Cogent Communications Holdings Inc (CCOI) 2011 Q2 法說會逐字稿

Good day, everyone, and welcome to the Cogent Communications Group second-quarter 2011 earnings conference call and webcast. Today's conference is being recorded. This call is available for replay at www.cogentco.com. Again, that's C-O-G-E-N-T-C-O.com.

At this time, I would like the conference over to Mr. Dave Schaeffer, Chief Executive Officer. Please go ahead, sir.

Thank you. Thank you and good morning to everyone. Welcome to our second-quarter 2011 earnings conference call. I'm Dave Schaeffer, Cogent's Chief Executive Officer. With me on this morning's call is Tad Weed, our Chief Financial Officer.

We are pleased with our results for the quarter. During the quarter we experienced sequential revenue growth, EBITDA margin expansion, significant increase in the number of buildings connected to our network, and a significant improvement in the productivity of our sales force.

Our revenue grew from the first quarter by 2.9% sequentially and grew on a year-over-year basis by 17.4% from the second quarter of 2010. Our EBITDA margin expanded by another 60 basis points sequentially to 33.6%. And it expanded by 430 basis points on a year-over-year basis from second quarter 2010. Our EBITDA increased by 4.8% sequentially from the first quarter of this year and by 34.6% from Q2 of 2010 to $25.4 million due to our revenue growth and the significant operating leverage in our business.

During the quarter traffic on our network grew on a year-over-year basis by 38% from the second quarter of 2010. However, traffic on our network sequentially declined by 1% due to seasonal factors. This is very similar to the effect we saw last year in second quarter. Since the end of the first quarter, we continued to expand our footprint by adding 60 buildings to our network, 400 metro fiber miles and over 2,700 inner city fiber route miles to our network. We generated positive cash flow of $1.9 million for the quarter, including the impact of these expansions.

Our operating income was $8.7 million for the quarter, an 18% increase from the first quarter of 2011. During the quarter our sales force productivity was 5.4 units of installed business per rep per month. Again, this was a significant improvement from our historical average of 4.1 units per rep per month and a 23% increase in rep productivity from the previous quarter in which our rep productivity was 4.4 units of installed business per rep per month in the first quarter.

Throughout our discussion we will highlight several operational statistics that I believe demonstrate our increasing market share, expansion of our network, increasing scale and, most importantly, the operating leverage of our business. We've assembled a unique set of assets, infrastructure, and operations that generate efficiently cash while selling a product at a lower price than our competitors with substantially better margins. And our product [use] demand is rapidly growing. We continue to believe there is a significant barrier for anyone to try to replicate the assets that we have assembled here at Cogent. And we are the lowest-cost, most efficient operator in our sector, focused on the most revenue-rich locations which we then bring on-net.

We are optimistic about our outlook for the remainder of 2011. We are encouraged by our results, recent trends, and order backlog, order signings that have been signed but have not yet installed on our network.

I will review in greater detail some of our detailed operational highlights and continued expansion plans. Tad will provide some additional details on our financial performance, and then following our prepared remarks we will open the floor for questions and answers.

Now I'd like to turn it over to Tad to read our safe harbor language.

Thank you, Dave, and good morning, everyone. This second-quarter 2011 earnings report and this earnings conference call discuss Cogent's business outlook and contain forward-looking statements within the meaning of Sections 27A and 21E of the Securities Act. Forward-looking statements are based upon our current intent, belief, and expectations. These forward-looking statements and all other statements that may be made on this call that are not historical fact are subject to a number of risks and uncertainties, and actual results may differ materially. Please refer to our SEC filings for more information on the factors that could cause actual results to differ.

You should also be aware that Cogent's expectations do not reflect the potential impact of mergers, acquisitions, and other business combinations or financing transactions that may be completed after today. Cogent undertakes no obligation to release publicly any revision to any forward-looking statement made today or otherwise update or supplement statements made on this call.

Also, during this call, if we use any non-GAAP financial measures, you will find these reconciled to the GAAP measurement in our earnings release and on our website at cogentco.com.

I'll turn the call back over to Dave.

Hey, thanks, Tad. Now for some highlights from our second-quarter results.

Hopefully you've had a chance to review our earnings press release. As with previous quarters, our press release includes a number of historical quarterly metrics as well as the current quarter's metrics. These metrics will be added to our website. Hopefully you will find the consistent presentation of these metrics informative and helpful in understanding our financial results and trending and our operations.

Our second-quarter 2011 revenue was $75.6 million, an increase of 2.9% sequentially and an increase of 17.4% on a year-over-year basis.

As a reminder, we evaluate components of our revenues based on product class, that is, on-net, off-net, and noncore, which Tad will cover in greater detail. We also evaluate our revenues based on customer type. Our customer type includes two major groups of customers, our net-centric customers and our corporate customers. Net-centric customers buy large amounts of bandwidth in carrier-neutral datacenters. Corporate customers buy band width from us in large multi-tenant office buildings.

Revenues and customer connections from both our corporate and net-centric customers grew in the quarter. Revenues from our corporate customers grew by 2.1% from the first quarter of 2011. Corporate customers represent 55% of our total number of customer connections at the end of the quarter and represent approximately 49% of our Q2 2011 revenues. Revenues from our net-centric customers increased by 3.7% sequentially and from the first quarter. Our net-centric customers represent 45% of our total customer connections at the end of the quarter and represent 51% of our Q2 2011 revenues.

Now for some discussion about pricing trends. Our most widely sold corporate product remains a 100 megabit connection, generally sold on a 3-year contract. And our most widely sold net-centric connection is a 10 gigabit connection. We continue to offer discounts related to contract term for both our corporate and net-centric customers, and we offer volume commitment discounts for our net-centric customers.

During the quarter, over 1,000 existing net-centric customers took advantage of our volume and term discounts and entered into longer-term contracts with Cogent, increasing their revenue commitment to Cogent by over $9 million. Customers also continued to express their confidence in Cogent by entering into longer-term contracts with us. Our average contract term increased by another 1.3% sequentially in the quarter. More than 1 out of every 3 customer contracts entered in 2011 was either 2 or 3 years in term. As a point of reference, multi-year contracts 2 years ago only represented 1 in 10 customer contracts.

Price per megabit of our installed base continues to decline, as anticipated. Our average price per megabit of the installed base declined by 3.4% in the quarter. The average price per megabit of the installed base declined from $4.10 per megabit in Q1 to $3.96 per megabit in Q2. The price per megabit of our new contracts in the quarter also decreased and was $2.96, a 5.7% decrease from the $3.14 per megabit for new orders sold in Q1 of 2011. Much of this decline was due to some of our promotional activities.

Our on-net ARPU decreased by 3.4% from the first quarter. Our on-net ARPU was $885 in the first quarter and decreased to $855 in the second quarter, due primarily to the decline in average price per megabit for our net-centric customers. Our off-net ARPU continues to increase and, in fact, increased by an additional 1.9% from the first quarter of 2011, as the size of our typical off-net connection continues to increase. Our off-net ARPU increased from $1,484 in the first quarter to $1,512 in the second quarter of 2011.

Before Tad gives you some additional details on the quarter, I'd like to address our results and our expectations against our announced long-term revenue and EBITDA targets. As a reminder, we have previously announced that we expect our top line to grow by 10% to 20%. For the second quarter of 2011 our revenues grew by 2.9% sequentially and 17.4% on a year-over-year basis.

As we announced on our last earnings call, we expect our EBITDA margins to expand by greater than 200 basis points full-year 2011 over full-year 2010. Our EBITDA margins for the quarter grew sequentially by 60 basis points and on a year-over-year basis expanded by 430 basis points. Our EBITDA margin for the first 6 months of 2011 grew by 470 basis over the comparable 6-month period in 2010. And our revenues for that same 6-month or half-year comparable measure grew by 17.2%. We continue to be comfortable with our revenue and EBITDA guidance as outlined.

Now I'd like to turn it over to Tad for some additional details on our results.

Thank you, Dave. And, again, good morning to everyone. And I'd like to thank and also congratulate the entire Cogent team on their hard work and their efforts for the second quarter.

I'll begin by providing some additional details on our revenue by product class, which is on-net and off-net and noncore. Our on-net revenue was $58 million for the second quarter, which was an increase of 2.2% from the first quarter of 2011. Similar to prior quarters, about 85% of our new sales for the quarter were for our on-net services.

On-net customer connections increased by 1,482 connections, or 6.8% for the quarter. And this was the largest increase in on-net customer connections since our second quarter of 2009. And we ended the quarter with 23,360 on-net customer connections on the Cogent network.

Revenue from our off-net business was $16.8 million, which was an increase of 5.2% from the first quarter. Off-net customer connections increased by 117, or 3.2% for the quarter. And we ended the quarter serving about 3,760 off-net customer connections in about 3,600 off-net building locations. We continued to experience an increase in the size of our off-net connections due to the increase in the availability of off-net Ethernet services.

Noncore revenues are about 1% of our revenues and about 600 customer connections.

Churn statistics -- our on-net and off-net gross monthly churn rates were essentially unchanged from the first quarter. And as a reminder, we have consistently reported our churn numbers on a gross basis. As a result, the customer who remains with us but enters into an amended Cogent contract is counted as churn and included in our gross churn rates.

Our on-net monthly gross churn rate was 2.4% for the second quarter, the same as it was for the first quarter. If you exclude the increase in these move/add/change contracts, our on-net monthly churn rate was also the same from the first quarter at 1.1%. Off-net gross monthly churn rate was 2.5% for the second quarter, and it was 2.4% for the first quarter. If you exclude the increase in move/add/change contracts, the off-net monthly churn rate was 1.6% in the second quarter and 1.4% for the first quarter.

EBITDA and gross margin statistics -- the operating leverage of our business continued to result in healthy gross and EBITDA margins. As Dave said, our EBITDA margin was 33.6% for the second quarter and that EBITDA margin percentage is the highest EBITDA margin percentage in the Company's history, beating the record we established last quarter. EBITDA as adjusted was $25.4 million for the quarter, which was a 4.8% increase from the first quarter, and a 34.6% increase from the $18.9 million of EBITDA we experienced in the second quarter of 2010.

Our gross profit margin decreased by 70 basis points from the quarter to 56.2%. And this is primarily due to our investment in the expansion of our network and the timing of those expansion costs. Our on-net revenues continued to carry 100% incremental direct gross profit margin, and our off-net revenues continued to carry an approximate 50% incremental direct gross profit margin.

Occasional lumpiness in our EBITDA and gross margins can and does occur. And if you examine our last 25 quarters of quarterly metrics that have been included in each of our press releases you will notice lumpiness in our quarterly margin expansion. That lumpiness can occur due to seasonal factors such as the resetting of payroll taxes, cost of living increases in the first quarter, which typically occurs then; the timing of our audit and tax services -- typically more of those in the third quarter when we're performing our stock services and in the first quarter when we're filing our Annual Report; and the timing and scope of our expansion activities. Our long-term trend has demonstrated increasing EBITDA margins and we expect that long-term trend to continue.

On earnings per share, our basic and diluted loss per share was a loss of $0.01 for the first quarter. But we experienced basic and diluted income per share of $0.05 for the second quarter of 2011. There was a nonrecurring amount in the results for the second quarter and included in the net income per share was a gain in the quarter related to the cancellation of an IRU capital lease of $2.7 million. That amount is approximately $0.06 per basic and diluted share for the quarter. So if you exclude that gain, the results for the quarter would have been a loss of $0.01 per share, the same as the first quarter.

Foreign currency impact -- about 30% of our businesses are derived from international operations. Consistent with prior quarters, during the quarter approximately 22% of our revenues were based in Europe and about 7% of our revenues related to Canada and our Mexican operations. Fluctuation in foreign exchange rates materially impact our revenues and, in particular, our net-centric revenues.

For the second quarter of 2010 to the second quarter of 2011, FX positively impacted our revenue growth by $2.2 million. In our first-quarter earnings call we estimated a positive foreign exchange impact on sequential quarterly revenues as an increase of $0.9 million, or $900,000. And in fact that turned out to be the actual impact on a sequential-quarter basis. Our second-quarter 2011 revenue increase of 2.9% as a result was 1.6% on a constant currency basis. And our second-quarter 2011 revenue increase of 17.4% from the second quarter of 2010 was 13.9% on a constant currency basis.

The average euro to US dollar rate so far in the third quarter is about $1.43. And as a reminder, revenues are translated at the average rate. This is a slight decline from the average in the second quarter of $1.44. If that average rate for the third quarter remains at $1.43 we will not have a material impact on sequential quarterly revenues in the third quarter from the second quarter.

On a quarterly basis we can and have historically experienced seasonal variations in CapEx and in our construction activities. Our quarterly CapEx are in part dependent on the number of buildings we connect to our network each quarter and the timing and scope of our network expansion activities. Typically we experience our lowest level of CapEx in our fourth quarter.

Our CapEx for the quarter was $13.2 million and it was $12.8 million in the first quarter. The comparable quarter from last year, the CapEx would be the same at $13.2 million for the second quarter of 2010.

As Dave said, we added another 60 buildings to our network in the second quarter. And year to date for 2011 we -- oh, I'm sorry -- since the second quarter of 2010 we've added 166 buildings to the network. We continue our network expansion in 2011. However, we expect it to be at a slightly more moderate pace than what we experienced in the last two years. However, we expect to add more than 120 buildings to our network based upon our results to date for the year. And we are on pace to achieve that goal.

Capital lease payments, including prepaid capital leases, were $4.4 million in the second quarter as compared to $3 million for the first quarter of 2011.

On balance sheet items, at the end of the quarter our cash and cash equivalents totaled $227.2 million. And, as Dave said, for the quarter our net cash increased by $1.9 million. Our operating cash flow increased substantially from the first quarter and was $19.2 million. It was a 42% increase from $13.5 million from the first quarter. And the operating cash flow of $19.2 million was partly offset by $13.2 million of CapEx and then $4.4 million of IRU capital lease payments, netting to the $1.9 million cash increase for the quarter.

We have about $92 million of our original $200 million of face value of our convertible notes outstanding. And those notes mature in 2027 and may be redeemed by us or put by the holders beginning in June of 2014. The notes are reported on our balance sheet at $73.9 million, which is net of their unamortized discount.

Our capital lease obligations are for long-term dark fiber leases and typically have initial terms of 15 to 20 years or longer and typically include multiple renewal options beyond the 15 to 20 years. These capital lease IRU fiber lease obligations were $135.6 million at the end of the quarter. And $8 million of that is a current liability. Our capital lease IRU obligations were $129.2 million at the end of the first quarter. We assumed $11.7 million of present value of new capital lease obligations in the quarter.

Our days sales outstanding on worldwide accounts receivable was 29 days at the end of the quarter, much better than our target. That was a slight increase of 1 day from the 28 days that was outstanding at the first quarter. However, the first quarter was unusually low. I want to again personally thank and recognize our worldwide billing and collections team for continuing to do a great job on customer collections and credit monitoring.

Further demonstrating our excellent customer collections performance, bad debt expense was only 1.4% of revenues for the quarter. That was an increase from the 0.9% from the first quarter which, again, was unusually low, but is an improvement from the 1.9% from the second quarter of 2010 and better than our historical rates.

As Dave mentioned, our average contract term continued to increase and increased by another 1.3% for the quarter. When we have increases in our average contract term, this can impact our revenue recognized under US GAAP. For example, increases in average contract term and estimated customer life result in a longer period over which to recognize our installation revenues.

I will now turn the call back over to Dave.

Hey, thanks, Tad.

I'd like to take a moment and chat a little bit about sales rep activity and productivity. We began the second quarter of 2011 with 243 sales reps and ended the quarter with 251 representatives selling our services. We hired 49 sales reps in the quarter and 41 reps left the Company during the quarter. Our monthly rep churn was 5.2% in the second quarter of 2011 versus 5.3% in the first quarter and below our historical averages.

We began the quarter with 230 full-time equivalents selling our services and ended with 232 full-time equivalents, or tenured reps, selling our services. Productivity on an FTE basis was in the second quarter 5.4 units per month. This performance was a significant improvement from the 4.4 units per FTE experienced in the first quarter, and better than our long-term average of 4.1 units per rep per month. As a reminder, our rep productivity rates are not based on contract signing, but are only calculated upon the installation of service and the beginning of billing for the customers.

Now to the scale of our networks. We added 60 buildings to our network in the second quarter and have 1,669 buildings on-net at the end of the quarter. This is actually a significant increase in the pace at which we've added buildings, but I view this as a relatively short-term increase. We also added to our network in both the metro and long-haul. Our network now consists of over 16,000 metro fiber miles and over 54,000 inter-city route miles of fiber. The Cogent network is one of the most interconnected networks in the world, with over 3,670 other networks connecting to our network.

We are currently utilizing only about 16% of the [linked] capacity in our network. We believe our network has substantial capacity to accommodate our future growth plans. We have expanded our network internationally, adding another country, Macedonia, and its capital, Skopje, to the network. We added significant fiber connectivity in Greece, extended our network from Thessaloniki, the second largest city in the country, into Athens, the country's largest city. And we added another datacenter to our European footprint, opening our first datacenter in London.

We continue to evaluate additional fiber routes in Europe and North America. We are evaluating and have taken advantage of certain opportunities presented in the volatile European markets. We may continue to be opportunistic and take advantage of certain of these opportunities to further extend our European network.

So in summary, traffic on our network continues to grow on an annual basis. Our revenues are growing faster than the growth rates of virtually all of our competitors, demonstrating our increasing market share gains. We believe that Cogent's strategy of being a low-cost provider and the value propositions that we offer our customers are truly unmatched in the industry. Our pricing strategy has continued to attract many new customers, improving our sales rep productivity, increasing our average contract term, increasing the volume and revenue commitments from our existing customers. It's important to remember that our business remains completely focused on the internet, which we believe is becoming an increasingly necessary utility for all businesses.

We evaluate how to optimize our balance sheet and look at the proceeds from our senior secured notes as a benefit to our shareholders. Our board of directors has authorized a $50 million buyback program similar to previous programs that we have had. We will be opportunistic about the deployment of this cash. At the end of the second quarter we had not yet purchased any stock under this program.

We continue to be encouraged by the results of our sales initiative and the size of our order backlog. We are committed to providing top-line revenue growth of between 10% and 20%, expanding our EBITDA margin and, most importantly, generating free cash for the benefit of our equity holders. We like and are confident with our network reach, the product set, the addressable market, and the operating leverage in our business. Again, as I've said in the past, in short, we like our business and our position in the market.

With that, I'd like to open the floor for questions.

Thank you. (Operator Instructions.) Dave Coleman; RBC Capital Markets.

I was wondering if you could talk about the success of the Q2 promotion that you're running in which you're targeting new logo net-centric customers, if that promotion is currently running or if you have anything new in the third quarter and what that may be.

And it sounded like Q2 on-net revenues were impacted a bit by the -- guess you said thousands of customers had repriced down to lower price points, taking volume and term commits. I was wondering how you'd expect that to trend in the second half of the year. Thank you.

Yes, sure, Dave. Thanks for the questions. First of all, we looked at some of the M&A opportunities in the market and the combinations and felt that this created an opportunity for Cogent to become a provider to companies that were buying diversity from companies that may no longer have diverse networks. So we actually are running a 3-month promotion that began in June -- so June, July, and August -- targeting these customers either directly or through the large resellers of these other networks.

I think it is fair to state that these promotions have been very successful. They had very little impact on our revenues in the second quarter because the promotions were not started until June and, while some of those were sold and installed in the quarter and did impact the average pricing calculation, they did not have a material amount of revenue impact, as generally it takes people a week or two, at minimum, to respond to an offer and sign and then a couple of weeks for us to install that service with our average provisioning window being about 10 days for on-net services.

We expect to see an increasing amount of results from this impacting our revenues in Q3 and feel very comfortable about the programs that we are offering. And we also anticipate that that should result in a sequential gain in traffic, even though we're in a relatively slow period, at least for the months of July and August. Typically we see a seasonal pick-up in September but we're actually seeing some pickup sooner this year. And I think it's in part due to these promotions.

Now, in terms of our existing customers, we continue to be encouraged by the fact that customers continue to extend their contract terms. And for net-centric customers, many of them are increasing their volume commitments. That did impact some of our price per megabit in the quarter. And I think if you look at the year-over-year rate of price per megabit decline in the quarter, it's been very consistent with what we've had over the past couple of years, which has generally been in the 23% to 25% range. We are committed to being the price leader in the market. And I think we continue to see less competitive response in the market. And we take advantage of the increase in our network scale and improvements in technology that I think are helping to stimulate new business models and grow internet traffic.

That's helpful. If I could just follow up on the on-net customer adds of about 1,500. It's a lot stronger than we were anticipating. Does that include a lot of the new logo net-centric customers that may have signed on with this promotion but that are not contributing to on-net revenues at this point?

Yes. And they will start to contribute later in the year as they will be billed for a full 90-day quarter as opposed to maybe a week or two of a quarter.

Got you. Thank you very much.

Frank Louthan; Raymond James.

Looking at the performance year to date relative to your guidance, you're saying you're still comfortable with the guidance, but you've clearly gone -- sort of surpassed that, at least on the trends you've surpassed that. Any thought to taking that up? Or is there going to be any sort of seasonality or is the second half going to come back a little bit so you'll sort of be in line with the guidance? Or any reason -- or why wouldn't you just go ahead and take the guidance up for the growth in the margin expansion? Thanks.

Sure, Frank. Thanks for the question. Our guidance is a long-term goal. We do not give specific quarterly guidance. And the guidance that we gave is not just for 2011, but 2012 and 2013 is kind of where we think the business is going to trend over a multi-year period. In the 25 quarters that we have reported as a public company, we have had sequential revenue growth each and every one of those quarters and have averaged about 15.1% year over year growth in that 25-year period. We think the range that we've set is very reasonable. While we have no reason to doubt that the trends that we've demonstrated in the first part of the year will continue for the remainder of the year -- we are not in any way concerned about any kind of slowdown in the latter part of the year -- we think that sticking with long-term guidance as opposed to kind of pointing people to short-term goals makes sense.

Now with regard to EBITDA margin, you are absolutely correct that we are trending substantially ahead of the 200 basis point floor that we've set. And while we think we will do better, we also know that, particularly in some of the countries that are experiencing some economic volatility, we're trying to take advantage of expanding the network. It's not an infinite expansion, but rather we have a target list that we're trying to grow into. And if we can be opportunistic and can pick up some of those network assets, that will have an impact on our cost of goods sold as we will pick up the operations cost before we pick up the actual revenue in those markets. So we want to be a bit careful and not necessarily paint ourselves into a box and promise something we can't deliver. But you could clearly see where the trend lines are and I don't think that we're going to really have any significant degradation from those trend lines for the remainder of the year.

Okay, great. And can you comment on what sort of made the difference in the rep productivity sequentially in the quarter? And is that -- we've got sort of a new base line here? I know it's not a perfect science. But what sort of changes have you made that have gotten that rep productivity back up where you'd like it to be and what should we think about for the next couple quarters?

I think some of it is Cogent; some of it may be more macro. Clearly we're very pleased. I mean, I think this is I think the second best rep productivity quarter in the Company's history. It's been a couple of years since we've had anything approaching this. We've done a lot in terms of rep training, rep management. I mean, these are long-term initiatives we've talked about over the last couple of years with investors. And I think we're seeing the dividends pay off. The promotional programs also are very helpful in getting that rep productivity up. I think also as the economy continues to grow at a slower rate, a lot of companies become more cost conscious and therefore really think of Cogent as a better solution. I think our credibility continues to increase in the market.

I can't necessarily say that this is a new baseline. The comp plans are all established with a quota goal of 5. We generally see rep productivity at about [4, 4.1] which says we've got our quota plans about correct. We are not changing our plan mid-year. Listen, the best news of all for me is I've got to write bigger commission checks to people. And I think Tad will agree that's the one cost he would like to see go up, commissions. But I think it's going to trend up. I think the base line average is going to trend up, but I'm not ready to put a stake in the ground, say 5's the new normal.

Okay, great. Just one last thing as far as CapEx goes -- seeing you may exceed a little bit the target for the buildings on-net. I know we've talked with investors in the past about sort of the finite number, or relatively finite number of buildings in North America that you have interest in going to. Are any of these, what you're saying, some opportunities in some of the European countries or the little bit faster base of buildings -- have you guys identified additional sites beyond that that you would go to, meaning that your overall CapEx may stay at little bit higher levels for a couple years longer? Or are we still kind of targeting the same amount of spend?

The answer is we have not identified any new buildings or new opportunities. Our CapEx for the quarter came in right in line with our guidance. We spent a little bit more on buildings and a little bit less on some other things. We took advantage of continued slowdowns in the construction industry and used that to get some very advantageous pricing in exchange for some volume commitments with some of our contractors, particularly those that operate in multiple cities. But we don't view the addressable market for buildings that generate an adequate return for Cogent increasing. We have a very good view of which buildings we're adding.

And as kind of a data point, our average building size is not degrading. We're still over 550,000 square feet for our North American buildings. And we do continue to be very selective. In terms of the datacenters, there it's very reactive. There has been capital that has flowed into that sector and there probably has been more datacenter construction activity than we anticipated. But our mix is roughly two thirds corporate buildings, one third datacenters. And that's remained pretty constant. And we feel pretty comfortable that our CapEx this year should come in slightly below what we did last year. And we anticipate over the next few years that capital number, in absolute terms, not just as a percentage of revenue, trending down.

Thanks, Dave. That's very helpful.

Colby Synesael; Cowen and Company.

Just two questions, somewhat similar to what's already been asked, but I just want to make sure I understand this. So the on-net ARPU is down about 3.4% quarter over quarter. That's a little bit more than what we were anticipating, although obviously the on-net buildings -- on-net customers were much better than anticipated. Was that a fairly even distribution or was it really -- the decrease, if you will, driven by the customers that you added associated with the promos and therefore we should see the on-net ARPU pick back up in the second half of the year?

And then my other question has to do again with the margins. Obviously you're looking at adding additional network in the second half and that could impact your margins. But based on what you're already added in terms of network in the second quarter, is there any residual impact on margin that we'll see in the third quarter? In other words, even if you don't go and do anything else other than what you've already done, we should still see some more compression on gross margins in the third quarter?

Okay. I'll take the ARPU one and Tad will probably take the margin question. On the ARPU, probably the biggest drag on ARPU on-net was with this promotion we were actually winning some smaller net-centric customers. And we are also, in some cases, winning net-centric customers where we are the backup or secondary provider because they were previously buying from two providers and now looking for diversity. But they may have contractual commitments where they could not shift the majority of their traffic to us. Some of it is price per megabit but some of it is just smaller, net-centric on-net connections.

I think as you go through the remainder of the year you will see actually more of an impact, not less of an impact, from the promotions, as most of the promotional customers will install in the third and possibly even fourth quarters of the year. But we anticipate our price per megabit rate of decline to remain about what it's been over the past few years.

And just kind of as a point of reference, our net-centric revenues on a year-over-year basis are up about 18.3%, while our price per megabit over that same period was down approximately 23.7%. And I think this demonstrates our ability to grow top line from that market segment specifically, even though price per megabit has declined. And during that same period, on a company-wide basis, because we don't really track margin by product -- it's virtually impossible with our network -- our EBITDA margins were up 430 basis points. We do not anticipate any margin pressure coming from this and we expect this to occur.

On the gross margin I'll let Tad --

And, David, if I could just add a little bit to that, so the net benefit obviously of these promos is even if the ARPU continues to see some pressure on the on-net side as a result of that, your overall revenue growth should accelerate in the back half of the year, I would think, considering you're adding so many more customers. So that essentially trumps obviously the pricing pressures. Is that correct?

That is what we are hoping for. Again, we've got our guidance out there. I don't want to over-promise. But that is absolutely part of our strategy. And that is exactly what has occurred in the past as we've run these types of promos.

And then with the margin question, I think from the stuff that we've already deployed I don't think there's going to be a lot of residual impact. Tad, I don't know if you want to talk about it a little bit more.

Sure. Stepping back from it, obviously our gross margin is impacted by the relative increase in our revenues versus our increase in our cost of goods sold. So what increases on a quarterly basis in our cost of goods sold is driven by the increase in our expansion activities, where we pick up new routes, we pick up the maintenance cost of that route.

As we add new facilities, Cogent, we categorize those costs since they're associated with revenues, as an increase in cost of goods sold. I'm not sure all of the companies in our sector treat that the same, but that's how we treat those costs. And then we also have the increase in costs associated with increases in on-net revenue, where about 50% of the revenue we have the associated cost for the tail circuit. So that's kind of the increase you see on a quarter-to-quarter basis. So the timing of the acceptance of a route or a new facility will impact on a quarterly basis the increase in gross margin, or the change in gross margin.

The margin for the quarter was 56.2%. That was a 70 basis point decline. If you look back to Q2 of 2010 it was 54.8%. We are anticipating an increase in gross margin on an annual basis. But if you look sequentially quarter over quarter, depending upon the timing of expansion activities, it can be lumpy.

And based on what you've done as of today, are you expecting that impact to pressure gross margins further in the third quarter?

We're not expecting an increase in pressure on gross margin in the third quarter, no.

Thank you.

James Breen; William Blair.

Couple questions -- one, again with respect to the CapEx number, it seemed versus our estimates you added about twice as many buildings this quarter than we had on there. I was wondering sort of what your estimate is for how many more you have to get to before you view that build-out as complete. And then maybe remind us sort of on an average CapEx basis how much it costs to get these [to the] buildings.

And then secondly, just with respect to Europe, we continue to see pretty strong numbers out of the co-lo guys there. Can you just talk a little bit about any geographical differences you're seeing in your business between the US and Europe? Thanks.

Thanks, Jim. On the corporate side, we're pretty comfortable with our target of about 1,300 multi-tenant office buildings that we will add to the network. We're at about 1,180 at this point, so about another 120. That number is not materially increasing. Our average cost to add a building, whether it be a datacenter or a corporate multi-tenant office building, is about $120,000. That did come down a little bit. It's actually -- in the quarter it was a little bit lower, because of some of these volume deals we were able to get from some of our constructors. And we felt that we had the capacity in the free cash flow that we're generating to kind of just turn up the rate a little bit at which we add those buildings. But we are not increasing the target number of buildings. So we again anticipate on the MTOB, or multi-tenant office building side, about a 10% expansion from our current footprint size and then a very slow pace from that point forward, just as new buildings are constructed.

On the datacenter side you are absolutely correct that not only in Europe, even here in North America there is a lot of capital going into that sector. There are a lot of datacenters being constructed. Customers such as the federal government are shutting down kind of proprietary centers and now moving into third-party or carrier-neutral type facilities. All this creates opportunity for us. And we react on the datacenter side, so in other words, we're in virtually all the centers that we want to be in that exist today. But as new centers get built we continue to go there. We anticipate that expansion continuing pretty much indefinitely as you continue to see more applications move to the internet, more cloud environment, all of those drivers of datacenter growth and therefore the derivative driver of our bandwidth growth.

In terms of geography, Tad, I don't know if you want to talk about it, but I think our mix has been pretty consistent.

It really has been. If you look at the margins, the relative mix of revenues, the growth rates. I'm kind of, I wouldn't say surprised, but it's proven that over the past 4 years, really, the rates of percentage of revenue of the business, of EBITDA of the business, of CapEx of the business, et cetera, have been relatively constant. As a reminder, those numbers are kind of like 22% of the business is in Europe. We did add Mexico, so we combine Mexico and Canada -- that's about 7% of the business. But we don't see tremendous geographical differences. There's a difference in customer mix, with Europe being primarily net-centric. But beyond that the geographical differences are not that significant.

Great. Thanks.

Robert Dezego; SunTrust Robinson Humphrey.

Just wanted to get a quick -- on the Global Crossing/Level 3 deal, [XOs] is out there saying that they think this is going to materially alter the market. Your comments from earlier today says that you're seeing an opportunity. Can you maybe talk about maybe what you see as the opportunities, but also the risks of that deal and kind of where you guys shake out on how you feel that deal should go through?

Yes. So we have not filed any formal comments on the transaction. We believe that these companies will combine, because independently they are struggling. And usually when businesses struggle they look for ways to inorganically fix those challenges and M&A is one of those. I'm not sure that it fundamentally changes the trajectory of those businesses. But it will allow them, at least in the short term, I think, better access to capital. And I think that's been proven.

We compete with about 10% of the revenue of each of those companies, the internet transmit market. We are the most interconnected network in the world, depending on how you measure traffic, number 1 or number 2 in the world. And we feel very comfortable that we'll continue to be a dominant player in that business. There are about 600 networks that buy today from the two of those companies, but not from Cogent. There are about another 700 networks globally that buy from the major resellers of those two companies, but not from Cogent. These are the companies that we have targeted, either as a diverse provider, or hopefully to become the primary provider.

Typically in a merger there is a number of issues that occur for a customer. One, they usually look to get a most-favored customer pricing at that point of merger. Two, they're looking for diversity. And three, and maybe most importantly, when you combine billing systems and you combine networks, you break some things. No matter how good you are, things fail. We've been through 13 acquisitions -- now it's true it's a few years ago, but we know how challenging these can be.

And quite honestly, we're in a position of strength and we feel with those 1,300 opportunities out there we need to be aggressive and capture as many as we can. Our goal is simple. We want to gain market share and we view that combination and other combinations in this industry as an opportunity for Cogent to continue to strengthen its market share.

In terms of regulatory impact, that's ultimately for Justice, the FCC and European regulators to look at. But my guess is it impacts much more the other parts of their business, since 90% of it has no overlap with Cogent.

Okay, thanks, Dave. And then if I can follow up with one other quick question. On the stock buyback and the debt you raised back in the first quarter, can you could talk a little bit about what's maybe holding you back here? And when do you think we can start to expect to see you basically buying back some stock? I think when you had talked about this in the fourth quarter you had talked about levering up and buying back a very big chunk of stock. And we just haven't seen anything yet. I'm just wondering if you could give a little bit of color on that.

Sure. Well, the goal in buying back stock is to borrow capital at a relatively low cost of capital and arbitrage that in buying equity when the market around us is in turmoil. Quite honestly, we've seen a challenging market, but our stock is up probably 50% since we announced the intentions to do this. Our goal is to buy as much of the company back as possible with the buyback and be as effective with those dollars as we can.

The money is not burning a hole in our pocket. We do understand the $1.1 million a month of negative carry. That's not lost on us. But we want to be opportunistic. We don't have an exact timeframe. I hope I'm wrong and I hope the capital markets do well, our stock appreciates and I buy at a higher price. But there is no firm timeline. It is an open authorization. We are committed to returning capital to shareholders. We are not frittering the capital away on unproductive investments and/or dilutive merger and acquisition opportunities. I understand we're not earning a lot on it while it's sitting on our balance sheet. But we're going to observe the macro world around us and try to take advantage of it, just as we have in the past.

That sounds fair. Thanks, Dave.

Barry McCarver; Stephens.

I wanted to talk, just kind of drill down on competition and specifically in Europe. Have you seen any change in your competitors on being any more aggressive on pricing? And kind of your thoughts on market share there.

Yes. I think in Europe, actually, what we've seen is a continued [de-emphasization] of this portion of the business by all of the incumbent telcos, whether they be FT, DT, BT, Telefonic or the major players, even some of the secondary players like Telia and KPN, Telecom Italia. I think all of them have de-emphasized internet transit as they have been under pressure by the European regulators, both for fiber deployment in their wireline networks and 4G rollout in the wireless. And they're all somewhat capital constrained, particularly as the capital markets there are in turmoil. So we don't really compete with them. We do compete with some of the other global players for the most part. But as I said, they've been a bit distracted with some M&A activity.

I would say the competitive environment is actually improving for Cogent, in the sense that there are less competitors. The competitors there seem to be a little less aggressive and are emphasizing other products and other services. So I feel good. And I think that's not just a European phenomenon. I think that's North America as well. But in Europe -- Europe is not quite as homogeneous as the US. And you've got to look country by country. But our goal is to the number one or two provider in each of those countries. And I think to the countries that we've been established for a long period, for the most part we're meeting that goal. For some of the newer countries, we've got a lot of opportunity ahead of us. We've got some work to do. But I feel very good about continuing to grow at the same kind of rates we have in the past across our European footprint.

Good, that's very helpful. Do you have the number of datacenters you're connected to at the quarter end?

Yes. Tad will give you the exact number. I think it's about 480 or something like that, but I see Tad's whipping through his chart. I've got 1,669 in buildings.

The other thing with datacenters -- I'm going to give you both a building number -- and actually the number of datacenters is usually about 100 higher than the building number. And the reason for that is sometimes within a building there will be multiple operators within the building and we have to connect to each one.

So, Tad, what's --

Yes. Cogent datacenters are up by 1 from 39 to 40 at the end of Q1 and then into Q2. And then carrier-neutral datacenters went from 505 to 518.

Okay. And then, just real quickly, thinking about those 1,300 buildings, to make sure I understand kind of the longer-term thought process, as you close in on that 1,300 over the course of the next several quarters, is the thought process still that coming into that number will actually be good for margins, as your sales force shifts a little bit more of its focus back to your current buildings? All things being equal, is that kind of what we're looking it?

I think it will improve margins as we improve penetration in the building. So actually, our average penetration per MTOB, or multi-tenant office building, declined a little bit in the quarter, because we added so many buildings at such a fast pace. So we're now at 8.9 customers per building, because virtually all of the new multi-tenant buildings the day they were added had no customers. This is out of an opportunity of 51 potential customers on average.

We feel very comfortable that in our more seasoned buildings we're up to kind of a 40% penetration or about 20 customers per building. We want to achieve that across the entire footprint. So we will get both an increase in EBITDA margin just because of the incremental contribution margin that Tad talked about, and we will see a decline in capital intensity as we stop adding buildings and a decline in the rate at which we're adding to our cost of goods sold, i.e., those operating costs of those buildings. So it should be accretive for both EBITDA margin and, most importantly to investors, cash flow margin.

Okay, that's great. Thanks a lot.

Donna Jaegers; D.A. Davidson.

Thanks for taking my question -- two of them really. I thought it was very interesting, your comments -- your point about that you were promoting to the 600 and 700 customers that are net-centric that you guys don't have yet. When I've talked to some of those customers, and obviously I haven't talked to a wide sampling of them, some people perceive your quality as lower than some of the big guys. How do you respond to that and how can you convince those customers that your bandwidth is as good as their bandwidth?

Sure. That's actually been a challenge that Cogent has been fighting for 10 years now. So we enter a market with a price that is only a few percent of the then-prevailing market price. We had hundreds of competitors and their response was to disparage the quality of our service. So they create doubts in the customer's mind. Now it is true today we're down to a much smaller universe of customers. Our competitors -- but for those customers who have not chosen Cogent, many of our competitors have convinced them that there is a difference, if you pay more you get something that's better. Because none of our competitors believe, I think, that they will go in and win based on price.

So the way in which we deal with that issue is we offer free trials to customers. Connect for a month and try us. Take us as a backup. Repeatedly I hear from management teams of customers -- you know, I really thought Cogent was bad, there was some issue with it. I connected to Cogent and now, not only was it as good, it's actually better than the other networks because you have better connectivity. Your network has more capacity. There's less constriction, more direct routing. It's actually a superior product. And we repeatedly end up winning the lion's share of that business over time.

Another measure of quality is service level agreements. If you look at our SLA guarantee, which is posted on our website, it looks identical to that of our other major competitors, whether they be ILECs or competitive carriers. The real measure, however, is how many credits we give against that. So for every hour of impaired service on a customer connection, we give 24 hours worth of credit. We monitor every connection, all 27,500 connections, every 5 seconds. If a failure occurs we generate a trouble ticket and begin the clock working on that SLA credit. On a revenue run rate of about $25 million a month, with a 24 to 1 multiplier, last month I think we issued $32,000 of service level credits. Yes, things fail. It's not that we're perfect. But that's a de minimis failure rate compared to our competitors.

And I think you should talk to some of our customers and I think those will tell you. Some of them were skeptical but I think they now will state that Cogent is probably as good if not better than anyone in the industry.

Thanks. And then one other quick follow-up. There's been a lot of questioning around peering charges because of what Comcast is doing with Level 3 and charging for sort of last-mile peering for CDN revenues. Are you guys seeing any change in what you're paying for IP transit peering?

Well, we pay nothing for it and that has not changed. So we have not paid. We do not pay for that service. We are not a CDN. We do sell to some CDNs but we also sell to access networks. We have today a very good relationship with virtually all of our peers. Of the roughly 3,670 networks that directly connect to Cogent, over 3,600 of them pay us for connectivity and 60, 70 of them are peers that we exchange traffic for free. Some of our peers occasionally have capacity constraints in their networks and they deal with that.

But I would say globally our relationships are good. We've had fights in the past with people. And I think our size has somewhat inoculated us. But I can't guarantee in the future. And I will say that some of the European incumbents are just very slow to move. But if you look, we've got great connectivity to most of the companies, the major access networks, all four of them here in North America, one of which, one of the larger ones, is a major customers of ours. And across Europe I think we have every large major incumbent operator either as a customer or we have a peering relationship. And we do not pay anyone for last-mile connectivity.

Okay. Thanks, Dave.

David Dixon; FBR Capital Markets.

Most of my questions have been answered, but I wondered if I could just tack onto the last question there and ask if you could provide some comment on just how the actual ratios have looked with the major peers and how that's perhaps been changing over time.

And then, secondly, just on the capital leases, if I heard you correctly, you assumed an $11.7 million in new capitalized leases. Could you just clarify that and the implication for capital leases going forward? Thanks.

Sure. I'll let Tad take the capital lease one, but I'll take the ratio question and our connectivity. So, couple points -- first of all, Cogent sells transit. It does not offer a paid peering product. Many of our competitors do, so people connect and only get a limited number of routes and pay for those. We do not offer that. You get a full routing table from Cogent.

Now, our corporate customers are all primarily pulling content as opposed to publishing content. But they represent only 2% of our traffic, yet they're 55% of connections and 49% of revenues. On the net-centric side, which is 98% of traffic, our business is almost equally divided between access networks and content publishers, whether they be CDNs, hosting companies, or direct publishers themselves.

If you look at our global interconnect ratio, we're about 1.6 to 1. So about 1.6 megabits outbound from our network versus 1 coming from other networks. Even the most stringent peering requirements are set at 2 to 1. Many CDNs run at 30 to 1, 40 to 1, even 100 to 1. We have no ratios anywhere like that. Now, some customers send to us like that, but we have offsetting customers that go the other way.

I would say that our interconnection agreements and peering policies are probably the most harmonious they've ever been. I think many of the access network operators have kind of acquiesced to our pricing strategy and have gotten comfortable and no longer want to compete with us. And remember, we are the only global Tier 1 provider that offers to furthest exit route, or run cold-potato routing. So the argument that we are somehow getting a free ride is absolutely off the table. We are more than happy to carry the bit the majority of the way. The average bit on the public internet travels about 2,700 miles. The average bit on Cogent's network travels about 2,400 miles. So we are absolutely carrying the majority of the load.

Most of the other Tier 1 providers do not configure that network in that way and are guilty of trying to get a free ride on someone else and therefore are more susceptible to these types of charges.

Dave, just on that, how would I reconcile the -- all the major carriers that we've spoken to in the US, indicating the ratios are up in the 4 to 5 range. Is that just balanced by inbound traffic from other areas, other jurisdictions?

Well, actually, we don't comment on specific, but none of the 4 major access network operators in the US have that type of ratio with Cogent.

All right. Well, thanks very much, Dave.

Okay. Tad, do you want to touch on the capital lease?

Sure. Good morning, Dave. For the quarter, as you mentioned, we assumed $11.7 million of present value of new capital lease obligations. If you look at the same quarter, comparable quarter, last year, so Q2 2010, it was $9.7 million, so slightly up. It just depends on, frankly, the length of the contract and where the contract is. If it's a long-haul agreement it's obviously going to have a larger payment. It's a longer route. And it will depend on the average term. I would say virtually every contract we sign has at least 15 years with renewals. Some of them start with a 20-year term drives that. If you look at -- for the 6-month period this year we've assumed $28.6 million in present value of new leases and the comparable number for last year was $15 million. Again, it depends on the mix of contracts, where they are and the length and term. But the amount that we assumed was consistent with what we expected, or assumed.

And the outlook, Tad, in terms of the direction for capital leases payments -- trend-wise?

Trend-wise, as we assume new obligations we expect it to moderately increase. You can see the amounts that we pay each period against the principal value on the cash flow statement. Frankly, the other amount that impacts the value on the balance sheet is going to be the foreign exchange rate, so all the euro-based leases will be translated at a higher rate at the end of the quarter if the euro to dollar happens to appreciate.

Okay, great. Thanks very much.

Michael Funk; Bank of America/Merrill Lynch.

Dave, [one thing,] all the color is very helpful. Maybe you can just help us think about the sequential trend in revenue. Obviously, timing of contracts coming on impacts any given quarter's revenue. And then you mentioned earlier, of course, terms discounts given can also create a step down. So first, maybe you could give us some comments on the timing of when the net adds came on during the quarter, I mean, specifically broken down to net-centric and enterprise. Then maybe just quantify the impact of the term discounts, too.

Sure. First of all, with regard to sequential revenue growth, we do not give exact quarterly guidance. We don't even give exact annual. As we said, we give long term. But I think the trends that you've seen in the last few quarters should continue for the remainder of the year.

In terms of timing I would say the corporate connections come on pretty much ratably over the quarter. And there's really no seasonality or degradation over the quarter or even over the year at the rate at which we add those connections.

On the net-centric side, you've really got a more complex equation. You typically will have people that have perhaps a [burst] contract, say, of a fixed commit and then they also have the ability to use more than that. Obviously if they're not using their full commit because traffic slows down in the summer, they're not bursting. They're not both paying that overage charge and likely to upgrade. You typically see heading into the fall people kind of reassessing their needs and increasing their commitments. So we get usually a lot of uptick from existing customers as global traffic just picks up seasonally. I mean, that's a common trend.

In terms of these contracts, the promotions that we ran started in June. It's a 3-month program, so I think we'll see the impact not only through this entire quarter but bleeding into next quarter as well.

To follow on Colby's question, a lot of times these are smaller commitments. To Donna's point about quality, sometimes we have to take a backup position and become primary after a quarter or two as people get comfortable with our service. We view this as laying the foundation for another 25 consecutive quarters of sequential growth and kind of that 15% average that we've achieved over the past 6 years going forward. I think next quarter should look a lot like the last quarter or two.

Great. Thank you very much, Dave.

Michael Rollins; Citi Investment Research.

I guess a couple questions; I'll limit it to two. The first is a clarification and then the second is more of a sales strategy question. So on the clarification, you mentioned in your prepared comments a couple of times about backlog. With that backlog that you were talking about heading into the third quarter, was that the revenue benefit that you were going to get, or are going to get, from the volume you sold from some of the promotional activity in 2Q that will then translate into better revenue in 3Q? Or is the backlog that you're referring to just stuff that's in the pipeline but hasn't really hit the system at all yet.

It's more the latter, but when you're saying in the pipeline, we're not counting things that we're actively working. We're talking about contracts that are signed, but not yet provisioned and built. So there really is two incremental benefits. There are the timing of some of the promotional activities and when they come on, and most of that was late in the quarter. You actually saw a similar impact when we ran some aggressive promotions in fourth quarter of last year and the majority of the benefit of that promotional activity occurred in first quarter. You also get the added benefit that those customers that signed for smaller commitments, as they get comfortable with the quality of the service, will then up their commitments or divert more traffic to us on a burst basis and we'll see added revenue. But the backlog, to be specific, we were talking about, are signed orders that have not yet provisioned.

That's very helpful. And then the second question was more in the sales strategy side. As companies are doing of the cloud and the idea, the hybrid cloud with network-neutral datacenters, a lot of the datacenters that you connect into, is your sales strategy changing at all to try to get some incremental share of that type of business, given the advantage you could provide in a very low cost of connectivity?

So I kind of vowed not to use the word "cloud," and I even broke that myself. I used it when I said the government is moving its cloud stuff into the datacenters, because I don't exactly know what that means to every customer. But it's clear that more and more computing is occurring remotely. I think it has two real benefits to Cogent. One, on the corporate side, our value proposition becomes even more compelling. They need our service. Their business has become more dependent on their connectivity to high bandwidth and over the public internet is the lowest-cost, highest-quality way to get to one or more locations.

On the net-centric side it really is a driver of demand, usually from some kind of software-as-a-service or content publisher who's pushing things out, and even to access providers who buy their connectivity from Cogent. So the net effect is as things move from the server room on site to a remote environment, they need to get connected. There are some businesses and some companies that believe private connectivity, i.e., closed networks, are the way of the future. We think that there is some limited applicability for that and we offered tunneled solutions with layer 2 products for those customers.

But the vast majority of connectivity to remote computing will occur over the public internet, therefore a public cloud, and we think that will benefit us. And not that it's so much changing the way we sell, what it is doing is making our pitch resonate much better with the customer, because they've now been educated by others, by software vendors, by hardware vendors. And they now are more likely to understand why they need the bandwidth.

For the corporate sale, probably the biggest objection we've ever gotten in the 10 years of the Company's history is -- I don't need 100 megs. It's too much. Sell me something smaller. We're getting a lot less of that now than we've gotten in the past.

That's very helpful. Thanks very much, Dave.

Jennifer Fritzsche; Wells Fargo.

I think most of my questions have been answered. I just -- and thank you for that color on David Dixon's question. That was very helpful. I just also -- and I'm not sure you mentioned this, Dave. But the IRU going down by about $2.7 million, is there an explanation there? Is that just regrooming or what-not?

Sure. I'll take that. You know, the accounting rules can be interesting to handle. This was actually an IRU that we ceased using back in the first quarter of 2008. So if you look back at those financials you'll see that we fully depreciated the asset side back then. Obviously when we pick up a capital lease IRU we record the asset and the corresponding liability. Under the rules for accounting for the liability they're a little bit more strict in terms of the timing as to when you can release that liability. And we had to wait for the statute of limitations to expire, which happened this quarter. So the corresponding liability, the one that we reversed this quarter, occurred this quarter while the asset was written down in 2008.

And to just add one thing to what Tad said, on this particular asset the problem came when the underlying provider could no longer meet its obligation. It actually had a right-of-way issue that it couldn't fulfill. And when that occurred we went to an alternate provider to meet that need. In many markets there are multiple providers. We then served the market we wanted with an IRU from a different provider. We impaired, stopped paying for this one. This one was not paid for up front. And we took the impairment charge, but now I guess 3-plus years later we're getting the reverse of that.

Great. Thank you for that color.

Donna Jaegers; D.A. Davidson.

Thanks for being so generous with your time. Just a quick question. Are you guys noticing any impact given Verizon is taking in Terremark now and Qwest is taking in -- or I'm sorry, Century is taking in Savvis? Are you noticing them being more-carrier/less-carrier neutral, so to speak and losing any business from those datacenters?

Sure. So we were never in the Savvis facilities, because they actually were never carrier-neutral. They would only sell Savvis bandwidth in those facilities. And that was also true of the Qwest facilities as well. We actually have been, and continue to be, a provider of connectivity to the Embarq part of Centurylink. And we've seen no real change in that behavior.

Now in the Terremark case, there are really two different relationships. We had sold bandwidth to their managed services business and continue to do that. It's a very small customer relationship, but I imagine at some point that bandwidth will be entirely on Verizon's network. That would make sense. And then in terms of being able to access their facilities, both here domestically and internationally, we remain in those facilities and have in our contracts terms and conditions that provide us unfettered ability to sell to any customer in those facilities. And we've seen no change in their behavior in allowing us to sell into those facilities which we view still as carrier neutral. That is different than organically developed Verizon data centers which have historically not allowed any other carrier in them. And we've never been in those facilities.

Great. Thanks, Dave.

It appears we have no further questions in the queue at this time. I would like to turn the conference back over to Dave Schaeffer for any additional or closing remarks.

Well, I just want to thank everyone. It was a long call, but a lot of good questions. And, again, we're very pleased with the quarter and we look forward to chatting with you guys next quarter. Take care. Bye, bye.

That does conclude today's conference. Thank you for your participation.