Telecosm - Ike Elliott

Disappearing Act: LEC Line Losses Continue

The trend of many years is continuing at the nation's biggest phone companies: people are cancelling their traditional landline phone service in droves.  This week, Verizon reported over 3.6M switched access lines lost in the past year, and last week AT&T reported a loss of 5M lines in the past year.  Qwest reports their 1st quarter results next week, so we'll have a complete 1st quarter picture at that time.  Meanwhile, here is a look at line losses in calendar year 2007 at the big three telcos:

LEC Access Lines	EOY 2006	EOY 2007	Lines Lost
AT&T	66,469,000	61,582,000	4,887,000
Verizon	45,100,000	41,441,000	3,659,000
Qwest	12,299,465	11,500,000	799,465
Total	123,868,465	114,523,000	9,345,465

9.3M lines lost in one year!  That's like the entire Chicago metropolitan area turning off their residential phone service in one year.  And that's assuming every man, woman, and child had their own residential phone line. 

Where are all of these lines going?  It's not like these folks are just deciding they don't need phone service any more.  Over the next few days we'll dive into the data on how many lines are going to wireless, cable, and VoIP services. 

Comcast Wins a BUNGL For HD Compression

The Associated Press is now reporting on the additional compression that cable companies are imposing on High Definition channels, that I reported on earlier this month.   The 3:1 compression creates a visibly inferior experience.  As one user put it:

"I couldn't really tell if what I was seeing was really better than what I saw on regular television"

The point here is that users are being ripped off, paying for high definition channels that they aren't really receiving, and for that achievement, I'm awarding a BUNGL to Comcast. 

This makes two BUNGLs for the cable industry, giving them sole position of first place! 

AT&T Exec Hypes Internet Growth

Dan Caruso highlighted comments from an AT&T executive's talk as an example of Internet growth hype in a post on his blog a couple of days ago.  Jim Cicconi, vice president of legislative affairs for AT&T, was speaking at Web 2.0 forum in London a few days ago when he made comments ranging from ludicrous to meaningless to merely believable.

Being that this is a blog, we'll focus on the ludicrous:

"In three years' time, 20 typical households will generate more traffic than the entire Internet today."

Really?  Let's do a little math to see what that means.  How big is the entire Internet today?  Let's give Mr. Cicconi the benefit of the doubt and assume that he was speaking only about the US Internet, which was somewhere between 750 and 1250 Petabytes per month at the end of 2007, according to MINTS. (A Petabyte is a 1 with 15 zeros after it) Again, giving Mr. Cicconi the benefit of the doubt, we'll take the lower of these two numbers, and assume that in three years, 20 typical households will be generating more than 750 Petabytes per month, meaning each household will generate 37.5 Petabytes per month in traffic.  Multiplying by 8 to get to Petabits, then each of these twenty households would need to generate

300,000,000,000,000,000 bits per month.

That's a lot of bits.  Even if each of these households watched 8 hours of uncompressed streaming HD video on the Internet every day, 30 days each month, they would be generating only on the order of

125,000,000,000,000 bits per month

...or about 0.04% of Mr. Cicconi's predication. 

Given that there were about 238 million Internet users in North America at the end of 2007, in order for Mr. Cicconi's prediction to come true, even if you assume the number of Interent users remains constant at 238 million users through 2010, the Internet would have to grow at a compounded rate of 23,350% per year for the next three years.  That's just a smidgen higher than the more typical 60% compounded growth rate the Internet has recently seen. 

Of course, all of this begs the question: why did AT&T let somebody with the title of VP of legislative affairs talk about Internet growth?

As for Mr. Caruso, I'm glad we can both spot hype when we see it!  It's good to be back on the same page.

New Music: Pinback

Telecosm reader Chris Jensen commented on last week's New Music post on Le Loup with a few suggestions for my listening enjoyment, including The Mars Volta, Pinkback, and The Sim Redmond Band.  I've listened to all three this week, and really enjoyed exploring the music of these bands, but have to say that San Diego-based Pinback is the real standout among these three.

I was playing some Pinback on Pandora while writing this post (more on Pandora next week), and my fifteen-year-old son Ben says "It's kind of loud."  I smiled and said "Yeah!" (with relish), and he asked if I could turn it down and I refused, of course.  I think in most families this Dad-to-teenager relationship is usually reversed.

I don't mean that Pinback is headbangin' metal music...they are anything but that.  This band makes "some of the most complex, postmodern indie pop happening on the West Coast."  They do some fascinating things with song structure and melody that keep my ear from getting bored.  The innovation with song structure reminds me of some of cooler songs by The Shins, but they don't sound like The Shins, really.  A good mix of mostly mellow electric guitar, synth, drums, with punk influences, and a relaxed vocal duet sound. I haven't heard a Pinback song I don't like.

Some of the songs that have caught my ear: Tripoli, Charborg, The Yellow Ones, and Offcell. 

The band isn't exactly new, having been around for about ten years.  They're just new to me! 

Thanks, Chris, for the recommendation...I'm going to enjoy exploring the rest of Pinback's music.

How Level 3 Gets to Positive Cash Flow

Yesterday I posted about how Level 3's positive earnings report, and their forecast of positive cash flow before the end of the year, could be due to couple of things:

• Cost reductions in both capital expenditures and SG&A
• More rapid growth in wholesale Internet access utilization and revenue

Meanwhile, I've been picking through the posts from the "anonymous analysts" over on the Investor Village Level 3 message board, and have discovered some additional factors that should figure in to an outstanding year for Level 3.  You've got to hand it to this particular message board: it has a number of very talented and thoughtful posters, and they turn over a lot of rocks, and they share their findings.  I often learn from them.

• Gross margin improvement: The company forecast a slight improvement in gross margins this year, but since gross margins came down a little bit in the first quarter, the company must be counting on improving gross margins more in the remaining three quarters.  This margin improvement should come from additional transitioning of services to on-net fiber, and from slight increases in wholesale voice pricing.  This could add on the order of $20M or so per quarter in additional free cash flow.
• Lower interest expense:  Level 3's bond interest expenses are seasonally concentrated in the 1st, 2nd, and 3rd quarters.  Less interest expense in the 4th quarter equals more free cash flow at that time.

Add these factors to continued improvements in SG&A forecast by the company, and you could get enough cash flow improvement to meet or beat guidance even without increasing sales much at all.  Meanwhile, the company is adding 100 salespeople in the first half of this year, so we should see some of those new hires begin to hit their stride in the 4th quarter.

Many thanks to a certain members of the Investor Village gang for this info!

Level 3's Wholesale Internet Business Turning the Corner

Listening to the Level 3 Communications quarterly earnings conference call this morning, CEO Jim Crowe offered the following forecast for the company's IP business for 2008:

Bandwidth Growth: 75-80%
Pricing Decline: 25-30%

Taking these two data points and plotting them on the table I published last week in my post on "Is An Internet Industry Resurgence Coming?", we can now plot a range for Level 3's IP and Data Services revenue growth for 2008:

Industry Internet Year over Year Revenue Growth
	Rate of Price decline in price per Mbps
Rate of Bandwidth Growth	20%	22%	24%	26%	28%	30%	32%	34%	36%	38%	40%	42%	44%
50%	20%	17%	14%	11%	8%	5%	2%	-1%	-4%	-7%	-10%	-13%	-16%
55%	24%	21%	18%	15%	12%	9%	5%	2%	-1%	-4%	-7%	-10%	-13%
60%	28%	25%	22%	18%	15%	12%	9%	6%	2%	-1%	-4%	-7%	-10%
65%	32%	29%	25%	22%	19%	16%	12%	9%	6%	2%	-1%	-4%	-8%
70%	36%	33%	29%	26%	22%	19%	16%	12%	9%	5%	2%	-1%	-5%
75%	40%	37%	33%	30%	26%	23%	19%	16%	12%	9%	5%	1%	-2%
80%	44%	40%	37%	33%	30%	26%	22%	19%	15%	12%	8%	4%	1%
85%	48%	44%	41%	37%	33%	30%	26%	22%	18%	15%	11%	7%	4%
90%	52%	48%	44%	41%	37%	33%	29%	25%	22%	18%	14%	10%	6%
95%	56%	52%	48%	44%	40%	37%	33%	29%	25%	21%	17%	13%	9%
100%	60%	56%	52%	48%	44%	40%	36%	32%	28%	24%	20%	16%	12%
105%	64%	60%	56%	52%	48%	44%	39%	35%	31%	27%	23%	19%	15%
110%	68%	64%	60%	55%	51%	47%	43%	39%	34%	30%	26%	22%	18%
115%	72%	68%	63%	59%	55%	51%	46%	42%	38%	33%	29%	25%	20%

It looks like 23-33% revenue growth for Level 3's IP and Data Services Business this year, thanks to a small reduction in the rate of price compression, and an uptick in traffic growth at Level 3. This is much better than 2007, when the company's IP and Data Services revenue grew at about 8 or 9%.  A higher IP and Data Services growth rate could be a big part of why Level 3 is comfortable reiterating its forecast of 8-13% growth in Core Communications Services revenue for 2008. 

If Level 3 is representative of the Internet industry as a whole, then maybe we are seeing the beginning of a "resurgence," as Dan Caruso calls it, in the wholesale Internet industry, here in 2008, in revenue terms.  Does Level 3's traffic growth of 75-80% match the industry, or is Level 3 expecting to take market share and grow faster than the industry?  I suspect that Level 3 is growing faster than the industry, but that we may actually be seeing an uptick on overall industry growth rates from the 50-60% we saw last year.

Jim Crowe pointed to video delivery over the Internet as the big driver, citing a stat from Comscore that 10 billion videos were viewed over the Internet in February, a 65% year-over-year increase.  A continuation or acceleration of that growth rate in 2008 could nudge the overall Internet's traffic growth above 60%, since video files are so large.

It's going to use data from Level 3 to track this in the future, though, because Level 3 has stopped reporting revenue by product group and is now reporting revenue by customer segment instead. 

LVLT: Crowe Predicts CFBE For Remaining Three Quarters of 2008

Level 3 Communications released a promising earnings report this morning, with the highlight being this statement from CEO James Crowe:

"...we expect to be free cash flow breakeven for the remaining three quarters of this year."

That's big news, even though Crowe only meant that the next three quarters would cash flow breakeven on a cumulative basis.  Prior to today, the company had only predicted positive cash flow, on a run rate basis, by the end of 2008.   Today's statement accelerates the march to positive cash flow by at least one quarter, and maybe by two quarters three quarters.

It appears that the company is achieving this milestone by managing its SG&A and capital expenditures downward.  SG&A for 1Q08 was $418M, a cut of $21M from the $439M in 4Q07.  Capital expenditures were down to $113M in 1Q08, compared with $153M in 4Q07, a drop of $40M. 

Cost-cutting your way to profitability by itself would not inspire celebration, but there was other good news in the report:

• CFO Sunit Patel reaffirmed EBITDA guidance of $950M - $1.1B for 2008
• Patel also reaffirmed the company's projection "that Core Communications Services revenue will grow 8 to 13 percent for the full year 2008."

There had been speculation that the company would need to cut its sales forecast for 2008, due to a worsening economy and lack of demonstrated sales momentum, so a reaffirmation of revenue guidance is a very good sign.  Growing sales for the remainder of the year is key for Level 3.

The only dark lining on this silver cloud was that 1st quarter results in core network services revenue were sequentially lower than 4th quarter 2007.  This high-margin segment declined to $774M in 1Q08 from $783M in 4Q07.  The company telegraphed that revenue would be down a bit in the 1st quarter on their last quarterly earnings call.

Overall, though, things seem to be heading in the right direction, and the market likes the news, too.  The stock is trading at $2.69, up 13.5% on the day, so far.

I may post more after the conference call.

Full disclosure: I own Level 3 Communications stock.

Mr. Caruso: Don't Bother With Dinner Reservations

If you have been following this blog for the past week or two, you will know that there is a running dispute between my good friend Dan Caruso and myself over whether the wholesale Internet access business is growing quickly, in revenue terms.  I say it's not, Dan says it is. 

Dan thought he had me cornered with a post on Sunday about Cogent Communications, noting how the company is growing quickly in revenue terms.  However, Dan is stretching when he claims that Cogent's revenue is evidence of rapidly-growing wholesale Internet access revenue, because Cogent's revenue is not limited to pay-per Mbps wholesale Internet access.  The company also operates data centers and has colocation revenue, and they also sell local loops, and these revenue streams are likely to be growing faster than their wholesale Internet access revenue, masking weaker growth there, too.  Plus, Cogent could be growing faster than the market, meaning they are taking market share as they grow.

How could Cogent's local loop revenue be growing faster than their wholesale Internet acccess revenue?  Let's look at a sample network to show how this can happen:

Transport Annual Price Compression	10%
2007 DS3 Monthly Transport Price	$      2,000
Internet Annual Price Compression	30%
2007 Monthly Price Per Mbps	$           25
Sample Network
Number of Sites	1000
2007 Peak Internet Bandwidth Demand/Site	38	Mbps
Annual Growth in Bandwidth Demand	55%
	2007	2008	2009	2010
Bandwidth Demand Per Site (Mbps)	38	58.9	91.295	141.50725
Total Mbps needed	38000	58900	91295