Tuesday, July 31, 2007

Is Telecom Italia serious about the Netherlands after all?

FTTH Update

Planet Multimedia reports that bbned, the Dutch wholesale unbundling operator owned by Telecom Italia, has bought InterNLnet, an ISP affiliated to Nijmegen University. InterNLnet's portfolio is geared toward both ADSL and FTTH networks. The latter include networks owned by Portaal (housing corporation) and GNA (Amsterdam Citynet).

I sort of expected a move like this. If TI do not sell the unit (as was rumoured at some point), why not add a service provider business to the network operator and get the broadband strategy aligned across Europe (Italy, France, Germany)? Also, bbned is among the unbundlers who are working out the details of 'regulation 2.0' (SLU) in the Netherlands with KPN, and they are already conducting a VDSL trial - with InterNLnet and SURFnet.

Taking this speculation one step further, one could envison TI beefing up its bbned operation even more. The market is consolidating, but there are numerous opportunities left:
  • ISPs, such as SURFnet, mentioned above (if it would be up for sale)
  • Reggefiber (once Dick Wessels would be interested in selling out or swapping)
  • Orange NL's broadband unit (once France Telecom finalises the deal to sell the company to T-Mobile)

That would add not only ISP capabilities and subscribers, but the inevitable road to FTTH as well.

Finally, an overview of recent FTTH activity (click to enlarge):

Thursday, July 26, 2007

Is Amazon's opportunity over twice eBay's?

Check this sheet, comparing Amazon.com and eBay, based on the 07Q2 quarter. (I had to estimate the eBay employee numbers.)

Is Amazon.com too expensive, or is there more to come? I suppose Amazon has a much bigger market opportunity (retail shifting online) than eBay (auctions must be much closer to maturing). However:

  • 1. Both show continued growth. Is the acceleration at Amazon going to last?

2. The same goes for their core revenue (Media at Amazon, Marketplaces at eBay).

    3. Margins at eBay are much fatter, but Amazon is showing a lot of growth. Where are Amazon's margins headed?

4. On most metrics, Amazon is about twice as expensive. How does that compare to growth?

I think the valuation differential is a little stretched, but I do believe Amazon.com demands a premium.

Friday, July 20, 2007

Yahoo! v. Google: converging strategies but very different results

Yahoo! and Google reported and showed some remarkable differences, even as in fact they have converging strategies.

Obviously, Yahoo! is playing both the search/ads game and the portal game, whereas Google is doing search/ads and applications. Besides, Yahoo! has income from fees and tries to add apps. Google on the other hand earns license fees and is growing its portal ambitions. Add to that that the boundaries between search ads and display ads are becoming more vague, and we can see companies evolving toward each other.

Put differently, it does make sense to compare them.

This Google Spreadsheet is self-explanatory, but maybe some remarks are needed.
  • Growth at Google is much steeper, despite its larger base.
  • Yahoo! is much less internationally diversified.
  • The 'other' source at Yahoo! (fees, mainly from broadband partner and HotJobs) is much more significant than the one at Google (mainly licensing).
  • The 'TAC to Affiliate Revenues' ratio peaked at Google at 85%. Being generous to partner sites in giving them by far the largest revenue share is obviously a good way to gain market share.
  • Margins are much higher at Google.
  • Both are generating lots of cash.
  • I used Amazon.com's RoIC definition. Here, they are pretty much alike. The TTM term (trailing twelve months) is also an Amazon thing.
  • Calculating the 'per query' ratios, one should use keyword related advertising income only. Since the companies do not provide this information (ad income isn't broken down along the search/display division), I had to make some assumptions. For Yahoo!, I assumed the division is 50/50 and for Google I assumed it is 100/0.
  • The 'per employee' ratios of Google are about twice as good as Yahoo!'s.
  • Google's P/FCF TTM looks particularly high, but look at the rate at which FCF TTM is growing!

Google takes another look at access network opportunities

Google is taking another look at the access business. Having invested in a worldwide fiber backbone and an extensive datacenter network, the 'only' piece missing from the netwok puzzle is an access network.

Today, they are investing in Ubiquisys, which manufactures femtocells. Basically, femtocells are an emerging technology, aimed at a better user experience (in-house coverage), as well as traffic offloading for mobile operators (traffic is offloaded through the femtocell and the home gateway onto the user's broadband connection).
It's not an easy sell, as Disruptive Dean has shown several times (how do you get the user to 'buy' it?; competition from the enhanced WiFi standard 11n for better in-home coverage of UMA handsets; the expected move to 900 MHz band, for better in-home coverage; etc.).

This adds to previous baby steps:

I am working on a story that could, at least in theory, very well be the next (big) step. More on that after the weekend week.

Tuesday, July 17, 2007

Yahoo! preview: was Q1 the bottom?

Growth numbers seem to suggest that Yahoo! passed the corner, 07Q1 having been the bottom.
Check out the table, based on sales of $1.24bn - the consensus number that slavishly follows guidance (i.e. toward the top end of the lower half of the original range).

Co-op gaining momentum

German alternative network operators, both wholesale (Vodafone's Arcor, Telefonica Deutschland) and integrated players (Versatel, Telecom Italia's Hansenet) are planning to jointly build a VDSL network.

This underscores a trend gaining considerable momentum: co-operation to better compete with the incumbent.

Comparable events:
  • Australia's G9 and Telenor's Swedish operation have similar plans.
  • Neuf Cegetel recently said it expected consolidation (FTTH).
  • In Nigeria 25 ISPs plan a joint WFi network.
  • SoftBank en eMobile plan a joint offer for a WiMAX license in Japan.

Monday, July 16, 2007

Will mobile network sharing lead to separation?

This news from Thailand again brings up the matter of separation in wireless networks. Thai Mobile seems to volunteer the building of the nation's first 3G network, to which it will offer open access to all operators.

Other recent news around mobile network sharing:
  • Vodafone UK and Orange UK plan RAN sharing (3G and 2G) to reduce capex and opex by 20-30%.
  • Hutch Essar and Bharti plan infrastructure sharing, to be supported by the regulator.
  • Nokia's solution is expanded to supporting up to 4 operators.
  • Yoigo (TeliaSonera) and Telefonica Movistar plan antenna sharing.

(Now, it is important to realise that sharing may be done at different levels: sites (towers), antennas, RAN, backhaul.)

Some observations, beyond the obvious cost savings target:

  • Similarly, operators are teaming for mobile TV (be it a shared DVB-H network, wholesale access to Qualcomm's MediaFLO or any other technology). So, why not for 3G as well - or for that matter: for 2G (not to mention 4G)? As in fixed, sharing and separation make a lot of sense in an IP-based world.
  • Differentiating by touting network coverage (as Verizon Wireless still does) will become a thing of the past.
  • So, operators will need to make sure they can differentiate on the services and applications level.
  • If network operations are to be separated, a new (natural) monopoly will arise. As long as existing service providers are deemed to have SMP (significant market power), this may give rise to new open access obligations at the network operator. On the other hand, the rise (not the fall) of MVNOs could preclude this (what will the difference be between MNOs and MVNOs anyway)? However, spectrum will always be much more of a scarce resource than anything equivalent in the fixed world (duct access, access to sewers, etc).

Monday, July 09, 2007

Google adds another element to its enterprise strategy

Today's Postini acquisition raises the issue of how serious Google is on its efforts aimed at the enterprise market. I think there is more to come.

So far, Google is almost entirely about the consumer market (if consumer and professional users can be separated at all), with advertisers to pay for it. 'Licensing & Other Revenue' was just 1.0% of total revenues in Q1 of 2007. Now, Google appears to be beefing up this tiny part of its business model.
Ultimately, the advertising market will prove to be a cyclical market, so why not do a little diversification while growth is still double digit? In Q1 of 2007 revenue growth was still around the 65% level and personnel numbers grew 80% yoy. Obviously this reminds us of Yahoo! and its diversification efforts, several years ago.

The licensing and other revenues stem from the Google Search Appliance sale and licensing of for-pay applications (applications that are free as long as they are not used in a commercial way, such as SketchUp and Earth).

Recent developments in this space include:
  • February 22: launches Google Apps Premiere Edition (50 $/year/user)
  • April 19: acquisition of Marratech (video conferencing software)
  • May 18: launches Google Apps Partner Edition (for ISPs)
  • May 30: launches Google Gears (offline application of Gmail, Docs & Spreadsheets, etc.)
  • June 20: plans slide presentations addition to Apps
  • June 27: additions to Docs & Spreadsheets (folders, etc.)
  • June 27: teams with Ingram Marshall (Search Appliance distribution)
  • July 9: acquisition of Postini (adds security and compliance solutions to Google Apps)
  • First, competition with Microsoft is intensified, especially from the ever increasing Google Apps suite and the addition of both Gears and Postini.
  • I suppose a major sales push of the for-pay services must be on the horizon, mirroring the Ingram Marshall deal for the Search Appliance.

Co-operation will drive separation and FTTH

This news out of Italy unites three of my favorite (and interrelated) trends: co-operation, separation and FTTH.

Apparently, alternative operators (Vodafone Italia, Fastweb, Wind, Tiscali, BT, Tele2, Welcome Italia and Eutelia) are calling for a break-up of Telecom Italia.

1. Co-op
Note that PTTs who resist full separation at home (BT, Swisscom), allow their foreign subsidiaries (BT Italy, Fastweb) a viewpoint different from their own. Deutsche Telekom to the contrary did not allow T-Mobile NL a divergent view when OPTA (the Dutch NRA) called for market response to KPN's All-IP network plans; T-Mobile, mirroring Deutsche Telekom's strategy in its home market, called for an end to regulation altogether in the Netherlands.
The new development in Italy is supportive of my call for a country-by-country approach, which will allow altnets to finally work together, no matter what their parents do or believe or say.

2. Separation
I am not sure what kind of separation the Italian market is headed for (anywhere between accounting and ownership regulation, but probably functional). Anyway, having truly equal access to TI's network must reduce the need for altnets to build their own infrastructure: embracing the wholesale market will drive the need for TI to build a broad portfolio of IP-based services, available to each altnet.
Another interesting development is that an incumbent like Telecom New Zealand openly lost interest in being a network operator and wants to be service provider.

Fiber being a natural monopoly and the TI network to some degree spun-off from the TI services organisations will, I believe, drive the newly created NetCo to aim for nationwide FTTH. Demand keeps rising and there is no such thing as the Broadband Incentive Problem in the wholesale market, where tariffs are usage-based (the BIP in the consumer market is a consequence of flat-fee tariffing and unsufficient pricing power). As a result, building FTTH will drive revenues and lower the cost base (after an intial capex hump, naturally).

Monday, July 02, 2007

The BCE bid implies 20% upside at KPN and 30% at BT

Pensioen funds and private equity are performing an LBO at BCE. Its valuation, 7.8 times EBITDA over the trailing twelve months, implies an equivalent bid level of EUR 14.93 at KPN (a 21% upside) and GBP 438 (a 31% premium) for BT.

Check out the underlying calculations in this Google Spreadsheet.

Adding some realism to WiMAX and xMax

How realistic are claims of both WiMAX and xMax when it comes to cost savings on network build-outs, relative to GSM or CDMA-based networks?

I suppose a good deal of common sense (if not scepticism) is needed.

Just to add a thought on each:
  • WiMAX: in this release it is revealed that the range is limited to just 0.75-1.00 miles when indoor-coverage is needed to deliver 5 Mbps. Remember that originally, WiMAX was supposed to deliver up to 120 Mbps over a distance of up to 50 km?
  • xMax: I was a little puzzled over the Frost & Sullivan award last week. That seems quite premature, as xMax is far from both technical and commercial reality (see this article): "But is it stretching the rules of economics, or those of physics?" The company behind the 'Flash Signal' technology, xG technologies, puts forward Prof Stuart Schwartz. But the article quotes Prof Ben Friedlander, who dismisses the technology. Who should we believe? Now, Frost & Sullivan seems to add some weight to the xG/xMax side, but the article delicately reminds us of a similar award in 2004 for Gaiacomm ("transmit a signal of any strength to all parts of the planet up to and including inner space and outer space"), which is dormant now. Lastly, the fact that Stuart Schwartz held an interest in xG must not be overlooked (check out this paper from last September, prior to the IPO).