The competitive metro provider with the most buildings on-net, TW Telecom (NASDAQ:TWTC, news, filings), posted its first quarter results today. Revenue growth accelerated at a higher rate than in the past and bested estimates at $332.5M, while earnings per share fell sequentially as expected to $0.08 due to seasonality in costs. Capital expenditures remained high, and the company continued to expand its on-net footprint very aggressively. Here is a quick tabular summary in the context of the past four quarters:
| ($ in millions) | Q1/10 | Q2/10 | Q3/10 | Q4/10 | Q1/11 |
|---|---|---|---|---|---|
| – Data & Internet Services | 129.1 | 134.1 | 138.8 | 145.1 | 152.2 |
| – Network Services | 89.6 | 90.2 | 90.0 | 89.5 | 89.5 |
| – Voice Services | 84.1 | 84.0 | 82.9 | 81.9 | 83.0 |
| – Intercarrier Compensation | 8.5 | 8.7 | 8.4 | 8.3 | 7.8 |
| Total Revenue | $311.2 | $316.8 | $320.3 | $324.8 | $332.5 |
| M-EBITDA | $114.2 | $114.4 | $115.5 | $119.4 | $121.4 |
| M-EBITDA Margin | 36.7% | 36.1% | 36.1% | 36.7% | 36.5% |
| Earnings per share | 0.08 | 0.10 | 0.11 | 0.11 | 0.08 |
| Revenue Churn | 0.9% | 1.0% | 1.1% | 1.0% | 1.0% |
| Capital Expenditures | $80.9 | $85.0 | $77.8 | $78.1 | $79.3 |
| On-net buildings added | 367 | 417 | 537 | 512 | |
| Free Cash Flow | $17.5 | $15.2 | $23.8 | 26.4 | $26.1 |
Revenue: As you can see, all the growth continues to come from the company’s Data & Internet services, while Network, Voice, and Intercarrier Comp remain flat. The first quarter normally sees slower growth, so to see an acceleration now bodes well for the rest of the year. The fact that they guided to the high end of their capex range for the full year supports that as well.
Costs & EBITDA: SG&A held firm despite seasonal pressures, but network expenses went the other direction. Thus EBITDA of $121.4 was probably inline with expectations. m-EBITDA margins fell slightly, but remained at the middle of the range we have been seeing for some time – 36.5%.
On-Net Buildings: TW Telecom continued its accelerated pace of adding buildings to its network, adding 512 more during the quarter – far more than anyone else amongst the competitive fiber operators. They also began reporting their full building count, now including the 1397 buildings without electronics and ILEC local serving offices they previously left out. That puts their full on-net building count at 13,742 as of the end of the first quarter.
Final Thoughts: The one thing missing from tw telecom’s performance over the past few years was a higher rate of growth. Things definitely seemed to shift into a higher gear this quarter, but of course it takes more than one quarter to call it a trend. I continue to hear M&A rumors involving the company, almost always as a buyer, but nothing yet definitive.
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!
Categories: Financials · Metro fiber
Join Our LinkedIn Group
The speed at which they connect buildings to their fiber makes them look like another construction company.
Rob, do you know how close their fiber rings are to these buildings that they are so quick to connect to? Do you know what % of the 12,345 bldgs. are owned vs. leased, if any, and how many are large facilities like co-lo or data centers? Do you know the avg. # of tenants in one of their connected bldgs.? Also, what capex additional is necessary for “without electronics” bldgs., and why do you think they feel compelled to report their ILEC serving offices now? Is this to say that ILEC serving offices(part of the 1,397 smaller #) are the only ones where ILEC access charges might be levied?
CarlK,
As requested, I took a shot at answering some of your questions here:
http://www.investorvillage.com/smbd.asp?mb=444&mn=103627&pt=msg&mid=10477669
Enron, I see you took a couple of pot “shots” at me as well. Did that feel good? I am sorry I make you feel as though I am the Alpha and the Omega, for I am not and am always looking to learn from investors like yourself. I have always been taught that, the only dumb questions are the ones that aren’t asked. We can modify that old adage somewhat to account for other variables, but for now it’s appropriate.
In the case of owned real estate which you have clarified, I had just wondered if the previous Time Warner Cable involvement would have contained commercial real estate assets for direct, or indirect use over time. In addition, whether like (3), they utilized large data centers or co-lo sites for bringing non traditional enterprise customers together, i.e., internet focused, etc.
For example, in our business we had owned the commercial building we ran operations from while renting to other multi purpose tenants at the same time.
What I conclude from your analysis is this, however. I would have to be a “desperate” seller in the marketplace to be accepting such “expensive collateral” in stock, if twtc was coming to my door making me an offer I would want to refuse.
Maybe that lead to Carl Icahn, we’ll see.
Thanks again, Enron.
Any time you reach 10K buildings, a large percentage of them are enterprises, as there are only so many large data centers. But you can find property ownership, such as it is in the 10-K for whichever company you like.
As for how close their fiber is to the buildings they are connecting, I’d say it’s probably a fairly standard number. Building laterals is heavily geographically dependent, but it’s not something for which there is an excessive amount of economic magic that some have and some don’t.
I’d disagree with en_ron that TWTC’s on-net buildings would have a very similar number of tenants per building as Cogent. Cogent focuses on a different type of enterprise building in a different type of city, on average. TWTC’s tenants per building would be smaller. For that matter, Cogent’s would be smaller too if they kept up their pace all the way to 10K buildings.
Rob,
Below are some comparisons between CCOI and TWTC. The numbers were taken from financials and the respective web sites. There may be some definitional discrepancies– for example “customer connections” compared to “customers”– and they likely have a different mix within those respective definitions of “buildings” and data centers. But it is probably instructive in a loosy-goosy sort of way.
CCOI, customer connections– 25,046
TWTC customers — 27,000
CCOI ’10 Revenue — $260 mm
TWTC ’10 Revenue– $1,300 mm
CCOI Rev/connection — $10,500 ($875 pm)
TWTC Rev/customer– $48,000 ($4,000 pm)
The lower MRR for CCOI reflects a number of factors:
— for sure far less complicated product set, IP data only
— for sure far far more aggressive data pricing per bit
— (probably) smaller average customer size (# of employees)
What I can’t conclude is that CCOI on average has lit smaller “buildings” — it may well be the case, but they do generally target the same domestic Metros and everyone sells into the same buildings. Anyway I’m not even sure that an answer to this small question gets us anywhere.
I did not mean to suggest that CCOI has lit smaller buildings, but rather that their customers are smaller and there are more of them per building. Part of it is that CCOI is so very specific about its target enterprise buildings, which are the very largest ones in downtown areas of tier 1 cities.
Rob,
In the above did you mean to say TWTC targets the largest buildings in Tier i cities?
Very good, Rob. Which brings us to the more important question:
How much Average Revenue Per Customer(ARPC) which Enron attempted to answer by mixing it with “COLOR” form Cogent as well.
CarlK,
The minimum ARPU which may justify running fiber to a building ($4,000 per month) wasn’t based on Cogent. It was a calculation based on an assumed $40-45,000 average cost and a required 30% IRR over a three year contract life. This is the hurdle rate identified by TWTC. The additional assumption is that the EBITDA margin on that revenue is 50%. Of course, once the building is lit incremental revenue from other customers is gravy.
You have done excellent analysis in determining why twtc is likely to sell more bandwidth vs. CCOI overtime as a result of the size of their customer sets inclusive of the types of commercial buildings they light. This might show up as a result of churn too, or what I’ll describe as the business sustainability factor(SF). Smaller businesses as a general rule have a much lower success rate which will lead to more “disconnects” inside CCOI’s world vs. twtc.
You have also proven another important factor which makes twtc less of a market leader, requiring a lesser multiple than (3), by example.
It is embodied in their IRR, or cash on cash return as well as territorial limitations where (3) provides global connections over domestic or regional.
This will be easy for anonymous to understand:
LVLT =six months-1.5 years cash on cash return on invested capital
twtc= 3 years cash on cash return on invested capital
The compounding effect for adding NEW CUSTOMERS, especially customers with large growing bandwidth needs is huge comparatively.
Now, what I am waiting for (3) to tell its owners to really catapult us ahead of these two silly companies you are in love with, Enron, is the speed at which they can return $1.00 if investment capital back in cash relative to the 100K enterprise buildings that their fiber rings are distanced just 500′ away!
There is an ongoing aggressive campaign-construction project-to get that done, and I expect that (3)’s payback for the same type of investment will trump twtc’s significantly.
Mileage does vary very much, business by business!
CarlK,
There is a fundamental and fatal flaw in your arithmetic which results in a fatal flaw in your conclusion.
I will give you 24 hours to identify this flaw and award you a red star if you are successful.
I’ll admit that I rarely, if ever, understand what CarlK is talking about, but that is a great question on ARPC, imo.
ARPC here would probably reflect customer focus more than anything else, wouldn’t it?
You’re right, Enron. twtc’s annual return is 10 percent(30/3) requiring paybacks of 7.2 years! That’s not much less than the time they have remaining on (3)’s “leased” dark fiber they purchased long ago and far away. No wonder they’re riding more (3) light since that Renesys report back in December, 2010.
I am still looking forward to (3)’s Storey for adding more “color” on their speed to return investment dollars from enterprise building connections.
I will also say that, while (3) has been slowly adding buildings yoy, they have never changed the six months to 1.5 year cash on cash payback return numbers. This leaves one to speculate that such “success based capital” will follow similar patterns in anticipation of them moving a 1000 strong “commission based” sales force building by building and leaving no prisoners by the time they’re done.
It will be very good for American Enterprise, as well as the American experience when it’s over, however.
CarlK,
You do not get the red star, in fact you are travelling farther down the rabbit hole of confusion and flawed conclusions.
A 30% IRR on success based CapX, as in my toy example, results in a cash on cash payback of ~ 22 months. It will be no different for LVLT who generally target the same ROIC for incremental capital.
However, I also must admit that somewhere in my process I pressed the wrong button. The required MRR to justify a newly lit building (assuming a $45K cost to build and a 30% IRR) is ~$2,400 (not $4K). Mea culpa.
However the cash on cash return is not three years and is not seven years. I don’t know how you inferred that but in doing so you inevitably reach the wrong conclusion.
I have overposted on Rob’s board so I will shut up.
I have my doubts about the TW Zayo rumor. I know someone who works with Zayo and heard that they are looking for new investors. I am sure they would sell to TW but I don’t think Zayo brings anything to the table for TW.