Good day, and welcome to the Cogent Communications Group, Inc. Second Quarter 2012 Earnings Conference Call and Webcast. Today's conference is being recorded and available for replay at www.cogentco.com. At this time, I would like to turn the conference over to your host, Chief Executive Officer, Dave Schaeffer. You may begin.

Great. Thanks, and good morning to everyone. Welcome to our Second Quarter 2012 Earnings Conference Call. I'm Dave Schaeffer, Cogent's CEO. And with me on this morning's call is Tad Weed, our Chief Financial Officer. We are pleased with our results for this quarter and are optimistic about the strength of our business and the outlook for the remainder of 2012. During the quarter, we experienced and returned to sequential revenue growth, EBITDA margin expansion and traffic growth. Additionally, we generated $5 million of net cash flow in the quarter, a quarter which included $1.3 million being spent for the purchase of shares of our common stock and the negative impact of $1.1 million on cash flow from foreign exchange. Our revenue growth from the first quarter was 1.2% and grew by 3% year-over-year from the second quarter of 2012. On a constant currency basis, our revenues grew from the first quarter of 2012 by 1.7% and on a year-over-year basis grew by 5.7%. On a constant currency basis and adjusting from the impact of the loss of our largest customer, Megaupload, in the first quarter in January, our revenues grew sequentially by 2.9%. An additional third item impacted our revenue growth, and it included the impact of recognizing our deferred revenue from our customers over a longer average contract period. Including this impact, our revenues grew sequentially 3%. On a constant currency basis, adjusting for the loss of Megaupload and the change in customer non-recurring install amortization our revenues grew, on a year-over-year basis, 12.1% from the second quarter of 2011. As we projected on our Q1 2012 earnings call, we would expect and did experience an extraordinary increase in our sequential EBITDA margin expansion. In this quarter, our EBITDA margin expanded by 330 basis points sequentially from the…

Thank you, Dave, and good morning, everyone. This second quarter 2012 earnings report and this earnings conference call discuss Cogent's business outlook and contain forward-looking statements within the meaning of Section 27A and 21E of the Securities Act. 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. You should also be aware that Cogent's expectations do not reflect the potential impact of mergers, acquisitions, other business combinations or financing transactions that may be completed after today. Cogent undertakes no obligation to release publicly any revision to any forward-looking statement made today or otherwise update or supplement statements made on this call. Also, during this call, if we use any non-GAAP financial measures, you will find these reconciled to the GAAP measurement in our earnings release and in our website at cogentco.com. Now I'd like to turn the call back over to Dave.

Yes. Hey, thanks, Tad. Now for some highlights from our second quarter results. Hopefully you've had a chance to review our earnings press release. As with previous quarters, our press release includes a number of historical metrics. These metrics will be added to our website. Hopefully you find the consistent presentation of these metrics informative and helpful in tracking and understanding our financial results and the trending from our operations. Our second quarter 2012 revenue was $77.8 million, an increase of 1.2% from our first quarter revenues and an increase of 3% from our Q2 2011 revenues. On a constant currency basis, our revenues increased sequentially 1.7% from the first quarter and 5.7% from the second quarter of 2011. On a constant currency basis and adjusting for the loss of Megaupload on January 19 and the increase in our average contract length, our revenue sequentially increased 3% from the first quarter and 12.1% from the second quarter of 2011. We evaluate our revenues based on product class, on-net, off-net and non-core, which Tad will cover in greater detail. We also evaluate our revenues by the type of customer who purchases our services. Our customer types include 2 groups: NetCentric customers and corporate customers. All Cogent customers are categorized 2 ways, either as corporate or NetCentric and is either on-net, off-net or non-core. NetCentric customers buy large amounts of bandwidth from us, generally in carrier-neutral data centers. Corporate customers generally buy bandwidth from us in large multi-tenant office buildings. Revenues from our corporate customers grew approximately 2.5% from the first quarter. Our revenue from Megaupload was 100% on-net, 100% NetCentric, so the loss in January, 2012, of that customer had no impact on our corporate customer growth. Our corporate customers represent 53% of our total number of customer connections at the…

Thank you, Dave, and again, good morning to everyone. I'd also like to thank and congratulate the entire Cogent team on their hard work and efforts during the quarter. I'll begin by providing additional details on our revenue by product class. Our on-net revenue was $57.3 million for the second quarter, which was an increase of 1% from the first quarter. We continue to add on-net customers to our network, and the impact of the loss of Megaupload negatively impacted our on-net sequential revenue results. If you exclude the impact of the loss of that customer, our on-net revenue increase was 2.6% from the first quarter. Similar to prior quarters, approximately 85% of our new sales for the second quarter were for our on-net services. Our on-net customer connections increased by 4.7% for the quarter, and we ended the quarter with about 27,500 on-net customer connections on our network in our 1,800 on-net buildings. Our revenue from our off-net business was $19.9 million for the second quarter, which was an increase of 1.9% from the first quarter as we continue to add off-net customers to our network. Our off-net customer connections increased by 3.5% for the quarter, and we ended the quarter serving about 4,100 off-net customer connections in approximately 4,000 off-net buildings. Our non-core revenues were approximately $0.6 million for the first and second quarter and now represent less than 1% of our revenues and about 500 customer connections. Related to churn. Churn rates for both on-net and off-net improved during the quarter. Our on-net monthly gross churn rate was 2.6% for the quarter as compared to 2.7% for the first quarter. Our gross churn rates, as a reminder, include customers who remain with us but enter into amended or so-called "move, add, change" Cogent contracts. If you exclude…

Hey, thanks, Tad. We're going to take a moment and chat a little bit about sales force activity and productivity. We began the second quarter of 2012 with 273 reps and ended the quarter with 269 reps selling our services. We hired 51 sales reps in the quarter, and 55 sales reps left the company during the quarter. Our monthly sales rep turnover was 7% for the quarter. This was the same rate as we experienced in the first quarter and is consistent with our long-term historical trends. We began the second quarter of 2012 with 269 full-time equivalent reps. We ended the quarter we 256 full-time equivalent sales reps. Productivity on an FTE basis for the first quarter was substantially lower than the 5.1 units that we were able to deliver in the second quarter. As a reminder, our sales force productivity rates are based not upon contract signing, but are based upon actual installations of customers onto our network and the beginning of the recognition of revenue from those customers. Now to the scale and scope of our network. The size and scope of our network continues to grow. We added 30 buildings to the network in the second quarter, and now we have over 1,800 buildings directly connected to our network. Our network consists of over 26,100 metro fiber miles and over 56,500 intercity route miles of fiber. The Cogent network is the most interconnected network in the world. Today, we connect to over 4,050 other networks. Approximately 40 of these networks are settlement-free peers. The remaining networks connected to Cogent are paying customers. We continually monitor the utilization on our network, and we upgrade sections of the network as needed. In the most recent quarter, we were utilizing approximately 16% of our width capacity. We have…

[Operator Instructions] We'll take our first question from Colby Synesael with Cowen and Company.

I have 2 questions, or 2 lines of questions. The first one has to do with the off-net revenue. I think you had mentioned in the past that you've been seeing some installation delays, particularly at Verizon, and the off-net was a little bit later than we were expecting this quarter. I was wondering if you could give us an update on that delay. And the other part of that question is you actually added 138 off-net customers, which is actually a very large number for you guys. Just curious what's driving that and if there's any concern that with such a big backlog now and potentially no change in the installation delays at Verizon, is that going to potentially lead to some issues down the road? And then I have one other question after that.

Sure. Let's take the off-net installation delays. We have seen those delays grow to over 105 days. They've shrunk back to about 94 days. It's still about 3 or 4 days above the long-term historical average, but we have seen an improvement from what we saw in the first quarter. We -- many of these installs came later in the quarter, so therefore, that was the reason why we did not recognize revenue for the entire quarter for many of those installs. Our sales force productivity was extremely high. We continue to sell a large amount of off-net, as well as on-net. As Tad mentioned, 85% of our sales are on-net, but 15% consistently have been in our off-net product. The final point that probably has impacted the growth rate in our off-net revenues is that we had and still have a base of TDM customers. Increasingly, almost all of our off-net sales are Ethernet. We appear to be through the rotation of our existing customers migrating from TDM to Ethernet, those that want to. Almost all of the new sales are Ethernet. And for that reason you saw a plateauing in our average revenue per user in those off-net circuits. So we do expect to see continued improvement in unit growth on the off-net business in Q3 and forward, particularly as these installation windows have normalized. And we also expect to see revenue growth. We may not, however, see continued increase in average revenue per user in the off-net business.

Okay. And then my other question, just in terms of the dividend. I was wondering if you could just talk about what the correlation going forward is going to be between free cash flow growth and potential dividend growth? And maybe add in to that conversation or explanation, what the company's view is on the amount of cash that you actually want to hold in the balance sheet -- on the balance sheet, and -- because I think you're actually at a much higher number than you've stated that you need. And how that could potentially impact the dividend going forward as well?

Yes, sure, Colby. So we are committed to a regular return of capital. In that return of capital, we have 2 tools in our toolbox, one being the dividend program, the second being the buyback. And the buyback allows us to opportunistically take advantage of market volatility. The $1.3 million that we spent in the second quarter allowed us to buy back stock at, I think, either at or near the lowest prices in the quarter on days of extreme volatility. We are committed to continuing to use buybacks to supplement the dividend program. As Tad mentioned, we had $237 million, or slightly over $237 million of cash on our balance sheet, which is far in excess of what we need to run the business. We are growing our sales force, our footprint, and we have continued to look at M&A opportunities but not seen anything that represented value that would be accretive to our equity. So we have not pursued any M&A. With that, we continue to build cash. As demonstrated, we have generated, even with the buyback, an extra $5 million this most recent quarter, and we expect that trend to continue going forward. What that means is we either need to be buying more stock back or increase the dividend. We think that we need generally about 6 weeks of revenue as a liquidity cushion to adequately run the company. That would be a number much lower than where we are today, probably in the $40 million -- $50 million range would be plenty of liquidity. We are committed to returning that capital, so I think the board will evaluate the volatility and stock and the ability to use the remaining buyback. And if we're unable to use that more aggressively, probably continue to see us increase the dividend as a way of returning the cash that we're generating from operations, as well as eating into the cash balance on the balance sheet.

We'll go next to James Breen with William Blair.

Just a couple of questions. One, with respect to the margins, can you just go through again -- there are a lot of puts and takes relative to the sequential growth in EBITDA margins. Can you talk about -- obviously, you saw a big jump this quarter. Do you expect those to come down? Are you just saying that you expect them to stay sort of flattish with current levels through the rest of the year? That's the first question.

Okay. Our margin this quarter was heavily impacted by the fact that in the previous quarter, we had the bad debt expense associated with the loss of Megaupload. Our margins globally are fairly consistent. So there has not been any real distortion in our margins due to FX. Our SG&A is pretty much at a good run rate now. As Tad mentioned, we have these very high operating leverage as we add incremental revenues, 100% gross, roughly $0.95 incremental EBITDA on on-net and about $0.45 of incremental EBITDA in that off-net business. We do expect that our margins will continue to grow next quarter at about 50 basis points. And as I said earlier, we view the expansion from full year '12 to full year '13, 200 basis points as an absolute floor. We actually think we can do a little better than that. In terms of puts and takes, probably the greatest volatility is usually in the first quarter where we have these extraordinary expenses. Other than that, things are actually pretty consistent in terms of margin expansion.

We'll take our next question from Jennifer Fritzsche with Wells Fargo Securities.

Just a question on new customer adds. With the Level 3 Global Crossing transition, has there been any -- or integration, has there been any opportunity to pick off new customers, and if you can offer some color there? Also Netflix seems to be a new customer for you, and I was wondering if you could offer any further color there as well.

Yes, sure. So first of all, in terms of customer additions, as you can see from our numbers, the net number of new customer connections increased by only about 800 in the first quarter over the fourth quarter of last year. And this year -- or in the second quarter over the first quarter, was up almost 1,400. So a material acceleration in the number of new customers that we have connected. Many of those customers came as a result of these promotions. As I have mentioned previously, we had targeted about 1,500 customers that bought from other leading companies that were going through merger and acquisition with some extremely aggressive pricing. That pricing is $0.75 a megabit. That pulled down our average new sale price in the quarter from $2.36 to $1.82. That did result in slightly lower ARPUs but lower prices per meg. Now the installed base did not fall as rapidly. It only fell from $3.81 to $3.44 in the quarter. We have had success, meaning we have connected about half of those targeted customers. That means there is still half to get. Most of those customers have not yet experienced any network disruption as a result of those networks trying to be integrated. Also it's important to note that many of those customers gave us only a very small portion of their business in the form of a limited number of ports. Some of them were under what we call "try and buy" programs where we actually give some free connectivity for a limited period of time, 30 to 60 days, to give customers the ability to test our network and get very comfortable with the quality that we deliver. That, I think, has helped us increase the number of customers and obviously showed up in…

We'll go next to Tom Seitz with Jefferies.

I guess it's a follow-up on what you just discussed. No secret, you're always cutting, shaving price. But given what you're doing with the Level 3 and Global promotion in securing contracts like the expansion of the Netflix business, are we in another period like a few years ago when you started the bigger term discounts? It's going to take a couple of quarters for volumes to catch up, and as a consequence, on a constant currency basis, revenue growth is likely to be at the lower end of the 10% to 20% guidance range for a couple of quarters, and then in 6 months or so, we'll see revenue acceleration again? Or was coming in at the lower end of your guidance range, or your typical sequentially quarterly growth, strictly a function of customers coming in late in the quarter and these promotions that you were just referencing and we'll see a pickup in third quarter?

Sure, Tom. So first of all, adjusting for foreign exchange, our NetCentric revenues grew 3.1% in the quarter as opposed to being flat. So the vast majority of our FX hits the NetCentric business. Two, we did install a large number of orders later in the quarter, which will have some impact. Again, we don't give specific quarterly guidance. There can always be some volatility. Each of these customers has their own business model, and many of them have commitments to other providers that need to unwind as they get comfortable in the savings that Cogent offers its customers. I don't necessarily think that there's going to be a material acceleration in the third quarter from the second quarter, particularly as Tad kind of outlined the fact that we've got, based on what we're at now, 800,000 of FX headwind that we kind of know about going into the quarter. Traffic can be a bit volatile, there's generally more traffic growth in the fall and winter months than in the summer months. But as you saw in this quarter, our promotional activities are working. I feel very comfortable with our long-term guidance, with the ability of NetCentric and corporate revenues to each grow long term in that kind of 3% to 4% sequential range, which gets us right at the midpoint of our long-term guidance, which is what we've historically done. I can't really comment exactly on next quarter or the quarter after, other than the trends are encouraging. Probably the one thing we know about today that's discouraging is the FX impact. And the rate at which video migrates to the Internet does have some impact in our NetCentric business. Actually, 2 ways: one, we directly benefit from the companies that are publishing the content, like the Netflix. But remember, we have over 75 PTTs around the world. Virtually every major cable operator in the Western world and hundreds of CLECs has access networks that draw down content. So we actually benefit 2 ways from this migration: one, from the publishers, like Netflix; but also from the networks that are drawing their traffic, who need to buy upstream connectivity. I just -- I'm not a good enough forecaster to tell you whether or not the next hot show is going to drive a lot of traffic next quarter, but I feel very comfortable in our long-term growth trajectory for the NetCentric business.

Yes. I mean, I don't think there's any question that given the customer growth, the revenue growth is going to be there. I guess I'm just looking for a little help so that you're not a victim of your -- of expectations. And whether or not this promotional activity could end -- could result in you guys being at the sort of lower end of the historical growth rate for a quarter or 2 before this influx of customers results in revenue re-acceleration?

It could, Tom. But in the past, when we ran different types of targeted promotions, some quarters, you're absolutely right. We saw a quarter or two lag. Other quarters, it hit us immediately. We're going to be as aggressive as possible. The best help we can get is for those companies that merge to try to achieve synergies and combine their networks. Because when that network combination occurs, that's when customers are going to feel the pain. It's important to remember that Level 3 and Global Crossing today still run 2 completely separate networks with completely separate autonomous systems, and they're not even interconnected today. So there's a lot of upside here. I just can't tell you that -- I just can't forecast. This could be a little slower this quarter or next. I just don't have that kind of visibility beyond the FX that I talked about.

We'll go next to Alex Sklar with Raymond James.

I guess, can you give us an idea what the year-over-year and sequential traffic growth was, excluding the impact of Mega? And then, I guess, a little bit more color on the sales rep productivity this quarter. What do you think is ultimately driving the increase? Do you think it's a result of less new reps coming on in the quarter or just the success with the -- from the promotional pricing? And then any comments if it's tracking at a similar level so far in third quarter?

Yes, sure. So our historical growth rate for the past couple of years, year-over-year, had been around 44%, 45%. Now that included Mega, that was growing at about the same pace as the remainder of our customer base. This quarter was obviously, on a year-over-year basis, heavily impacted by the loss of that customer. Probably if you removed Mega from both numbers, we actually saw an acceleration in our year-over-year traffic growth, in part due to this promotional activity and picking up these other accounts. So probably in the 55% range. If you want to look on a quarter-over-quarter basis, we were up 10%, but the first quarter had 19 days of Mega traffic included in that. If we took that out, we were probably up closer to 18% or 19%. The traffic measurement that we use is very straightforward. We literally measure every port on our network every 5 seconds. We take the number of petabytes, meaning actual traffic transferred, not the speed or the throughput, but the actual bits transferred in and out over the course of the entire quarter, and we report that number. So it is an absolute objective number measuring every single port. Let me switch to the second question, which is rep productivity. We still have a goal of growing the sales force, we're actively recruiting, we're actively hiring people. Our rep turnover has been pretty consistent. I do think the promotions that we are running contribute to our unit productivity. At 5.1, that is substantially above the 4.1 that we have averaged historically over the past 7.5 years as a public company. We are continuing to see improvement. It is not as a result of having less reps. In fact, we still have huge amounts of our addressable market that are not yet being adequately penetrated by our sales organization. So we still desperately need more reps. In terms of the third quarter, we are encouraged by what we are seeing in the quarter so far in terms of rep productivity, and it should be above our long-term average.

We will go next to Michael Funk with Bank of America Merrill Lynch.

I guess, first, [indiscernible] of your view on the opportunity for acquiring assets in the current market. Obviously, valuations have been bid up. Just wondering what you're seeing out there today? And then maybe, just a little bit more color. You've provided in the past about your view on use of capital. I know you provided some comments on the call about the share repurchase and obviously the dividend. But maybe just some thoughts on whether it's better to be a buyer or seller in the current environment given those valuations that I mentioned?

Yes, sure. So Cogent had been an extremely active acquirer in the 2001 to 2005 time frame. We have not participated in any material way in M&A since then. We have added a few data centers to our footprint. We have opportunistically bought more dark fiber. We have a very simple filter that we use in looking at acquisitions, and it's a two-pronged approach. The first approach is, by acquiring the whole company, are we acquiring assets that we need to run our business less expensively through the acquisition than we would through either building out those assets ourselves or just buying the assets and not the whole company? We have never been a significant builder of network and have no intentions to do that, other than connecting buildings to our network. And then in terms of acquiring whole companies, it's been far more prudent for us in the past 7 years to just acquire the assets. The second filter that we apply is if we buy the ongoing business, is it, whether we pay cash or issue stock, accretive on a free cash flow per share basis? That's a very objective measure. We have not been able to find any businesses that trade at a -- in our sector, that trade at a lower free cash flow multiple than our business based on looking at the growth in our free cash flow. So we will continue to look at opportunities, but I will assure shareholders that I take the job of capital allocation seriously. Tad does, the entire Board does, and we think about it. In terms of the second way to approach that question, should we be a seller? I tell shareholders all the time, the company is for sale everyday. It's bought and sold in the public markets, and the markets determine our price. Ultimately, the value though of this business is its future stream of free cash flow and what the appropriate discount rate you apply to that is, it's somewhat subjective. If someone paid a premium, we would always listen. But at the end of the day, we are very comfortable in where we sit in the ecosystem, in the value chain, in the competitive landscape, and intend to execute our business on a standalone basis. In terms of use of capital, I couldn't be any clearer. We have too much capital on the balance sheet. We are committed to returning it regularly in a dividend. We will evaluate how we need to grow that dividend to get that cash back. But buybacks are tax efficient, they compound and they allow us an additional degree of freedom where we can be opportunistic. And you saw us in the last quarter take advantage of that. I think the Board, at least today, wants to preserve both tools as ways to return capital.

We'll go next to Michael Bowen with Pacific Crest.

A couple of things, and I'm sorry if I missed this. But you've mentioned in the past that corporate pricing has been very solid. Can you tell us what you're seeing now? And then secondly, on the EBITDA improvement that you're looking for, I think you said over 200 basis points 2013 over 2012, can you give us a little idea on how much of that do you think might be coming from the cost side versus the pricing and revenue side?

Hey, sure, Michael. I'm going to take the first part, and Tad will take the EBITDA piece of it. On the corporate pricing, let me break that into 2 pieces. Roughly 60% of our corporate business, just corporate, is on-net. 40% is off-net. For our corporate on-net product, 100 megabit symmetric, non-oversubscribed Internet access service for $700 a month has been our price for the past dozen years, and we have not modified that at all. Now we do occasionally run promotions that could impact it. Buy a 2-year contract and get the 3-year pricing or buy 12 months, get the 13th month free. Part of sales is always having some excitement and call to action. But there's been no significant degradation in the competitive landscape or the pricing dynamic. And we expect that our corporate on-net pricing still represents the most compelling value, whether compared to cable, to telephone or CLEC offerings. And we hear consistently from customers when we do surveys that price is almost too good to be true. So on the corporate on-net, no change whatsoever, and we see a very stable environment. Off-net, we also see stable environment. As I commented earlier, the ARPU at $1,643 is plateauing as we've migrated to Ethernet, but we see a very stable environment there. And I think increasingly, customers realize that T1s or TDM-based services are no longer reliable or large enough for most corporate users. And wherever possible, they are trying to migrate to an Ethernet-based product. Tad will take your margin questions.

Yes, Dave, real quick before Tad gets to that. On the promotions. Promotions tend to have almost a negative connotation across the communications sector. Have you looked at historically, when you run the promotions, is there any way to look at kind of like a payback or percentage benefit on those promotions? Is that a good way to look at it?

Yes, we actually do look at it that way. So we say, if we give up something, either waive an install or we give away a free month or we give a longer-term discount for a shorter-term contract, those are kind of the most common flavors of these corporate promotions. How many incremental orders would we get, what is the margin of those orders and what is the payback? Well, we generally see paybacks very quickly. Generally, 3 to 6 months when we run those promotions. But what you've got to be careful of is that you don't want promotions to become the new standard price. You still need to create in the customer's mind, some variability in the offering, some excitement and some call to action. So we constantly probe the market. One of the interesting promotions that we ran -- actually, some of it was run last quarter -- was if a customer showed us an invoice from a pre-existing provider, we would actually agree to buy them out of their contract. Meaning, we would give them free service on Cogent for the remainder of their contract. So they'd only pay one bill, as long as they signed a multi-year contract with us. What that did is, one, it spurred activity; but two, the second thing it did that was very interesting is it gave us, on the corporate side, a much more granular and precise view at the competitive landscape. We get that all the time on the NetCentric side, but on the corporate side, it really opened our eyes to just how much better Cogent's value prop is than virtually anyone else in the market. Let me let Tad take that EBITDA question.

Sure. So on our plan, if we look at full year 2013 over 2012, again, we are very comfortable with the revenue guidance of 10% to 20%, and we are very comfortable with the EBITDA expansion of at least 200 basis points. So that's kind of from a high-level view. Granular on that, really, don't expect any material changes in trends on the revenue front. And this forecast, our updated forecast, includes the most current rates of foreign exchange that are included within that. If you look at our cost side, really, the variability that we have on the cost side is when we add each off-net connection, about 50% of the revenue increase adds to cost of the tail circuit. We do have some costs associated with expansion related to maintenance on the fiber routes and lease costs for facilities that are added, expect that pace to be slightly lower than it has been historically. As we mentioned, CapEx we expect to be about 10% less, though the rate of expansion in terms of those cost would be similar. And then lastly, we have a CPI on many of our contracts, and those tend to kind of rollover on the first of the year. So that's kind of the cost of goods sold line. If you look at the SG&A component of our costs, besides the seasonality that we went over on the call in the first quarter, on an annual basis, the changes we see are about -- historically, about 1% of our revenues wind up as bad debt expense. So 99% of the amounts we bill are paid in cash, and the remainder we need to write off as uncollectible. And then the increase in our sales force is baked into that as we expect sales force to continue to grow. And then you do have a CPI again on SG&A expenses. So that kind of, if you look at our run rate and if you're modeling and how we look at changes in costs, both cost and SG&A when we are forecasting forward.

We will go next to Donna Jaegers with D.A. Davidson.

Nick here for Donna. Just a quick question on your comment about taking advantage of volatility in Europe. Can you give -- provide a little bit more color on that? What kind of opportunities are you talking about? And are you looking to increase your exposure in Europe then?

Yes. Hey, sure, Nick. Our ability to take advantage of volatility in Europe hits us 2 ways. First of all, it's purely to allow us to purchase fiber assets to serve additional European markets. We by far and away today, already operate the densest Pan-European network. If you go look at our map against any of the other global network operators, we have far more pops in more locations in more cities in Europe. We operate in I think 25 of the 27 EU countries and 3 or 4 non-EU countries in Europe. Our ability to take advantage of the volatility is twofold. One, as the dollar gets stronger relative to the euro, things just get cheaper in dollars. So that's a good thing for us as a U.S.-based company. And then secondly, some of the owners of fiber who may have been more reluctant to sell in better times become more interested in cash. Probably the most extreme example of that we witnessed was our entry into Greece probably about a year ago. There were no other global network operators that had a dark fiber network in Greece. It was all owned by Greek national companies. We were able to go into Greece and establish the first facilities-based network in Greece that is not run by a Greek -- or a company controlled by a Greek national company. So it's that ability to take advantage of volatility. On the customer side, we're trying to sell all customers. You can argue that as people become more concerned about their economic viability, they may be more interested in buying from Cogent just because we're a lower-cost solution. So I think as many of our customers are experiencing pressures in their business, they are more open to buying from Cogent. And I think that's part of the reason why we've been gaining market share. Remember, our European business is almost exclusively NetCentric. We virtually don't have a corporate European footprint.

Okay. Just a quick follow-up on that last point. Obviously, there's some very meaningful ties between some of the wireless companies and the PTTs. Is there desperation for profitability? Are you seeing any better success in sort of prying away and becoming second source for some of those wireless carriers?

Yes. We have a number of wireless carriers in Europe, where we are the upstream provider. In some cases, we're the second provider. In some case, we're the third. A couple of the incumbent operators try to operate something that resembles a global network. I would argue none of them really have a true global footprint. D&T, FT, Telefonica. So it's a little difficult sometimes to be the primary to those companies, but we do sell to some of the subsidiaries as a secondary position. But for the non-affiliated wireless carriers, we actually do very well in that market segment.

We'll go next to Michael Rollins with Citi.

Two questions. First question is, did you mention the capital lease inception for the quarter? If we could just get the number? And then secondly, just a follow-up on some of the promotions. Dave, is there a way to think about what the promotional discounts, free service promotions, et cetera, what that total amount represented in the second quarter in dollars or as a percent of revenue? And maybe comparing that to some past quarters to try to understand the built-up potential to get more revenue out of the customers that you've already signed up and added to the network?

Sure. I'm going to let Tad take the capital lease question first.

Sure. For the 6 months, we added net present value of new capital leases of $5.7 million. For the quarter, that number was $3.4 million.

And then our capital lease payment in the quarter, actual cash out the door was $1.7 million.

Our principal payment was $1.7 million. That's correct.

Yes, in the quarter. To the promotional activity, I do think it was probably more aggressive than our historical run rate. To put a dollar value on that is actually difficult because, in many cases, particularly on the NetCentric side, we give customers 1 or 2 ports, 10-gig ports at no charge for a period of 30 to 60 days to test us. Those customers have almost unanimously entered into some contract with us. But in many cases, they're under pre-existing obligations to their incumbent provider and have far more than 20 gigs of traffic that they could send us and, therefore, pay us for. And the transition from the test to the initial revenue to kind of a stable state of revenue may take 6, 9 or even as much as 12 months for us to fully recognize. I think just as a "back of the envelope" way to look at it, you can see the extreme increase in the number of customer ports installed versus the more moderate or normal rate of revenue growth. And those are pretty good indicators. So as I said earlier, we went from adding 800 net ports to 1,400 net new customer connections in the quarter. But our revenue growth sequentially, subject to the adjustments that I mentioned, was 3%. So it was kind of an abnormal quarter in ports, a normal quarter in terms of revenue growth. So I think there's definitely some catch-up that will happen. It's just that I don't want to put a number out there, because some of it is dependent on the contracts that customers have.

Dave, one just other follow-up. How are you assessing the risk today versus maybe when you have addressed this question in the past? The risk that at some point, the volume growth doesn't offset the price declines within the NetCentric segment, do you feel any differently about that risk or managing that total revenue growth than you have in periods past? And what would contribute to that answer?

Yes, sure, Mike. So as I've stated numerous times to investors, the total addressable market in dollar terms in the NetCentric universe has been flat at about $1.5 billion dollars now for about a dozen years. We have not seen any significant degradation in that market. Clearly as the dollar has strengthened a little and some of that market is denominated in other currencies, in nominal terms it may have shrunk a little. But the market in local currencies for Internet transit globally is flat. And we expect that to continue. I am very conscious of not being at the forefront of driving down prices and destroying that addressable market, but I'm balancing that in our pricing decisions against the 84% of our width capacity that is being unsold everyday at the high incremental margins associated with it. We continue to be the most aggressive provider in the market. Some independent third parties -- I know TeleGeography just did a study, and it said that the rate of Internet transit pricing declines had accelerated and actually noted that it was a result of some targeted promotional activity. We have always been at the forefront of leading that pricing reduction. Prices will go down forever. But at the end of the day, the number that we're solving for is maximum revenue for Cogent. And we're getting growth when others are not by capturing market share in a flat market. We have not necessarily seen a need to shrink the market, i.e. drive down prices even faster to increase our revenue even faster. We haven't seen any correlation. We think that our discount structure is sufficient, that when we survey existing customers, we find that price is not a major issue. When we go and try to talk to people who don't buy from Cogent, it's almost never that, "We think Cogent's too high." There is usually some perception issue that we need to overcome, and that's the reason for these promotional programs that we talked about. And I think that's a better tool for us to continue to gain share. So I feel very comfortable that our NetCentric revenues over the next 5, 6, 7 years for Cogent will grow very similarly to the rate of growth that they've experienced in the past 7 years we've been public, right at the midpoint of our range. If you look historically, our revenues have been relatively equally split between corporate and NetCentric. That hasn't changed at all. We think that'll continue going forward, and we feel very comfortable that we're not reducing our revenue opportunity, but we're also not necessarily destroying the market. We're trying to take a balanced approach here. But quite honestly, I don't care about my competitors. I care about Cogent's shareholders.

And our final question comes from Barry McCarver, Stephens.

This is actually Mac Tatman in for Barry. I just had a quick housekeeping question that I think I missed earlier. What was the bad debt expense as a percent of revenue in the quarter?

Sure. Let me give you the precise number here, just bear with me one second. We did have unusual bad debt expense in the first quarter related to Megaupload. It was 1.1% for Q2 2012. As I said before, kind of our -- relatively consistent with our historical average, absent the Megaupload impact.

And there are no further questions at this time.

We thank everyone and appreciate your interest in Cogent, and we'll talk on the next quarter. Take care. Thanks.

That concludes today's conference. We thank you for your participation.