PlusNet - Washing dirty linen in public

The problems at PlusNet are well documented, but last night the management team published a new re-organisation plan to try and solve the problems. In true PlusNet style, they have washed their dirty linen in public.

In short, PlusNet seem to have sacked the COO, Danny Sullivan and head of Customer Service and Networks, John Pickerskill: both of whom arrived at PlusNet in the 1st half of 2005 from BT.

It is admirable to see a company actually firing senior executives when things go wrong and seemingly really putting in the effort to fix problems. I suspect the bigger problem is the “not-so-free” broadband offers in the market. The share price certainly doesn’t look pretty dropping around 75% since the CPW offer kicked off the UK Broadband Wars.



On a side note, Plusnet still have all LLuing on hold with their wholesale provider, Tiscali, whilst problems are ironed out.

Thumbs Up for BskyB - Thumbs Down to TalkTalk (yet again)

It looks as if BSkyB have been busy bees upgrading the Easynet backbone network as well as fitting out lots of BT Exchanges. They have been installing the biggest current beast in the jungle, the Cisco CRS-1 Router, in their network. They now have a IP/Multiprotocol Label Switching (MPLS) network with a core ring capacity of a huge 1Terabit (Tbps), and is scalable to 4Tbps when needed. They installed 50 racks of equipment in 31 different locations in just 4 weeks.

Mind you, the Easynet security leaves something to be desired as someone has currently hacked the PR section of their website posting an offensive message to Israel and the USA.

On the other side of technical competence (ie zero) it looks as if the comedians at TalkTalk have also been busy unbundling my line at the BT exchange – the reason I know is that the voice quality has noticeably deteriorated and even worse is that my current (notso) broadband speed is “Downstream 443KB and Upstream 217KB with a limit on the upstream of 2MB.” On a traceroute, I only see Opal/TalkTalk IP addresses and the BT addresses are nowhere to be seen - ie I'm 100% on TalkTalk kit. I am getting really, really angry: especially with the prospect of spending tomorrow evening waiting on their pathetic call centre.

On another related point Mickey Mouse could have configured the TalkTalk DNS servers to provide a better service than the TalkTalk engineers. Unbelievably for such a basic and well known service, the DNS servers are up and down more often than a yoyo. Also, currently they are intercepting incorrect url requests and serving up adverts via a company called 404direct.com. I seem to remember in the distant past another ISP getting in trouble for intercepting url's and serving up adverts. It is a good job that I know the addresses of a few other non-useless DNS servers.

TeliaSonera: Bare Faced Cheek

How can a company one month proclaim to the world that conditions have changed and it now has a really sound project to launch a new network in a saturated market and the next month appeal to the regulator to reduce interconnect rates by 40-60% in order to allow it to launch?

The only rational reason is that TeliaSonera is trying its’ luck with the Spanish regulator starting off negotiations with outrageous demands. The regulator should reject the demands and tell the TeliaSonera executives to sort out its' own mess in Xfera.

Tiscali: Partner Pending?

There was an interesting comment at the weekend by the Tiscali CFO to an Italian newspaper where he said Tiscali was not ruling out seeking a partner.

The problem Tiscali have got is that they are short of cash and the German and Dutch ISP’s are not performing well. Of course, the UK is under the spotlight after the launch of not-so-free offers onto the market and the purchase of HomeChoice. The UK is especially important for Tiscali as it is the largest operation and contributes the majority of EBITDA – in Q1 2006 the UK contributed €15.7m out of a total group EBITDA of €29.2m.

In Sept 2006, Tiscali have to redeem an approx. €210m bond part of which will be repaid via new equity (approx. 7% issue of new shares) and partly through cash approx €130m. The cash element will be paid via an extension to a borrowing facility from a hedge fund. The total borrowing will be €220m at an incredible Euribor +8% (approx. 11%). Given that overall Tiscali is still burning cash, it will be extremely interesting to see how Tiscali gets out of its’ current cash squeeze.

It is in this context that I believe the decision by the owner of Homechoice to invest at the Tiscali UK level and not the Tiscali Group level should be seen. The majority owner of Homechoice was an ex-early Microsoft employee called Chris Larson who is probably sitting on big losses for his Homechoice investment having allegedly pumping in around £100m through the years. When he bought into Homechoice in 2001, he transferred the IPR to an offshore company which I suspect he still owns today with Tiscali having to pay a licence fee for the use of Video on Demand technology.

Tiscali itself is one of dotcom survivors having been formed by a Sardinian, Renato Soru who still has a significant 27.9% shareholding but has retired from the company to become the president of Sardinia. I suspect he will still exert significant influence on the company. In 2000, Tiscali meged with a scandal ridden Dutch based ISP called WorldOnline who happened to have €1.3bn of cash on its’ balance sheet and so became the original plan to build a pan-European network. In five years and after many years of wheeling and dealing the end result is no cash left, networks in the Italy, UK, Netherlands, Germany and Czech Republic and still burning cash.

Smaller shareholders in Tiscali include the Sandoz Foundation with 9%, who came along with the WorldOnline merger and are behind the Interoute Trans-European Fibre network and Kingfischer with 3% who can along with the French LibertySurf ISP purchase which has since been disposed of to pay past debts. The Sandoz board members resigned in May at the time of the agreement for the rearranged financing.

It is this context that I believe that Tiscali are struggling to find a suitable new partner for Mr Soru. Personally, I think Tiscali is lacking the financial firepower to compete in the new reality of the UK broadband market and will suffer from the lack of strategic direction from the Tiscali Group in the coming year. A cash strapped ISP will struggle to compete with BT, BSkyB, Orange and O2 who are prepared to sink many, many millions into the UK - £400m in the BSkyB case.

WiFi: Struggling Handling Handoff’s

Apologies to my long suffering readers who have been awaiting a daily dose of TeleBusillis, but I’ve been struggling to catch up since my holiday and fortunately have spent most of the day laughing rather than catching-up: a rather obscure paper on unusual behaviour in a heavily congested WiFi network has tickled my fancy.

To place this into context we have to travel back in time to a period where mobile handsets resembled bricks and a group of engineers were struggling in relative privacy to implement the “Great Software Monster” which has been abbreviated to GSM in today’s language. If my memory serves me well: the hardest nut to crack was the problem of handing over a call from one base station to another without the user knowing and the call dropping.

Step forward in time a decade or so and a different set of engineers were working on a different technology this time in the full glare of publicity on cracking a rather similar problem – handing over a call from a UMTS base station to a GSM station and back again without a call dropping.

I was reading today about the wireless network traffic at the 65th IETF meeting in Dallas, Texas in March 2006. It is well known that the current family of 802.11 standards does not perform well in heavily congested environments and 500 IETF engineers with laptops in a single room is probably as congested as you can get. The main conclusion of the paper was the abnormal level of handoff’s adding significant traffic to a congested network as clients seek better bandwidth from an alternative Access Points. Furthermore, different vendor implementations of the standard had different handoff performance with Apple appearing to have the best algorithm with Intel and Cisco vying for the worst. Even worse was the treatment of a “Level 3 handoff” which basically closed down all network sockets on the client and reassigned a new IP address before re-establishing sockets: the average time taken to detect the loss of connection was 2.5secs with the average handoff time being 1.2secs. In other words 3.7secs of “dead time” - hardly acceptable for a voice call or video watching for instance. It seems pretty obvious that the 802.11 standards aren’t really designed to deal with handoffs efficiently.

Now with the blueprints for the next greatest technology (Wimax or 802.16) being argued about (sorry discussed) in the IETF corridors, I wonder if the biggest problem they’ll have to address is handoffs or will they brush it to the side like WiFi?

By the way the paper was “IEEE 802.11 in the Large: Observations at an IETF Meeting” by Andrea G. Forte, Sangho Shin and Henning Schulzrinne. Professor Schulzrinne is seen by many as the Godfather of the SIP protocols so I seriously doubt that anyone can doubt his impartiality.

Mobile UK: Time for Distribution Surgery

The recently appointed T-Mobile CEO, Jim Hyde, who originates from the USA is certainly shaking things up in the UK distribution channels.

So far, the best thing he has done is banning outbound telesales companies from marketing the T-Mobile products. These companies are a curse on the industry and I can’t believe they generate loyal customers for the cellular companies. Personally, I have received calls both on my landline (despite being a member of the normally excellent TPS scheme) and on my mobile: each time the calls have been selling the 3 UK products. I wish all the other network operators would follow the T-Mobile lead and ban outbound telesales. However, I doubt this will happen anytime soon.

In my opinion the most important thing he is trying to achieve is to reign in the power of largest phone retailers: Carphone Warehouse (CPW) and Phones4U. In order to achieve this, he is trying to improve his shop footprint and more importantly dropping commissions on T-Mobile products. He openly says that the power of retailers is due to the shortsightedness of the networks which is a position that I completely agree with.

In my opinion, we are at a point in time where power can be easily taken away from CPW and Phones4U:
• there is the closure of The Link which means there is less competition on the high street;
• there is the change in ownership at Phones4U which possibly will mean a change in direction and perhaps an opportunity for the operators to renegotiate deals;
• there is the rapid development of 3 UK’s, T-Mobile’s and Virgin Mobile’s retail presence;
• there is the tendency of the mobile networks to offer special deals through the online and inbound telesales channels supported by heavy advertising campaigns; and
• Most importantly is the CPW push into broadband via TalkTalk.

The push into broadband basically gives Orange and O2, which are big competitors to CPW in the broadband space, a big incentive to stop feeding CPW mobile commissions. I am totally bewildered that currently Orange’s products are heavily featured on the CPW websites and therefore Orange is probably paying CPW the largest commissions.

CPW are basically reliant on the short termism of the mobile operators: the network operators all make noises about reducing dealer commissions and then with the first sign of a reduction in gross additions they offer CPW higher commissions to push their products. Until I hear Charles Dunstone eating humble pie and saying that commissions are reducing, I won’t believe that the operators have altered their current suicidal short term strategy.

The latest trick that I see CPW is employing is the bundling of other products with the handsets: for instance bundling a Nokia phone with an Xbox (doesn’t Nokia see Microsoft as its’ biggest long term threat?) or even a Samsung and an iPod (aren’t phones nowadays offering music capabilities?). I see this strategy driving the handset manufacturers crazy. If Charles Dunstone doesn't start being a little more careful he will be left with zero friends in the whole telecommunications industry which admittedly would be quite an achievement, but not something that the CPW shareholders should be happy about.

Beeb – “Give me the Digital Dividend”

I always thought that the aim of the switchover from analogue to digital TV in the UK would bring new spectrum onto the market which would be auctioned off by the UK government leading to either innovative new services or a reduction in price of existing services.

In todays Sunday Times was a lengthy article about what could happen to the new spectrum, the BBC’s position was clear:

“We believe that public-service broadcasters should have access to the spectrum they need to get their services to audiences. Charging them for that spectrum simply takes money out of services and programmes.”

There it is: 20 years of UK spectrum policy is thrown out of the window because some public sector monopoly insists that they should have as much spectrum as they think they deserve. Fortunately, the UK regulator OFCOM is currently performing a review of what should happen to the spectrum which is due to finish in Q3 2006 and they will publish the results in Q4.

However, I do not believe the BBC will accept the results and seeing they have managed to exempt themselves from regulation from OFCOM, the inevitable result will be furious lobbying to politicians and the undermining of OFCOM.

My personal position can be summarized as follows:
• There is no case for High Definition TV being a mandatory public service requirement served via digital terrestial television (DTT). It can be adequately served to interested members of the UK population via. services from Sky (satellite) or ntl (cable) distribution platforms whether as license-fee paid for BBC or more overtly commercial stations;
• OFCOM should auction the digital dividend spectrum to all and sundry without any technological constraints on the use of spectrum; and
• Public Service broadcasters such as the BBC or Channel4 should be allowed to bid for the spectrum as any private sector company, whether existing or start-up. If the BBC won and needed to increase its' tax (oops sorry – the license fee) then so be it.

In order to be perfectly clear, I think that both the following extreme scenarios are perfectly OK:
• Intel bid for and won the entire available spectrum for WIMAX deployment. Intel then built a network and licensed a couple of companies (eg BT) to provide service; and
• Qualcomm bid for and won the entire available spectrum to provide MobileTV service via its’ MediaFlo technology and not the European DVB-H standard. Qualcomm then built a network and signed up one customer, Sky, to provide service via the mobile operators and various MVNOs.

My underlying concern is that rational economic arguments are going to be ignored as the BBC will start playing emotional cards in order to get the spectrum without going to auction ie "free". If the BBC succeeds, it will only mean that more costs are hidden from the British public about the true cost of the BBC, which is exactly the BBC's strategy.

Broadband Britain: More fallout

Euro1Net has had its’ umbilical cord cut by BT Wholesale for non-payment of invoices. This leaves around 5.5k broadband customers in deep trouble. A notice on the company website says it can’t issue customers with MAC codes and therefore they are prevented from swapping broadband suppliers.

The worst of it is that Euro1Net’s unique selling point is that it allows its’ customers to prepay broadband costs for 12 or 24 months, so customers will be faced with the huge difficulty of getting money back from a company teetering on the edge. We shouldn’t laugh at the afflicted, but a BBC News reporter seems to have been caught out.

OFCOM claims to be watching the situation closely, but I’m not sure what OFCOM can do apart from change the rules regarding issue of MAC codes to allow BT Wholesale to issue them in the event of a broadband supplier going bust. I suspect this type of problem will become a feature of the market rather than the exception as the market consolidates.

Another small UK ISP, BrightView, with around 40.5k customers is also heading for trouble: the parent company Inox who released annual results this week have a division up-for-sale and are renegotiating with Barclays Bank about its’ overdraft. Inox has an interesting history seemingly jumping from one fad (ringtones) to another (scratchcards) without ever making a real impression or return for shareholders. The purchase of the BrightView ISP in 2004 for £25m looks like another case of unfortunate timing. The shame is that BrightView are actually considered quite a good ISP with brands such as Madasafish winning awards and Brightview being trusted with the Waitrose supermarket brand for reselling broadband. The directors expect consolidation ahead in the market and perhaps this is a sign that the ISP is also up-for-sale. They better be quick because Inox’s market capitalization has fallen by 80% over the last 12 months to approx £5.3m which is far less than they paid for BrightView and they raised £15m in equity at the same time!

I suspect that the small independent consumer ISPs are heading for a really harsh winter.

Hutchison 3G – I don’t believe a word…

Hutchison Whampoa (HWL) released its interim results this week and the obvious conclusion is that the 3 Group is still eating cash, albeit at a slower rate than last year. I estimate the ballpark cash outflow at £676m versus £1,268m last year. The Sunday Times have today tried to explain the harsh realities facing Hutchison 3G, but I think it is extremely difficult to draw any conclusions due to what I consider persistent “extreme KPI gaming” and I think it is more fun to play – spot the games.

Game 1 – Customer Figures

HWL seem to have a reporting cycle all of its’ own – the customer figures are the number registered on the day before the results are released rather than the closing figure for the reporting period. For example in the six months to 30th June 2006, released on the 24th August, the customer numbers relate to the 23rd August rather than the traditional 30th June.

I also note that The Sunday Times quoted an internal source saying that as many as 40% of 3 UK prepaid customers are inactive. H3G UK has certainly had a torrid time in the prepaid segment and I’m sure that today they are wishing they never bothered with the segment.

The net effect of this is that comparisons are impossible to make across the sector. Also standard “per customer” ratio’s for the 3 Group are impossible.

Game 2 – No Churn Percentages

The 3 Group does not release churn percentages: there have been several rumours during the year of huge figures for 3 churn in the UK as high as 60%. HWL seems to have addressed this problem by stating that “churn for the first-half was higher than expected…churn has been progressively reduced and was approx. 3.6% for the month of July”

It was obvious to me right from the start of the huge push in 2004 that 3 UK would have a huge problem with churn. I thought the problem was that effectively 3 UK were giving away a handset and a huge bucket of minutes and texts for “free” over the contract length. Even worse was the mechanism that 3 choose to give away the “free” service which was via a “cashback” mechanism, thus allowing 3 UK to book revenues and booking the discount in the “customer acquisition cost” line. This was a recipe was charlatans and crooks to enter the industry and for extremely dissatisfied customers. It is hardly surprised that recently 3UK have been making noises about building its’ own distribution channel.

Game 3 – New Definitions of Profitability

I know a few accountants who think EBITDA is quite a controversial measure of profitability when used as a valuation metric, because it doesn’t reflect depreciation and amortization charges. Every accountant I know thinks “EBITDA before Customer Acquisition Costs” is a total travesty as a serious metric. HWL dreamed up “EBITDA before Customer Acquisition Costs”.

The only positive about the HWL definition is that it places a focus on the importance of Customer Acquisition Costs on overall profitability. For instance, the 3 Group spent approx. £210m on acquiring prepaid customers in the first six months despite saying that they are not that interested in this segment. Given that total prepaid revenues for the six months were £473m – customer acquisition costs represent a huge percentage (44%). Net prepaid customer numbers grew around 500k in the six months (assumption subject to game 1 trickery) giving a net cost for a prepaid add of £420. For a normal operator who releases churn figures we would be able to work out the % relating to growth (ie net adds - 500k) and the % relating due to ongoing business (replacing churners) with HWL we are clueless and therefore the cynics assume the worst.

Game 4 – Ridiculous ARPU Figures

It only takes a little look at the UK prepaid figures to realize how ridiculous the 3 Group figures are: the UK& Ireland had prepaid revenues of £66.665m, given that there is no growth in the subscriber numbers over the period and the prepaid customer numbers were 1.57m on the 22nd March then in the six months the ARPU was approx. £7.07. The published ARPU (in the trailing twelve months) was £15.49. There is obviously a huge game going on here.

Given the huge percentage discrepancy above and the fact that 3 ARPU’s are declared before promotional discounts, I don’t trust the figures at all, even with the huge incoming interconnect charges. The sudden increase of 20% in UK APRUs should be seen in this context.

In my opinion, the only innovation that 3 has introduced in the UK is the concept of “cashback” and I believe this is the most self-destructive pricing policy ever designed encouraging dubious distributors and fraudulent schemes. “Cashback” schemes fit perfectly into the HWL business model of ignoring acquisition costs and churn.

Game 5 - Non-Voice Revenues

It is almost impossible given the fog of war relating to “cashbacks” and the bundles on offer by 3 UK to have any faith in the non-voice revenue percentages.

For instance, currently the Music & Talk & Text 700 offer has been reduced from £37.50 to £18.75pm for a 12 month contract, in this deal is:
• 400 mins x-network voice (non-geographical extra)
• 250 texts
• 25 UK Video call minutes
• 25 Picture/Video messages
• 5 Music Track downloads
• FREE internet surfing powered by Yahoo! Search

What is the percentage of data revenues in this bundle?

Furthermore, the music companies who are including “free” music in the bundle are running a big risk of devaluing their “download” product, in much the same way that 3G UK has created a problem with the quality of their customer base by giving away service. However, the music companies are well known for their lack of foresight and nothing surprises me about them. You would have thought that with the music industries current problems of illegal “free” downloads; they would not be supporting another “free” scheme.

Game 6 – Capital Expenditure

I was extremely surprised with the level of Capex in the 3 operation - £289m for the six months, down from £474m in the previous year. This seems extremely low to me for a company that should still be building out its’ network in the UK, Ireland, Italy, Sweden, Denmark, Austria and Australia and adding new services (eg HSDPA). For a comparison, Vodafone across the whole of 2005/6 spent £665m in the UK and £541m in Italy.

Two main conclusions can be drawn from this, either:
1. HWL are constraining Capex to improve cashflow figures; or
2. Vodafone are overspending.
I have a little sympathy with both of these answers and there is probably a grain of truth in both. Vodafone capitalize a lot of internal work on IT systems and Network Design and Build to be written off during many years below the EBITDA line in the “Depreciation” line of the accounts, thereby “hiding” a lot of staff costs.

There is a third more potentially harmful conclusion and this relates to the mega-outsourcing deal done with Ericsson in the UK, Italy & Australia for IT and Network Operations. I’ve been thinking that perhaps some of these outsourcing deals in general are done on a “PFI”-type basis, which really is just another expensive type of off-balance sheet funding. This would fit in with some of Ericsson declarations that profits were not as high because of the timing of large outsourcing deals in their results. I would think this type of deal to be reflected in contingent liabilities, but I can’t see any signs in the HWL accounts, so perhaps my fears have no basis. I really hope that my thoughts are incorrect and this is not a growing trend in the industry: I can still remember the difficulties caused by vendor-financing of network builds in the bubble years.

Game 7 – Purchase / Sale of Assets

When the 3 Italia IPO failed in February, HWL announced that it had sold 10% of the company privately to a group of private investors which valued the company at €9bn. It also revealed that with the deal came an option to sell-back the stake to HWL. Now, it is revealed that this deal has to be treated as debt under Accounting Standards and therefore not a sale at all. Personally, I think the transaction was more about saving face than realizing value in the 3G subsidiaries.

Conclusion

It seems at first glance that things are improved albeit slower than the targets that HWL set. However, given the games the HWL seem to play personally I don’t believe a word or number that HWL publish relating to the 3 Group operations.

On Holiday...

...be back in a week.

AWS-1 Auction Day 8 – SupaLicenses

There was no bids at all today on the Supa-Licenses; yesterday, there was only one bid. It therefore seems sensible to review the status before the Auction moves into Stage 2 and possibly the action starts to pick up again.



The first thing to notice is that the licenses are dominated by T-Mobile and Verizon Wireless. Currently, T-Mobile has full lower-48 state coverage with 20MHz in 3 of the areas and 10MHz in the other 3. Verizon Wireless has 20MHz in 4 of the 6 markets. Effectively, T-Mobile can still bid on all the regions and have enough bidding units to push for 20MHz in all the regions if they so desire.

However, Verizon Wireless does not have enough bidding units left for 20MHz for each lower 48 market. Currently Verizon Wireless is pretty quiet - they haven't made a single bid in 3 days or 12 rounds. On Monday, I think the auction will move into Stage 2 and because of the change in activity rules, Verizon Wireless will be forced to reveal its’ hand. In stage 2 each bidder is required to bid 95% of eligible bidding units rather than the 80% currently permitted. I believe Verizon Wireless has five options:
i) Stick with the licenses and spectrum they currently have and/or defend the licenses if someone else bids;
ii) Bid on 20MHz in the Central region;
iii) Bid on 10MHz in Central and 10MHz in the West;
iv) Move into specific markets in the CMA and BEA bands; or
v) Look to make an acquisition outside of the auction.

Effectively the decisions made by T-Mobile and Verizon Wireless will be driven by the other potential regional license holders:
• Cingular holds the largest amount of spectrum already and whatever decision they make will not make them a bigger threat than they are already. Cingular doesn’t have enough bidding units to prevent T-Mobile getting the nationwide license.
• MetroPCS is financed by the Private Equity industry and is a relatively niche player with around 2m customers in Atlanta, Miami, Tampa, San Francisco, Sacramento and Detroit. I am really surprised they are bidding in the supa-regional licenses and even thinking of a major build-out in a new area for MetroPCS like the North East. As well as bidding US$908m on two 10MHz supa-regions, MetroPCS has also bid US$248m for the biggest 10MHz BEA license available in New York and New Jersey. It is anyone’s guess how much money the Private Equity people are prepared to pump-in, but I think the incumbents will want to keep another new player out of these markets. There is another wild-card is as much as I’m sure the Private Equity people will also quite happily trade the spectrum or even sell the company to a buyer. MetroPCS is a CDMA operator.
• Leap Wireless is also interesting because they normally operate in secondary markets and target a different market segment than Verizon Wireless - they basically serve the “value conscious consumer” or someone looking for a bargain. They are a quoted company with US$2.6bn of turnover, made US$25m of profits in the first six months of 2006, have low overheads and also operate a CDMA network. In short they are hardly a threat to Verizon, Cingular and T-Mobile and they could be a potential purchase for Verizon in future years. Somehow, Leap is getting a start-up credit for bidding in the name of Denali.
• The final bidder left on the Supa-Licenses is a strange start-up called Barat who are actually controlled by US Cellular who in themselves are a pretty strange company and already operate out of Chicago – and again hardly a threat.

In summary the only bidder left in the Supa-licensees who could potentially upset the current status-quo is MetroPCS. The auction will probably move into Stage 2 on Monday where Verizon Wireless will be forced to reveal its’ hand for the end game.

Reseller Double Play Summer Madness

Virgin Mobile and Virgin.net, both owned by ntl and yet resellers on the T-Mobile mobile network and BT wholesale broadband network have teamed up to offer yet another twist on multiplay market.

For £15/month you basically get:
• 300 mins Geographic calls to any mobile or fixed network
• 300 Texts
• Up To 8Mbps download speed
• 6GB/month download allowance

Notice that in the bundle you do not get:
• Subsidised mobile handset – its’ a SIM-card deal
• BT line rental (an extra £11/month)
• Any BT Calls
• Non-geographic Calls

Even crazier is that the Virgin Mobile Sim-free deal doesn’t come with a contract length – so if you’re bored of your handset after 6 months and want a new one for free in another type of bundle - you can leave both services

Personally, I cannot possiblly see, given the charges ntl will have to pay to BT and T-Mobile, how Virgin can possibly make any money whatsoever on the deal. Virgin.net and Virgin Mobile must be desperate for customers. The offer is only available under the end of October when either people will come to their senses or another crazy deal for the Autumn will be dreamed up.

AWS-1 Day 7 Summary

Unfortunately, I was out and about yesterday evening and unable to analyse the days activities. I will catch-up this weekend.

In brief, it looks as if bidding was slow with the differentials further declining and a lack of activity in the "F" band. Mind you the US Treasury still managed to earn US$768,724,100 to take Total Net Bids to US$10,890,284,600

Of the major's Cingular appears to have dropped its' bidding units further. Of the medium players, it looks as if Charles Dolan is pulling out.

I expect today to be another quiet one with everyone looking forward to the weekend which will give an opportunity for the players to potentially revise strategy and decide upon how they are going to play their cards in the end game.

OFCOM: Facilitating Broadband Churn

A few unscrupulous ISPs have been playing dirty when customers have requested to change suppliers for broadband services. These methods vary from the reluctance to giving out the MAC code required for a clean, uninterrupted service provider change to placing tags on lines to prevent new owners of houses moving to a different provider. The MAC code system is completely voluntary at the moment and OFCOM has no jurisdiction. This scenario is just for the standard BT Wholesale service.

In we add to this, the undoubted chaos that will ensue when the early adopters of LLU are at the end of their current contract. We have a situation made for regulation and OFCOM plan to fill the gap. Normally, I’m completely against regulation, but it is apparent that the ISPs have brought this one upon themselves as the Mobile Operators did before them.

The only thing I’m extremely disappointed about is that OFCOM plan to exempt ntl from the legislation. Personally I think this is completely unfair and delivers an advantage to the largest broadband player.

AWS-1 Auction - Day 6 Summary

Today, there was very little action on the supa-regional D/E/F bands: 0 bids on the “F” band; 3 on the “D”; and 6 on the “E”. Verizon Wireless didn’t make a single bid all day. However, the total number of bids on Day 6 was higher than on other day apart from the initial day. This is because today was all about reducing the price differential between the A/B/C bands and the D/E/F bands. With the differential on the key A and C bands still standing at 6.06x and 3.9x, as opposed to 11.2x and 7.5x yesterday, realistically I expect tomorrow to be another day of “differential reduction”

The U.S. Treasury will not be crying too much - they still raised an additional US$1,641,242,350 in the day.



In terms of Cingular, they have joined Verizon Wireless in having less bidding units left than is required to buy 20MHz of mainland USA license. So only T-Mobile and SpectrumCo are left with more than enough bidding units left to buy a nationwide license.

We are getting to the stage in the auction that the major bidders remaining (ie Verizon Wireless, Cingular, SpectrumCo and T-Mobile) are looking at the medium sized bidders and trying to figure out how much they can really afford to continue bidding.



I suspect that overall Cingular and Verizon Wireless are well within budget and can afford it if the prices go a lot higher. I also suspect that T-Mobile who with their current availability could gain at a maximum around 40MHz of nationwide spectrum can afford the price even on the “F” exposure figure. The wild card is SpectrumCo – who knows the details of the partnership agreement, let alone the strategic thinking of the partnership.

Leap Wireless the backer of Cricket and Denali announced today it would raise an addition US$250m in equity. However, its exposure is much greater than this and Leap is already heavily geared. The big players will have already spent a lot of money doing an analysis of Leap Wireless and the other medium sized bidders – how much can they really afford? When I say afford, I mean not only spectrum costs but also rollout and launch costs in new markets.

After all, when Nextwave bid a huge amount in previous auctions and ultimately went bankrupt and the spectrum went unused for many a year, the only real winners were the incumbents in the Nextwave regions.

LLU can seriously damage a companies health

Last Friday a medium sized error prone UK ISP called PlusNet placed its’ LLU migration on hold. PlusNet does not have its’ own LLU equipment in the BT exchanges, but instead uses Tiscali as a wholesaler. In total, PlusNet claims to migrated 12,000 customers.

This is after the original UK LLUer, Bulldog, had such severe problems that OFCOM was called in to investigate. Wanadoo/Orange, a LLU wannabe, has revealed severe problems in the past as well.

According to the todays BT Openreach figures, we are now up to 682,432 unbundled lines and the weekly amounts have dropped significantly in the last couple of weeks (the last 4 weeks have been 10k,2k,17k,27k). It is still taking 10 days to complete an order.

This is all before vast backlog of Carphone Warehouse unbundling occurs – if the troubles are the same, Carphone current problems could pale into insignificance to what lies ahead…

AWS-1 Auction - Day 5 Summary

A wild day – methinks the beginning of the end of the auction.



Wireless DBA aka Echostar / DirectTV / Rupert Murdoch / John Malone are out!

This is despite the placing of the biggest deposit of US$$972,546,000, they will be exiting with absolutely nothing and they have American Cellular Corporation to thank that they have not been left with 15MHz covering Alaska. American Cellular Corporation is part of Dobson Communication Corporation and already offers service in Alaska. Theoretically Wireless DBA are still in the Auction, but in practice they will exit in Round19.


The next biggest story of the day is that Verizon Wireless took a waiver at the start of the day in Round 13: T-Mobile won every “F” license. In Round14, Verizon Wireless came back strongly and won every “F” license. Then in Round 15, T-Mobile didn’t bid on every “F” license. Strangely in Round16, Verizon Wireless didn’t bid enough to maintain its’ activity level and now does not have enough “bidding units” left to win a license covering the whole of mainland USA. I thought when Verizon Wireless used its’ waiver first thing in the morning that it was going to come back with bids on the smaller BEA and CMA licenses which are still relatively cheap. It is almost as if Verizon Wireless were bluffing to get T-Mobile to pay as much as possible for “F” bands and it has all gone wrong. T-Mobile still have a huge number of bidding units available equating to over 40MHz of nationwide spectrum – the only thing stopping them now buying a huge amount of spectrum and securing the future is the size of the parent’s wallet.

SpectrumCo have now moved into the BEA and CMA bands and are still bidding heavily and have the equivalent of 37MHz of nationwide spectrum left to bid on. Cingular have come off slightly as well, but can still, just, buy a nationwide 20MHz license.

Tomorrow, I expect the bidding on the supa-regional blocks – D/E/F to slow to a trickle and the action to move to the CMA and BEA blocks – A/B/C. But I could be wrong (again)…

Box Breaking and Counterfeiting

It is interesting that that in the USA people are being arrested for blatant box breaking on the suspicion of being terrorists. They may be annoying a few operators and avoiding a few taxes, but I doubt that this can be called terrorism.

The box breaking phenomena has blighted the sale of prepaid phones in the UK for many a year now. It creates huge problems especially with the customer numbers, churn figures and ARPU reported by the operators, not to mention the unprofitable nature of subsidizing prepaid handsets and paying agents commissions for “non-sales”. I suspect that some of the prepaid customer figures in the States also are suffering from this type of inflation. I actually saw the video footage on Fox news and the phones which were densely boxed were Tracfone GSM phones, which no doubt were heading off to a non-subsidy based third world country. The use of SIM cards in GSM phones actually encourages this type of trade, whereas I’m sure there is a very limited market in non-SIM based iDen and CDMA phones.

An even more dangerous example of box breaking is also starting to appear in UK market. This time it is more dangerous to the unwitting buyers health than the operator’s pocket. This is innocuously called “mixing” or “blending” but involves the replacement of a manufacturers battery with a counterfeit battery in a new phone and subsequently selling the manufacturers battery as an original accessory. The problem is that with margins so slim for the distributors and for instance with an original Nokia 6230i battery selling for around £7 and a fake for £2. The margins on counterfeiting are higher than on selling the phone. Add to this that Nokia are allowing distributors to package the product themselves (eg the insertion of an operator SIM card and operator specific packaging) and you have the recipe for unscrupulous distributors to cause damage to the whole value chain. Again the use of a SIM card, actually encourages this type of behaviour.

My own opinion is that with the VAT fraud problem and also counterfeiting, it will not be too long before a manufacturer decides there is just too much reputational risk is dealing with distributor and try to build a network themselves cutting out the distributor’s altogether.

AWS-1 Auction – Day 4 Summary

What a day!

I wouldn’t be surprised if even The President of the USA finds out that today the FCC raised US$3,269,449,800 for the US Treasury.



The absolute net figure raised of US$5,673,103, 700 is only the start of the story, the supa-license band “F” has got to such a level that if the other bands sell for the same price per population per MHz: the total value raised at the auction would be US$12,005,208,000.

The differentials now are:

• 1.13x (1.21x day-3) for the combined REA D/E band;
• 7.5x (5.77x – day3) for the BEA B band; and an incredible
• 11.33x (6.15x – day3) for the CMA A band.

I am amazed that the bidders are not yet coming off the top bands and bidding up the values for the CMA and BEA licenses. I expect people to start seeing some sense overnight and run the risk of “stranded” licenses explained previously.

The record raised in an auction was US$16.857bn in 2000/1 for Auction #35 for the “C” and “F” bands. We are almost certain to beat this absolute figure now, but I doubt we will beat the price of MHz/pop as much more spectrum is on offer this time around.

AWS-1 Auction - Bidding Units

The following graphic explains the mechanics of the Auction Rules and why some parties are bidding on a lot of licenses and others seem to bidding hardly at all. Furthermore, it explains why the amounts reported as "Provisionally Winning Bids" are also meaningless. The bidding units themselves are proportional to the population of the region: smaller regions have smaller amounts, for instance CMA547 only requires 8,000 bidding units. 20MHz of mainland USA is equivalent to 254,002,000 (or roughly the population at the time of the 2000 census)

Each party started off with an original amount of bidding units and paid a proportional deposit. So Verizon Wireless elected to start the auction with 255,262,000 bidding units - just enough for a nationwide 20MHz license. Wireless DBS elected to start off with enough for the equivalent of a nationwide license covering an incredible 51MHz! Even accounting for the fact they need Alaska and Hawaii as well as mainland USA, 648,364,000 seems a little excessive to me.



At this stage of the auction, everyone has to bid at least 80% of the remaining units or else the bidding units are gradually reduced in proportional to amount of activity and therefore these parties can only bid for less spectrumin subsequent rounds. If you have a license in the previous round this counts in your bidding units.

It seems apparent to me that Verizon Wireless wants a 20MHz nationwide license and is bidding accordingly. Because of the 80% rule, they do not have to bid on every one of the supa-regions in every round and keep the ability to bid on everything, but they must bid on the biggest region, the Great Lakes, because that accounts for more than 20% of the total. Verizon Wireless have no interest to speed up the auction, so don’t have to bid on the most expensive license. This seems to be the real reason for behaviour last week - I have to admit it looks like I was completely wrong on Friday. In Round 9, when SpectrumCo raised the ante on the West supa-license to £442m, Verizon immediately stopped bidding on this license to allow the other regions to catch-up. Perhaps Verizon Wireless will swap to bidding on Bands D&E (in total 20MHz) when the differential between D+E compared to F gets too great.

The real reason that there so much life in the auction at the moment and the reason that the money is shooting up so quickly is because the Top 5 companies in the list are still bidding on 160MHz when in total there is only 90MHz on offer. And this is before the "majors" start bidding on the "minor" licenses…

Tiscali / HomeChoice and Zen

The Broadband market consolidates further with the acquisition of HomeChoice by Tiscali UK, however I see the deal as more adding Video options to the Tiscali product set rather than a standard Be or Easynet LLU capability type acquisition.

Homechoice has been around for about 15 years and has developed its’ own IPTV/Video on Demand system. It is interesting that Tiscali are entering into the world of designing proprietary equipment after years of selling out-of-the-box-routers. This is not as unusual as first might appear: FastWeb in Italy and Illiad/Free in France also sell there own designed CPE equipment. The first problem I can see for Tiscali is that the Homechoice box has an in-built ADSL modem, has two SCART connections (for TV & VCR) and two Ethernet connections: so the typical way of building a home network is slightly changed (ie only need to connect a WiFi router to the STB rather the typical USB modem or a straight-forward WiFI ADSL Router) However, the good thing for Tiscali is that the Homechoice Video on Demand channels with its’ self-design approach for Kids Channel, Music Channel, Film Channel and Entertainment channel seemingly exptremely popular with the subscribers. I don’t believe the uptake of premium subscription TV is very popular with only 4,400 subscribers to the Sky TV channels (according to BSkyB 2006 Accounts) or 10% of HomeChoice’s 45k subscribers.

Although it is a big risk for Tiscali to be involved with designing complex electronics and really complex computer software middleware, something that the much larger BSkyB (OpenTV, NDS) and BT (Microsoft) have avoided, it also presents an opportunity to differentiate. I am wary about saying it is a great strategy because BSkyB and BT both allegedly looked at buying HomeChoice and before deciding against. It certainly differentiates Tiscali in the wholesale market compared to the Cable&Wireless WholeNoSaling of the old Bulldog Network.

It will be interesting to see how the various video options actually play out over the next 12 months with Sky, BT, ntl and now Tiscali offerings. Tiscali might need the differentiation to overcome being an also run in a crowded marketplace, however the penetration of Homechoice amongst it’s 2.6m potential households running at around 1.7% doesn’t bode well for Tiscali.

Most intriguingly is the way the deal has been structured with a share swap at the Tiscali UK level rather than the Tiscali Group level. This may be giving Tiscali more options if any future break-up of the group occurs.

On a brighter note, Zen Internet has picked up the pace in its’ LLU efforts – it is now up to a grand total of 3 exchanges with another one planned for this week. Although, people may laugh, I think Zen is showing the way forward for ISPs who wish to remain independent - via. differentiation through technical excellence and predominately through serving the SME sector. Did anyone ever hear the story of the hare and the tortoise?

AWS-1 Auction – Day 3 Summary

Another day, another US$952,823,305 into the US government coffers.



The price of a mainland US licence in the “F” band has crossed the threshold of US$1bn. The price differential between the equivalent “A” band and “B” band is growing huge at 6.15x and 5.77x.

It is difficult to analyse bidder intentions during the progress of an auction because some of the moves may just be bluffs. It is akin to analyzing a poker game without the privilege of seeing the hands that the players hold. Mind you, the Monday Morning Quarterbacking that goes on post-auctions is also fraught with difficulties as people re-invent history and intentions to prove that they are brilliant after all. It is almost as if everyone takes a cognitive dissonance pill on the night of auction exit and wake in the morning with a whole new view on the world.

I think Verizon Wireless and Atlantic Wireless are exhibiting strange behaviour. First of all, Verizon Wireless is bidding on every supa-region of mainland USA in the “F” band except the North East. Atlantic Wireless is similarly only bidding in the “F” band and only on the North East region. Each is just bidding by the minimum increment each round. This is probably just a weird statistical freak because collusion is not permitted. Atlantic Wireless is a new company which qualifies for the 25% discount for new entrants and reputedly is financed by the same people behind Aloha partners who already hold some spectrum in the WiMax bands.

In 2000, one of the Verizon Wireless partnership, Vodafone, employed a similar tactic in the UK 3G auctions. Everyone knew before the start of the auction that Vodafone was by far and away both the strongest financially and had the biggest market share in the UK. Vodafone just sat on the license with the highest amount of available frequency (30MHz) and let the rest of the incumbents fight for the remaining 20MHz licenses on offer. Vodafone never budged from the 30Mhz license during the auction process. If it wasn’t for BT who was determined to force Vodafone to pay a realistic price for the extra spectrum, Vodafone would bought the extra frequency at a big theoretical discount. In the end Vodafone got the license at between 1.47x and 1.49x the price of the smaller licenses. Only time will tell whether Vodafone bought too much frequency but the Monday Morning Quarterbacks who egged all the bidders on during the auction now claim all the bidders were idiots who vastly overpaid.

There was also a rather grand gesture by T-Mobile this evening in Round 8 who bid more than the minimum (a 20% increase) on every “F” license that it was permitted to bid on and therefore won every bid. Personally, I think the tactic is so that T-Mobile will be featured in all the weekend press as the auction leader. Also, it meant the total spend of T-Mobile was just over a headline grabbing US$1bn. T-Mobile isn’t bluffing me – it is desperate for the extra spectrum and will pay whatever is necessary. The T-Mobile mothership, Deutsche Telekom, announced on Thursday that USA cash flow was around US$1bn for the first six months of the year. All of this is at risk if T-Mobile doesn’t buy enough spectrum to launch next generation services in the key US markets. Therefore, T-Mobile will continue bidding until enough people pull out and leaves it with the necessary spectrum. Then the T-Mobile and Deutsche Telekom executives will breathe a huge sigh of relief and hope that too much damage hasn’t been done to the shareprice of the mothership.

Rather than the Peacock-esque behaviour of the mega-bidders, the more interesting fights at this early stage are breaking out in the rural areas. I’ll try to look at some of smaller players over the weekend.

AWS-1 Auction – Day 2 Summary

End of Day 2 and the US government is US$553,014,500 richer or less poor depending upon your personal way of thinking about the US deficit.



We witnessed the departure of 5 companies from the runners and riders to leave 163 still bidding.

PCS Partners, L.P. (US$3m upfront payment), Hawaiian Telcom Communications, Inc. (US$2.155m), Leaco Rural Telephone Cooperative Inc (US$712.5k) and Arapahoe Telephone Company d/b/a ATC Communication (US$136k) all exited without making a single bid. If anyone can explain the rationale of going through the whole FCC procedure of registration and actually making a deposit without even bothering to make a single bid, I’d be extremely interested to listen. Yes, they are going to get their deposits back but it all seems a nihilistic existence to me.

At least Craw-Kan Telephone Cooperative, Inc. (US$434k) bid on five licenses in Round 1 and then saw that they faced a minimum of two other bidders on every license and used their proactive waivers to see how the bidding developed before calling it a day in Round 5.

CenturyTel who outbid Craw-Tel on two rural licenses are probably the story of the auction so far. CenturyTel placed a huge US$59m deposit with the FCC which was the 11th largest and is now left bidding on a mere US$9.162 “bidding units” having automatically lost 2 waivers because of low activity in round 1 & 2. CenturyTel currently have the winning bids on 32 licenses at a total price of $7.7m. I feel CenturyTel have paid a much bigger deposit than necessary and probably have a really poor bidding strategy.

The heavy hitters have followed the same course as yesterday. If everything continues on Friday as the first two days, I’ll be able to reveal what I think is the first “signal” and furthermore place it in a historical context with the infamous UK 3G auction of Mar-Apr 2000.

AWS-1 Auction - Day 2 Observation 1

There is currently reluctance for the big players to bid on the smaller CMA and BEA licenses.

For an illustration of the problem, look at the Top 50 markets represented by CMA001 to CMA050. 26 of these licenses currently have no bid. Los Angeles has had 2 rounds of bids and only 1 – New York have people bothered to bid in every single round. All of the other 22 licenses have only had 1 bid. T-Mobile is in the lead in 19 of the markets having bid in round 1.

The reason for the reluctance for bidding is explained on a small scale with this example from Hawaii.

Say you want 20MHz in Hawaii and bid $100m in the REA008-F, someone in the next round bid $110m for the spectrum, but you notice that the cumulative bids on CMA050 ($20m), CMA385 ($10m), CMA386 ($10m), CMA387 ($10m) are only $50m so you place a bids on each of the individual licenses. As the bidding evolves the price of CMA050 rises to say, $150m, whereas the total price for the region rises to $160m. This is perfectly rational if the bidders value the “rural” areas of Hawaii at next to nothing. It could also mean that you could be left with “stranded” licenses in the rural areas which are only worth something when “bundled” together with the major license CMA050.

It will be interesting to see when the big players (ie Cingular, Verizon, T-Mobile and Wireless DBS) stop bidding on REA-F licenses and start moving into the smaller pool. This will be some sort of signal that they believe the value of REA-F has peaked.

For some like Wireless DBS, it may be pointless actually owning the smaller licenses as the value is only if then have spectrum which they can offer to all its’ base, hence the reason it seems to be the only major bidding in Alaska....

UK Altnet Consolidation Continues

Another UK fibre network looks as if it is about to be consolidated.

Fibernet is a small company with a market cap of around £35m (with net cash balance of £10.4m) – it has a UK fibre network and a couple of MANs in London and Frankfurt. It has UK turnover for the most recent six months of £23.7m and is slightly loss-making at £1m. The losses will be easily sorted out by a few executive redundancies and a big write-down of the network (thereby reducing depreciation and amortization charges)

The appeal to a potential purchaser if an altnet is reducing the number of bidders on contracts or perhaps a cheaper way of adding expansion to their network rather than building themselves.

Sometimes I wonder whether these mysterious bid approaches actually happen or it is an easy way to put the company up for sale without losing too much face. I also wonder how many other small independent UK networks are hoping for a similar fate?

AWS-1 Auction – Day 1 Summary

Two rounds into the expected marathon auction and the sparring has begun, no-one has landed a heavy punch and certainly no blood has been drawn yet.



As expected there is a lot of action around the Regional Licenses with everyone who can be bothered bidding at this early stage placing the minimum amount.
• Verizon Wireless has only bid on the big F Block, but not in their original home turf of the North-East;
• Cingular is also focusing on the big F block, adding Hawaii to mainland USA and also experimenting with the BEA blocks for Chicago and New York;
• T-Mobile is bidding in all the Regional Blocks and also a few of the main metropolitan areas in the BEA and CMA bands;
• Wireless DBS is also only bidding on the Regional Blocks;
• Perhaps the only surprise is that SpectrumCo (ie the Cable companies) are bidding on the big regional licenses rather than their home turf.

Also, there seems to be a lack of initial interest in the smaller blocks with a lot of the big metropolitan areas not having received a single bid yet.

But there is a long, long way to go…

LLU Britain Update

I posted an analysis of the state of LLU Britain on the 20th June, so I thought it was time to do an update.



O2 have been busy on the acquisition of Be rather than unbundling, they have recently added RFS dates for a whole string of exchanges well into the future (max. date Dec ’07). O2 seem to be taking their time and not rushing into anything rash.

Bulldog has actually had their number of exchanges substantially reduced from 628 to 478. This is because The Broadband Resource takes the data from the Bulldog residential checker rather than the wholesale checker. They also seem to halted unbundled after the announcement they were to focus on wholesale and the business sector.

D-day for TalkTalk is rapidly approaching with the 1,018 exchanges due to unbundled only 22 days away now.

Easynet are by far and away the busiest unbundler actually completing 135 exchanges in the last 50 days.

Pipex have started to unbundle with 11 exchanges completed in the month.

Wanadoo have seemed to have momentarily paused, the 87 exchanges due to be unbundled are on the 30th Sept.

AOL are missing from the list and keeping their data close to their chests (unlike some other more important data). Apparently they have approx. 300 exchanges already unbundled.

Many thanks for The Broadband Resource for the data and answering my questions. The Broadband Resource is without doubt the best starting place to see which ADSL services you can get from your local exchange.

Sprint: Wireless Broadband Decision

I’d do not want to be drawn into the debate whether Sprint was right or wrong in going for Mobile Wimax. Personally, I think they have made the wrong decision.

Instead, I want to focus on the wonders of the US spectrum licensing system, whereby anyone can choose any technology to deploy. In the USA, they now have: T-Mobile and Cingular committed to deploying 3GPP solutions, Verizon Wireless committed to 3GPP2 and Sprint committed to IEEE 802.16e and this is before the forthcoming AWS-1 auction. I can quite easily envisage one of winners deploying either the UMTS-TDD solution, despite the manager sounding a little like a sore loser with the Sprint decision and even the Flash-OFDM solution.

In my opinion, this has in the past and will in the future lead to greater consumer choice and probably better technology worldwide as one standard fights against the other for dominance. With the size of wireless market, there is room for more than one standard and inefficient technology tends to die off pretty quickly, witness the fate of TDMA.

ntl: feeling the pain

The ntl q2 results looked really bad to me.

First of all in terms of market share, it looks like ntl lost ground in all 3 major areas: TV, Telephony and Broadband.
• In the TV area, it looks like ntl actually lost 22.8k customers to take the total to 3.293m customers. This compares with 7.749m Sky customers and 77k net adds for Sky. That is a swing in market share of just under 1% for quarter. Apart from the poor performance in the market place the amazing thing for me is that ntl still has 456.9k (or 13.8% of TV base) customers on Analogue TV, approximately 6 years after introduction. This does not bode well for the integration of ntl/telewest and the launch of new products.
• In Telephony, again a loss of customers, this time 35.1k to take the total to 4.233m customers. The fact that ntl have more telephony customers than TV customers translates to me that people bought the telephony product for cheaper prices than BT. Ntl are far more exposed to declining telephony revenues than a typical US-based cableco.
• In Broadband, there was actual net adds with 80.6k on-net customer additions taking the total to 2.902m customers and 12.1k off-net (ie virgin.net bt resale) customers to 205.3k total. Although it is good to see some growth – it is noticeable that the ntl figures were well behind the BT Retail net adds of 158k implying a loss of market share.

Overall, ntl penetration is 38.9% of homes passed (taking at least 1 product) and 37% of these taking the triple play or around 15% of homes passed. I'd love to see the BT Vision business plan, because ntl have been offering the triple play now for over 6 years and do not seem to making fantastic in-roads into the market.

Other Divisions include:
• The Business Division is struggling as all alt-nets in the market currently seem to be and is slightly down y-o-y to £160.1m revenues. This division is really important as business customers tend to fill the pipes during the day when the residential customers are not using them.
• The Content Division looks as if a real improvement was made. This comprises of three areas – the first is Flextech which is 10 channels including Bravo, Trouble, LivingTV & Challenge seems to have improved to £34.4m in the quarter up 17.8% y-o-y, due to increased multi-channel penetration and increased audience. Sit-Up TV which is some sort TV retailing venture via auctioning is flat y-o-y at £45.1m. UKTV which is the 50%/50% joint venture with the BBC isn’t consolidated and doesn’t seem to report figures, but I would suspect has similar growth to Flextech. This is the division that ntl put up for sale when it originally merged with Telewest which is surprising as it seems to be performing the best.

Overall revenues are up 2.8% (£24.4m) y-o-y to £884.3m with all the increase coming from the content division. All in all there is nothing in the merged ntl/telewest which is showing that the company will grow fast.

All of this is nowhere near ntl’s biggest problem – this is the amount of debt ntl owes: £5.4bn net. Furthermore ntl aren’t anywhere near starting to pay down this debt. According to the cash flow statement released Net Cash provided by the operations in the first six months was £243.8m whilst Capital Expenditure was £263.4m – in other words ntl was eating cash even before the interest charges kicked-in.

Also the debt has actually increased with the post-q2 big event - the Virgin Mobile deal with gross debt pro-forma going to £6.3bn from £5.8bn – a lot of this seems to be linked to LIBOR (ie variable with bank rates). The recent 0.25% rise in interest rates will have added £15m p.a. interest charges, approximately the same as savings realised from the ntl/telewest merger. Although to be fair, ntl claim a lot more savings to come.

Virgin Mobile is in itself not without problems with margins shrinking on pre-paid products as the MNOs drop pricing and Virgin trying to move into contract sales to offset these which requires a big investment in subsidies.

All in all, I feel ntl was in a lot of trouble without the UK Broadband Wars breaking out and is probably too weak to do much in the marketplace. It is an extremely good soundbite to be the only quadruple player in the market, but when the triple-play is doing so badly, you have to wonder about the strategy.

AWS-1 Auction 66

Two days to go and I’m getting pretty psyched – all the ingredients are prepared and thrown into the mixing bowl - I'm ready for a fiercely fought and therefore expensive auction.

Frequency Band

This is not the same frequency band as the UMTS spectrum licensed in Western Europe and Japan. So dual-mode phones will be needed for UMTS roaming services to work cross continents. Personally, over time this is not a big problem as quad-band GSM phones have overcome this problem in the past. I’m sure that the proportion of total Bill of Material cost for this extra functionality will be minimal in the overall scheme of a multi-baseband 3G phone.

Anyway, the US internal market is so big who really cares about roaming?

I’m sure that some companies will be considering deploying non-standard 3G technologies (ie apart from EV-DO and UMTS) – hopefully, we may see some companies trying to deploy OFDM based technologies.

Market Consolidation

The argument goes that as AT&T Wireless and Nextel have disappeared as independent companies from the market then there will less demand for the spectrum. I really, really doubt it. Spectrum is the oxygen that wireless companies breathe: the more they have the more things they can do. Any new entrant wanting to offer a wireless service using exclusive spectrum has to compete in these auctions: otherwise they have to rely on unreliable unlicensed shared spectrum.

More importantly, after all the expense of consolidation, I’m 100% positive that the huge wireless companies will pay huge premiums (in the form of license fees) to keep their club as small as possible: Verizon Wireless, Cingular and T-Mobile will fight to keep anyone else from owning a nationwide license and more importantly to keep potential customers out of the key Top 10 markets. Sprint is a more complicated matter and more of that later.

Potential “Once in a Life” Opportunity

All other previous spectrum auctions have been based upon a patchwork of licensees across the USA covering as many as 714 areas or at a minimum 176 licenses. Although this policy has allowed the emergence of small rural providers, it has meant that not one single wireless provider covers 100% of mainland USA.

This auction is different: mainland USA can be covered by only winning 6 regional licenses! There are 2 blocks of 10MHz and 1 block of 20MHz – expect these blocks to be fiercely contested.

It should not be forgotten that these licenses are as valuable for Verizon Wireless and Cingular as they are for a new entrant. For instance if Verizon got 10Mhz in all these 6 blocks – it could:
• slowly but surely build out the network in areas where they currently have high roaming charges with 3rd parties;
• as the Top 100 markets become saturated move into new smaller markets where before Verizon has had no presence leveraging its’ economies of scale and pushing the current small rural players into submission – a sort of WalMart Wireless strategy.
In other words, it would have the capability of both reducing ongoing costs and improving market coverage and therefore revenue and profitability growth. A similar story applies to Cingular.

The same arguments apply for T-Mobile however because they are currently spectrum challenged, they need to buy as much spectrum as possible. If T-Mobile don't buy enough spectrum in enough markets, they will be forced to be a pure voice and sms seller. Verizon and Cingular will know the T-Mobile weak spots and poor coverage: they will fight to keep the weak spots weak and thereby keep T-Mobile weak.

Sprint Nextel is someone who I believe isn’t really interested in this auction, mainly because of bigger internal challenges. I think the process is being driven by TimeWarner and Comcast. I expect to see SpectrumCo (the name of their JV) to bid strongly in the markets where the cables companies have licenses, so if the holy grail of triple and quad plays actually work they have spectrum to fight AT&T/Cingular and Verizon without all of the proceeds going to Sprint Nextel.

The most intriguing bidder is DirectTV, EchoStar and LibertyOne aka Wireless DBS. As we see from exploits of Murdoch in the UK via BSkyB – satellite is going interactive and I think Auction 66 is one of the few possibilities of having a return path in the USA. I suspect if successful, the whole US Satellite industry will be restructured and the inclusion of John Malone’s LibertyOne signals to me that they are prepared to bid more than DirectTV and EchoStar could afford together. The placing of a near US$1bn deposit signals to everyone that this partnership is serious and going to bid strong.

That is enough players to ensure that the Regional mega-licenses are sold for a lot of money: Verizon, Cingular, T-Mobile and Wireless DBS cannot carve up 4 nationwide licenses when there are only 3 on offer.

In addition in huge auctions there are always silent players sat on sidelines waiting for an opportunity. An example of this was the huge bidding in the UK 3G auction by TIW of Canada which remained a mystery until it was revealed that Hutchison of Hong Kong were behind them. There will be plenty of wireless arbitragers sat on the sidelines, watching and willing to buy a seat on the table in the middle of the poker (oops sorry auction) game if necessary.

Smaller License Areas

One of the great things about the US auctions is the amount of players who are willing to risk a lot of capital, including the major players named above there are 168 bidders qualified: 9 have placed deposits over US$100m; 8 have placed deposits between US$80m and US$10m; 30 have placed deposits between US$10m and US$1m; and 121 less than US$1m.

For instance, Leap Wireless aka Cricket have placed a deposit of US$255m on the auction. Leap Wireless specialise in building low cost networks in smaller cities - that is a big deposit for purely small cities. Perhaps they have a mystery backer for moving into the bigger cities?

In fact, there are professional players in the US Auctions who are basically just shopping for bargains to be later re-sold. The existence of these players is enough to ensure that prices are kept high. There is more than enough bidders to keep the bidding even on the smaller licenses fair and square. In we add into the equation that the majors will want to keep minor players out of the Top 100 markets means bidding will be fierce even on the blocks A-C.

No-one wants 3G services

I’ve heard more than a couple of people say 3G services are a big disappointment and therefore no-one will bid a lot for the licenses.

The truth is that Verizon have just reported that their quarterly non-SMS data has climbed above US$500m per quarter and that by my calculation Free Cash Flow was about US2bn in the first six months of the year. Cingular have also just reported record profits. That is a lot of profit both to go after and more importantly to protect.

Cellular Companies learnt from the UK Auctions not to overpay

Total Rubbish – for evidence, look at the recent auction of some African assets.

The US government estimates that they will receive around US$15bn from the auction – that is a real low figure and if this is all they receive the Big Cellular companies will be extremely happy. After all, it is just 4 years of Free Cash Flow for Verizon: in other words Verizon could buy every single license on offer and without any additional sales will be in the money after 4 years of a 15 year license.

The reason that auctions realise a lot of money is from the auction design – a lot of people with a lot of money and a lot to lose need to compete for a scarce resource.

That is the environment with Auction 66 and I’ll be extremely surprised if in a couple of week’s time everyone (apart from me) isn’t amazed at the prices being discussed.

US$30bn is the minimum I’m expecting…

Cellular in the Central European Markets

I’ve always been fascinated with Central Europe; primarily because when I was a youth the majority (excluding Austria and part of Germany) was behind a mysterious Iron Curtain and more recently due to the tragic break-up of Yugoslavia and the entrance of several countries into the EU club.

The large Cellular companies have wasted no time in taking large stakes within Central Europe. We have not quite reached the end-game but are getting quite close.


In a few countries the invasion of the Western Cellular Companies has not quite finished: Bosnia is due to sell its’ three state owned companies over the next 12 months; Slovenia still has to privatize its’ state-owned company; the third Bulgarian license, Vivatel is actually owned by the Icelandic company behind the infamous Saunalahti MVNO in Finland; and I can’t see any long term strategic reason for Tele2 holding a license in Croatia. I believe all of these licenses will be sold to the majors over the next couple of years to consolidate the Central European market further.

Also, there is a possibility of several countries licensing a third player: Slovenia, Serbia, Macedonia, Montenegro and Albania fall into this category.

I believe the final end game will be when the smaller regional players start selling out or making alliances will the larger players: we see this happening with Vodafone / Telekom Austria network partnership deal. I can quite easily envisage an environment where Telefónica and OTE do a similar network partnership deal to Vodafone/Telekom Austria.

Several of the big players such as Vodafone, Orange and T-Mobile face difficulties in taking control of several venture (eg Polkomtel) and squeezing out minorities. These problems I can see dragging out for several years.

As per normal, the wildcards are Orange and Telenor. I can’t see Orange being happy with its’ extremely weak position in the fastest growing part of Europe, but given that they are cash constrained at the moment their options are severely limited. An obvious solution to this difficulty is some sort of merger of Central European assets given the lack of overlap of properties: in Austria both own minority positions in the same operator, One. It is difficult to understand Telenor current strategy, so who knows what will happen.

The three key variables in looking at attractiveness of cellular markets are population, GDP and size of countries: Germany is by far and away the key market. A strong position in this market will pay for all sorts of expansion elsewhere.

Bad Week for the Austrians

Someone once said “2 out 3 ain’t bad” – well that’s if you’re winning. 2 out 3 ain’t good at all if you’re losing, especially considering that the 3rd hasn’t been decided yet. I’m talking about Telekom Austria’s attempted Central European expansion into Serbia (lost), Slovakia (lost) and Bosnia (yet to be decided)

The loss in Serbia would have been a huge blow for the Austrian’s: I have been following the saga for the last couple of years and never have I thought that the Austrians would lose. Basically, the problem lay with the sordid recent past of Serbia. Mobtel was the largest cellular operator in a duopolistic market in Serbia, it was owned 49% by the state and 51% by Bogoljub Karic who appears to be one of Slobodan Milosevic’s cronies. Mobtel seems to have been run exclusively for the benefit of Karic and his family, tricks he performed included: taxes not paid to the government, mysterious purchases of property which were sold to members of family at discounted prices, dodgy network sharing deals in Kosovo. I’m sure this list is just the tip of the iceberg. Once the Milosevic era collapsed, one of the biggest tasks that the new government faced was the clean-up of Serbia business practices – Karic was target number one. He also didn’t help his cause with his interfering in politics. In 2005, Karic fled the country and sold his interest in Mobtel to a group of Austrian businessman led by Martin Schlaff. Schlaff has a really murky past and seems to specialise in being a honest broker in the sale of East European companies to Western Companies. In fact, the Telekom Austrian purchase of Mobitel in Bulgaria was brokered via. Martin Schlaff. I’m sure at this stage, Telekom Austria was 100% sure of winning Mobtel. However, the Serbia government stepped in and grabbed all the assets of Mobtel, renamed the company Mobi 63 and proclaimed they would auction it. Schalff somehow wheeled and dealed himself back into the deal with the help of the Austrian government and a compromise was reached.

On Monday the auction took place in Belgrade live on TV and the result was a Telenor victory with a final bid of €1,513m. They won by a meagre €20m from the Austrians who bid €1,493m. The Austrians have a chance of returning with a 3rd cellular license to be issued by the Serbian government, but the minimum price of €320m seems extremely steep to me.

The Austrians moved onto Slovakia where a “beauty contest” was ongoing for a third cellular license and the main candidates were Telefónica and Telekom Austria. The Slovakia market is basically a duopoly between Orange and T-Mobile and with quite high penetration. The capital of Slovakia is only 60km from Vienna and only the Danube separates Austria from Slovakia. However, Telefonica runs the old state PTT in the Czech Republic and has deep ties in Slovakia. So on Wednesday, Telekom Austria lost again. The embarrassing part is that Telekom Austria alledgedly outbid Telefonica with a bid of SKK250m (€6m) compared to SKK150m (€4m), however as the bid amount only 5% weight in the evaluation criteria, Slovakia had ultimately lost.

In both of these auctions, Telekom Austria could have won if they would have moved earlier: first in Serbia by buying the stake from Karic rather than allowing Schlaff to get involved and second in Slovakia by bidding in the previous two auctions when the government failed to sell the licence.

Next, the Austrians are moving onto Bosnia, let’s hope they have better luck this time…

BBC Radio

BBC Radio released its’ viewing figures for the 2Q and they are monopolistically huge – having a weekly reach of 54.7%. However, this still means that 45.3% of radio listeners don’t listen to the BBC and still have to pay for the service, this is in addition to 100% of non-radio listeners - hardly a fair practice.



It is obvious from the figures that the Digital Only Stations are limping along and need a shot in the arm. Obviously, I can’t solely blame Nokia for this sorry state of affairs, although taxing their analogue chips should help with the promotion and awareness of digital radio. I think the BBC needs to make a grand gesture and move Radio 1, 2 & 4 to being Digital Only – this would really up the penetration and make the day of analogue turn-off and re-use of the spectrum much closer. Isn’t this the type of gesture that would make public broadcasting great again?

AOL Europe Sale

So, it looks like AOL France is going to be sold to Neuf Cegetel for around €300m for 600k broadband customers. No details of revenues or profits are given. This will act to further consolidate the French consumer market around 3 providers: France Telecom, Illiad and Neuf Cegetel. Apparently, the French business of Tele2 is also up for sale. LLU is acting now as a catalyst in several markets to make the fixed-line division more of an oligopoly. James Enck wrote a great piece yesterday about the consolidation going on the Netherlands. I’m less depressed than him over the situation and feel oligopolies are a natural economic consequence of any industry which needs large infrastructure investments. The only way around the situation is to break-up the telcos with regulated NetCos and free market ServeCos. In the long run, I believe the EU will try to legislate this as the industry structure. The Telco2 Disintermediators have lined up an EU bureaucrat to speak at a forthcoming conference, although I'm sure he will be holding his cards extremely close to his chest.

In the UK, the exit of AOL from the access business also spells further consolidation with the purchase by any of the three short-listed parties: CarphoneWarehouse, Orange and BSkyB. As a ball-park guesstimate if we double the French figure for 1.2m customers which gives a sale price of around £400m. I can’t see CPW affording this without putting more equity into the business – will TimeWarner take shares as payment? Otherwise I think CPW will be dangerously leveraged and I think some of the players will be quite happy to exploit this weakness. Also, I can quite easily see Orange buying Bulldog’s customer base and LLU infrastructure if they are unsuccessful in the AOL process.

I really don’t know enough (currently) about the German market to comment.

PlusNet: Soon to be NegativeNet

PlusNet, a small UK Quoted ISP, published its' interim results yesterday and only added a miserly 3k broadband customers in the quarter. PlusNet is the #8 UK Broadband provider and now has 195k customers.

It really is between a rock and a hard place. Although profitable, PlusNet does not have the sort of money to allow it to go building LLU infrastructure and actually has signed a wholesale deal with Tiscali. It cannot afford to go head-to-head with the big guns with marketing spend. I actually feel sorry for the position PlusNet is in, after all it is a Yorkshire company. However, I think investors should think of doing the same as the original founder, who sold up in 1999 to start ebuyer.com which is now one of the UK’s biggest online retailers.

Unfortunately as well as the market moving against it, PlusNet inflicted a couple of self-inflicted wounds during the quarter: one was the loss of customer email data and the other was the sending out of 20k of its’ customer addresses (as if the churn wasn’t bad enough)

Even the Marketing Director admits that “PlusNet is not going through a great time right now...”

Tele2: Feeling the Pain

Tele2 is really suffering and I suspect that Kinnevik who own 28.3% (as well as 38.3% of Millicom who have already reported) is being dragged down by the poor performance. Tele2 reported a drop in net profit from 676SEK to 246SEK y-o-y, which together with a cash outflow after capex of 383SEK spells a lot of trouble for the company.

Historically, Tele2 used the cashflow from the Swedish mobile operator, Comvig, to expand into mainline Europe. It built up a huge set of revenues and smallish profits reselling voice minutes cheaper than the regulated incumbents in big European Markets (eg Germany, France, Italy, Spain, UK, Poland) and occasionally bought some infrastructure assets eg. Baltic Mobile Operations. This plan was working well until LLU and VOIP players meant that the typical arbitraging reseller was no longer the cheapest game in town. So, Tele2 has basically spent the last year trying to remodel its’ business model.

Tele2 is trying to do three expensive things at once:
- replace fixed reseller revenues with direct access and LLU business
- build a MVNO Mobile business in certain markets
- expand into Russia

I think Tele2 is trying to do to much and this could potentially bring the whole mini-empire to its’ knees. If I was in charge of Tele2, I’d certainly be thinking of withdrawing from the whole of UK&Benelux & Souther Eurpe and possibly Central Europe. This would leave the company focussed on Nordic & The Baltics with Russia as the growth driver.



Kinnevik are obviously thinking about how to reshuffle its’ pack of cards with the recent aborted sale of Millicom. The performance of the shares has not been awe-inspiring this year. If a sale of Tele2 is being considered, I suspect that the only people who would be interesting in buying Tele2 would be someone who was willing to do a break-up, something that Kinnevik should be more than capable of doing itself. Telecoms (via Tele2 & Millicom) account for more than 50% of the asset value of Kinnevik and they need to do something sooner rather later before more value is destroyed.

Verizon Wireless v Cingular: 2005 Q2

A quick comparison between the Verizon Wireless (VZW) and Cingular 2Q reports highlights the big part that cost control plays in the far better performance by VZW. The individual line items in the operating expenses do not match, but it is clear that overall VZW has the lower cost base, even on a cash basis (ie before depreciation and amortization)

VZW’s revenues are catching up fast with Cingular and VZW will in all probability overtake Cingular before the end of the year, even in terms of service revenue.



As I mentioned before Cingular are currently in the process of turning off its’ TDMA and Analogue network which are both due to be turned off in early 2008. I’m sure that this will reduce their cost base, but don’t think by enough to catch up with VZW. The Cingular marketing brains have come up with a rather unique way of forcing people to move networks – by raising the bar for monthly subscriptions – upping it by $5/month. I’m sure that this will force people to move networks, however I’m not sure how many will remain as customers of Cingular and knowing the USA will probably trigger a few more class action lawsuits.

Broad Progress for Tiscali

Tiscali announced a shortened form of half year results yesterday – full results are due in mid-September. They revealed they had 1.2m broadband customers in the UK a gain of 265k for the half year. According to my maths that is around 20% of the total market adds which is not bad at all, especially seeing that Tiscali does not have the marketing clout of the big guns. It is certainly a lot better than Orange/Wanadoo/France Telecom. The slight complication is that the 1.2m includes both retail & wholesale customers so we do not know where the growth is yet.

The big question for the Tiscali board is whether it is the time to sell up: with operations in the UK, Germany, Italy and the Netherlands it is a perfect fit for a certain cellular company and with a market cap of less than a €1bn, Tiscali is easily digestible.

For Sale: A Really Dumb Pipe

AOL want to keep its customers using its’ web services – this makes pretty good sense to me. According to last week’s Financial Times in a big article on Time Warner, AOL saw as key that they remained in control of the email addresses as this drives a lot of traffic (and therefore advertising revenue) to its’ portal. They also want, according to this mornings Guardian, that routing to the AOL portal remains – again this makes sense for the new advertising driven AOL. This effectively turns the customer into purely a purchaser of internet access with any upside being in the sale of telephony products. This definitely runs contrary to Sky’s content domination plans, but also France Telecom and Deutsche Telekom have similar plans. That only leaves one potential purchaser – Carphone Warehouse - unless o2 or Vodafone surprise us all and it provides an immediate solution to the lack of content on the TalkTalk offering.

I believe that these conditions of the auction would have been known a long time ago and they are now only being revealed now because the auction is going to realise a lot less money than the figure originally bandied about in the press - £650m.

Anyway, we don’t have to wait a long time to know the final price. I expect the results of the auction will be made public with the AOL results – along with the sale of the French and German auctions.

Also, if the purchaser is Carphone Warehouse, they do not have a lot of spare cash available on the balance sheet. On 1st April 2006, CPW had net debt of £273.4m – another biggish purchase (for Carphone Warehouse) would put them under a lot more pressure.