Cogent Communications Holdings Inc (CCOI) 2015 Q2 法說會逐字稿

Good morning, and welcome to the Cogent Communications Holdings second-quarter 2015 earnings conference call. As a reminder, this conference call is being recorded, and it will be available for replay at www.CogentCo.com. I would now like to turn the call over to Mr. Dave Schaeffer, Chairman and Chief Executive Officer of Cogent Communications Holdings. You may begin.

Thank you, and good morning to everyone. Welcome to our second-quarter 2015 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.

Despite substantial foreign currency headwinds, we are pleased with our results for the quarter, and are optimistic for the strength of our business and the outlook for the remainder of 2015. During the quarter, we experienced accelerated constant currency revenue growth on a year-over-year basis, decreased sales rep turnover, and increased productivity out of our sales reps that was again far better than our historical averages.

During the quarter, we returned a total of $38.1 million to our shareholders through a combination of dividends and stock buybacks. We purchased 579,000 shares of our common stock for a total cost of $19.1 million, at an average price of $32.98 during the quarter. At quarter's end, we had a total of $10 million available under our previous stock authorization buyback program. The Board authorized an additional $50 million under a new program, and both programs will be extended through December 2016.

We continue to remain confident in the growth potential of our cash generating capabilities. As a result, and as indicated in our press release, we announced another increase on our regular quarterly dividend from $0.33 a share to $0.34 a share, our 12th consecutive increase in our regular quarterly dividend. Since we repurchased $19.1 million of our common stock in the second quarter, under our return of capital program that was greater than our minimum commitment of $12 million, and as such, we will not be making a special dividend payment in our third quarter, only our regular dividend.

Our return of capital program was originally planned to continue till our net debt to EBITDA as adjusted for asset related gains ratios reached 2.5 to 1. This ratio increased to 2.77 as of June 30, 2015, from 2.45 at the end of the first quarter of 2015. Our Board has authorized a continuation of our return of capital program, and changed our targeted net debt to EBITDA leverage ratio to now be within a range of 2.5 to 1, to 3.5 to 1. Please note our return of capital program is subject to change, as conditions warrant.

Now for a comment on the continued progress on the Open Internet Order. On June 12, 2015, the Open Internet Order became effective. We are fully supportive of this order, and the classification of the internet as a Title II service. We believe that the Open Internet rules strike a reasonable balance between ensuring consumers can access all lawful content while limiting the regulatory burdens on the multiple entities that comprise the internet. The inclusion of peering or interconnection in the Open Internet rules is essential to delivering on the promise of an open internet. Without interconnection, there would be no access to the entire internet, and only access to an ISP's own networks. We hope all ISPs will abide by these new rules, and support them.

Now, for a couple of comments on revenue. On a constant currency basis, our second-quarter 2015 revenue grew on a year-over-year basis from 2014 by 9.5%, which is an accelerated rate of growth from our first-quarter constant currency annual rate of revenue growth of 9.3%. On a constant currency basis, our Q2 2015 revenue grew from the first quarter of 2015 by 2%. During the quarter, traffic grew on our network sequentially by 5% from the first quarter of 2015, and growth accelerated on a year-over-year basis to 33% from the second quarter of 2014.

Our sales force rep productivity increased to 5.6 units per full-time equivalent per month, a rate that is significantly above our long-term average of 4.8 units per full-time equivalent rep per month, and a 6% increase from the rep productivity rate of 5.3% per full-time equivalent that we experienced last quarter. Our sales force turnover also improved, and was the lowest in several years at 3.4% per month, a rate that is significantly below our long-term historical averages for turnover in our sales force of 6.1%.

Throughout this discussion, we will highlight a number of operational statistics. I will review in greater detail certain operational highlights and trends, Tad will provide some additional detail on our financial performance, and then following our remarks, we'll open up the call for questions and answers. We have shortened our prepared remarks to save additional time for questions and answers. Now, I'd like to turn things over to Tad to read our Safe Harbor language.

Thank you, Dave, and good morning everyone. This earnings conference call includes forward-looking statements. These 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 facts 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. Cogent undertakes no obligation to update or revise forward-looking statements. If we use any non-GAAP financial measures during this call, you'll find these reconciled to GAAP measurement in our earnings release, which is posted on our website at CogentCo.com. I'll turn the call back over to Dave.

Thanks, Tad. Hopefully you have had a chance to review our earnings press release. Our earnings press release includes a number of historical metrics, and we have actually expanded the metrics that we report on, based on input from investors.

Expectations against targeted guidance. Our EBITDA margins for the second quarter increased by 40 basis points from the first quarter, which was 30.9%. We continued to anticipate our full-year revenue growth for 2015 over 2014 on a constant currency basis will be within our guidance range of 10% to 20%. We also anticipate that our gains on equipment transactions for 2015 will be approximately $5 million, as compared to the $11 million we experienced in 2014.

Gains on equipment transactions were $700,000 in the quarter, as compared to $1.5 million in the first quarter of the year, and $2.7 million in the second quarter of 2014. We expect our EBITDA margin expansion, excluding asset gains for 2015 over 2014, will be over 100 basis points, and this will be partly dependent on the level of legal fees we will incur related to net neutrality. We do believe that the fees we will incur in 2015 will be below the fees that we incurred in 2014 in fighting for net neutrality. Tad will now cover some additional details related to our results for the quarter.

Thank you, Dave, and again, good morning everyone. I would also like to thank and congratulate the entire Cogent team for our quarterly results, and for their continued hard work and efforts during another busy and productive quarter for the company.

Corporate and NetCentric revenue and customer connections. We analyze our revenues based upon product class, which is On-Net, Off-Net and Non-Core. We also analyze our revenues based upon customer type. We classify all of our customers into two types, NetCentric customers, and corporate customers. Our NetCentric customers buy large amounts of bandwidth from us in carrier-neutral data centers, and our corporate customers buy bandwidth from us in large multi-tenant office buildings.

Revenue from our corporate customers grew by 3.5% from the first quarter of 2015 to $57.1 million, and grew by 15.3% from the second quarter of last year. Revenue from our NetCentric customers decreased by 0.9% to $41.7 million for this quarter, and decreased by 7.5% from the second quarter of last year. The decrease was primarily due to the sequential negative impact of foreign exchange of $300,000, and year-over-year negative impact of foreign exchange of $4.8 million. Our European revenue is almost entirely NetCentric revenue, and vulnerable to the impact of foreign exchange.

Revenue and customer connections by product class. Our On-Net revenue was $72 million for the quarter, which was a sequential quarterly increase of 1.1%, and an increase of 2.3% from last year. Our On-Net customer connections increased by 3.1% sequentially, and increased by 12.3% from the second quarter of 2014. We ended the quarter with over 42,000 On-Net customer connections on our network, and our 2,191 total On-Net buildings.

Our Off-Net revenue was $26.5 million for the quarter, which was a sequential quarterly increase of 3.1% and an 11.2% increase from last year. Our Off-Net customer connections increased sequentially by 3.4%, and increased by 20% from the second quarter of last year. We ended the quarter serving over 6,500 Off-Net customer connections in over 4,500 Off-Net buildings, primarily in North America.

Comments on pricing. Consistent with historical trends, the average price per megabit of our installed base and our new customer contracts decreased for the quarter. The average price per megabit for our installed base declined by 5.5% from $1.73 from the first quarter to $1.63 for this quarter, and declined by 20.4% from $2.05, which was the price for the second quarter of last year. The average price per megabit for our new customer contracts was $1.12, which was a 7% decline from $1.20 from the first quarter, and a 14.9% decline from $1.31, which was the price for new customer contracts that were sold in the second quarter of last year, but that rate was actually better than our historical year-over-year average rate of decline. These average price declines were also impacted by FX rates, as about 50% of our traffic is in Europe, as that's primarily a NetCentric business.

Comments on ARPU, both On-Net and Off-net ARPUs declined sequentially from the first quarter. Our On-Net ARPU, which includes corporate and NetCentric customers, was $580, which decreased by 1.6% from $590 from the first quarter. Our Off-Net ARPU, which is comprised of predominantly corporate customers, was $1,365 for the quarter, which was a decline of 1% from the first quarter.

Churn rates. Our unit churn rate for our On-Net customers improved during the quarter, and our Off-Net churn rate increased, due to our planned reduction of T1 customers. Our On-Net unit churn rate was less than 1% at 0.9%, which was a slight improvement from 1% from the first quarter. Our Off-Net unit churn rate was 2.1% for the quarter, again impacted by the planned reduction of T1 customers, which was an increase from the 1.2% from the first quarter. Our gross profit margin decreased by 70 basis points from the first quarter. Our gross profit margin was 57.2% for this quarter, 57.9% from the first quarter, and 58.3% for the second quarter of last year.

Comments on EBITDA and EBITDA as adjusted. These are all outlined in the table in our press release. Our presentment of EBITDA, which excludes asset-related gains and includes all of our net neutrality, legal and economic analysis fees. EBITDA is reconciled to our cash flow from operations in all of our press releases historically, and was $30.6 million for the quarter, and our EBITDA margin improved sequentially by 40 basis points, and was 30.9% for the quarter.

EBITDA, as adjusted by including asset-related gains previously called EBITDA as adjusted, is EBITDA, and includes gains on equipment transactions. Our EBITDA as adjusted by including these gains was $31.3 million for the quarter, and the margin was 31.7%. EBITDA, as adjusted by including asset-related gains and adding back our net neutrality fees was -- the asset-related gains were $700,000 for the quarter and we incurred $1 million of net neutrality fees, so that EBITDA was $32.2 million for the quarter, and the margin was 32.6%. Finally, EBITDA as adjusted by only adding back net neutrality fees of $1 million, and those fees are included in our SG&A expense. That amount was $31.5 million, and that margin was 31.9%.

Asset-related gains and net neutrality fees have a material impact on our EBITDA as adjusted calculations and can also vary materially from quarter to quarter. Additionally, we have seasonal factors that typically impact our EBITDA, and primarily our SG&A expenses. They include the resetting of payroll taxes in the US, the cost of our annual sales meeting, annual cost of living increases including increased medical insurance costs, and the timing and level of our audit and tax and other professional services. Key seasonal factors typically increase and did increase our SG&A expense in our first quarter, and our SG&A expenses declined by $800,000 from the first to second quarter, essentially as planned, despite our net neutrality fees not declining as rapidly as we had hoped. On EPS, our basic and diluted EPS was $0.02 for the quarter, compared to a loss of $0.04 for last quarter and $0.03 earnings per share for the second quarter of last year.

Some comments on foreign currency. The impact of foreign currency has reduced the proportion of our business reported in US dollars outside of the US to about 23%. About 18% of our second-quarter revenues were based in Europe, and about 5% of our revenues were related to our Canadian, Mexican and Asian operations. Continued volatility in foreign currency exchange rates can materially impact our quarterly revenue and financial results. The average euro to USD rate for the first quarter 2015 was $1.13, and that reduced to $1.11 for this quarter. The average euro to USD rate for the second quarter of last year was substantially greater, at $1.37, so the foreign exchange impact on our revenues sequentially was a $300,000 decrease to our revenue, and on a year-over-year quarterly basis the impact was much more material and was significant at $4.8 million.

The average euro to USD rate for this quarter, so the third quarter, so far where we stand is about $1.10. Should the average foreign exchange rates remain at current levels for this quarter, being the third quarter of 2015, we estimate that the FX conversion impact on sequential quarterly revenues will have a negative impact of about $300,000. The average euro to USD rate for the third quarter of 2014, so last year, was $1.33. Should the average exchange rates remain at current levels for this quarter, we estimate the impact on year-over-year quarter basis will be a negative impact of $4.5 million.

Customer concentration. We believe that our revenue and customer base is not highly concentrated. For the second quarter of 2015, no customer represented more than 1.9% of our revenues, and our top 25 customers represented less than 7.9% of our revenue. CapEx. Our CapEx decreased 16% sequentially, and decreased 32% year-over-year basis. Our CapEx for the quarter was $10.9 million versus $12.9 million for the first quarter, and $16 million for the second quarter of last year.

Capital lease principal lease payments. They're for long-term dark fiber IRU agreements. These payments were $7.3 million for the quarter, $3.7 million for the first quarter, and $4.8 million in the second quarter of last year.

Some balance sheet amounts. At the quarter end, our cash and cash equivalents totaled $224.5 million. For the quarter, our cash decreased by $35.6 million, as we returned a total of $43.7 million of capital to our stakeholders. During the quarter, we paid $19 million for the second-quarter dividend, $19.1 million was spent on stock buybacks, and we spent $5.6 million on an interest payment on our debt. Cash flow from operations was $20 million for the quarter as compared to $18.4 million for the first quarter, and $28.4 million for the second quarter of last year. Capital lease IRU obligations again are for long-term dark fiber leases, and typically have terms of 15 to 20 years or longer, and often include multiple renewals after that. Our long-term and short-term capital lease obligations were $128.7 million at quarter-end.

Some debt ratios. Our total debt, including capital lease obligations, was $584.3 million at quarter end, and our net debt was $359.8 million. Total gross debt to trailing last 12 months EBITDA, adjusted for asset gains ratio was 4.50 at June 30, and our net debt ratio was 2.77. Bad debt expense for the quarter was again less than 1%, and was 0.9% of our revenues, the same as the first quarter, and slightly better than 1% of our revenue for the second quarter of last year. Finally, days sales outstanding was again at an historically low level, and was only 25 days at June 30, which was the same as the end of the first quarter, and again, our worldwide billing and collection team just continues to do a great job on customer service and customer collections, so I thank them again, and I turn the call back over to Dave.

Thanks, Tad. Now, for a couple of comments on the scope and scale of our network. We have over 818 million square feet of multi-tenant office space, directly connected to our network. Our network consists of over 28,000 metro fiber miles, and over 55,000 intercity fiber route miles. Our network remains the most interconnected network in the world, where we directly connect to 5,435 networks, of which approximately [45] of these networks are settlement free peers. The remaining networks are Cogent customers.

We are currently utilizing approximately 30% of the lit capacity in our network. We routinely augment capacity in portions of our network, so we can maintain these low utilization rates. Cogent operates 50 Cogent-controlled data centers globally, with approximately 555,000 square feet of raised floor space.

In summary, we believe that Cogent is the low-cost provider of internet access and transit services, and our value proposition remains unmatched in the industry. Our business remains completely focused on the internet and IP connectivity, and data center colocation services. We provide a necessary utility to our customers. We expect our annualized constant currency revenue growth and EBITDA margin expansion trends to be consistent with our historical rates, that is 10% to 20% of revenue growth, and we will achieve annual EBITDA margin expansion of over 100 basis points, inclusive of the extraordinary legal expenses that we continue to incur.

We have incurred material legal and economic analysis fees of $1 million in the second quarter of 2015, in supporting net neutrality and preparing for potential litigation. We believe that we will continue to incur additional fees related to net neutrality in 2015, but at a lower level than 2014. We will be opportunistic in the timing and purchase of our common stock. After the increase in authorization at the end of the quarter, we had $60 million remaining in our buyback program, which now is authorized to run through the end of 2016.

Our Board of Directors has approved another increase in our regular quarterly dividend to $0.34. Our dividend increase and our return of capital program demonstrates our continued optimism regarding the increasing cash flow generating capabilities of our business. We remain committed to returning increasing amounts of capital to our shareholders on a regular basis.

With that, let me open the floor for questions.

(Operator Instructions)

Colby Synesael, Cowen and Company.

Dave, it's Greg sitting in for Colby. Thanks for taking our questions. Two if I may?

One on gross margins, they came down despite the revenue increasing. The general rule of thumb you have messaged before is that 50% contribution margin Off-Net, 90% On-Net, and that rule didn't seem to hold. Hopefully you can walk us through the Off-Net and On-Net margin, and how to frame that going forward.

And second, on the revenue deceleration, specifically on the corporate side. NetCentric, I understand that with seasonality and FX, but it looks like corporate went down from 3.8% to 3.5%, and you have been excited about the productivity and the sales tenure of your guys, and just wondering if you can help us explain the quarter-over-quarter downtick, and what we can expect going forward? Thanks.

Sure, Greg.

Let me take the gross margin question first. Our historical EBITDA margin contribution rates have been around 47%. That is the blend of our On-Net and Off-Net sales.

Our On-Net sales carry a 95% EBITDA margin contribution, Off-Net about 45%. And the differential to get to the blended rate is the increase in our fixed cost of operations.

In the quarter, we saw our cost of goods sold increase, primarily as a result of the large number of data centers that we acquired last year fully coming online, and picking up the fully-loaded cost of those data centers, many of which have very low occupancy. As we continue to fill up those data centers, we will see a return to our historical rate of contribution margin, and a continued improvement in our gross margin.

If we look at EBITDA margins, the impact has really come from two things on the SG&A side: One, still spending more in legal expenses than we had hoped to spend, and while we continue to remain hopeful that we will not have to file any litigation, we have prepared briefs and motions for litigation against several parties that have been unwilling to upgrade their interconnection capacity, as required under the Open Internet Order.

And while we continue to negotiate, we felt that it was necessary to do so from a position of strength. And therefore, we have incurred the expense of preparing the filings, the pleadings, and the briefs necessary for us to file those litigations, if those negotiations do not progress.

Finally, and we have commented on this previously, we began a program of investing in our NetCentric sales organization about six months ago. That continued investment is occurring, and that has had a slight drag on our EBITDA margins by increasing SG&A.

Now to your second question about the growth rate in the corporate business, we feel very comfortable with that growth rate. It is above our historical averages. It had gone up for six consecutive quarters sequentially, to an all-time high of 3.8% sequential growth.

In the most recent quarter, 3.5% still is well within or above the range that we've historically experienced, and there will be some lumpiness, but we feel very optimistic that we'll continue to see elevated levels of corporate growth. And in fact, you saw the penetration rates in our On-Net buildings go to an all time high, particularly as we turned up several large BPLS, orders in the quarter with multiple sites On-Net for those customers.

So we remain very encouraged about our corporate business on a year-over-year basis. The growth from that business actually slightly accelerated from 15.2% to 15.3%. So overall, we think the corporate business is doing well, we think the NetCentric business is poised for significant improvement as we continue to see port capacity being made available for the problem ISPs.

Great, thanks.

Nick Del Deo, MoffettNathanson.

Hi, Dave. Thanks for taking the question.

Back to the gross margin question, my understanding was that a lot of those new data center costs like the underlying leases, would have started to have been incurred last year, when you actually assumed them. And lapping those is part of what would help to drive margin expansion this year. What were the data center costs that only started to be incurred this quarter?

Sure. So on several of the data centers, and we did acquire seven data centers last year, and about 125,000 square feet of raised floor space. In this quarter, we actually acquired one data center, and it was relatively small, it was only about 7,000 square feet. The incremental cost was rent cost. In a few of those deals we had several months of free rent, until we started having to pay that rent, and now we're incurring those additional costs.

Okay and then maybe as a follow-up, if my math is right, you are probably going to need to see something like 300 basis points of margin expansion in the back half of this year, relative to the second half of 2014, to achieve the margin guidance that you reiterated earlier in the call.

What, specifically, gives you the confidence that you can achieve that? Is it really still realistic?

Sure. So I think there are two things, Nick, that give us confidence. The first is on the revenue side.

We have seen significant improvement in port congestion with three major ISPs. Those being Comcast, which began in the first quarter, and quite honestly, was congestion-free by the end of second quarter. AT&T and Verizon, we entered agreements in the first quarter, but -- first and second quarter, excuse me, but did not see any port capacity made available in the second quarter, but are rapidly seeing both of those companies fulfill their obligation to make those ports available.

We've also seen two international carriers agree to additional port capacity, one of them has actually turned that up, Orange France Telecom. The other, Telefonica has agreed, but has not yet completed those port augmentations. With that, we will see a significant acceleration in traffic and revenue from our NetCentric customers, and because that business is On-Net, it carries a 100% gross margin contribution.

So I think that's on the revenue side, the biggest thing that gives us confidence. I think the corporate business will continue to perform above trend line. And then, on the SG&A side, we expect that our legal expenses will come down.

Again, we sincerely hope we do not have to file the several motions and pleadings that we have prepared. So we have spent the money and hope that the courtesy copies that we provided to the counter parties will show them our resolve. But again, we don't have 100% clarity on that, but we really do believe we'll see an improvement on the SG&A side, as well.

Okay. And just to be clear. All the data centers you've assumed, those are now at, the full rents are in the OpEx line?

Yes, they are.

Okay. All right, thanks.

Frank Louthan, Raymond James.

Yes. So looking at the sales, the headcount for the sales force, went up a little bit. Can you tell us which bucket those are in? Was that more of the corporate side, or the NetCentric side?

Then what are your plans for ramping the sales force on the NetCentric side, now that you've got more port capacity? When should we see that start to accelerate, and what is the target headcount there?

Sure. So Frank, our aggregate sales force will grow by between 7% and 10% on a year-over-year basis. The relative mix of sales people has historically been two-thirds corporate, one-third NetCentric. That continues to be generally true.

On the -- we did do more hiring on the corporate side in 2014, and probably that ratio slipped a little bit to something like 72-28. So most of the hiring that will occur in the remainder of this year will probably be on the NetCentric side.

The other two things that we've done on the NetCentric side is revamped the account alignment for those individuals, and reprioritized accounts to more tenured and less tenured reps, the larger accounts going to the more tenured reps. On the corporate side, that change had a material impact on rep productivity, and we're starting to see some of that on the NetCentric side, as well.

And then finally, just as we had implemented a very extensive training program for the corporate sales teams, we began a few months ago with that same accelerated level of training and management for the NetCentric. And we should see continued improvement, both because of the availability of interconnection capacity, and by having both more NetCentric reps, and having those more productive.

So I think by the end of the year, we should return to roughly two-thirds, one-third mix between corporate and NetCentric reps.

So just to clarify. Can you give us the breakdown within the quota-bearing reps which -- how many you've got today? And on the corporate side? And how many you think that might end up -- gross numbers, it might end up by the end of the year?

Sure. Today the breakdown is 72% corporate, 28% NetCentric. And by year end, we expect to be up full year by between 7% and 10%, so that's probably around 380 to 390 total quota-bearing reps. Roughly 92% or 93% of those should be full-time equivalents, and the mix by year end should be roughly 67% corporate, and 33% NetCentric.

All right, thank you.

Michael Bowen, Pacific Crest.

Thank you, good morning. Thanks for taking my questions.

The sudden drop in the sales force churn was a pleasant surprise, particularly versus historical levels. So can you give us your thoughts on what drove that? And how we should think about that percentage going forward? And then can you also let us -- I'm sorry if I missed this, but can you let us know the split now between corporate and NetCentric as a percent of revenue?

And lastly, to the extent that you can, can you talk about legal fees for the rest of the year? You have spent about $2.4 million thus far. Can you let us know, or give us an idea of how much has been spent thus far in the third quarter and handicap that for us? Thanks.

Yes sure, Michael. Let me try to take those.

So in terms of the sales force turnover, it's really a direct result of better training and better management. And we continue to refine our training programs, and I think you can see that both in the lower turnover, but again the increased productivity. And the productivity is both on -- both improved sequentially, but also is still substantially ahead of where it has historically been, and I think again, there's room for continued improvement.

There's also probably a bit of a drag on the unit productivity, because we have started to see an increasing amount of NetCentric customers wishing to take 100-gig interfaces as opposed to multiple 10-gig interfaces. And since productivity is measured on a per port basis, if you have one port replacing 10, that would actually lower unit productivity, and we are actually still seeing it at an elevated level.

So we feel pretty comfortable that we'll continue to see good productivity, even with these larger ports being installed. And I think the training on the corporate side is now fully in effect, and on the NetCentric side, is still showing improvement.

I'm going to let Tad take the revenue split question.

Sure. For the quarter, corporate revenue was 57.8% of the revenue for the quarter, and NetCentric, 42.2%. With respect to connections for the quarter, corporate was 50.8% of the connections, and NetCentric, 49.2%.

And then I'll jump back on the legal forecast.

Last year, we spent $5.4 million. We were very pleased with the results of the expenditures and the rules that were adopted by the FCC, and have at least sustained their initial court challenge. We will not be a participant in those further challenges, other than providing an amicus brief to support the FCC's position in the December hearing at the appeals court level. This is really around the validity of the rules, and the classification of the internet as a Title II service.

Now, under those rules, there is a clear requirement to have unfettered interconnectivity between networks for mass-market ISPs. Most of the ISPs that had been exhibiting anti-competitive behavior have changed their behavior, and we've announced agreements, and named names.

There are several, though, that we continue in negotiations with. Those negotiations are still active. But because there doesn't seem to be the same will to resolve the problem, we felt it necessary to go ahead and prepare for litigation.

And, quite honestly, it was a material expense in the second quarter. It was, the majority of that $1 million and preparing for that litigation, we hope it's litigation preparation we never have to use. But if we do, most of the money has been spent, and we still believe that for the full year, we will spend less in 2015 than we did in 2014.

Okay thanks.

One quick follow-up. Can you give us a little bit of your thoughts on how the rest of the interconnection agreement negotiations are going?

There are active discussions, we believe that -- again, we hope all of the parties come to the table. We believe there's at least a couple that it appears possible that we may have to file litigation.

But again, hopefully as they read through the content that is included in our filings, or our proposed filings, they may understand the strength of our argument, and come to the table. Clearly, the largest three ISPs in the US have understood this, understood their obligation under the law, and have been good corporate citizens, and opened up interconnections.

All right, thanks, Dave.

James Breen, William Blair.

Thanks for taking the question.

Dave, on the interconnect side, you talked about Verizon and AT&T and Comcast working on their port capacity now. It seems like the three of them will be pretty well caught up by the end of the third quarter, and then a couple of the international guys.

Just wondering, can you just talk about, as you look at your NetCentric traffic, how much of that is impacted by those large three? And just give us a picture of even if only half of them come to an agreement with you, the ones that you have reached an agreement with, is that the majority of your traffic to date? Thanks.

Sure. Again approximately 62% of our traffic goes to our paying customers, so it was never impacted by this. 38% of our traffic goes to our peers, approximately 20% of that 38% was going to peers that never exhibited this type of anticompetitive behavior.

So at peak, we had 18% of our traffic impacted. AT&T, Verizon and Comcast, in total, represent about 55% of the 18%, so about 10% of the problem gets resolved with those three. That leaves a remaining 8% of aggregate traffic that is still with providers that are experiencing congestion.

And as I said, there are at least two international carriers, one of which has actively been turning up ports, that being France Telecom, the other, Telefonica, has agreed and those additional ports are in progress. If those two additional providers bring up adequate port capacity to relieve the congestion, we will have about another 3% of the problem resolved, and be down to probably less than 5% of our total traffic, with the remaining three problematic ISPs; those being Time Warner Cable in the US, Deutsche Telekom, and then CenturyLink, which is the smallest of the three.

Great. Thanks.

Tim Horan, Oppenheimer.

You might have said this, but just repeat what your NetCentric revenue growth was on a normalized basis in the quarter? When you resolve all these interconnection issues, where do you think that revenue growth can go back to?

And I think historically, you used to talk about 200 basis points improvement per year in margins. When these issues are behind you, do you think you can get back to that historical improvement? Thank you.

Yes, so let me take the margin one first. The answer is absolutely yes, we will get back to annualized at least 200 basis points of EBITDA margin expansion. And again, over the 10-year history of Cogent, we have delivered approximately 200 basis points per year of margin expansion.

Our margin expansion basically stopped with the problems that we began experiencing on port congestion, due to the refusal of these ISPs to upgrade connections. Now that we are seeing resolution for the majority of these providers, and we're starting to see some relief in the legal expenses that we spent, this year, we have said 100 basis points. But then going forward, we expect to continue to see about 200 basis points.

I'm going to let Tad take the NetCentric revenue growth one.

NetCentric sequentially quarterly declined by just under 1%, 0.9%, and then year-over-year quarterly basis was a decline of 7.5%. Again, much of the FX impact, impacts the NetCentric business, since Europe is almost entirely NetCentric.

So if you look at it on a constant currency basis, we basically had about a little less than 1% sequential growth in NetCentric. And on a year-over-year basis, again, netting out FX, about 3% growth.

The long-term run rate of NetCentric growth for Cogent has been about 10% year over year. And we do believe, with the port congestion issues resolved, we will return to that rate of long-term 10% year-over-year NetCentric revenue growth, and continued traffic growth substantially faster than that of the market.

Dave, lastly on the interconnectivity, I know you can't talk about specific agreements, but can you just talk about, at a high level, are you paying for any termination traffic, are you paying for any CapEx to enable the port capacity? Thank you.

So we are paying for our own CapEx. So the ports that we install are our own equipment. The counter-party has picked up their own costs.

What I can say about these agreements, as each one is covered by a non-disclosure, but what I can categorically say, is Cogent has never purchased or sold peering, and today, does not pay for peering, and has no intention of paying for peering and it is not contemplated under these agreements. And we have no intention of selling peering as a product. We sell transit, which is connectivity to the whole internet, rather than just our portion of the internet.

Thank you.

Scott Goldman, Jefferies.

Thanks, good morning, Dave and Tad.

I guess just to ask the interconnect question a little bit differently, Dave. One of the questions we've always had, as we have been going through this process, has been around what the pent-up demand is. So you've talked about how Comcast has been active turning up ports. I think a few others, AT&T and Verizon, more coming online in 3Q and some of the international stuff.

Maybe you could just help us understand a little bit what that looks like, as these ports are getting turned up? How much pent-up demand is, or are the initial expectations for ports being filled up quickly, and perhaps waiting to achieve additional ports to that, or what that looks like?

And secondly, maybe a broader industry question. Obviously, a lot going on in the media world these days around over the top, and more and more traffic or video distribution, perhaps, working its way there. Obviously, I think you would be pretty well positioned to capture that traffic.

Wondering if you've seen changes at the margin over the last two or three quarters, as more and more of these content programmers are putting more of their content available through over the top distribution platforms? Thanks.

Sure. So just to frame your interconnection question, last year, on a sequential basis, traffic was actually down Q1 to Q2, just by 0.5%. It was essentially flat, where it was up sequentially this year 5%. That was primarily as a result of having some interconnection capacity with Comcast turned up in the second quarter, and continued acceleration in over the top video, which is the key driver of traffic growth.

So just to remind investors, these issues do not directly impact our corporate business, as there is adequate port capacity for those customers. For our NetCentric customers, this is a material part of the traffic that they are sending us, and an issue of grave concern. For pent-up demand, what we saw is a rapid rate of port augmentations by Comcast throughout the second quarter, and it was only at the end of the quarter that we saw ports not running at full capacity.

Now we met with Comcast, we provided our forecasts, and there are additional ports being turned up in this quarter. But at this point, there are sufficient ports that there is no congestion between our two networks, and we hope and expect that to continue, going forward.

So we did see a surge in traffic, as a result of having that port capacity available to us, and that was partially responsible for the better sequential performance this year versus last year. With AT&T and Verizon, there were agreements executed, but no additional ports in the second quarter.

In the third quarter, both companies have been cooperative, and rapidly turning up ports. However, the ports still remain congested, and there are more ports in queue to be turned up. We hope and expect that by the end of this quarter, that we will be in the same position we are currently with Comcast, that is, more ports and no congestion, and a road map for additional ports in subsequent quarters.

It is really too early for me to prognosticate how much pent-up demand there really is. I think we're continuing to see, literally, daily, records set in traffic growth, and I think that is driven by over the top. And third-party studies have indicated that we are now seeing OTT consume more than 20 minutes per day, which is about a 35% or 40% increase from just where it was last year.

Great, and then just one other question.

Just wondering maybe if you could talk a little bit about expectations for CapEx as we go to the back half of the year? It's obviously come down nicely year over year, and in the first half of the year. Maybe what your outlook is for CapEx for 2015 and 2016?

We expect CapEx for the remainder of this year to continue to trend substantially below where it was in 2014. As we explained to investors in 2014, we made significant investments in the data center footprint that we acquired. Many of those investments were offset by concessions by the landlords, but recorded as CapEx by us.

We fully expect our CapEx to be substantially below 2014, and even 2013 run rates. When you add CapEx plus principal payment, under our capital leases together, that total number will be below the total number incurred in 2013, and investors will clearly see that 2014 was a bit of an anomaly.

Great. Thanks for taking the question, Dave.

Eric Pan, JPMorgan.

Thanks for taking the question.

Now that you have raised your net leverage target, how should we think about the pace of the buyback concerning the acceleration pace in this quarter?

So we want to be opportunistic about taking advantage of market conditions. What the Board did is give us a little more flexibility by not having a rigid target, but rather a flexible target.

Our business does naturally delever, both based on the top line growth, which again, we expect to accelerate for the entire business, as well as for the NetCentric portion of it, and continue to expect operating leverage. So we have a, I think, very consistent dividend program in place.

We have a minimum commitment for buybacks or a special dividend, and we just want to remain flexible on additional buybacks. So it's difficult for us to give you an exact target.

Got it. Thank you.

Georgios Kyriakopoulos, SunTrust.

Dave, one more question on the connection. Just trying to understand how contracts with your bigger content provider customers are stacked through?

Do these customers require that you resolve interconnection issues with all service providers before coming back to Cogent? Or there could be a more segmented approach, where you provide service for the resolved connections like Verizon or Comcast, while the customer is still using agreements with other providers who have not opened their ports yet?

So, in all of our agreements, we have no provisions for connectivity to any specific network. We always went to sell the best product we possibly can, and have full connectivity, unfettered and uncongested to all networks.

However, our customers are free to migrate portions of their traffic, and in fact, that's what we have seen as a pattern of behavior. In fact, our sales team routinely sits down with large content customers and shares with them traffic flows of how much traffic that customer is sending us, what networks that traffic is destined to, at what locations, and where there is congestion, where we have to reroute traffic to other locations, or there is no ability to reroute it.

And customers, constantly, I think, throttle their traffic to meet the limitations we and others have on those interconnections. But for the most part, as ports get turned up, the traffic comes in and fills those ports up.

Okay, and then, one question on sales productivity. You mentioned 5.6 units per units this quarter which rebounded nicely from the last two quarters.

I was wondering if you could give us an indication how the productivity levels look like for both your NetCentric and corporate business? And also, what is an achievable target level for both businesses, and when do you think you can get there? Thank you.

Sure. Productivity levels are generally higher for the NetCentric than for the corporate reps because they oftentimes sell multiple ports at the same location. As I indicated, as customers migrate from multiple 10-gig interfaces to 100-gig interfaces, that may slightly dampen that productivity.

The corporate productivity tends to be very consistent, and generally is -- been flat on a per-rep basis, based on rep tenure. But as we have been bringing more reps on, and as the average tenure has increased, we have seen that rep productivity increase, as well.

Thanks.

James Moorman, DA Davidson.

Yes thanks for taking the question. I'll keep it real quick.

First in terms of legal. The fees that you said you've already prepared for to litigate, or possibly to have to litigate if things don't get better, is that the bulk of the costs in the quarter? I know you may have been involved with Charter and other things. Was that any impact, or was most of this preparing for those potential -- dealing with those people that won't upgrade?

And also, in terms of NetCentric, when you start to, if you resolve all this and you convert an over the top scales like we have heard DISH and others talking about, it's really taking off. Can you see going back to like YouTube type of growth rates in the teens? Or what do you expect? Thanks.

So two different questions there.

First of all, fortunately, we have spent nothing with Charter on lawyers, because Charter is a Cogent customer, has reaffirmed its commitment to net neutrality, and we were able to interact directly with management, without lawyers involved. And the answer is yes. The vast majority of the $1 million we spent in the quarter was related to preparing for litigation against several parties, and then sharing with them some of the work that we've done and the nature of our litigation. Not in a threatening way, but just showing them that we were serious about this, and we expected all players to abide by the law.

Now, we're still in discussion with the parties, and it is like you don't want to throw gas on a fire and go to court, if you are still having productive discussions. But it doesn't appear that these other parties really embrace the open internet principles, and believe it is the law of the land. So, it is possible that we may have to actually file. And the good news is, we spent most of the money to be ready to do that.

And then, on the growth rate question, the answer is yes. We will see, I believe, an accelerating rate of unit volume growth for the internet in its entirety as a result of more over the top products being made available to customers. And the internet has historically grown in the high 20s in a unit volume growth.

We have grown at about twice that rate, growing at about 55% year over year, over the past decade, in unit volume. We are below that. We only grew 33% on a year-over-year basis this quarter. I think as these ports become uncongested and as more content becomes available, we'll return to historic market growth rates, even though the base is much larger, and Cogent is uniquely positioned to capture the majority of that business, based on our pricing model.

Great. Thanks a lot.

Walter Piecyk, BTIG.

First question is on the leverage target, 3.5. Can you just give us some historical perspective of when you went to 2.5 and I guess some of the factors that the Board considered in providing greater flexibility, just to give us a framework going forward, to understand when that flexibility may extend further to 4, 4.5, or whatever?

So, well we're definitely not going to go to a higher number until we get to where we said we are going to be.

Understood. Yes understood.

So first of all, Cogent was very fortunate, coming out of the telecom meltdown, to emerge with a large network, with virtually no debt. We took on our first debt in 2007 by selling some convertible bonds, and when the convertible market imploded in 2008, we actually bought back the majority of those bonds, and remained with very low debt levels, and continued to use buybacks as our sole means of returning capital.

At the end of 2010, we decided to put senior secured debt on the balance sheet, and began a two-pronged return of capital that is continuing buybacks, as well as using dividends. And we've grown that dividend every quarter.

Now in doing so, we have ratcheted our net leverage up from about 1.5 times to about 2.7. The 2.5 was a target that was substantially below the comparable Company universe, and we felt it was appropriate.

Going forward, we look at the comps, we look at our cost of capital, and I think, what the Board looked at was the growth rate in the business, as well as the operating leverage of the business, and felt comfortable that we can continue to return increasing amounts of capital through this -- these two methods, and stay within that range. If it appears that we want to go above that range, the Board, I think, will take another look at it. But there's no promises or commitments, or even implications, at this point. We feel this gives management adequate flexibility.

Okay. Then the second question is on your data centers, having seen a couple of them, it seems like there is a significant opportunity to add some incremental revenue. I'm just curious. Some of these things you've owned for -- or some of these centers you've owned for a year now.

What do you think is going to drive your ability to add customers? I know you're focused on Cogent customers, but what do you think the turning point is going to be to leverage those experiences and layer on some incremental revenue there?

Sure. So it's really training the sales force on selling that footprint, and going back to our existing customer base, as well as new customers, and making sure they understand the footprint that we have.

Data centers have never been our core business. It still remains an adjunct to our primary business, which is selling bandwidth. But we think that we've done the training of the sales force, we've invested in bringing these centers up to current industry standards, and we continue to use our same pricing strategy on our data center footprint that we do in our bandwidth, which is to undercut our competition. And we feel quite comfortable that we'll see increased occupancy rates and very high incremental margins from those customers in the data centers.

Okay and then my last question is just on the corporate side, do you attribute the success there, the more recent success there, to the sales groups within existing buildings? Are your sales guys identifying new, more attractive buildings than you have had before?

And is any of this related to any change in how the incumbents that you are effectively stealing, or at least stealing their growth from, are acting in the market?

I think it's better growth in existing buildings. We do continue to add buildings, but at a continued slowing rate, and we have seen our penetration in buildings continue to rise to an all time high. And I think the incumbents increasingly have switched their focus to their residential businesses, and to their wireless businesses, and are deemphasizing their business customers, and we do not generally see cable in the multi-tenant office buildings the same way you would see them in the suburban campus environments, because of the costs involved in pre-wiring the buildings. And our substantial footprint in the buildings, I think, acts as somewhat of a deterrent for new entrants.

I would think with the numbers that Amazon put up in their AWS business that a lot of your corporate customers are going to be moving more stuff to the cloud and that you'd be seeing more interest in maybe an incumbent customer with a lower bandwidth moving to your services. But is Verizon just not trying to capture that with their own cloud services? Just foregoing both market opportunities, both to you guys as well as to AWS, Microsoft, whoever?

That's really a question you would have to ask to Verizon.

Understood.

What I can say is that our average corporate customer has increased their utilization to now 16% of the 100 Mbps that they have purchased.

Interesting. Thank you.

(Operator Instructions)

Michael Rollins, Citi.

Dave, when you are in selling bandwidth to NetCentric customers, what percent of the sales want just unfettered access to the internet? In order for us to buy more capacity from you, we just need to be able to get to the whole cable uncongested, versus those that say, well we understand you might have an issue with this ISP here, or this ISP there. Therefore, we're okay with that, we don't need the whole cable, we're going to buy from you specifically, where right now you can get to.

And clearly, I think as part of this, is there like a line where it just has to get to a certain level to make customers more satisfied with buying that bandwidth? Thanks.

Yes, sure. I think every customer wants the entire routing table and full unfettered access to the internet, and that is our desire to deliver that. Secondly, customers understand that Cogent is not unique in having interconnection issues. In fact, we're in much better shape now than most of our competitors, in terms of having more parts of the internet with no congestion than they have.

Customers will selectively route traffic. The preference is not to have to do that, but they also understand the realities of the market. And it really is a two dimensional analysis. Part of it being the quality of service, the second being price.

And that's why we have been able to grow faster than the market. And I think, as customers continue to see more and more of the interconnection ports become uncongested, we'll get more and more traffic.

And how do you look at the growth in the number of OTT providers as a catalyst for you to grow volume and revenue in NetCentric? And from what you see out there in customer conversations and press announcements to date, where are we, on that path, for you to get more potential revenue and volume out of your NetCentric business?

I think we're at the very early stages. So OTT represents only about 20 minutes per day per capita in the developed world out of 300 minutes a day of total video consumption. I think the market continues to become -- accelerate -- it's developing at an accelerating rate. As more content becomes available, it's becoming more fragmented with more players.

Overall, I think the business models are starting to coalesce around subscription models, and we will see, I think, at some point, almost a tsunami-type effect as it becomes mainstream, and the majority of video goes over the top. I just can't predict, is that a year away? Is it six months away? Or is it two years away? But it appears that all of the pieces are coming together.

And I think the regulation was the linchpin that really gave everybody the confidence that these models will work. And I am convinced that with our pricing strategy and our quality, we will continue to capture a disproportionate share of the market.

Thanks very much.

(Operator Instructions)

I show no further questions in the queue at this time. I would now like to turn the call over to Mr. Dave Schaeffer.

Thanks a lot, everyone. We appreciate everyone's time. We look forward to chatting, both at conferences and on next quarter's results, and again, we feel very encouraged by the trends that we're seeing in our business, both on the corporate and NetCentric side.

Thank you all very much. Take care. Bye-bye.

Thank you, ladies and gentlemen, for attending today's conference. This concludes the program. You may now disconnect. Good day.