who share their
installed modem, Wi-Free connectivity is dependent on the amount of homespots that are
connected to the network at a certain point in time and at a certain location. As a
consequence, Wi-Free provides only fragmented and non-homogenous connectivity. This
makes Wi-Free a complementary service for its users, rather than a standalone wireless
broadband service. Beneficiaries of the regulation, such as Mobistar, will dispose of the
same network bandwidth as Telenet. and will therefore be able to offer the same homespot
service to their customers as Telenet without incurring an extra costs to that effect. Mobistar
will in fact have the ability to offer a much more performing mobile broadband connectivity
on the basis of its 3G/4G network.
The geographic markets used in the benchmark and their corresponding geographic
hotspot and homespot/modem network coverage differ considerably from the service that
Telenet provides. For instance, FON, which is also offered in Belgium by Proximus, provides
worldwide coverage, integrating more than 16 million hotspots.
54
BIPT, Consultatie betreffende aanpassingen aan de tariefsimulator, 10 oktober 2014, p. 3: “Schrapping
van de vragen in verband met de gewenste functionaliteiten: de vragen die bitter weinig relevantie
hadden inzake de e-mailadressen, antivirusprogramma's, antispamprogramma's en de persoonlijke
webruimte zullen worden geschrapt om de gebruikersinterface te verlichten”.
29
EUI-1200160743v1
15 JULY 2015
The Regulator’s methodology also presupposes an adoption rate of Wi-Free of 100%.
This does not correspond to the market reality as Telenet’s Wi-Free only has an adoption rate
of [confidential]% for the homespot and [confidential]% for hotspots customers (as regards
login rate), which amounts to [confidential] customers of which [confidential] also use the
Telenet hotspots. This amount is in fact not even a good proxy for calculating the value of
the service as it does not accurately reflect the users that would actually be willing to pay € 2
for Telenet’s Wi-Free service. Therefore, the adoption rate of the service would be even
lower if Wi-Free would only be offered as a standalone paying service.
2.
2nd screen
Telenet offers a 2nd screen service under the name “Yelo TV”. This service is
always included in a bundle. The Regulator’s benchmark analysis (Draft Decision, § 60-64)
values a standalone 2nd screen service at € 5. Yelo TV is a service that users can only use at
home, via their home Wi-Fi network. It allows them to watch a limited TV content offer on a
tablet, a PC, or other Wi-Fi-enabled device.
As for Wi-Free, Yelo TV is not really a “free” service: the related data usage is
counted and imputed to the customers’ monthly download allowance. The Draft Decision
fails to distinguish this service from other types of Yelo TV (Yelo Play, Yelo Play More),
which are richer in terms of features. In the Draft Decision, 2nd screen service is analyzed as
a generic service irrespective of its specific features and geographic coverage.
The services which the Regulator uses as a benchmark for Yelo TV are not
comparable:
 In Belgium, Proximus’ “TV Overal” (€ 4,99) is only a paying service for
Proximus’ legacy packs. It is currently included for free in all Proximus “packs”
with TV and internet. 55 Using “TV Overal” as a benchmark is completely
arbitrary and damaging for Telenet, because if Proximus would decide to increase
the retail standalone price of its TV Overal offer, this would have a direct
negative effect on the wholesale tariff applicable to Telenet. Moreover, “TV
Overal” differs substantially from Yelo TV as it can not only be used via the
customer’s home Wi-Fi network, but also be used on Proximus’ mobile 3G and
4G network. This has an impact on the copyright costs, as there is an additional
charge for “out-of-home” rights. Moreover, TV Overal also includes 1 GB of
mobile data, which should also be taken into account in the valuation of this
service.
 In Belgium, Numericable’s “ON Multiscreen” (€ 5) was launched only recently.
As the number of customers using this service is likely to be marginal, this service
is not an appropriate benchmark. Moreover, this service is not offered on a
standalone basis but is only available to triple play customers. The pricing
structure of Numericable’s triple play bundles and additional services also differs
significantly from Telenet’s and therefore does not offer a comparable basis. ON
55
All of Proximus’ current “Start”, “Comfort” and “Maxi” packs include “TV Overal” without extra
charge (“Gratis TV Overal”).
For Proximus’ product information fiches see
https://www.proximus.be/nl/id_cr_infofich/particulieren/onze-producten/r-orphans/product-informatiefiches.html.
30
EUI-1200160743v1
15 JULY 2015
Multiscreen is only included for free in Numericable’s “Triple Play Max” bundle,
and is a paying service for “Triple Play Start” and “Triple Play Extra” subscribers.
 In Germany, Deutsche Telekom’s standalone “Entertain to go” service (€ 6,95)
includes a much richer offer than Yelo TV that can also be used out of home (e.g.,
40 channels, 10 hours of cloud recording, access to sport channels, shift to TV,
Mobile options, etc.). The inclusion of cloud recording functionality has an
impact on the related copyright costs. As Deutsche Telekom offers add-on cloud
recording at €4,95 for 20 additional hours of recording, the included 10 hours of
cloud recording alone could be valued at €2,475.56
 In Luxemburg, “Zattoo” (€ 2,99) is not a product by Post Luxemburg, but an a
service with a coverage in six countries. Besides in Luxemburg, Zattoo users can
use this service in five other countries, which results in additional copyright
costs.57 This is the only real standalone product in the benchmark that is not used
as an upselling incentive. However, contrary to Yelo, “Zattoo” can also be used
over mobile data connections, which as noted above also has an impact on the
copyright costs.
 In Poland, operator UPC provides a 2nd screen service which the Regulator has
ignored in its benchmarking exercise. “TV Horizon GO” is offered as an
additional service by for € 1,2058, which is far below the Regulator’s 2nd screen
price of € 5.
It follows from the above that the services that are being used in the benchmark differ
considerably from a service and costs perspective with Yelo TV.
Moreover, it is inappropriate to include offers from potential access seekers (such as
Proximus and Numericable) in the benchmark, let alone to base the value of the 2 nd screen
service on the pricing of such potential access seekers, given the distortive effect on
competition of such an approach.The limited significance of Telenet’s 2nd screen in the
bundle is evidenced by its low adoption rate (only [confidential] customers used the Basic
Yelo functionality during the month of January 2015 out of a total broadband customer base
of 1.544.000, which amounts to an adoption rate of only [confidential]%). The amount of
customers that uses the service is even an overestimation as it does not accurately reflect the
amount of users that would actually be willing to pay € 5 for Telenet’s Yelo service assuming
it would be charged on a separate basis.
Because of the above, Telenet considers that it would be more accurate and
representative to take into account the actual costs related to the provision of the additional
services. Taking into account the amount of customers that can use the basic functionalities
of Yelo TV, the actual value of the service amounts to €[confidential] per customer per year,
or €[confidential] per customer per month.
56
For
more
information,
go.html#esr=&stagetab=1.
57
Zattoo Live TV is available in Germany, Switzerland, Spain, Denmark, Luxembourg, and the UK, see
http://zattoo.com/int/.
58
For more information, please refer to http://www.upc.pl/pdf/upc_cennik_uslug_dodatkowych_dtv.pdf.
see
http://www.entertain.de/startseite/funktionen/entertain-to-
31
EUI-1200160743v1
15 JULY 2015
3.
E-mail
The Regulator’s benchmark analysis (Draft Decision, § 65-68) attributes a value of €
0,05 per e-mail address, based on the price charged by Telenet for additional e-mail
addresses outside of the bundle (i.e. in addition to the 10 e-mail addresses already included in
all of Telenet’s offers with broadband). For 10 mail boxes included in the Telenet bundle, a
value of € 0,50 is retained.
This approach grossly overestimates the value of this service and is arbitrary. The
Regulator does not explain why Telenet’s retail price outside of the bundle is an appropriate
benchmark, while Numericable’s retail price outside of the bundle (€ 7,44/month) is not.
The Numericable example shows that basing the value of the mail boxes included in the
bundle on the price outside of the bundle is absurd and invites the Regulator to request the
usage made of Numericable’s stand-alone email box offering.
The Regulator recognizes that its benchmark confirms that e-mail services are
“usually” either included in an internet subscription or not offered to the customers at all.
The Regulator could not provide a single example where this service was offered only on a
standalone basis for an additional charge. The Regulator also fails to present any evidence
that it has investigated the actual cost and profit margin in relation to the e-mail services.
Customers are not willing to pay for e-mail services, which is evidenced by the high
adoption rates of free e-mail service providers, such as Google or Yahoo and the low
adoption rate of Telenet’s own/additional e-mail services. On average, less than
[confidential] e-mail addresses per Telenet customer are actually used, i.e. approximately
[confidential] million e-mail addresses for a total broadband customer base of about 1,5
million. Taking into account this actual usage rate, Telenet has provisioned only
[confidential] million mail boxes ([confidential]%). The Regulator’s valuation is therefore
totally unrealistic because it assumes an adoption rate of 100% (i.e. all 1,5 million broadband
customers of Telenet using all 10 included e-mail addresses, or 15 million mail boxes in
total).
This adoption rate of the service would also be much lower if the e-mail service
would not be included in the bundle. There are internet service providers (e.g., Tiscali 59 and
Libero Infostrada60 in Italy) that offer their e-mail services for free also to non-customers.
The fact that Telenet uses its e-mail services as means to advertize, as the Regulator points
out, shows that Telenet’s service model can also be compared to that of free service providers
such as Google or Yahoo. In any case, this cannot justify attributing a high retail value to the
e-mail service, since this is an advantage for the provider and not for the customer (who pays
the free email providers in “data currency”).
Moreover, e-mail services that are offered on the market are priced much lower than
the Regulator’s benchmark. For example, internet service provider “OneCom” provides
unlimited e-mail accounts for € 1,50/month. This package also includes many other services
such as advanced calendar and contact functions, web hosting, web editor, traffic statistics, a
domain name, etc. As explained below, many web hosting providers include a large number
59
For more information, please refer to https://mail.tiscali.it/.
60
For more information, please refer to https://registrazione.libero.it/.
32
EUI-1200160743v1
15 JULY 2015
or unlimited e-mail addresses in web hosting subscriptions that cost far less per MB of web
space than the retail price retained by the Regulator (see point 4 hereafter).
Because of the above, Telenet considers that it would be more accurate and
representative to take into account the actual costs related to the provision of the e-mail
addresses. The production cost per e-mail box for Telenet amounts to €[confidential] per
month. 61
4.
Web hosting
With regard to web hosting for personal web pages, the Regulator’s benchmark
analysis (Draft Decision, § 69-70) retains a retail value of € 1 for 50 MB of web space, i.e. 2
cent per MB per month.
The Regulator’s tariff is based on Telenet’s retail price of € 2,05 to extend the
standard 50 MB of web space, which is included in the broadband subscription or bundle,
with 100 Mb.
Webhosting services that are offered on the market on a standalone basis are much
cheaper per MB than the Regulator’s benchmark. For example, Versio offers 1.5 GB (1500
MB) of storage and unlimited e-mail addresses for € 0,5 per month, and 7,5 GB (7500 MB)
of storage with unlimited e-mail addresses for € 2,25 per month.62 Both offers work out to
0,03 cent per MB per month. Another Belgian provider, Easyhost, offers 5 GB (5000 MB) of
storage with unlimited e-mail addresses for € 3,99 per month (which works out to 0,08 cent
per MB per month). 63 Moreover, service provider One.com, which is mentioned as an
example in the Draft Decision (§ 67), offers 15 GB of storage with unlimited e-mail
addresses for € 1,5/month, which is equivalent to € 0.005 for 50 MB per month, or 0,01 cent
per MB per month. 64 This package also includes many other services such as advanced
calendar and contact functions, web hosting, web editor, traffic statistics, a domain name, etc.
The Regulator fails to take into account the low adoption rate of the web hosting
services. The adoption rate of Telenet’s webhosting service is only 8%, with only 124.000
customers using the free web space out of an broadband customer base of 1,5 million.
In view of the above, Telenet considers that it would be more accurate and
representative to take into account the actual costs related to the provision of additional
services. The production costs for webspace for Telenet amounts to € [confidential] per
month per user. This amount covers total CAPEX investments of € [confidential]. These
investments relate to the last 3 years and are amortized over a period of 3 years.
61
This amount includes (i) the yearly maintenance costs for the platform (€[confidential] per mailbox
per month) and (ii) the one-time license costs (amortized over a period of 3 years, which amounts to
€[confidential] per month per mailbox).When determining the production cost, the amount of
provisioned e-mailboxes is taken into account, being [confidential] million, regardless of whether they
are effectively used or not.
62
For more information, please refer to http://www.versio.be/webhosting.
63
For more information, please refer to http://www.easyhost.be/en/shared-hosting/.
64
For more information, please refer to http://www.one.com/be.
33
EUI-1200160743v1
15 JULY 2015
5.
Premium content
With regard to Telenet’s premium content offer “Sporting Telenet”, the Draft
Decision (§ 71-77) notes that this is a paying service for all customers, but that subscribers to
“Internet en TV”, “Whop” and “Whoppa” receive a discount. For this component, the
Regulator took into account usage, which it ignored for the other “additional services”. This
confirms the artificial approach for the other services.
Telenet notes that the benchmarking for other cable operators such as Numericable
seem to be inconsistent and arbitrary. For Numericable, for example, the Regulator considers
the increment price charged to end users with an existing content package to take an
additional content package. No justification is provided for this approach. It shows that the
comparison between the bundle components cannot be made and confirms that the
objectivity which the Regulator has sought to achieve through the benchmarking analysis is a
purely facial exercise.
Premium content is not a regulated product, but is nevertheless taken into account in
the calculation of the wholesale price. This will have a negative impact on retail price
competition and product innovation.
Moreover, if this precedent is also followed for other non-regulated components that
may be added to the bundle in the future, this will eventually deprive the regulated operators
of their retail pricing freedom.
6.
Fixed telephony
Before entering into detail on other elements regarding the Fixed Telephony component,
Telenet would like to refer to the meeting held on 15 June 2015. During the meeting Telenet
mentioned that the installation costs that were reported during the update exercise did not
take into account the fact that part of the installation costs should also be allocated to fixed
telephony. In its current product portfolio Telenet only offers standalone products (TV,
Internet and Fixed Telephony) or “triple play” offers (TV, Internet and Fixed Telephony
combined with a bundle discount). As Telenet does not have a “dual play” offer, the
Regulator took into account a combination of standalone products TV and Internet and the
triple play offers to cover all broadband profiles that Telenet currently offers to clients. In
order to determine the wholesale price of dual play wholesale offer including one of the
broadband profiles that Telenet offers as part of its triple play bundles, the Regulator had to
construct reference retail prices excluding the fixed voice telephony component of those
triple play bundles. The wholesale price is then determined by applying the minus to these
fictive reference retail prices (excl. VAT).
The Draft Decision does not explicitly deal with the way in which the fixed voice
component of the bundles is deducted from the reference retail price. However, it appears
from Analysys Mason’s “Rekenmodule”, which is attached to the Draft Decision, that the
approach which the Regulator followed to construct a reference retail price without the fixed
voice component is as follows. As a starting point, the Regulator took the retail price excl.
VAT of Telenet’s standalone fixed telephony offer, FreePhone Europe, i.e. € 17,93. Then,
34
EUI-1200160743v1
15 JULY 2015
the Regulator applied the inflated bundle discount65 to this standalone price (e.g. for “Whop”,
€ 17,93 – 40% = € 10,7566) and deducted the resulting value from the retail price of the triple
play offer. In other words, the Regulator followed essentially the same approach for fixed
telephony as for the “additional services” or “value-added services”.
The Regulator’s methodology based on the standalone retail price of fixed telephony
leads to a gross overestimation of the economic value of the fixed telephony component of
the bundle.
First, the Regulator disregards the fact that the standalone retail price of the fixed
voice subscription includes an access line component. In the triple-play bundles, this is a
common component to all three services. It is incorrect to impute the access line component
entirely to the fixed voice service. This leads to an excessive deduction from the retail price
and means that this cost should not be recovered at wholesale level from the access seekers.
Second, the Regulator does not take into account the limited commercial relevance of
the fixed voice component as compared to the other components in an age of universal
mobile phone ownership, low mobile rates and the availability of OTT voice telephony
services (e.g., Skype). The Regulator did not take into account (i) the low standalone
telephony subscriber base on cable; i.e. [confidential] stand-alone fixed telephony
subscribers, (ii) the low overall voice telephony usage among 3P customers and (iii) the fact
that nearly [confidential]% of Telenet’s 3P customers have no inbound or outbound calls at
all. In this respect Telenet can confirm that the use of fixed telephony has dropped with
approximately [confidential]% since 2011 (from an average of more than [confidential]
minutes per month in 2011 to less than [confidential] minutes per month in 2015), and this
irrespective of the additional functionalities and value that Telenet has brought to its fixed
telephony offer (unlimited calling fixed numbers in Belgium, 2000 free minutes for mobile
numbers or calls within Europe, etc).
[confidential chart: call minutes to domestic geographical numbers per subscriber per
month]
This also shows that there is hardly any price elasticity left for this product, meaning
that even if Telenet would lower its prices for fixed telephony, it is expected that the usage of
the service will not increase. When benchmarking this average usage of the Telenet
Freephone Europe with other OTT providers such as Skype, it appears that the value of the
Telenet fixed telephony service can be compared with a offer of 3,50 EUR (incl. VAT) of
Skype.67 In other words, by focusing on the standalone retail price of fixed telephony to
65
I.e., the difference in percent between the sum of the prices excl. VAT (or benchmark values) of the
standalone components of the bundle (TV, internet, 2nd screen, etc.) on the one hand, and the bundle
price excl. VAT on the other hand.
66
As the bundle discount is not the same for “Whop” (-40%) as for “Whoppa” (-42%), the Regulator’s
methodology also leads to the illogical result that exactly the same service (FreePhone Europe) is
valued differently depending on the broadband profile it is combined with (€ 10,75 as part of “Whop”
and € 10,38 as part of “Whoppa”).
67
[confidential] the use can be compared with the Skype offer of “120 minutes mobile + fixed numbers”
of €3,50 per month, incl. VAT.
35
EUI-1200160743v1
15 JULY 2015
assess the value of the “voice component” in the bundle, the Regulator overblows its value
leading to an artificial reduction of the wholesale price.
The Framework Decision did not consider that the regulation of the voice component
of the bundle was necessary (but limited itself to the resale of broadband). The Court of
Appeal retained this analysis, insisting, however, on the fact that the regulation of such
connected services was subject to strict conditions given the principle of proportionality and
the fact that only those services for which SMP is found can be regulated68. This concern
confirms the excessive nature of the proposed methodology which artificially reduces the
value of wholesale prices because the SMP-operator includes voice telephony in the bundle.
This is precisely the concern expressed in Commission’s Revised Recommendation on
Relevant Markets regarding potential regulatory spill-over in assessing the replicability of
bundles:
“What is important in this respect is that NRAs are able to ensure that the vertically
integrated SMP operator's regulated elements of the bundle can be effectively
replicated (in terms of both technical and economic replicability) at the retail level,
without an implicit extension of regulation to other components which are available
under competitive conditions. Moreover, it has been argued that, in cases of the
provision of the fixed voice service with broadband access and/or IPTV, bundling at
the retail level is rather a phenomenon of continued provision of a declining fixed
voice service alongside broadband access and/or IPTV, rather than an economically
significant offer that alters the competitive dynamics over a longer period.”69
The regulation seeks to enable alternative operators relying on the cable operators’
wholesale inputs to compete effectively on the retail market. To achieve this, it is
proportionate and sufficient to take into account the production cost of the fixed voice service
as an avoided cost. Telenet estimates its own production cost for the fixed telephony
component of its bundles at € [confidential].70 This is enough to enable an equally efficient
operator to add its own fixed telephony service to the bundle using VoB/VoIP technology.
Telenet recalls that it has no SMP for fixed telephony and that its fixed telephony
pricing is not regulated. There is therefore no objective reason why its pricing for fixed
telephony should result in a reduction of its wholesale margin for television and broadband
services. This is all the more so in a context where Proximus, the incumbent provider of
fixed telephony services, does not have a similar constraint on its retail fixed telephony
pricing (since Proximus’ wholesale margin on its BRUO and BROBA offers is wholly
independent from its retail pricing for standalone fixed telephony) and, in addition, is a
potential beneficiary of Telenet’s wholesale services (or at least claims to be so).
It is important to keep in mind that the only two operators who have requested access
to Telenet’s wholesale services – Mobistar and Proximus – do not suffer from any significant
competitive handicap compared to Telenet with respect to telephony services. This is
68
Brussels Court of Appeal, 2011/AR/2289, 14 November 2014, § 155-156.
69
COM(2015) 192 final, p. 18.
70
This includes the share of telephony in the cost of the modem (assumption: 50%), interconnection costs
and the cost of the network proxy.
36
EUI-1200160743v1
15 JULY 2015
obvious in the case of Proximus, which benefited from monopoly rights and has an inherited
customer base for these services. Mobistar too can not only bundle its television and
broadband services (provided on the basis of the cable operators’ wholesale inputs) with
mobile services that it provides using its own network infrastructure, but it is also able to
implement a VoB/VoIP solution on the back of its broadband access at limited cost.
Mobistar can also play the fixed-mobile convergence card by providing users with an app on
their mobile phones allowing them to make VoIP calls over their home Wi-Fi network (cf.
Telenet’s own Triiing app) and by implementing EAP-SIM technology (allowing users to
switch seamlessly between mobile and fixed networks). None of this requires high
investments. It should also be noted that Mobistar reports an existing customer base for fixed
telephony of over 200.000 subscribers.71
C.
Modem
In accordance with the CRC Decision of 1 July 2011, the modem is not part of the
regulated wholesale offer and the beneficiaries of the wholesale remedies have to supply
their own modems.
Telenet does not charge its broadband customers for the modem, regardless of
whether the broadband service is purchased as part of a (triple play) bundle or not. In the
CRC Decision of 11 December 2013, the costs related to the modems and set-top-boxes
(STB) were factored into the minus as avoidable costs, whereas the revenues related to the
modems and STB were factored into the ARPU. 72 This approach was based on the
assumption that the modem-related costs were implicitly covered by subscription revenues.
In the Draft Decision, the Regulator abandons this approach. It excludes the modem
and STB from the calculation of the minus, both on the costs side and on the revenues side,
but deducts the “value” of the modem and STB from the retail reference price to which the
minus is applied.
As none of the three regulated operators charges extra for the modem and there is no
observable retail price for the modem, the Regulator uses a benchmark to determine the
“value” of the modem. Telenet understands that the Regulator has determined the monthly
“rental value” of the modem by multiplying the price Telenet charges to customers who
request a new modem (€ 75) with the ratio between the monthly rental price of the STB (€
8,7) and the purchase price of the STB (€ 249), expressed as a fraction. This method of
determining the rental value of the modem is arbitrary.
First, the Regulator mentions four different possible benchmarks (Kabel Deutschland,
Orange France,73 Proximus and the above-mentioned benchmark), without explaining in any
way how it chose among these possible benchmarks. This shows that the Regulator is
cherry-picking among the listed benchmarks.
71
Mobistar First Quarter 2015 Results – Analysts & Investors Presentation, slide 47, available at
http://corporate.mobistar.be/documents/Mobistar%20-%20Q1%202015%20Results%20%20Analysts%20%20Investors%20Presentation%20FINAL%20Roadshows1.pdf.
72
CRC Decision of 11 December 2013 concerning the Dutch language region, § 72-73.
73
These foreign benchmarks.
37
EUI-1200160743v1
15 JULY 2015
Second, the price of € 75 is not representative. Telenet charges this only to customers
who request a new modem even though they do not need it to use the services they subscribe
to. In all other cases, the modem is provided free of charge. For instance, if a customer’s
modem does not support Docsis 3.0 and the customer upgrades to a product that requires a
Docsis 3.0-enabled modem (e.g., Whop or Whoppa), Telenet replaces the modem free of
charge. The € 75 is the price of the replacement of the modem including installation by a
Telenet technician. It is therefore wrong to use this amount as representing the value of the
modem as such.
Third, the Regulator’s choice to use the ratio between the monthly rental price and the
purchase price of the STB (i.e. of a different product) is wholly arbitrary and not justified in
any way. The Regulator does not explain why it chose this novel method rather than simply
amortizing the value over a period of 36 months (as it does in the preceding example
concerning Proximus’ modem). The fact that the newly proposed method yields a higher
value (€ 2,08), is another illustration of the goal-oriented nature of the Regulator’s analysis.
Fourth, the foreign benchmarks mentioned in the Draft Decision are not
representative. To Telenet’s knowledge, all national operators and nearly all operators in
neighboring countries do not charge anything extra for the modem. This explains why the
Regulator found only a few examples to use as a benchmark. Telenet has carried out its own
survey of offers by operators in France, Germany, the Netherlands and the United Kingdom
and found that all broadband subscriptions (single play or in a bundle) include at least a basic
modem (see table in Annex I). The price of € 2/month for Kabel Deutschland, mentioned in
the Draft Decision, concerns a modem with wireless functionality (“WLAN Kabelrouter”).
All Kabel Deutschland subscriptions include at least a basic free modem without wireless
functionality (“Kabelrouter”). Moreover, even the wireless modem is included for free in
most of Kabel Deutschland’s subscriptions (Internet & Phone 25, 50 and 100 and Internet
Business Kabel 50 and 100). The price of € 3/month mentioned in the Draft Decision for
Orange France also includes the rental of the STB (decoder) and one could therefore just as
well consider that the modem is included for free and the customer only pays for the STB.
Only Orange Luxembourg charges € 3/month for the modem as indicated in the Draft
Decision (for subscribers with a 24-month contract), but the wireless modem in question
(“Fritzbox”) includes advanced voice telephony functionality 74 and is therefore not
comparable to standard internet modems.
Once again, rather than deducting an arbitrary value from the reference retail price, a
far more objective and proportionate way to take into account the modem would be to
include the cost of the modem as an avoided cost in the calculation of the minus. This is
sufficient to ensure that an efficient alternative operator can compete with the cable operators
on the retail market, even though it has to procure its own modems. Telenet estimates the
monthly production cost of the modem at € [confidential] (i.e. [confidential] € purchase
price amortized over 5 years75). It must, however, be emphasized that this is a conservative
74
75
Fritzbox modems are not only wireless cable/xDSL modems but also DECT base stations supporting
multiple ISDN/analog phones and/or VoIP phones.
It is incorrect to use the same depreciation period for the STB and the modem. Modems are much more
robust equipment than STB and do not need to be replaced so often. A five-year amortization period is
therefore more appropriate than the three-year depreciation period which the Regulator applies to
Proximus’ b-box.
38
EUI-1200160743v1
15 JULY 2015
approach (and therefore an overestimation) as it disregards the fact that the modem also
includes a fixed voice service component. Therefore, for the purpose of estimating the
production cost of the modem, a part of these costs should be allocated to the fixed telephony
service on a RGU basis (in the charts above, Telenet has allocated [confidential]% of the
cost of the modem to telephony).
Finally, as already mentioned, the Regulator also makes a double-counting error by
deducting from the reference retail price both a value for the Wi-Free service and a value for
the modem, since Wi-Free is a functionality of the modem.
D.
Broadband profiles
The Draft Decision includes a formula to compute the applicable price for specific
broadband profiles requested by the beneficiary. The formula is based on an intrapolation
method seeking to correlate (i) download speed, upload speed, volume and (ii) the wholesale
price.
For Telenet, this results in a formula only bearing in mind the download speed and
volume of a specific profile. Based on the detailed calculation sheet Telenet concludes that
the formula is not fit for purpose and relies on erroneous factual data:
 There are factual errors in the measuring points included in the calculation sheet.
For instance the upstream speed for the Telenet Fibernet XL product is currently
set at [confidential] (and not [confidential] as included in the calculation sheet).
 Correcting the values for the upstream speed significantly impacts the outcome of
the formula.
The initial values of the formula for Telenet are the following:
[confidential table]
When adjusting the values of the measuring points with the correct upstream
value for Fibernet XL to [confidential], the values of the formula change to:
[confidential table]
By correcting the values for Fibernet XL the volume characteristic appears to be
no longer relevant for determining the price for specific broadband profiles. This
finding undermines the formula as such given the importance of volume both
from a commercial perspective and from a pure network management perspective.
 The input is limited to [confidential] Telenet broadband profiles of which
[confidential] profiles have nearly similar characteristics. A limited number of
measuring points entails that any change to the measuring points will have an
important impact on the outcome of the exercise and change the formula.
For instance, when correcting the upload speed of the Fibernet XL (see our
comment above), the formula for Telenet is no longer based on download speed
and volume, but rather on upload speed and volume.
39
EUI-1200160743v1
15 JULY 2015
Competition requires Telenet to constantly monitor its product offerings and
regularly change product specifications and pricing. A formula which is so
sensitive to changes and relies on limited and isolated measuring points runs
against the general principles of access regulation.
 The calculation sheet includes wholesale prices which the Regulator has
computed based on Telenet 3P legacy tariffs which it has updated with inflation
percentages, etc. Telenet cannot align these tariffs with the tariffs it applies. The
tariffs should be replaced with the tariffs effectively applied.
Also for computing the wholesale price, the Regulator has taken into account
certain elements of its revised methodology, such as 3P legacy tariffs as reference
tariffs (even if a 2P reference tariff is available), adjustments for additional
services, etc. Telenet refers to its argumentation set out above on these points.
Furthermore, since the Regulator uses the wholesale prices following the adjusted
retail minus methodology, this also impacts the outcome of the calculation for the
formula. A possible pricing strategy (included in the retail prices) will be
undermined by the corrections of the Regulator when determining the wholesale
prices. Therefore it is possible that there is no link between measuring points due
to the wholesale prices determined by the Regulator. This will thus result in a
formula where certain characteristics (download speed, upload speed, volume)
will not be taken into account.
Telenet agrees that there should be a methodology in place to determine an adequate
pricing for specific broadband profiles requested by the Beneficiary, but finds that the
formula is inadequate, unreliable and unpredictable.
As an alternative, Telenet proposes that a specific process be established for
determining the price with a central role for the Regulator to confirm the pricing proposition
made by Telenet. In order to suggest a specific price, Telenet will need to map the specific
profile requested by the Beneficiary against its own product offering (active and legacy
offerings) and will subsequently suggest a wholesale price to the Regulator bearing in mind
all other aspects of the retail minus methodology. The Regulator will review the price
proposition of Telenet and will subsequently confirm whether the price can be considered as
adequate for the Beneficiary or not.
 Using standalone prices as a benchmark for additional services included in a bundle
is generally not appropriate.
 The national and foreign offers against which the Regulator has benchmarked the
value of the additional services are not comparable in terms of features and
underlying costs.
 The choice of benchmarks is arbitrary (cherry-picking) and the methodology used
inconsistent.
 The fixed telephony component is grossly overvalued, inter alia because the
Regulator disregards that the retail price of the standalone fixed telephony service
includes an access line component which should be allocated to all three components
40
EUI-1200160743v1
15 JULY 2015
of the bundle.
 A cost-based approach to the additional services leads to much more proportionate
results.
 The proposed broadband pricing formula contains factual errors and is unreliable.
It should be replaced by a procedure whereby Telenet’s would be proposing a
wholesale tariff based on its product offering to be validated by the Regulator.
IV.
THE DRAFT
FRAMEWORK
DECISION FRUSTRATES THE POLICY OBJECTIVES OF THE REGULATORY
The Draft Decision goes against the policy objectives of the EU regulatory
framework for electronic communications (as enshrined inter alia in Article 8 of the
Framework Directive) in several ways:
 First, it distorts competition by promoting inefficient market entry;
 Second, it frustrates consumer interests by stifling product innovation
(particularly through bundling of services) and discouraging retail promotions and
discounts;
 Third, it favors service competition over infrastructure competition and
undermines cable operators’ ability to continue to invest in their networks;
 Fourth, it negatively affects regulatory predictability by failing to ensure a
consistent regulatory approach over an appropriate review period.
A.
The Draft Decision promotes inefficient market entry
As detailed above, the Draft Decision provides excessive entry assistance to the
beneficiaries of the wholesale obligations. The only two operators that have manifested their
intention to make use of these obligations – Mobistar and Proximus – have a sufficient
critical mass and do not require any special entry assistance. This is more than obvious in the
case of Proximus, the incumbent fixed line operator, with a 45% market share for broadband
and a 34% market share in the digital television market (corresponding to over 1,6 million
customers).76 Mobistar too is an established operator in the Belgian market, with an installed
base of nearly 3,8 million mobile customers to which it can cross-sell its fixed-line offers.
By overcompensating for the supposed scale disadvantage of the beneficiaries of the
wholesale measures, the Regulator risks promoting inefficient entry, which may seem
beneficial for consumers in the short run but is detrimental for consumer welfare in the long
run. Only efficient entry can benefit consumer welfare in the long run. As the Belgian
Competition Authority has stressed in relation to the regulation of MTR:
76
Proximus Q1 2015 presentation, slide 20.
41
EUI-1200160743v1
15 JULY 2015
“[h]et positief effect van een systeem waarbij (het cliënteel van) een operator met
aanmerkelijke marktmacht (het cliënteel van) andere operatoren met een zwaardere
kostenstructuur als het ware “subsidieert”, ligt zacht uitgedrukt niet voor de hand
en vereist dus minstens een uitvoerige en op de specifieke nationale context
toegespitste motivering. De Raad wil geenszins bij voorbaat uitsluiten dat dergelijk
positief effect zou kunnen bestaan en dus valt af te wegen tegen andere evenmin uit te
sluiten negatieve effecten, maar mist in het document waarover advies wordt
uitgebracht een afdoende motivering voor de omstandigheid dat het door het BIPT
gehanteerde model België binnen Europa in een unieke, geheel tegen de Europese
tendens ingaande, situatie zou brengen qua tariefverschillen voor interconnectie
tussen operatoren.”
The European Commission has also considered that MTRs “should normally be
symmetric” and that asymmetry might be justified “in exceptional cases by objective cost
differences which are outside the control of the operators concerned”, but that this should
then be duly motivated in the Regulator’s decision. 77 We can infer from this that any
obligation for SMP operators to subsidize specific entry assistance for new market entrants,
as an exception to the rule, should be supported by circumstantial reasoning. Such reasoning
is lacking in the Draft Decision.
B.
The Draft Decision does not serve consumer interests
The Draft Decision stifles retail product innovation by discouraging cable operators
to enrich their offers with additional services. Each time a cable operator adds an additional
service to the bundle, this may have an immediate (and, due to the calculation method
applied, disproportionate and unforeseeable) impact on the wholesale prices. The Draft
Decision even creates a strong economic incentive for cable operators to remove valueadding services from their offerings and to offer only “bare-bones” internet and television
services, which would actually lead to a reduction in consumer welfare. In a converging
environment, there is, however, a strong interest in bundling services and consumers can
benefit from increased and potentially disruptive competition when operators seek to broaden
their services by engaging into cross-selling which was acknowledged for example by
BEREC in its Report on Convergent Services. 78 Telenet’s Wi-Free service is a good
example of a converged service which is disruptive for mobile broadband services (which are
generally considered expensive in Belgium). The methodology is penalizing Telenet for
offering this service in its bundle disregarding the evident pro-consumer benefits resulting
from this practice.
77
Letter from the European Commission, Case BE/2006/0433 Voice Call Termination on Individual
Mobile Networks in Belgium, Comments pursuant to Article 7(3) of Directive 2002/21/EC, p. 5.
78
“§ 45. Convergence underpins more efficient and economic bundling of products and services,
although convergence could also arise in a standalone service offer. A common trend in Member
States, as identified in the 2009 ERG Report on the discussion of the application of margin squeeze
tests to bundles, is the increasing preference of consumers for buying products in bundles which
convergence makes more economical and easier. As consumers buy multiple offerings from one
provider, the number of services required to compete effectively in retail markets increases. On the
other hand, as described above, convergence could lead to enhanced products which could easily
replace traditional ones.”
42
EUI-1200160743v1
15 JULY 2015
Likewise, the Draft Decision stifles retail price competition by discouraging discounts
and promotional offers, which are reflected in the wholesale prices, in some cases even with
retroactive effect. The retail price competition will be paralyzed as a result of the
burdensome notification procedure, the “freezing” of the retail pricing for certain
components of the bundles (additional services, decoders and modems), and the uncertainty
created by potential retroactive adjustments of the wholesale prices. These factors will
severely impair the ability of cable operators to react swiftly to pricing decisions by
competitors, and in particular by Proximus, which is not hampered by the same handicaps.
Rather than promoting price competition, the proposed regulation may therefore paralyze it.
It could also undermine the cable operators’ competitive position, not to the benefit of new
players, but to that of Proximus. Such developments would obviously not contribute to
consumer welfare.
C.
The Draft Decision undermines network investments and infrastructurebased competition
By eroding the cable operators’ wholesale margins, the Regulator undermines the
cable operators’ ability to continue investing in their network infrastructure. The cable
operators are forced to sacrifice wholesale margin to subsidize (excessive) entry assistance
by new service-based competitors which will create a temporary but potentially destructive
aggressive price competition. The Regulator thus promotes service-based competition over
infrastructure-based competition, which runs counter to the objectives of the EU regulatory
framework for electronic communications (see Article 8(5)(c) of the Framework Directive).
The regulation not only risks promoting inefficient new entry, but it also does not
include any incentives for the beneficiaries to become more efficient or to move from
service-based competition to infrastructure-based competition. Instead, the regulation will
most likely be self-perpetuating as the beneficiaries are far better off remaining dependent on
the cable operators’ wholesale inputs than becoming infrastructure-based operators in their
own right. As the Commission notes in its explanatory document accompanying the
Communication on a Digital Single Market Strategy for Europe, regulation may reduce the
investment incentives of alternative operators “if regulated wholesale access is made
disproportionately attractive, i.e. access seekers’ build-or-buy decisions rendered
economically inefficient”.79 In this context it is also noteworthy that Belgium leads in highspeed fixed broadband access, but lags behind in mobile broadband.80
The best regulatory outcomes are to be expected from spurring competition between
existing network infrastructures, rather than weakening it for the benefit of mere servicebased competition. Any regulatory intervention that discourages or slows down investments
in NGA networks, even if it is designed to promote additional service-based competition (the
success of which is uncertain), is likely to result in a net loss of social welfare. As the
79
80
Commission staff working document, A Digital Single Market Strategy for Europe - Analysis and
Evidence, accompanying the document Communication from the Commission to the European
Parliament, the Council, the European Economic and Social Committee and the Committee of the
Regions: A Digital Single Market Strategy for Europe, 6 May 2015, SWD(2015) 100 final, p. 36-37.
Digital Economy and Society Index (DESI) 2015, Country Profile Belgium, p.2
43
EUI-1200160743v1
15 JULY 2015
European Commission notes, “the social return from investment in higher quality networks
tends to be greater than for the individual operator”.81
The Regulator also failed to consider whether it would not be desirable to modulate
the intensity of the regulation in function of the intensity of infrastructure-competition that is
already present. As the European Commission observes, “little full "infrastructure
competition" has emerged in fixed-line networks, except in very densely populated areas,
where cable networks were already present, or where local authorities have been active.
There is a need for simpler and more proportionate regulation in those areas where
infrastructure competition has emerged at regional or national scale. The deployment of very
high capacity networks needs to be encouraged while maintaining effective competition and
adequate returns relative to risks”.82 In the case of Belgium, intense infrastructure-based
competition already exists at the national level, since both the DSL network and the cable
networks (combined) cover more than 80% of the population, and there is therefore “full
infrastructure competition” in most of the country (and certainly in practically the entire
Flemish and Brussels regions). The Regulator should therefore reassess whether the
proposed, heavy-handed, regulation in support of service-based competition is a desirable
and proportionate means to achieve the objectives of the regulatory framework.
D.
The Draft Decision undermines regulatory predictability
Article 8(5)(a) of the Framework Directive provides that NRAs shall promote
regulatory predictability by ensuring a consistent regulatory approach over appropriate
review periods. Regulatory stability is also one of the main objectives of the Commission
Recommendation on relevant markets susceptible to ex ante regulation. 83 The CRC
framework decision also emphasizes that, when making any adjustments to the regulation,
“BIPT shall take into account the need for stability on the market for electronic
communications”. Regulatory predictability and regulatory certainty are especially important
at a time when network operators are expected to make important investments in NGN to
meet the needs of consumers and businesses. Regulatory predictability is a crucial
consideration in operators’ investment decisions and the lack thereof seriously hampers
investment and innovation.
The Draft Decision tries to ensure wholesale price stability for the beneficiaries of the
obligations by imposing a “freeze” on the additional services and on the pricing for modems
and STB. However, this does not work both ways as the Regulator reserves the right to
impose retroactive wholesale price adjustments or to bring components into the calculation of
81
Ibidem, p. 36.
82
Communication from the Commission to the European Parliament, the Council, the European
Economic and Social Committee and the Committee of the Regions: A Digital Single Market Strategy
for Europe, 6 May 2015, COM(2015) 192 final, p. 10.
83
Commission staff working document – Explanatory note, accompanying document to the Commission
Recommendation on relevant product and service markets within the electronic communications sector
susceptible to ex ante regulation in accordance with Directive 2002/21/EC of the European Parliament
and of the Council on a common regulatory framework for electronic communications networks and
services, 9 October 2014, SWD(2014) 298, p. 5: “[The Recommendation] seeks to ensure that broadly
the same product and services markets will be subject to a market analysis in all Member States and
that market players will be aware in advance of the markets to be analysed, in line with the principles
of regulatory predictability and legal certainty”.
44
EUI-1200160743v1
15 JULY 2015
the wholesale prices that have not so far been considered relevant. Hence, there is a lack of
predictability with respect to the wholesale prices that the cable operators can expect to
receive, which further undermines their incentives to invest and innovate.
 By overcompensating the supposed scale disadvantage of the beneficiaries
(disregarding the actual situation of the operators that have so far sought access),
the Regulator distorts competition and promotes inefficient entry.
 The Draft Decision frustrates consumer interests by discouraging price competition,
pro-competitive bundling practices and other forms of product innovation.
 The Draft Decision promotes service-based competition over infrastructure-based
competition and risks discouraging NGN investments.
 The regulatory unpredictability created by the Draft Decision will further
undermine incentives to invest and innovate.
45
EUI-1200160743v1
15 JULY 2015
ANNEX I
MONTHLY RETAIL VALUE OF THE MODEM – INTERNATIONAL COMPARISON
Country
Operator
Device
Price (monthly rental fee)
Source
Netherlands
KPN
Wireless
modem
Included in subscriptions (+ includes access to FON
network)
https://bestellen.kpn.com/
Netherlands
Tele2
Wireless
modem
Included in subscriptions
https://www.tele2.nl/thuis/tarie
ven/
Netherlands
UPC
-
See Ziggo
-
Netherlands
Ziggo
Wireless
modem
Included in subscriptions (+ includes access to 2 million
“WifiSpots”)
https://www.ziggo.nl/alles-in1/
France
Bouygues
Bbox
Wireless modem and decoder (2 separate devices)
included in subscriptions
https://www.bouyguestelecom.
fr/offres-internet/bbox
Bbox Miami
Wireless modem and decoder (1 device) included in
subscription
https://www.bouyguestelecom.
fr/offres-internet/bbox-miami
France
Free
Freebox
Wireless modem, decoder, Blue-Ray player, game
console and Femto-cell (all-in-one device) included in
subscriptions
http://www.free.fr/adsl/index.h
tml
France
SFR
Box Fibre
Wireless modem and decoder with hard disk (all-in-one
device) included in subscriptions
http://static.ssfr.fr/media/brochure_box.pdf
46
EUI-1200160743v1
15 JULY 2015
Country
Operator
Device
Source
France
Orange
Livebox Play
€ 3/month for wireless modem and decoder
http://abonnezvous.orange.fr/residentiel/offre
s-internet-ADSL-VDSL2.aspx
Germany
1&1
1&1 WLAN
Modem,
1&1
HomeServer
http://dsl.1und1.de/?linkId=hd.
subnav.dslflatratesuebersicht&ucuoId=PUAC:def
ault.EUE.DE20150702113100768B986642FD30003853C058
E0B9CDF1.TCpfix114b&ac=
OM.BR.BRd09K41503T7073a
#pakete
Germany
Deutsche
Telekom
Magenta
Zuhause
http://www.telekom.de/privatk
unden/zuhause/internet-undfernsehen/magenta-zuhause-s
Germany
O2
HomeBox 2
http://dsl.o2online.de/provider/
content/segment/anbieter/produ
kte/dslhome/?exclusivId=epo2p_3x3w_home-dsl-vdsl
Germany
Kabel
Deutschland
(Vodafone)
Kabelrouter
http://www.kabeldeutschland.d
e/internettelefon/kabelrouter.html
47
EUI-1200160743v1
15 JULY 2015
Country
Operator
Device
Source
WLAN
Kabelrouter
Wireless modem € 2/month but included in Internet &
Phone 25, 50 and 100 subscriptions as well as in
Internet Business Kabel 50 and 100 subscriptions
http://www.kabeldeutschland.d
e/internet-telefon/wlankabelrouter.html
Germany
Unity Media
Highspeed
modem
http://www.unitymedia.de/priv
atkunden/internet/basisinternetzugang/
United
Kingdom
BT
BT Home Hub Included in subscriptions
http://www.productsandservice
s.bt.com/products/broadbandpackages/?s_intcid=con_fly_pn
s_bbpkgs
United
Kingdom
Sky
Router
http://www.sky.com/shop/broa
dband-talk/broadbandunlimited/
United
Kingdom
TalkTalk
Advanced
Technology
Router
https://sales.talktalk.co.uk/prod
uct/broadband/simplybb
United
Kingdom
Virgin Media
Super Hub
http://store.virginmedia.com/bi
g-bundles.html
48
EUI-1200160743v1
BIPT’s review of wholesale cable access
rates
RBB Economics, 15 July 2015
RBB Economics
Page 1
1.
Introduction
Telenet asked RBB Economics to respond to BIPT’s draft decision relating to its review of
wholesale cable access rates. BIPT’s review concerns a further implementation and detailing of
the wholesale rates set by BIPT at an earlier stage.
More specifically, Telenet has asked us to assess the way in which BIPT in the draft decision
and within the proposed retail minus approach deals with (a) additional, not regulated, services
and (b) promotions.
In this report we will assess these issues from BIPT’s draft decision from an economic
perspective.

Section 2 discusses the economic underpinnings of retail minus regulation.

Section 3 summarises the BIPT draft decision as concerns additional services and
comments on this part of the decision.

Section 4 aims to do the same for promotions: we will first summarise BIPT’s approach
as regards promotions within the retail minus framework, followed by our economic
assessment of BIPT’s approach.
RBB Economics
Page 2
2.
Retail minus regulation
1
Retail minus regulation is derived from the Efficient Component Pricing Rule (ECPR). The
ECPR is aimed at situations in which firms active in a retail market are, for the provision of retail
services, dependent on (wholesale) inputs from a vertically integrated firm with which they
compete at the retail level. The purpose of ECPR (and retail minus) based regulation is to allow
for efficient competition at the retail level.
This should be contrasted to cost based regulation aimed at setting a cost based price for the
wholesale input: retail minus regulation concerns the setting of a wholesale price that allows
efficient retail competitors to compete at the retail level with the vertically integrated firm.
Firms which are equally or more efficient than the vertically integrated firm will be able on the
basis of a wholesale price set in accordance with the ECPR to offer retail prices that allow them
to compete with the vertically integrated firm.
The ECPR, and by implication retail minus regulation, is specifically not aimed at facilitating
retail competitors that operate less efficiently on the retail market relative to the vertically
integrated firm.
In the retail minus variant of the ECPR the wholesale price is set on the basis of the following
formula:
𝑊ℎ𝑜𝑙𝑒𝑠𝑎𝑙𝑒 𝑝𝑟𝑖𝑐𝑒 = 𝑟𝑒𝑡𝑎𝑖𝑙 𝑝𝑟𝑖𝑐𝑒 𝑇𝑒𝑙𝑒𝑛𝑒𝑡 − 𝑐𝑜𝑠𝑡 𝑜𝑓 𝑟𝑒𝑡𝑎𝑖𝑙 𝑎𝑐𝑡𝑖𝑣𝑖𝑡𝑖𝑒𝑠 𝑇𝑒𝑙𝑒𝑛𝑒𝑡
The idea behind this is rule is simple: if retail costs (including a charge for retail profits) are
taken out of the retail price of the regulated firm to obtain the wholesale price, other firms can
compete with the regulated firm if their own retail costs are equal to or lower than the retail costs
of the regulated firm. This is because such a wholesale price allows these firms to set a retail
price which is equal to or lower than the retail price of the regulated firm.
A wholesale price which would be too high would not allow competitors to offer a competing
retail price without making losses. A wholesale price which would be too low would result in
entry of inefficient firms that would be able to operate profitably despite higher costs of their
2
retail activities. Both results would be economically inefficient.
For a proper application of the retail minus methodology, and for economically efficient
outcomes, it is hence very important to accurately determine (a) the retail price of the regulated
firm and (b) the costs of the retail activities of the regulated firm.
1
See for example Armstrong (2001), The theory of access pricing and interconnection, MPRA Paper No. 15608.
2
See for a more extensive discussion of the various aspects of the ECPR (and retail minus): Baumol, Ordover en Willig (1996),
Parity pricing and its critics: necessary condition for efficiency in provision of bottleneck services to competitors.
RBB Economics
Page 3
3.
Additional services
3.1. BIPT’s draft decision
Below we will summarise BIPT’s draft decision in as far as it is relevant for the way om which
BIPT suggests to consider additional, extra services which are incorporated in the services
offered by Telenet.
The main issue that BIPT seeks to solve in the draft decision is to determine the retail price and
the minus for the regulated cable access service, in view of the fact that the retail offers of the
regulated cable operators, including Telenet, include additional, non-regulated services.
According to BIPT these services are not part of the regulated wholesale access service, and
the wholesale access service is not a necessary input for these additional services.
BIPT considers that the retail price includes these additional services and considers that it is
reasonable to assume that these additional services result in additional costs for the regulated
cable operators, in addition to the costs of the retail offer of the regulated (naked) service.
In other words, if the retail offers of the regulated cable operators would not include these
additional services, the costs of the retail offers of these operators would have been lower,
according to BIPT. A competitor which would want to include comparable additional services in
its retail offer would hence, just like the regulated cable operator, face additional costs.
BIPT considers that there are two ways in which these additional services can be taken into
account when setting the wholesale rates:
“Their production costs could be determined in order to be added to the
avoidable costs that make up the minus. This approach results in an
increase of the minus percentage.”
Or
“Their value can be subtracted from the retail price before the minus is
applied. This approach results in a decrease of the base for the application
of the minus”.
BIPT argues in the draft decision why it considers the second approach to be the correct
approach. In essence BIPT provides the following reasons for this:

Priority should be given to an approach that limits the basis for the application of the
minus because:
o
RBB Economics
The additional services are not part of the wholesale service and it would be
more appropriate that their value does not influence the level of the wholesale
prices.
Page 4
o
Discounting for additional services through the minus, would result in a higher
minus, but such minus would be applied to a base influenced by the existence
of the additional services. As a result the comparability of the wholesale and
retail prices would no longer be guaranteed.
o
BIPT considers it important to guarantee that firms do not need to pay for
facilities which are not necessary for the service required.
The value of the additional services can be determined, according to BIPT, by (a) the price
applied by the cable operators themselves, the price that other operators charge for such
services (a benchmark) or (b) on the basis of a methodology that uses production costs (in the
absence of a benchmark).
On the basis of this methodology, BIPT reduces the retail price of Telenet, for the purposes of
calculating a retail minus based wholesale price with:

€ 2 per month for Wi-Fi access. This price is based on the price of Telenet for its standalone hotspot service and a benchmark price for the use of homespots, as the latter is
included in the standard offers (and prices) of Telenet.

€ 5 per month for the second screen service. This price is based on a benchmark price
as this service is included in the standard offers (and prices) of Telenet.

€ 0.50 per month for the e-mail service. This price is derived from the price that Telenet
charges for additional e-mail addresses (€ 0.05 per address) on top of the 10 addresses
included in Telenet’s standard packages.

€ 1 per month for the webhosting service included in Telenet’s standard offer. This price
is based on the price charged by Telenet for expansion of the standard webhosting
space offered.

An amount for Sporting Telenet, as this service is offered as part of different
subscriptions for different prices. To determine the value, BIPT first adds the net
receipts from customers that receive Sporting Telenet at a reduced price as part of a
bundle, to the retail price based on average usage within the subscriber base of
Telenet. In a next step BIPT subtracts the stand-alone “value” of Sporting Telenet which
is based on the price that stand-alone digital TV customers pay, again based on
average usage. The resulting rebate to the retail price is hence the difference between
the stand-alone price and the average price paid for subscribers for Sporting Telenet as
part of their package.
In the calculation BIPT applies a bundle rebate equal to the bundle rebate that BIPT derives for
the relevant bundle. Hence, the rebate received by Telenet customers e.g. for buying a TV and
broadband bundle, also appears to be applied to the value of the additional services that is
subtracted from the retail price on which the minus is based.
Below we will assess BIPTs methodology and its consequences.
RBB Economics
Page 5
3.2. Comment on the draft decision as regards additional services
In and of itself it is understandable, within the logic of BIPT’s retail minus methodology, that
BIPT seeks to implement a correction for additional extra services offered by Telenet that are
not part of its regulated wholesale offer, as the regulated wholesale offer does not include the
(wholesale component of) the additional services.
The question is however whether the way in which BIPT seeks to do this, results in an
economically efficient wholesale price, this is in accordance with the underpinnings of the
chosen regulatory retail minus approach.
To assess this, we take one of the items for which BIPT has determined a “market value” that is
deducted from the retail price, and assess the implications for the wholesale price. In particular
we assess whether the impact of the methodology could carry a risk that wholesale prices are
set too high or too low, in which case the purpose of regulation – to set an economically efficient
wholesale price – is not achieved.
The conclusions that can be drawn from this, will apply to all additional services, as for each of
the additional services BIPT has followed in essence the same approach, i.e. in none of the
cases BIPT used production costs for setting the value: in each case BIPT has either used the
stand-alone price or a price based on a benchmark.
3.2.1. The e-mail service as an example
The “value” of the e-mail addresses that would be incorporated in the retail services of Telenet
is set at € 0.50, because Telenet customers can buy an additional e-mail address at a price of €
0.05 per month, and because each internet subscription allows for a maximum of ten e-mail
addresses. Therefore, in the logic of BIPT, the value of the retail offer should be decreased by €
0.50 (subject to further adaptions to account for the relevant bundle rebate and VAT).
This however is a gross overestimation of the true value. Dit is een grove overschatting van de
werkelijke waarde. On average, usage of the e-mail addresses by Telenet customers more
closely to 1.5 rather than 10. Telenet does indeed charge a notional stand-alone price for
additional e-mail addresses, but the law, the market assessment nor the draft decision provides
for a justification for the use of this stand-alone price to determine what the “value” should be
within the bundle of services offered.
In addition, it can be safely assumed that the costs related to offering this additional service are
very limited, and hence that the profit margin of the stand-alone product – despite the low price
– would be very high (there are various providers, like Google, that offer e-mail addresses for
free). Telenet charges a price for additional e-mail addresses in order not to be forced to allow
subscribers unlimited numbers of e-mail addresses. In practice there are no or virtually no
users that in fact buy the stand-alone service, precisely because most Telenet customers will
only need a very limited number of e-mail addresses (a factor that is not investigated by BIPT at
all). This makes it highly unlikely that the price charged by Telenet for the stand-alone service is
a good benchmark for the value of the ten e-mail addresses which are included in the
subscription.
RBB Economics
Page 6
This has the following consequences:

A ‘value’ for the e-mail service that is too high results in a deduction from the retail price
of the regulated service that is too high, which results in the relevant retail price of the
regulated service to be set at a level which is too low.

The value of the additional, extra, services includes a profit margin of the regulated firm
when a service is sold on a stand-alone basis (or when a benchmark is found). If the
profit margin on additional services would be greatly in excess of costs, than these
margin eat into the true margin of the regulated wholesale service: this is because a
high margin on the additional service results in a higher deduction from the retail price
and hence to a lower resulting margin on the wholesale service of the regulated firm.

Without an assessment of the true (stand-alone) costs and margins of the additional
services, there is high risk that the application of such reference values results in
reducing the retail prices by too much.

In a regulated setting this would result in incentives for Telenet to no longer offer
additional services as part of its packages, since these additional services for
customers, that allow Telenet to differentiate itself from other operators, necessarily
result in lower wholesale margins within BIPT’s methodology.

This is particularly relevant in view of the fact that the additional services as defined by
BIPT are part of the services of Telenet, but cannot, like for example broadband internet
and Sporting Telenet, be considered as proper stand-alone services, separate from
other services. Telenet only offers e-mail addresses in function of its broadband internet
service.

BIPT’s approach with respect to additional services in combination with bundle rebates
also does not result in very consistent outcomes. In the Internet + TV bundle of
Telenet, the additional services have a combined value of € 4.96, which after
application of the bundle rebate results in an amount of € 4.32, which is deducted from
the “reference retail price”, resulting in a decrease of this reference price by 15%. In the
Whoppa bundle, a bundle rebate of 42% is used on the same set of additional services
resulting in a value of these services of € 2.88 and a corresponding decrease of the
reference price by 7%.

BIPT does not address this, but the costs of the additional services, and in particular an
additional service like the e-mail service, should be regarded more appropriately as
retail costs which are strongly related to the principal service, and should hence be
taken into account when determining the minus, rather than determining the reference
retail price.
Taking the (assumed) market value out of the retail price appears primarily targeted at reducing
the wholesale price as much as possible: BIPT does not in any way motivate why the chosen
methodology would result in the appropriate value of the relevant retail service, with the risk that
the wholesale price is set at too low a level. As we will show further below, this results in
perverse incentives and inefficient entry.
RBB Economics
Page 7
3.2.2. Avoidable costs avoid perverse incentives
Competitors that use cable access may offer additional services just like Telenet in order to
provide a comparable set of services in their packages. The can also decide not to do so.
To allow a competitor to be competitive in the retail market on the basis of the regulated
wholesale input, it is “only” necessary, in relation to the additional services, to exclude the
avoidable costs of these services from the wholesale rate, it is not required for this purpose to
subtract the (hypothetical) revenues from these services from the retail price.
This is because what matters is to ensure that the – equally efficient – competitor would not
have a cost disadvantage relative to the regulated firm. It is sufficient for this purpose to account
for the avoidable costs of the additional services – such as the e-mail service – in the minus.
This prevents that:

Regulated operators would be given an incentive to unbundle and offer ‘naked’ services
only, because including additional services would otherwise result in lower wholesale
charges.

Competitors would be offered too low wholesale prices, resulting in promoting inefficient
entry and allowing these competitors to offer similar packages at much lower prices
relative to the regulated firm.
A simple numerical example can illustrate this. Suppose that the retail price, including additional
services is € 30 and the minus € 10, resulting in a wholesale charge of € 20. Assume further a
package of additional services that is offered as an option for € 5, maar is also included in
bundles. The avoidable costs of this package are € 1, and the margin € 4.
In BIPT’s approach, the stand-alone retail price is deducted from the retail price, resulting in a
decrease of the wholesale rate with € 5. In an avoidable cost approach the € 1 in avoidable
costs is added to the retail costs, resulting in a decrease of the wholesale rate of € 1.
Leaving aside bundle rebates shows that the differences between these approaches results in a
considerable difference in the resulting wholesale rates.
If the avoidable costs of the package of additional services for the access seeking competitor
would also amount to € 1, then this party could in BIPT’s approach replicate the package of the
regulated firm for € 15 + € 1 = € 16. The difference with the scenario without additional services
is identical to the margin on additional services when sold on a stand-alone basis. This margin
on the additional services hence eats into the wholesale margin of the regulated firm.
RBB Economics
Page 8
Retail price including additional services
€ 30
Retail costs
€ 10
Wholesale rate
€ 20
Additional services retail
€5
Avoidable costs
€1
Margin
€4
BIPT approach
Naked retail price (€ 30 - € 5)
€ 25
Retail costs
€ 10
Wholesale rate
€ 15
Avoidable costs approach
Retail price
€ 30
Retail costs
€ 11
Wholesale rate
€ 19
To the extent more ‘extra’ services are unbundled on the basis of a hypothetical retail price for
these services, the more dramatic will wholesale prices decrease, without the regulated firm
saving costs in a comparable way.
The result of this could furthermore become unreal, in the sense that, the longer the list of
additional services with a high margin would be identified, the methodology could result in
wholesale rates below the costs of the wholesale service of the regulated firm. In an approach
based on avoidable costs, such a result would not be possible by definition, as the wholesale
rates would only decrease as a result of cost decreases for the regulated firm as a result of
stripping down the regulated service.
3.2.3. Methodological remarks
BIPT states in defence of its methodology that “the extra services are not part of the wholesale
service and it would be more appropriate for their value not to influence the level of wholesale
prices”.
It is questionable however, in particular for additional services, whether quasi-market prices are
a good proxy for their value, in particular if the purpose of such services is mainly to increase
the attractiveness of the principal service. In any event, competitors should be considered well
capable to add similar services to their bundles on the basis of an avoidable cost approach.
In addition, BIPT states that “discounting through the minus would result in a higher minus, but
this minus would be applicable to a base that is influenced by the existence of the extra
services, as a result which the comparability of the wholesale and retail price would no longer
be guaranteed.”
RBB Economics
Page 9
BIPT appears to mean here that it has chosen to account for extra services in the retail price,
because discounting for these in the minus results in a higher minus percentage, that would
however be applied to a base price which would be too high according to BIPT, as the extra
services would be included in the base.
The cause of this issue is however mainly that the minus is determined as a percentage of the
retail price, rather than an absolute value that is deducted from the retail price. If the
consequence thereof would be that the comparability of the wholesale and retail prices would
no longer be guaranteed, as stated by BIPT, then this problem appears to be caused by BIPTs
own methodology, not because a minus percentage would necessarily follow from a retail minus
methodology.
Therefore, BIPT’s reasoning does not appear very convincing, in particular as this results in a
‘solution’ that is strongly to the disadvantage of the regulated firms. If BIPT’s methodology
results in a wrong conclusion, it seems more appropriate to change the methodology itself.
Finally, BIPT states that it consider it “of importance to guarantee that firms do not need to pay
for facilities that are not required for the desired service.” This argument is also difficult to
understand. Accounting for costs avoided by Telenet would guarantee that competitors do not
need to pay for services they are not supplied with.
The point is that BIPT appears to use its methodology to lower wholesale rates and facilitate
entry for third parties, without taking account of the proportionality of its approach and without a
clear reasoning for the methodology it follows. The draft decision does not clarify in any
meaningful way how the changes fit within the purpose of a retail minus methodology and how
these changes are aimed at facilitating efficient entry at the retail level.
An approach based on the avoidable costs of the additional services fits much better with the
logic of a retail minus approach and would be closely tied to the European framework for the
assessment of exclusionary abuse of dominance, in which the starting point is also that entry
should be possible for equally efficient competitors:
“With a view to preventing anti-competitive foreclosure, the Commission will
normally only intervene where the conduct concerned has already been or is
capable of hampering competition from competitors which are considered to
3
be as efficient as the dominant undertaking.”
It is also telling that BIPT assesses the additional services one by one, deducting the
hypothetical market value of each of them from the retail price, but that it does not in any way
assess whether the outcome of the application of its methodology – the sum of the individual
steps – leads to reasonable and proportionate results, even though the consequence of the
individual step is a considerable decrease of the wholesale price.
3
Guidance on the Commission's enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct
by dominant undertakings, Communication from the Commission, OJ C 45/7 of 24 February 2009, paragraph 23.
RBB Economics
Page 10
4.
Promotions
4.1. BIPT’s assessment
When assessing the question how promotions should be considered for determining the minus,
BIPT states that the regulated firms have high market shares resulting in important scale
efficiencies.
In BIPT’s methodology this would result in relatively low avoidable promotion costs as these are
spread over the entire customer base through a lower ARPU. Therefore BIPT suggests to (a)
treat promotion costs as avoidable costs, resulting in a higher minus (rather than a lower ARPU)
and (b) to (temporarily) incorporate promotion costs in the minus with reference to new
customers only. As a result, the minus increases and the wholesale price decreases. These
steps also result in a change in the formula that BIPT applies to calculate the minus.
In the decision of 11 December 2013, the minus is determined by the following formula:
𝑚𝑖𝑛𝑢𝑠 =
𝑡𝑜𝑡𝑎𝑙 𝑎𝑣𝑜𝑖𝑑𝑎𝑏𝑙𝑒 𝑐𝑜𝑠𝑡𝑠
𝑡𝑜𝑡𝑎𝑙 𝑟𝑒𝑣𝑒𝑛𝑢𝑒𝑠 − 𝑝𝑟𝑜𝑚𝑜𝑡𝑖𝑜𝑛𝑠
In the draft decision, the formula suggested (after the start-up phase) would be:
𝑚𝑖𝑛𝑢𝑠 =
𝑡𝑜𝑡𝑎𝑙 𝑎𝑣𝑜𝑖𝑑𝑎𝑏𝑙𝑒 𝑐𝑜𝑠𝑡𝑠 𝑝𝑟𝑜𝑚𝑜𝑡𝑖𝑜𝑛𝑠
𝑡𝑜𝑡𝑎𝑙𝑒 𝑟𝑒𝑣𝑒𝑛𝑢𝑒𝑠
As indicated above, account would be taken of scale disadvantages in the start-up phase by
determining costs and revenues for new customers only.
4.2. Comments
4.2.1. Change of the formula
Other than that the change in the formula would result in a higher minus percentage, BIPT does
not show why a change in the formula would be correct and why the original formula would
have been wrong.
De base formula to determine the wholesale price in the retail minus methodology is:
𝑤ℎ𝑜𝑙𝑒𝑠𝑎𝑙𝑒 𝑝𝑟𝑖𝑐𝑒 = 𝑟𝑒𝑡𝑎𝑖𝑙 𝑝𝑟𝑖𝑐𝑒 𝑇𝑒𝑙𝑒𝑛𝑒𝑡 − 𝑐𝑜𝑠𝑡 𝑜𝑓 𝑟𝑒𝑡𝑎𝑖𝑙 𝑎𝑐𝑡𝑖𝑣𝑖𝑡𝑖𝑒𝑠 𝑇𝑒𝑙𝑒𝑛𝑒𝑡
In this formula it does not matter whether promotion costs are deducted from the retail price or
would be added to the costs of the retail activities. In both cases the promotion costs result in an
identical reduction of the wholesale price.
Let’s assume that total revenues are 100, the avoidable costs 20, promotion costs of 10 and 10
subscribers.
RBB Economics
Page 11
In the standard formula, the calculation of the wholesale price in the original approach of BIPT
would be as follows:
𝑤ℎ𝑜𝑙𝑒𝑠𝑎𝑙𝑒 𝑝𝑟𝑖𝑐𝑒 =
100 − 10 20
−
10
10
In the new approach this changes to:
100 20 + 10
−
10
10
In both cases, using absolute values, the resulting wholesale price is 7.
Using the same numbers the minus percentage was calculated as follows in BIPTs original
approach:
90
20
∗ (1 −
)
10
90
Again the resulting wholesale price is 7.
In BIPT’s new approach this would be:
100
20 + 10
∗ (1 −
)
10
100
And again the wholesale price is 7: the change in formula suggested by BIPT should hence be
neutral and should in principle not impact on the wholesale price. What does happen however is
that a change in the formula results in a higher minus percentage: in the example from 22.2% to
30%.
BIPT itself consider that the change in the formula does “not allow for correcting the dilutioneffect in full”. That should indeed be the case because the change in formula should be neutral
to the wholesale price. Normally, and as shown above, the increase in the minus will be
compensated by the higher retail price to which the minus percentage is applied (and vice
versa).
This does not appear to happen however as the change in the formula is not accompanied by
change in the retail price to which the minus percentage is applied: in both the old and the new
approach, the minus is applied to the same retail price.
A change in the formula to correct for a dilution-effect is, considering the above, an inconsistent
approach that results in a lower wholesale price, when in fact the change in formula should
have no impact on the wholesale price. Moreover, the change of the formula comes on top of a
separate correction for scale which will be discussed below.
RBB Economics
Page 12
4.2.2. Compensation for diseconomies of scale
The right starting point for retail minus regulation should be that the regulatory regime facilitates
efficient entry, taking account of the avoidable costs of the regulated firm.
A (temporary) additional rebate on the wholesale price to account for diseconomies of scale is
as such contrary to this staring point, as such rebate aims at taking account of the costs of new
entrants that would use the regulated wholesale input. Total costs for the market would hence
increase in this approach, as new entrants would be allowed higher costs than the efficient
operator.
The purpose of this change is hence not to properly set the avoidable costs of the regulated
firm, but to facilitate entry. As BIPT itself states: “(…) in this way a balance is created between
the operators with significant market power and new entrants (…).”
A relevant question in this context is whether potential buyers of wholesale access services truly
face scale disadvantages, and with that whether a ‘bonus’ or entry allowance to facilitate entry
is proportional. BIPT’s reasoning would only be valid (though still not in accordance with a retail
minus methodology) insofar as new entrants would also indeed have scale disadvantages in
practice.
But this implies that the need to change the methodology, in line with BIPTs reasoning, should
be dependent on the identity of the potential entrant that would buy wholesale access services
from Telenet.
It is commonly known that the most likely entrant that would want to use the option of access to
Telenet’s network is Mobistar. Mobistar is an established mobile operator that itself already has
considerable scale, and for which it is unlikely that it would have significant scale disadvantages
if it would use the option of wholesale access. In the extreme, BIPT’s reasoning would even
allow Proximus an entry allowance based on assumed scale disadvantages, even though
Proximus is the largest operator on the Belgian market.
To give a significant entry allowance should hence be made dependent upon the specific
characteristics of the new entrant, or could be determined on the basis of an approach in which
a “reasonably efficient operator” would be defined.
Apart from the fact that giving a specific advantage within the retail minus methodology does not
fit well with facilitating new entry to equally efficient operators, it is not shown in the draft
decision why an entry allowance would be required. It is also not clear from the draft decision
why not giving an entry allowance would in practice be a significant impediment to operators to
make use of Telenet’s wholesale offer.
BIPT should have proven why the change in its own methodology would be necessary, and in
any event, this change should not be of general application, in particular because the change
assumes no scale efficiencies at all for a new entrant.
As this is an adaptation of BIPT’s own methodology, it would seem reasonable for BIPT to
distinguish between true new entrants, and firms that are already active in the Belgian market,
RBB Economics
Page 13
and to show in each case why an entry allowance would be required and why such allowance
would not be contrary to the principle that less efficient firms should not be facilitated.
The specific conditions on the Belgian market have resulted in a choice for a retail minus
methodology, instead of more intrusive regulation based on cost oriented wholesale price. The
way in which BIPT suggest to account for promotion costs risks that BIPT uses retail minus as a
veil to take steps that go in the direction of such cost orientation.
RBB Economics
Page 14
5.
Conclusions
As we have shown in this report, BIPT lowers the reference retail prices artificially by deducting
hypothetical market values for additional services, resulting in considerably lower wholesale
rates.
On top of that BIPT changes its methodology to compensate for assumed scale disadvantages
that new entrants would have, resulting in a further reduction of wholesale prices.
In combination with other elements in the draft decision (not discussed in this report), this
results in a decrease of the wholesale rates, relative to BIPT’s earlier decision by between 20
and 40%.
The draft decision does not show that BIPT has in any way looked at the combined effect of the
chosen approach: BIPT does not substantiate or motivate why such a downward revision of the
wholesale rates is proportional and justified, and whether the resulting wholesale rates are (still)
reasonable.
The draft decision therefore strongly suggests to have merely been aimed at decreasing the
wholesale prices by as much as possible and to facilitate new entrants to the largest extent
possible, rather than to facilitate efficient entry at the retail level in accordance with the retail
minus methodology.
RBB Economics
Page 15

Version non confidentielle 1 M. Dominique Vosters Président

Transcription

Similar documents

Retail Leasing Opportunity

list

Listing Flyer - Former Crusty`s Pizza

your card! - Benefits Corporation Helps Chambers Grow Their

IBF Customer Management Presentation - Daymaker

ForMer liNeNs `n tHiNgs @ gAteWAy stAtioN

sweet floam - imagearea.co.uk

1001 Chemin Valiquette, Saint-Agathe-des-Monts

Prime Retail Location For Sale/Lease Former Ken Jones Ski Mart 15

crescent village - Prime Commercial, Inc.

Telenet, at the heart of your digital future