xDSL Summit 2002
by John Latta
Wave Issue 0233 9/6/02
July 1-3, 2002
Nice, France
The WAVE Report has gone to Several IIR Telecoms conferences before and they were quite good. This was no exception. These events are usually not attended by individuals from the US and that is unfortunate. As the WAVE Report readers will see, there were many insights into the differences between broadband in Europe and the US. Developing an international market requires understanding the local conditions and this is especially the case in broadband Europe.
This event was about xDSL from a European perspective. It strongly reinforced the reality that if one brings a US mind set to Europe either it will be wrong or miss the point. We will review the similarities and differences in detail later in this report.
Here are some interesting statistics from the presentations.
For every 5 homes in Europe,
1 will have cable and DSL available,
2 will have just DSL available and
2 will not have any broadband.95% of those who buy broadband in France do a self-install;
In France there are 600,000 DSL lines. Only 600 are not provided by FT.
The objective of the DSL Forum in Europe is to drive towards 100% self install DSL based on retail purchase of modems.
Europe Compared to the US
During many of the presentations, similarities and differences highlighted the European environment compared to the US market.
Similar
The depressed telecommunications sector is also depressed in Europe. Many of the CLECs have disappeared. Only a few are left standing who will offer either or both broadband and voice.
The PTTs, which are similar to the ILECs, vociferously protect their monopolies. As the German regulator stated "Every time we issue a rule DT takes us to court."
Many of the PTTs are now public companies and their actions are very much driven by the markets, value of the company, ability to raise funds for CapEX and similar considerations to those that directly impact the ILECs in the US.
The broadband CLECs are very anxious to offer voice as it is felt this will lower churn; however, it remains unclear which service, VoDSL or VoIP will win in the long term. By every indication the only viable technology today is VoDSL.
Although not discussed specifically here, but which I picked up in conversations, the broadband providers do not know what will drive broadband adoption once the market has gone to those only seeking speed and always on. This is very consistent with what we observed at Cable 2002.
Pricing of the wholesale offering of local loop access is the critical element in making CLEC services profitable or a loser. This remains a major contention point, as the PTTs seek high rates, and the regulators and CLECs seek just the opposite. The net result is time is lost in gaining market access while the PTTs press ahead. Yet, a CLEC from Denmark, a country with the lowest ULL rates, stated they cannot make money with residential service.
DifferentEurope is DSL committed - this is the broadband technology of choice or necessity. The US is cable committed, this being the dominant broadband delivery infrastructure.
Increasingly the EU is playing a major role in setting competitive telecommunications policies, in much the same way that the FCC does. However, the EU is setting polices over its member countries, which are fighting many of the rules that are intended to break the role of the PTTs.
The emphasis in regulation for broadband delivery is DSL and practically no attention is paid to cable, given its low penetration. The key regulatory action is ULL - unbundling the Local Loop (ULL).
The PTTs have engaged in a long struggle to deny the CLECs access to the actual copper to the home or business. That is, they seek to provide equipment in front of the local loop. This can have the practical impact of limiting the bandwidth and possible services provided. In Germany, at least, the courts have turned down two such an appeals to limit access to the local loop.
While the EU seeks to set in place an open market approach to telephone access and ULL, there is a fear that in the 12 months before the actual rules go into effect in each member country, using the NRA - National Regulatory Authority, the PTTs will have totally dominated the market for DSL based broadband. One CLEC stated in response to my question, "By the time the EU implements deregulation we might be dead as companies."
Business plays a major role as customers for DSL and broadband offerings using the ULL. In fact, a CLEC from Denmark stated, "It is very hard to make money on residential customers." It appears that the PTTs are dominating the residential market as they are the total level of installs.
With the strong emphasis by the CLECs on using DSL to service business customers, this undercuts the PTTs and their cash cow SLA communications services such as E-1.
For reasons not clear it was stated that the performance of DSL is very good. In particular, one CLEC in the UK stated they were getting 2Mb/s second reliably. There was no discussion on lack of coverage or bandwidth limitations.
Deregulation of Telecommunications at the EU
Leo Koolen spoke, from the European Commission, Information Society Directorate - General. This is the agency that regulates telecommunications policy for Europe. He was a very good speaker and the talk was eye opening vis-à-vis the developments in Europe to foster broadband deployment. His talk was another indication of the role that the EU is playing in deregulating the European market. Key points made include:
With uncharacteristic speed the EU decided at the Lisbon Summit in March 2000 to make ULL a part of eEurope. By December 2000 regulations were decided upon to implement this.
In May 2002, five member states were cited by the EU for lack of a complete reference order which implements the regulations (Denmark, Germany, France, Finland, Ireland and The Netherlands). Germany was also cited as not implementing shared access on the LL.
By May 2002 there were 750,000 ULL access lines.
By May 7, 2003 all member states are to have rules in force for ULL and these will be based on market studies in the individual states.
It remains unclear how much choice a consumer will have with the ISP on broadband services. The EU is doing a study now of this issue which they described as the horizontal vs. vertical approach to broadband content supply. Early study results are bifurcated. That is, some broadband suppliers feel it is important to provide a complete food chain in the broadband offering, while others are only interested in providing connectivity.
In an interesting comment, Leo stated that "Digital television over mobil is closer that one thinks."
Germany Regulator View
Points made include:
There are 52.28m telephone lines in Germany and 1.58m are provided by new entrants.
In some cities, such as Berlin, up to 4 carriers are available for telephone service.
There are 730,000 ULL lines as of 4/02.
There are 35 ADSL providers, 9 national and 26 regional. DT has 2.3 m ADSL lines, is expected to grow to 5m by 2005 and 20m by 2010. However, all other DSL providers have only 70,000 lines. Even more narrow is that only 30,000 households have broadband provided by cable operators other than DT.
The wholesale cost of a ULL in Germany is 12.48 euros, while the European average is 13.2 and Denmark is only 8.2 euros.
Telenor
Telenor ran a VDSL trial which they called: Stavanger. In preparation they had to lay fiber and the total number of homes which could be served was 15,000 with 17 fiber nodes. The actual number of homes which participated was 750. Key points from the test were:
There were four bundles which began with just broadcast TV, Internet access and telephony. The last bundle added content on demand, interactive TV and an electronic program guide.
The results were quite consistent with many tests in the US. High speed Internet was the most popular, especially the always on feature, the users would pay for multiple PCs and multiple television streams, however, in spite of the ability to offer interesting content, Telenor found it does not mean the consumers would pay for it.
The harshest words came for VoD. Telenor saw this as competition with DVD players and content in the home. The high penetration of DVD and reasonable cost of content makes it difficult for VoD to complete, especially when the content offerings on VoD lag so far behind even DVD. They stated that the typical release window for VoD is 40 months after theatrical release.
Of the 750 users, 450 chose the premium offerings which included Internet at $33/month (41% of the total) and all television channels for $44/month (26% of the total).
Telenor was positive on the adoption of the services but concerned about the outside plant issues including installing cabinets to reduce the distances to the home and in-home wiring. In this later case the installers spent 4 hours setting up the cable, STB and television connection.
UK Environment
Much has been made about the UK competitive environment due to the policies of BT to frustrate the market. There are only two competitive carriers left: BullDog Communications and Easynet. They both gave presentations. One of the most interesting was Easynet's approach to the market. They only seek ULL when the market has been proven. The method Easynet uses is to go into a market and sign up subscribers near a CO, nearly all business subscribers. This is done as a reseller of BT services, thus, the profits are low but there is no investment required. When there are 25 - 50 subscribers this is described as a "hot spot." Easynet will then convert these subscribers to ULL and the margins go up significantly. Thus, the model that they use minimizes capital expenses and maximizes the gross margins on those loops which they covert. They were quite proud in claiming that this is the opposite of the "build it and they will come" approach to the market.
BT Wholesale
Very consistent with the comments we heard during this conference, BT, once they supported the wholesale concept for DSL services, leapt in with both feet. Marcus Jackson, of the BT Wholesale division gave an interesting talk. Points include:
The cash cow at BT are the 20m fixed line customers. It has targeted that by 2006 around 5m of those will have DSL. However, with the same chorus we have heard before - it is not clear what applications will drive adoption and usage.
Under Broadband England BT has an obligation to make broadband available to every home in the UK. This is a considerable challenge given the rural areas. Thus, they are evaluating both satellite and fixed wireless to augment ADSL.
By June 2002 BT had already covered 66% of the households in the UK and 75% of the Internet using households with 1,115 equipped exchanges.
BT ran a trial of VDSL with 10 users and was quite satisfied with the results.
A challenge that BT faces is to not only implement DSL and VDSL to its customers, but to offer attachment to the equipment to competitors and still make money. In a refreshing turn, BT described how they are working on this basis to make BT Wholesale a profitable and viable entity in BT. It is unfortunate that this attitude does not exist with the ILECs in the US.
Across Many Talks
A portion of the presentations covered two areas: broadband services and EoVDSL. Rather than cover each talk we will extract some key elements.
Business Model
The most successful provider of ADSL is DT with 2.7m lines. Yet, it is not making money in the business.
MorganStanley reported on March 13, 2002: "Basic DSL is a defensive product, which safeguards against customer loss, but creates little value as a stand-alone business."
McKinsey and Company also reported that: "Broadband services in Europe will require:
Improved Local telephone economics;
Strengthened CATV industry; and
Open the gates to attackers.The main threat to DSL in Europe is triple play: voice, data and video.
A number of service providers spoke of the ability to provide bandwidth on demand. That is, from a web site be able to increase the bandwidth to ones own DSL connection in 5 minutes or less by just doing point and click on a web site. The same service can be downgraded in the same way when the bandwidth need is lessened.
All of the speakers see a need to go beyond a pure broadband play. Yet, there is just as much uncertainty in Europe as in the US on what will be accepted by the consumers. The key word is ARPU - Average Revenue Per User. A number of speakers spoke of the triple play but the cable industry in the US remains much more sophisticated than the service providers here in Europe.
Nothing was presented on home networking or home services as a means to increase revenue.
One speaker was critical of the operator mentality that says - "bring your services to me and I will take the income." The prevailing view is that since your application runs on my network I deserve the income. Yet, the speaker also credited NTT DoCoMo with shaking up this narrow mentality. DoCoMo takes 6 - 7% of the revenue stream and as a result they have 30,000 to 50,000 developers creating new services. The i-mode service is due in Europe in the fall and it remains to be seen how well it will do.
One speaker said "We cannot get video content for VoD," and this mirrored what we found at DSLcon. Unless there is a resolution of the copy protection issues being demanded by Hollywood it is unlikely this segment of the market will take off.
Going beyond ADSL
The telecommunications sector is in a dilemma. New technologies such as G.SHDSL offers disruptive price and performance. Ethernet over VDSL offers lower cost for operations. Yet, if CapEx is involved no one will spend the money.
Much to my surprise Ethernet as an Access solution got quite a bit of traction. This is not only about making Ethernet pervasive but increasing the bandwidth. Cisco gave a good presentation on the topic and is pushing this technology. They have a solution from the cloud to the home - typical of Cisco. A point that Cisco made is that 1Mb/s is hardly broadband and we must work to drive the bandwidth up. Ethernet as a first mile copper solution does not make sense unless the bandwidth increases. Thus, the distance for EoVDSL is < 1Km when copper is used.
The goals in the IEEE standards effort are:
50/30Mb/s to 300m
36/12Mb/s to 1000m
15/3 Mb/s to 1500mCisco leads the IEEE working group on this.
Cisco stated that pre-standard products are available today.
A point made during the panel is that the EoVDSL is causing a rethinking about the implementation of broadband on the copper plant. As a result the TIA is relooking the issue and other standard bodies. All of this spells trouble for the incumbents, as long as there is an ULL. We regard Europe a leader in the competitive landscape.
G.SHDSL was cited multiple times in the presentations. We wondered why is this important in Europe?
Our answer:
G.SHDSL goes to the heart of the business of the PTTs. For example, BT makes $30b a year in business services and G.SHDSL allows the CLECs to directly attack this market. For example, BT claims that G.SHDSL will cause interference in the telephone physical plant and are pushing very hard to have it blocked. However, many on the opposite side do not believe this is the case and studies show it does not. BT is taking a very conservative position and this not only protects its network but more important its business. The CLECs want G.SHDSL because they can go after the cash cow of the telecommunications business.
During the meeting some interesting views came from the DSL Forum, during off-line conversations.
ADSL is much more complex that DOCSIS, and it is wished, at the DSL Forum, that a similar activity had been undertaken to create one standard for DSL. Currently, there is Annex A (US and UK), Annex B (ADSL over ISDN) and Annex C (Japan). This makes it nearly impossible to have one implementation of DSL. Further, there are regional variations of DSL which keep the same implementation from working over two "similar" phone systems. This is a real problem for mobil workers seeking to have one modem which works as one travels around.
Intel is pushing very hard to have one implementation of ADSL in silicon that will allow notebook users to roam and use DSL no matter where they are. It has been questioned if this will be possible on a notebook but it could be done on a motherboard.
Virtually all the ADSL modems, even if sold at retail, are verified by the PTTs. This means that even in retail the PTTs control the market. The problem is that the PTT testing process is slow and complex. For example, BT has tested over 100 DSL modems and found that they all passed. From this BT found that there are three key tests and if a modem passes these it will pass all the tests.
The DSL Forum is seeking to avoid these issues of PTT tests as a gate to the market and is establishing certified testing labs. The PTTs would really like to get out of the testing process as it is time consuming and costly. The DSL Forum would like to have this operational by the end of the year.
On a worldwide basis the only competitive DSL market is Korea. There is no real competition in Europe as the PTTs dominate the residential market.
Role of the CLECs in Europe
In the US the DCLECs, with the exception of Covad, have disappeared. In Europe the ranks have been greatly thinned. However, there are what appear to be a number of viable players, who presented at this conference. Yet, these companies are focused on the business market. They have left the residential market to the PTTs who are struggling to make money. But common between the PTTs and the CLECs is the technology - all are using ADSL as a means to deliver bandwidth. The CLECs get copper plant from the PTTs and sell high quality services. The CLECs, being very aggressive about finding new customers and stemming cash flow, are going for the business customers. As we reported on Easynet, it has an innovative approach to getting business customers by reselling BT services before making an investment.
Powerline for Broadband Delivery (PLC)
One of the speakers was mentioned he went to a powerline conference in Germany the prior week. His comments included:
Two years ago I would have assessed that powerline for broadband distribution was dead. Not the case now. Based on what I heard in Germany powerline is a done deal. I believe the modem issues have been solved. It is possible to serve all homes with broadband with powerline distribution. One point remaining is the back haul from the transformers.
Some of the power companies in Germany are predicting 20% to 25% penetration. The power companies, compared to the telecos which are poor of capital cash, have money to invest. A critical issue is that the power companies are seeking to retain their customers. Customer retention is so important that the power companies are willing to take a short term loss in order to keep customers.
Notes: Europe is unbundling power and with this retention of customers is very important to power companies. Offering broadband to the home is one way to retain customers. It is interesting that this argument is very similar to the ones offered by the cable companies for why they offer telephone over cable - to lower the churn rate. Thus, there is a huge incentive for the power companies to offer broadband in Europe, which does not exist in the US. In fact, the deregulatory environment for power in the US has made major steps backward due to the bad California experience. This is further compounded by Enron and its role in trading power in California.