Good morning, and welcome to the Cogent Communications Holdings Second Quarter 2016 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.

Hey. Thank you, and good morning. Welcome to our second quarter 2016 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 and continue to be optimistic about the strength of our business and the outlook for the remainder of 2016. For the quarter, we achieved sequential revenue growth of 1.5% and year-over-year revenue growth of 11.3%. For the quarter, we achieved year-over-year quarterly traffic growth of 41% and sequential traffic growth of 2% in a traditionally slow seasonal part of the year. Our EBITDA, as adjusted, increased by 25.9% from the second quarter of 2015 and by 10.6% from the first quarter of 2016 to $39.4 million. Our sales rep productivity was 5.9 units per full-time equivalent rep per month, a productivity rate that is significantly above our long-term historical average of 4.9 units per full-time equivalent rep per month. Now, for a couple of comments on our leverage ratios and builder baskets. During the quarter, we purchased $10.8 million of our $200 million notes at par. And, combined with our increased EBITDA, we were able to reduce our gross leverage to 3.94. Our leverage ratio is now below 4.25 to 1, which is the maximum threshold allowed under our indentures governing our notes for the release of $155 million of accumulated builder basket that has essentially been trapped in our operating companies. This builder basket is now available for us to transfer funds to our holding company to be used for dividend and for stock buybacks. At quarter's end, we had a total of $47.8 million available for stock buybacks under our stock buyback program, which is authorized to continue through December of 2016. During the quarter, we returned a total of $16.7 million to our shareholders through our regular quarterly dividend. We remain confident in the growth potential and cash-generating capabilities of our business. As a result, as indicated in our press release, we announced yet another increase in our regular quarterly dividend from $0.37 per share per quarter to $0.38 per share per quarter. This represents our 16th consecutive increase in our regular quarterly dividend. Throughout this discussion, we'll highlight a number of operational statistics. I will review in greater detail certain operational highlights and trends. Tad will provide some additional details on our financial performance. Following our prepared remarks, we'll open the call for questions and answers. Now, I'd like 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 will find these reconciled to the GAAP measurement in our earnings release, which is posted on our website at cogentco.com. Now, I'll turn the call back over to Dave.

Hey. Thanks, Tad. Hopefully, you've had a chance to review our earnings press release. Our press release includes a number of historical quarterly metrics. We continue to anticipate that our full-year growth rate for 2016 over 2015 on a constant currency basis will be within our long-term guidance range of 10% to 20%. On a constant currency basis, our quarterly revenues increased by 0.09% from the first quarter of 2016 and by 11.1% from the second quarter of 2015. Our quarterly revenue increased sequentially by 1.5% and by 11.3% on a year-over-year basis, as reported. Our EBITDA, as adjusted for Q2 2016, increased by 25.9% from the second quarter of 2015, an increase by 10.6% from the first quarter of 2016 to $39.4 million. Our EBITDA, as adjusted, margin increased from the second quarter of 2015 by 410 basis points and increased from the first quarter of 2016 by 290 basis points to 35.8%. Our gains on equipment transactions were $4.4 million in Q2 2016 as compared to $1.9 million in the first quarter and $700,000 in Q2 of 2015. We anticipate the gains on our equipment transactions in the third quarter will be relatively minimal. However, we do anticipate that our total equipment gains for full year 2016 over 2015 will be slightly greater. For the first six months of 2016, we achieved EBITDA, as adjusted, margin expansion of greater than 200 basis points on a year-over-year basis. We will expect that we will achieve approximately 200 basis points EBITDA expansion in margin for full year 2016 versus 2015. Our revenue and EBITDA guidance are intended to be long-term goals rather than be used as specific quarterly guidance. Our EBITDA, as adjusted, is impacted by the amount of equipment gained and also impacted by the legal fees that we have been spending to support net neutrality. Tad will now cover some additional financial details related to our quarter.

Thank you, Dave. And, again, good morning to everyone. And I'd also like to thank and congratulate the Cogent team for their results this quarter and their hard work and efforts during another very busy and productive quarter for the Company. Corporate and NetCentric details. We analyze our revenues based upon product class, which is on-net, off-net, and non-core. And 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 multitenant office buildings. Revenue from our corporate customers grew by 3% from the first quarter of this year to $66.5 million and grew by 16.5% from the second quarter of 2015. Revenue from our NetCentric customers decreased from the first quarter by 0.6% to $43.4 million for the quarter and increased by 4.1% from the second quarter of 2015. On connections, corporate and NetCentric, corporate connections were 28,596, which was a sequential increase of 2.6%, and NetCentric connections were 28,967, which was an increase sequentially of 5.4%. Revenue and customer connections by product class. Our on-net revenue was $79.5 million for the quarter, which was a sequential increase of 1.1% and an increase of 10.5% from last year. Our on-net customer connections increased by 4.2% sequentially and by 17.2% from Q2 of last year. We ended the quarter with over 49,200 on-net customer connections on our network and over 2,297 total on that multitenant office buildings and carrier-neutral data centers. Our off-net revenue was $30.1 million for the quarter, which was a sequential increase of 2.7% and an increase of 13.7% from last year. Our off-net customer connections increased sequentially by 4.1% and…

Hey. Thanks, Tad. Now, for a couple of comments on the scale and scope of our network. Our network continues to grow. We have over 840 million square feet of multitenant office space in North America directly connected to our network. Our network consists of over 28,800 miles of metropolitan fiber and 56,100 miles of intercity fiber. The Cogent network is one of the most interconnected networks in the world. Today, we directly connect with over 5,700 networks. Approximately 30 of these networks are settlement-free peers. The remaining 5,670 networks are paying Cogent customers purchasing transit from us. We are currently utilizing approximately 28% of the lit capacity in our network. We routinely augment capacity on portions of our network to maintain these low utilization rates. We operate 51 Cogent data centers, comprising approximately 565,000 square feet of raised floor space. We are currently operating these data centers at approximately 30% capacity utilization. So, in summary, we believe that Cogent is a low-cost provider of internet access transit services, and our value proposition is unmatched in our industry. Our business remains completely focused on internet and IP connectivity and data center colocation services. We provide a necessary utility to our customers. We expect our annualized constant currency long-term growth rate to be within our historical range of 10% to 20% revenue growth and our long-term EBITDA, as adjusted, annual margin expansion to grow by approximately 200 basis points. We have incurred material legal and economic analyses fees defending net neutrality. We spent a total of slightly over $1 million in the second quarter of 2016, and we spent $3.7 million for the full year of 2015. We will continue to incur additional legal fees in 2016, but we do expect those fees to be slightly less than the full year 2015 expenditures. Our board of directors has approved yet another increase in our regular quarterly dividend to $0.38 per share per quarter. Our dividend increase continues to demonstrate the optimism we have about our business and the increasing cash flow-generating capabilities of our business. Both growth in revenue and operating leverage are showing through in our business. We will continue to be opportunistic about the repurchase of our outstanding common stock. At the end of the quarter, we had a total of $47.8 million remaining under our current buyback authorization, which runs through the end of the year. We are committed to returning an increasing amount of capital to our shareholders on a regular and systematic basis. Now, I'd like to open the floor for questions.

And I'm showing our first question or comment comes from the line of Colby Synesael with Cowen and Company. Your line is now open.

Hi. Great. Thank you. I wanted to talk first about CapEx. So you spent $14 million in the second quarter. I think, year-to-date, you've spent $29 million. And if I'm remembering correctly, I think the guidance was that you were going to spend something less than the $36 million that you spent in 2015. So I'm trying to understand if that's still expected. And, also, just a little bit of color on what exactly it is you're spending that $14 million or $29 million year-to-date on. Your metro miles, for example, haven't increased, and I think you already had taken care of the arrangement with Cisco late last year. And then, my other question is, your top customer went from 2.2% of revenues in the first quarter to 1.3% this quarter. It seems like that may have been almost the sole reason for the slight miss on the top line relative to expectations, particularly on on-net. I was wondering if you can address that and if there's something that's structural that we should assume that that's kind of the new run rate level to which they'll be spending with you. Thank you.

Sure. Thanks, Colby, for the question. So, first of all, when we look at our capital expenditures, they are primarily for equipment in our network as well as the construction activities associated with connecting buildings, the laterals into the buildings and the riser work in the buildings. In the quarter, we added 16 multitenant office buildings. We also did actually add almost 500 miles to our metro footprint. So we did go up from approximately 28,300 miles at the end of last quarter to 28,800 miles at the end of this quarter. Also, it's important to look at the combination of principal payments on capital leases along with CapEx. And there, the number was virtually identical for the first half of this year versus the first half of last year. And we do expect that combined number to be lower this year than it was last year. Our capital spending tends to be front-end-loaded in the first part of the year. Typically, the fourth quarter is always our lightest, as most of our construction activity gets done in advance of certain building moratoriums. So I think when investors look at that combined number of CapEx plus principal payments, it will be lower this year than it was last year, and that guidance remains in place. We have not specified exactly how much lower, but it will be lower. Now, switching to your second question. We had gotten a number of questions previously about the moderating or appeared moderating rate of price decline per megabit. And that had been hovering around 16%, 17%, which is below the historical average of about 22% year-over-year. In this quarter, certain of our large customers were able to reach volume thresholds, which then lowered their price per megabit. That resulted in a much more precipitous drop in our price per megabit, about 31% year-over-year, which is, again, above trend line, making up for some of the lost ground. As a result of that, the relative mix or concentration of revenue from certain customers changed. While I'm uncomfortable in commenting on very specific customer contract terms – I think they're proprietary between us and our customers – we continue to see growth from all of our customers. Our traffic growth was up 41% year-over-year and 2% in what was a sequentially slow quarter. And the reduction in our percentages is a result of people getting lower prices per megabit, primarily. So hopefully that answered your question.

Okay. Thank you.

And our next question or comment comes from the line of Timothy Horan with Oppenheimer. Your line is now open.

Great. Just following up on the volume growth, Dave, can you give us some color kind of what you're expecting for the second half now that some of the traffic is kind of getting unblocked? And, I guess, the traffic growth was a little bit stronger year-over-year in the first quarter. And then, secondly, any more color on the ARPU declines on the corporate customers? Are you seeing any pricing pressure there? Or any other color there would be great.

Yeah. Sure. Thanks for the questions, Tim. So, first of all, in terms of volume growth in traffic, if we go back and look at the past several years, typically, our sequential growth rates are, in fact, the slowest in Q2 because schools are out, people are spending more time outside. And, in fact, there are some quarters – or some years in which we experienced zero sequential traffic growth in the second quarter. So there was nothing abnormal here. It is true our year-over-year growth declined slightly from 45% to 41%. I view that as more noise than anything, indicating any trend. We do continue to experience significant improvement in port augmentations with all of the various carriers in North America that had been congesting ports. And we are virtually congestion-free with all but one of those carriers, CenturyLink, which is in the process of continuing to augment and add capacity. In Europe, we have gotten material improvements in connectivity with Telefónica. While we are not completely congestion-free in all markets, we are seeing significant improvements. And both companies are committed to adding capacity until there is no issue with customers. We have gotten modest increases in connectivity with both FT and Deutsche Telekom, although the capacity increases have been too small to alleviate the congestion. And since there is a binding agreement with Deutsche Telekom that requires them to add additional capacity, we actually chose to litigate, and many of those expenses of that litigation flowed through in this quarter. And we are set for trial in Germany later this year. With regard to traffic growth in the latter part of the year, we expect sequential growth to re-accelerate heading into the cooler months, and we also expect gradual improvement in our year-over-year growth rates, as we have still not returned to our historical normal growth rate of approximately 52% year-over-year. Today, we're at 41%, so we do think there is room for improvement, and we'll expect to see that continue throughout the second half of the year. Now, with regard to the corporate ARPUs, the decline was relatively modest and in line with historical averages, at about 2.9%. We would expect to continue to see our corporate business performing very well. We grew sequentially 3% in that business. We grew year-over-year 16.5%. Now, it is true part of that 16.5% did include USF fees, but we also do expect to continue to grow the business sequentially at comparable rates, meaning roughly 3% and roughly 13% or 14% year-over-year rates. The actual corporate ARPU was flat. It was on-net ARPU that was down 2.9%, and a lot of that was driven by NetCentric.

Great. Do you think the Olympics and OTT launches in the second half will help drive overall volume growth? Any color there?

I think it will drive volume growth. And many of those broadcasters are, in fact, coaching customers. But with any one specific event, whether it be carrying traffic for Amazon Prime Day or carrying traffic for Pokémon GO, we see bumps up from specific customers. But because our base is so broad and so diversified, I think it's rather the confluence of all of these things as opposed to any one specific event. And remember, the Olympics only last two weeks.

Thank you.

And our next question or comment comes from the line of Eric Pan with JPMorgan. Your line is now open.

Hey, guys. Thanks for taking the questions. So first, asset sales were much greater than expected this quarter. Can you give us some color what kind of equipment you're selling, how old they are, who's buying them? And since this seems to be a regular occurrence, should we think about depreciating these assets at a slower rate?

So two different questions here, Eric. First of all, these asset sales have been occurring for the last several years. They probably will occur, going forward. We have continued to find buyers for some of this older gear. The gear that is primarily being sold is a combination of Cisco routers that are older generations, 7206s, 7609s, and some 6500 switchgear. We continue to find buyers. I think it'd probably inappropriate to comment on who those buyers are. And, typically, this equipment has been fully depreciated by Cogent and, therefore, has no book value. And it's when we receive the money for equipment or receive value for that equipment that we then have to record the gain. As Tad indicated, we will experience some additional sales this year, but they will be, I think, at a much lower pace in the latter part of the year than they were in the first part.

Got it. And then, now that you have the access to the restricted cash, should we expect share buybacks to resume in the third quarter? And will you set a minimum buyback level like you did previously?

So we do have an authorization in place. I think it's inappropriate for us to signal in advance of buying stock to not distort the model (36:28). We do have a buyback. We have a demonstrated history of spending over $220 million in buybacks and purchasing $10.2 million shares back over a 10-year period. We will be opportunistic. We saw our net leverage decline, actually, by 10 basis points in the quarter. We have a goal of being between 2.5 and 3.5, we're at 2.88 now. So we will use a variety of tools to return capital to shareholders and stay within that guidance range. In terms of a specific guaranteed buyback program, I think the board's considering that but has not made a final decision on it at this point.

Got it. Thank you.

And our next question or comment comes from the line of Nick Del Deo with MoffettNathanson. Your line is now open.

Hey. Good morning. Thanks for taking my question. I want to return to Netflix. Are you still getting your fair share of traffic from them? And how should we think about that $1 million sequential revenue decline in terms of this split between pricing and traffic and seasonality?

So traffic was up for the quarter, sequentially, 2% in a quarter that is typically characterized by flatter or even sometimes negative traffic growth. Two, we, I think, continue to remain Netflix's primary supplier of transit services. They are committed to Cogent and will continue to grow. We have in place for most, if not all of our large customers, volume-tiering, where when they hit certain unit volume thresholds, they then are entitled for their entire base to get a lower price per megabit. While we do not comment on price per megabit for specific customers, typically, the more a customer buys, the lower the price gets. We anticipate that Netflix and all of our OTT customers will continue to see increasing traffic growth. And we actually think that rate of traffic growth, the second derivative, will also continue to accelerate as poor congestion issues continue to get resolved in Europe, as they have been here in the U.S., and as more OTT content becomes available. We feel that the average price per megabit decline of about 22%, which we've experienced over the past 12 years, is about the right way for our customers and our investors to think about it. And if you actually go back, we disclose that rate of decline every quarter. And for the last several years, we've had episodes where customers hit volume thresholds and got lower prices per megabit. That will continue going forward, so it's not perfectly linear. But if you go back and look, in Q2 of 2012, our price per megabit declined 38.5%. In Q3 of 2013, it was 32.2%. And the 31.5% decline this quarter is in line with that. So there are these kind of episodic periods where people hit volume thresholds. But the average rate of decline is pretty consistent. And just as investors got excited, we had a couple of quarters of what appeared to be more moderate declines, it was really based more on customer mix than it was the underlying rate of decline changing.

Okay. I mean, it seemed like if traffic was consistent, it seems like that step-down in unit pricing would have been pretty steep. If we look out over the next several years, and assuming Netflix continues to grow their traffic with you at a rate consistent with what they've done in the past, should we anticipate another sort of step-down at some future volume threshold?

Well, so two things are happening. One, there are pre-existing volume thresholds that are already offered to customers; and two, as technology improves, those thresholds will probably come down. So I think what you need to think about is, one, we have a very diversified customer base. While Netflix is a very important customer, we have many other large and important customers, whether it be Apple or Amazon or other customers, CDN operators or direct content producers. And we will expect the average rate of price decline to be in that 22% range, and it will be episodic. I cannot predict for a specific customer when they're going to hit their next volume threshold.

Okay. Thanks for the color, Dave.

Okay.

And our next question or comment comes from the line of Scott Goldman with Jefferies. Your line is now open.

Hey. Good morning, Dave. I guess, a couple of questions. I want to go into the growth this quarter. And, obviously, a lot has been talked about the pricing and large customers. And we also can see salesforce productivity a little bit lighter than it was last quarter. So maybe if you could just walk us through some of the drivers for the deceleration we saw in constant currency growth this quarter. And as we think about 3Q, I think that's when we start to anniversary some of these USF excise taxes. And so, that's obviously been a little bit of a help on growth. Wondering how we think about what growth could look like in 3Q as you anniversary that, but should start to see maybe traffic growth re-accelerate and get some of these benefits from some of the one-time events that we've been talking about? And then, secondly, just on the margins, I guess, I like to think of things in a world where we don't worry so much about asset gains or litigation expense. And maybe at some point in time, we'll get to that. But just wondering if you could walk us through how you think about the underlying margins ex those two items, whether you feel that if we strip out that stuff on a normal – any given year, could we see 200 basis points of margin on that basis? Thanks.

Yeah. Sure. So in terms of the growth drivers, let's start with rep productivity. It was 5.9 units per rep per month. That actually is the best second quarter rep productivity number we have ever had. Two, we continue to see a rep productivity of 5.9 units be substantially above the long-term average of 4.9 units. So we feel very good about the number of reps and their productivity and their continuing ability to build a funnel. Now, the majority of our reps are corporate. Corporate sales performed well, with 3% sequential growth, 16.5% year-over-year. And you are absolutely correct that there was a material impact on the corporate part of our business in USF. And kind of ex that USF, which did not impact the sequential growth rate but did impact the year-over-year growth rate, the 16.5% is above trend line, and more like 13% or 14% is what we do kind of ex those effects. So for Cogent's total revenue growth, it is the combination of corporate and NetCentric. We're roughly 60% corporate, 40% NetCentric. And we expect both portions of our business to continue to perform well and around historical averages. So that means kind of 14% growth on the corporate side and 10% on the NetCentric side. So we feel very comfortable that over the next several years – not just next quarter but next several years with no USF impact -- that we will continue to see growth rates in our 10% to 20% range. Then, in terms of margin, remember, the USF had a detrimental impact on our margins, both on gross and on EBITDA, impacting us with approximately a 90 basis point headwind – or a 100 basis point headwind on the gross side and a smaller headwind on EBITDA, but still probably about 60 basis points. Ex those things, we grew on a sequential basis, including the net neutrality expenditures about 0.7% or 70 basis points and about 90 basis points year-over-year. I'm going to let Tad go into a little more color on how we think of it ex these items. But whether if they – with the gains, which clearly were better than 200 basis points, or ex the gains, but also ex the net neutrality fees and some of these extraordinary things like USF, we feel the operating leverage of 200 basis points, which we've got a 12-year track record of demonstrating, will continue. Tad, maybe, a little more color.

Yeah, if you look at EBITDA and then adjusted for the legal fees Q1 versus Q2, so EBITDA net of net neutrality fees, and the net neutrality fees were $1 million this quarter and $500,000 last quarter, the margin for this quarter would have been $32.7 million EBITDA margin and $31.5 million the first quarter, so that's a 120 basis point increase sequentially. And as Dave mentioned, what's the impact on USF, particularly year-over-year? So adjusting the EBITDA margin for that, it would have been $32.5 million versus $31.8 million, and the second quarter of last year $30.9 million. So that's a 160 basis points increase, net of USF, for EBITDA. So I hope that helps.

Yeah. That's very helpful. And then, Dave, just a follow-up. As we go back to talking the productivity you mentioned, I think you're happy with where the size of the reps are. It looks like it was relatively flat this quarter. Is that a result of less hiring and you feel comfortable with the levels now? Or how you think about what the size of the salesforce is through year-end?

So, again, we have guided to an annual growth in our salesforce of between 7% and 10%, the number of total reps and full-time equivalents. We expect that to continue going forward. And we will continue to hire through the year. We are comfortable on where we are now, but we also know that the number we're at today is not adequate for the end of the year, and we will have more net reps. And we're careful in balancing the rate at which we add those reps and the ability to maintain these very high productivity levels.

Great. Did you give a salesforce churn this quarter? I may have missed it.

We did not give that number. It was actually up this quarter, and it was about 5.9%. That is above where it had been running the last several quarters, at about 3.5% to 3.7%, but it is still below our long-term historical average rate of turnover.

Great. Thanks for taking all the questions, guys.

Hey, thanks, Scott.

Our next question or comment comes from the line of Walter Piecyk with BTIG. Your line is now open. Pardon me, Walter, your phone might be on mute.

Can you hear me now? Can you hear me now? Hello.

Yeah. Yeah. Just like a Verizon ad.

Sorry about that. My headset went dead, I guess. Dave, can you give a breakdown on the $43.4 million of NetCentric revenue as far as on-net versus off-net?

Yeah. I'll let Tad give you that. Yeah, grab that. He's just running over and getting that out of our binder. That's normally not (50:30)

(50:30) emailed that to you earlier. And, I guess, my question is going to be if, in fact, it's the off-net stuff – because I think that was relatively – you saw some declines there last quarter year-on-year – does all the stuff you talked about, as far as pricing and Netflix, does that relate to off-net, or is that volume that they're generating more hit the on-net than the off-net?

It's definitely on-net. So customers...

Yeah.

...like Netflix and all of our large NetCentric customers typically buy on-net services. We do have about 7% or 8% of our NetCentric business that is off-net. And what that is, is customers that are still buying from us in an on-net location, but we have to buy the cross-connect to serve them because they do not have a physical presence in that data center. And the data center operator will not sell them the cross-connect. So because we are buying an element of the network from a third party, we treat that as off-net, even though it's delivered in an on-net facility.

So you just said 7% of the $43.4 million was off-net?

Of the NetCentric is off-net, that is correct.

Okay. That's a big drop from the first quarter then, right? Because if I'm looking at the first quarter number, let me just take a look, I think it was more like 15% or 16% of the NetCentric revenue was off-net. Or maybe you're talking about traffic?

I'm actually talking about connections.

Connections, sorry. So can we get the revenue breakdown of the $43.4 million?

Actually, I don't think we have that broken down that way.

Okay. On the NetCentric business – I'm sorry, on the net neutrality expenses, last quarter, I think your comment was -- let me just look for the quote here -- the decline to be less than full year 2015. I mean, your wording was exactly the same, except for the fact you put this time around slightly less than full year 2015. So is there a reason that your net neutrality expenses, because I think you're pretty precise with your words, are trending higher than maybe what you previously thought? And maybe you're going to be closer to the 370s (53:01) than last year?

Yeah. And part of it is the inexperience with our German counsel and their billing. So the court case against Deutsche Telekom was moved from the Eastern District of Northern Virginia to Germany. We are continuing that prosecution and feel comfortable it was the strength of our case is actually what pressured Deutsche Telekom into adding some capacity, although not enough. So we are continuing our litigation. I feel comfortable that we will be lower, exactly how much lower I do not know because I just haven't received enough billing experience on an hourly rate from our German counsel. I've got pretty good feelings about what -Boies Schiller who has represented us here in the U.S. spell, but our German counsel just doesn't have enough track record. So I'm just trying to caution. I feel very comfortable we'll be below, but how big that decline will be, I just don't know at this point.

Okay. And on the call, I think you guys referenced that, excluding this, things would have declined by $700,000. But the seasonal expense in the December – excuse me, in the March quarter, the first quarter of the year, I thought, typically, enables a much larger – not much larger but a larger sequential drop than just $700,000, even if you excluded these incremental net neutrality expenses. So why didn't SG&A go down further in the second quarter?

I think it was primarily due to some of the payroll offsets not completely rolling off. So you, basically, get a reduction in your FICA and FUTA matches and some of the higher-paid employees end up rolling off in the first quarter. Some of the lower-paid employees bleed through to the second and even a little bit to third quarter. So we do expect that number to continue to come down.

Yeah, we did have a higher level of employees on a weighted average basis. So that also impacts, obviously, the SG&A expenses for the period.

So looking at third quarter you've got this balance of maybe you get some incremental on those employees rolling down or that expense rolling down, but we have no idea, basically, where the net neutrality expense is. So, from an SG&A standpoint, it sounds like we won't see this type of seasonal drop that we've seen in past years?

We believe that the $1 million that we spent in the second quarter was a high watermark, and we will definitely spend less than that in the third quarter and fourth quarter because if we didn't, we would not be under for the full year.

Okay. And can you give – sorry, you already did on CapEx. Sorry. Last question then. My favorite, the Netflix question. I'm going to ask it a different way, I guess, because I think you had a lot of commentary earlier on, like, pricing. And with your focus on the pricing aspect, I get that. Growth, just data growth from Netflix, did growth accelerate, decelerate, or remain flat in Q2 versus Q1?

Okay. So, again, I feel uncomfortable in picking them out as a specific customer. I think they should comment. They continue...

How about your largest customer then?

All of our large customers continue to grow. Our year-over-year growth rate slightly decelerated from 45% to 41%. We expect that year-over-year growth rate for the remainder of the year to improve and get back up towards our long-term trend line of about 52% year-over-year. That growth continues to come from a broad base of customers, dominated by OTT providers. And the bigger the OTT provider, the bigger their growth rate is. And Netflix is the largest provider. I think I'm indirectly answering your question without probably stepping out of line.

Yeah. I think that's fine, Dave. Thank you very much.

And our next question or comment comes from the line of Michael Bowen with Pacific Crest. Your line is now open.

Okay. Thanks for taking the question. Dave, so – and I think you may have touched on this, but I wanted to just make sure I understand this. If you look at your SG&A expenses, excluding stock-based comp, the first quarter and second quarter this year did step up. I mean, should we expect a step-down in the second half? And, again, I apologize if you've already touched on that. And then, second question. Can you give us a bit of an update on the re-org on the NetCentric sales side? Because I know that your unit productivity may have come down, but we're focusing more on revenue per rep. And given the fact that the reps were flat, I would assume that the way the math works, obviously, is your revenue per rep went up sequentially. But can you help us break that down on whether the improvement was on the NetCentric side or the corporate side? Thanks.

Sure. So let me take the first one. You should expect moderate declines in SG&A in the second half of the year, although we will have additional head count partially offsetting the underlying declines. In terms of the NetCentric salesforce, the account realignment between our premier and general NetCentric sales teams is complete. We are adding reps to both portions of our business, and you are correct that the revenue productivity per employee improved in the quarter, and we would expect to see both more reps and revenue productivity improving throughout the remainder of the year. We feel very comfortable that the NetCentric programs that we are putting in place are paying results. We saw a material spike up, for example, in the number of ASs connected incrementally in Q2 versus Q4 to Q1, and that is a direct effort of the NetCentric salesforce. Now, they are going after smaller accounts. Those smaller accounts generally pay a higher price per megabit. But, again, that's being offset by some of the very large accounts really seeing material improvement in traffic growth and, therefore, getting lower prices per megabit, which also tends to drag down rep revenue productivity on a dollar basis without the rep doing anything. There are just customers who're sending us a lot more traffic, and as a result of that, they've worked their way down to a lower price tier.

I mean, is it still fair to say though that most of the revenue per rep productivity gains are coming from the corporate side?

So the salesforce is dominated by corporate reps. It's roughly 73% of the salesforce. And corporate growth was pretty much on track with historical averages. And I do think the majority of the productivity comes from them because they're the majority of the reps. But we are seeing improvements on the NetCentric side as well.

Okay. Great. Thanks for taking the question.

And our next question or comment comes from the line of James Breen with William Blair. Your line is now open.

Thanks for taking the question. A lot have been answered, but can you just talk about the CapEx trends here? And, obviously, last year, I think CapEx was higher in the first half of the year and then scaled down a lot in the back half. Do you expect to sort of see those same trends this year as well?

Yeah. Hey, Jim. So CapEx – again, I want to tell people to work both the CapEx and the principal payments of capital leases. It was effectively flat year-over-year, even though revenue was up 11.3%. So as a percentage, it went down. But in absolute terms, it was flat. You are absolutely correct that our capital spending in both categories tends to be lower in the second half of the year. That will absolutely be true this year. And as I said, we will be below 2015 full-year numbers for both CapEx and principal payments in 2016 versus 2015.

Great. Thank you.

Thanks.

Our next question or comment comes from the line of Frank Louthan with Raymond James. Your line is now open.

Great. Thank you. Can you comment a little bit on the corporate part of the business? Are you seeing any changes in trends from customers, either from what they are purchasing or their expansion plans? And can you comment on sort of your long-term target for the payout ratio for the dividends? It's a pretty high-level. I understand you've got growth and so forth, but how comfortable are you with where that's trending for 2017?

Okay. So let me take the first one. We are absolutely seeing a change in our corporate customers' behavior in two respects. One, we continue to see an increase in the average peak utilization of our corporate customers, utilizing about 17 megs of the 100 megs that they have purchased. That's up tenfold in 10 years, and that's driven by remote computing and cloud applications. We think that's going to continue, going forward. Secondly, starting a couple of years ago, we began to emphasize our VPN services, basically point-to-point connectivity services. That has grown to be approximately 14% of our revenue. Those are the services that are subjected to the somewhat confiscatory USF tax rate of nearly 18%, resulting in the roughly $2.2 million of USF charges that we charge our customers and pass on to USAC to be distributed. We do expect to see those VPN services continue to grow as well. We, different than other providers, price our services identically to our dedicated Internet access services. It is a large addressable market. It's actually a larger addressable market in dollar terms, according to the FCC, than the DIA market, and we continue to gain market share. So many of our corporate customers, actually, buy two connections in the same location, one for DIA and one for VPN services. We expect those trends to continue, and that will drive our ability to grow corporate growth rate in that 13%, 14% year-over-year rate. With regard to our dividend, we have had a roughly balanced approach to returning capital, both through buybacks and dividends. We have now, including this dividend, returned about $465 million, about $245 million through dividends, about $220 million through buybacks. We actually saw our net leverage fall in the quarter. We have decided to take on leverage to accelerate our return of capital. We were able to move $155 million because of our ability to go below the 4.25 times leverage at the operating company level. So we actually have more cash to return to shareholders than just cash flow from operations. So we actually expect, for the next couple of years, that our return of capital to shareholders will be greater than 100% of cash flow generated in the business. Now, that return will be through a combination of dividends and buybacks. And we have not said what the exact mix will be going forward. We'll try to be opportunistic. But we are absolutely committed to returning more cash for the next couple of years than we produce, as long as we remain in a low interest rate environment.

Okay. Thank you. One quick follow-up on the salesforce. Where are you as far as your long-term mix of the NetCentric sales folks versus the corporate side? Do you still need to ramp up? And can we assume that the hiring going forward is going to still overindex to the NetCentric side?

So more of the turnover is on the corporate side than the NetCentric. We do need to net have more NetCentric than corporate, in that our historical mix has been roughly 66/34, 66 corporate, 34 NetCentric. And today, we're at about 73% corporate. So we will hire net more NetCentric reps. On a gross basis, we generally hire more corporate because they tend to be younger, lower paid, and much higher turnover rates.

Okay. Thank you.

And our next question or comment comes from the line of Colby Synesael with Cowen and Company. Your line is now open.

Sorry, I just had one quick follow-up. Is it my understanding that to reignite the buyback, you actually have to go back to the board or the board has to kind of sign off on that? Or are you free to start buying back the stock as we sit here today? And with that said and you being below your leverage target, is it fair to assume that we don't know if it's buyback or dividend, but is it fair to assume that you'll start that extra component, if you will, to the recurring dividend up again at some point in the third quarter? Not necessarily trying to quantify size here, but that component, now that you can kind of get back to it, will kind of start immediately?

So with regard to the buyback, we have a current authorization in place, and we do not need any additional authorization from the board. Secondly, we will use the buybacks opportunistically. If we decide to return excess capital through dividends, we will, in fact, go back to the board. And those dividends are typically distributed on a quarterly basis. So I think it would be unreasonable to expect an intra-quarter dividend distribution. That would be an unreasonable expectation as we've never done that. And we are committed to a gradual return of capital over a multiyear period. We are committed to, in this interest rate environment, a net leverage range of 2.5 times to 3.5 times, and we're trending towards the lower end of that range, delevering in the quarter. And we will make sure that we return capital at an adequate pace.

Would you let the leverage dip below 2.5 times, even just for a temporary period of time? Or are you guys pretty committed to that being the floor?

We've set a range, and we've always tried to stay in the range. So I can't predict the future, and I can't predict what will happen to interest rates. But I think it's fair to assume we – in this environment, we would not go below that level.

Great. Thank you, Dave.

Hey, thanks, Colby.

And our next question or comment comes from the line of Michael Rollins with Citi. Your line is now open.

Thanks, and good morning. I just wanted to follow up a little bit on the NetCentric side. Hey, Dave, can you help us just get a little more color on what the constant currency sequential trend was in the NetCentric business? And given the seasonality that you see in volumes, should investors expect that there is going to be more seasonality, maybe, in revenue to come, where revenue of NetCentric just in any second quarter now might be down sequentially just because of the way volumes move through your business over time? Thanks.

Sure. So, first of all, our NetCentric revenue on a reported basis was sequentially down about 0.6% or roughly $250,000. Virtually, all of the FX benefit that we got hit the NetCentric portion of the business. So it's fair to assume that the rate of decline on a constant currency basis would be greater. It would be the 0.6% plus the FX impact. In terms of the NetCentric business over the past 12.5 years that we have reported data, always reporting NetCentric and corporate as we do, we've had a number of quarters – probably about 40% of the quarters, have actually had negative sequential NetCentric growth. There is a great deal of volatility in that growth. But the underlying trend line has been a 10% year-over-year growth rate. This most recent quarter, we actually only grew an underlying reported number of 4.1%. And since there was virtually no year-over-year FX impact, it was pretty much 4% growth rate, which was below trend line. We do expect the proliferation of OTT video and a number of different providers to continue to grow. We do always offer volume discounts. We expect the rate of price decline to be about 22% a year, coupled to a low 50s unit volume growth rate, and, as a result of that, our ability to grow our business over the long term on our NetCentric business at about 10%. But it's very difficult to predict a specific quarter. And it's probably even more difficult to predict exactly when a certain customer will experience a surge in traffic and therefore the ability to get to lower pricing.

Thanks very much.

Hey. Thanks, Mike.

And our next question or comment comes from the line of Walter Piecyk with BTIG. Your line is now open.

Dave, just one follow-up. Can you give us any sense now where CapEx will look next year? I mean obviously this year you're saying it's going to be down a little bit, but what about in 2017?

In 2017, CapEx will be, again, below 2016 number. The magnitude of that decline, we have not yet disclosed. And that will be the combination of CapEx plus principal payments on capital leases.

Yeah.

Or if you look at just CapEx by itself, it will also be down.

Okay. Thanks.

Thanks.

And our next question or comment comes from the line of Jonathan Atkin with RBC Capital Markets. Your line is now open.

Yeah. Sorry if I missed this earlier, but the off-net portion of NetCentric, can you tell us what the margin profile is of that?

Yeah. So on the off-net NetCentric, we are purchasing a cross-connect and a data center or a local loop from a landing station for a company that either does not have a presence in that data center or does not have the license to buy in that particular country. It is typically better than the 50% margin that we have on our typical corporate off-net, where we go out, price a loop, and then double it. Typically, those, either cross-connects or local loops, they can be all over the board. But they're probably resulting in kind of an 80% gross margin, about 20% cost of goods sold.

Thank you.

And at this time, I'm showing no further questions or comments. So with that said, I'd like to turn the conference back over to the Chairman and CEO, Mr. Dave Schaeffer, for closing remarks.

Well, thank you very much. Thanks, everyone, for taking time. We did run a little longer, but we wanted to answer all the questions. Hopefully, we were complete in those answers. And, again, we are committed to growing our revenue, expanding our margins and, most importantly, accelerating and continuing to return capital to shareholders. So 10% to 20% top line growth, 200 basis points year-over-year margin expansion, declining capital intensity, these are long-term goals. Leverage targets of 2.5 to 3.5 times on a net basis and, most importantly, getting that capital back to shareholders in the most tax-efficient way possible. Take care, everyone. Thanks a lot, and we'll talk soon.

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a wonderful day.