Cogent Communications Holdings Inc (CCOI) 2006 Q3 法說會逐字稿

Good morning and welcome to the Cogent Communications Group third-quarter 2006 earnings conference call. As a reminder, this conference is being recorded and it will be available for replay at www.cogentco.com. That again is www.cogentco.com.

I would now like to turn the call over to Mr. Dave Schaeffer, Chairman and Chief Executive Officer of Cogent Communications. Please go ahead, sir.

Thanks very much, and thank everyone for joining the call, and good morning. Welcome to our third-quarter 2006 earnings conference call. I'm Dave Schaeffer, Cogent's Chief Executive Officer. With me this morning is Tad Weed, our Chief Financial Officer.

We are extremely pleased with our results for the quarter; and I again would like to take the opportunity to personally thank the entire Cogent team for their efforts.

Throughout our discussions, we will continue to focus on the results of our on-net business. This is, in fact, our growth business. As a reminder, our on-net services are provided to customers who are located in buildings that are directly and physically connected to our network. We will also highlight several operational statistics we believe that will demonstrate our expanding scale and tremendous operating leverage of this business.

I will start the call today with a review of certain operational highlights. Tad will provide additional details on our financial performance for the quarter and walk through our expectations for fourth quarter of 2006 as well as our guidance for 2007. Following our remarks, we will then open the floor for questions. Now, I would like to turn things over to Tad so he can read our Safe Harbor language.

Thank you, Dave, and good morning, everyone. This third-quarter earnings report and this earnings conference call discuss Cogent's business outlook and contain forward-looking statements. The specific forward-looking statements cover Cogent's expectations for revenue, EBITDA as adjusted, our percentage of on-net revenues, and our operating results and loss per share for the fourth quarter of 2006 and fiscal-year 2007. These particular 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 in 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 as defined by SEC and Reg. G, you'll find these reconciled to the GAAP measurement either in our earnings release or on our website at cogentco.com. Now I would like to turn the call back over to Dave.

Thanks, Tad. I would like to provide some highlights for our third quarter. Hopefully you have had a chance to take a look at the Company's earnings press release and review some of the key metrics in that release.

As with previous quarters, the press release hopefully includes several historical metrics that will help you track the trends in Cogent's business. These metrics will be added to our website. Hopefully, you find these metrics informative and helpful in understanding the results of our operations. As always, if you have any further suggestions on additional metrics or ways in which we can refine these metrics, please let us know.

Our third-quarter 2006 revenue of $38 million was at the midpoint of our guidance range of 37.5 to $38.5 million provided to you on our last earnings call. Our EBITDA of $6.9 million for the quarter was above the top end of our range for guidance for the third quarter, which was between 5.5 and $6.5 million. Our gross margin percentage continued to improve with over 450 basis points of expansion, from 44.5% to 49% from Q2 2006 to Q3 2006.

We are -- continue to be pleased with the growth of our on-net business. Our on-net revenues grew by over 9% quarter-over-quarter from Q2 2006 to Q3 2006. We continue our investment in our sales and marketing efforts and network expansions, which are helping us gain market share and driving our increased on-net revenues. The significant operating leverage of our on-net business is driving the expansion in our EBITDA and gross margin.

Now, I would like to give you a highlight of some of our operational trends as well. Total service revenue for Q3 2006 of $38 million versus the $36.2 million of Q2 2006 represents a 5% increase in our top-line revenue.

Traffic on our network grew by 20% Q2 2006 to Q3 2006. We continue to believe that this traffic growth statistic demonstrates that being a low-cost provider and industry price leader is resulting in continued gains in market share. While some of our competitors have decided to discontinue the practice of providing traffic statistics to investors, we believe that these statistics are informative and important and will help continue to provide an insight into our continued operations. We intend to continue to issue these statistics on our regularly scheduled earnings conference calls.

On-net revenues grew by $2.3 million or approximately 9.2% quarter-over-quarter. On-net revenues continue to increase as a percentage of our total revenue, growing from 69.5% of service revenue in Q2 of '06 to over 72.4% of revenue in Q3 of 2006. On-net customer connections grew by 14.3% for the quarter, growing from approximately 6,100 to almost 7,000 customer connections on-net by the end of the quarter.

Revenues from our legacy off-net business declined approximately 3.3% for the quarter. In addition, our [off] representing a percentage decline of 24% from Q2 of 2005 or Q3 2006. Total customer connections were 11,372 at the end of the quarter.

Our pricing policy remains unchanged. We remain committed to being an industry price leader. Pricing for our most widely-sold product, 100 megabits for $1,000, remains unchanged at $10 per megabit.

On-net customer connections growing at over 14% and exceeding the growth in our on-net revenue of approximately 9.2% resulted in a decline in our net ARPU of approximately 4.7% -- that ARPU going from approximately $1,480 to $1,412. This change in average revenue per user is attributable to a larger sales force that is allowing us to target smaller customers for our NetCentric services, who typically purchased lower bandwidth services.

Our pricing policy, again, remains intact. For off-net services, our average revenue per user remained effectively unchanged at approximately $810 quarter-over-quarter.

Our EBITDA was $6.9 million for Q3 2006, an increase of approximately 52.8% from the $4.5 million of EBITDA that we reported in Q2 of 2006. EBITDA increased largely due to gross margin expansion that is coming from the increasing proportion of our business that is derived from our on-net services, as well as the aggregate growth in those on-net services.

Our incremental margins from our on-net business remain approximately 100%, while our incremental EBITDA margins for this on-net business remain over 95%.

Tad would now like to cover some of our additional operating details for Q3 2006, as well as give you a window into our guidance for 2007. Now, I would like to turn things over to Tad Weed, our Chief Financial Officer.

Thank you, Dave, and again, good morning, everyone. I also would like to thank our team for their hard work and the performance we achieved this quarter.

Regarding loss per share, our loss per basic and diluted common share was a loss of $0.24 for Q3 2006. That compares to a loss of $0.34 for Q2 2006, or an improvement of $0.10 per share.

We had a couple of nonrecurring, but non-EBITDA-impacting, gains during the quarter -- a capital lease restructuring of $255,000 and a reduction to an unused facility lease accrual of approximately $400,000. That added about $0.01 to the EPS for the quarter, but again did not contribute to EBITDA, as both these amounts are below the line.

Our loss per basic and diluted common share of $0.24 for the third quarter was better than our range of third-quarter guidance of between a loss of $0.30 and $0.35.

Our weighted average shares outstanding increased by 3.4 million quarter-over-quarter from 4.5 million for Q2 to 48.5 million for Q3. This is attributed to the weighting of the shares that we issued in our June 2006 secondary offering. As a reminder, that was approximately 4.4 million shares. They are outstanding for the entire third quarter, as opposed to being outstanding for just a portion of the second quarter. As of September 30, we had 48.8 million common shares outstanding.

Statement 123(R), which requires us to expense stock option grants, increased our Q3 2006 loss by about $140,000. That amount is similar to the amount we experienced in the second quarter.

Regarding the impact of foreign currency, the euro-to-dollar impact positively impacted our quarter-over-quarter revenues, but only at about $100,000. As a reminder, approximately 20% of our revenues are based in Europe, and about 5% of our revenues are based in Canada.

Regarding capital expenditures, we added 18 buildings to the network in the third quarter. These buildings were above our average of about one per week. That largely attributed to CapEx totaling $6.1 million for the third quarter, which is above our target of about $5 million per quarter.

As a reminder, we do experience some seasonality in our construction activities. As an example, the fourth-quarter construction is expected to be less than the third quarter, largely due to construction moratoriums in certain cities. However, we expect to continue to add on average about one building a week to the network. We do expect to have over 1,100 buildings on-net at year-end. As Dave mentioned, we have 1,094 on-net buildings on-net at the end of the quarter.

Cash and debt. As of September 30, our cash and short-term investments totaled $49.1 million. We still have a $20 million credit facility that was unused during the quarter. Our true debt remains at $10 million; and those are 7.5% Allied Riser Notes that are due in June 2007.

We do have capital lease obligations. Our capital lease obligations totaled $89.1 million at September 30; and $6.3 million of this amount is considered a current liability. Again, as a reminder, these obligations will be paid out over a remaining weighted average life of approximately 13 years.

Days Sales Outstanding on accounts receivable was approximately 41 days. That is an improvement from the 46 days that we had at June 30, 2006.

Operating cash flow. Positive cash flow from operations was $1.5 million for the third quarter. The reduction from the second quarter, which was $4.9 million, is attributed to changes in working capital, largely quarterly payments for accounts payables.

Regarding guidance for the fourth quarter of 2006 that we're providing now based upon the current run rates of our business, and our expectations for sales rep productivity, an increase in sales and marketing programs, and also network expansion projects. We expect revenue for the fourth quarter of 2006 to be between 40 and $41 million.

The components of that revenue. We expect on-net revenue quarter-over-quarter growth of approximately 10%. We expect off-net quarter-over-quarter revenue to decline approximately 5%. We expect noncore revenues to decline approximately 10%, again quarter-over-quarter.

We expect gross margin to expand to the low 50s from the 49% we experienced in the third quarter of '06. We expect EBITDA as adjusted to grow to between $7.5 million and $8.5 million.

We expect SG&A to be between 30 and 33% of our revenues. As a reminder, this amount excludes equity-based compensation expense, which is classified as SG&A on the face of our 10-Q. That amount, which is disclosed parenthetically on the face of the income statement in the 10-Q was $2.5 million for the third quarter of '06. We expect equity-based compensation expense, including the impact of stock option expensing, to be approximately 1.5 to $2 million for the fourth quarter of '06. The option expense impact is expected to be about $300,000 of that amount.

We expect depreciation and amortization to be between 15.5 to $16.5 million for the fourth quarter. We expect a loss per share for the fourth quarter of approximately $0.20 to $0.25 per share. That loss per share assumes weighted average shares of 48.8 million outstanding for the fourth quarter.

I'm now going to provide 2007 guidance, which again is based upon the run rates of our business and expectations for sales rep productivity, and sales and marketing, and planned network expansion projects. We expect revenue for 2007 to be between 180 and $190 million.

The components of that revenue. We expect on-net revenue to grow approximately 35 to 40% year-over-year. We expect off-net revenue to decline approximately 7.5 to 10% year-over-year. To we expect noncore revenues to decline approximately 25 to 30% year-over-year.

We expect gross margin to expand to the mid to high 50%s. We expect EBITDA to grow to between 45 and $50 million.

We do anticipate becoming true cash flow positive in the first quarter of 2007, and this is the strictest definition, meaning more cash in the bank at the end of the quarter than at the beginning of the quarter.

We expect SG&A to be between 30% and 32% of revenues. We expect equity-based compensation expense -- which again includes option expensing of approximately 3.5 to $4 million for 2007 -- and the option expensing component of that amount to be between 1 and $1.5 million. We expect depreciation and amortization to be between 70 and $75 million for '07.

Lastly, we expect loss per share to be between $0.55 and $0.65 per share. That assumes 48.8 million shares outstanding. Now, I will turn the call back over to Dave.

Thank you, Tad. Let me give you a little more depth into our network expansion. In the quarter, we added 18 buildings to our network, compared to the 23 buildings that we added in Q2 2006. That represented a lower amount of capital spend. As Tad indicated, our projections for the fourth quarter indicate a further reduction in the number of buildings that we add, getting back to our goal of about one building per week.

As of September 30, 2006, we had 1,094 buildings on-net. The multitenant office buildings represented about 475 million square feet of rentable space. We currently have approximately 17 buildings that are in various stages of construction that will be added to our network. We expect to have over 1,100 buildings on-net by December 31, 2006.

Sales force productivity. This is a key metric that management follows as we continue to monitor our ability to grow our on-net revenues. As of November 1, we had 132 sales reps. We began the third quarter with 103 sales reps and ended the quarter with 120 sales reps. This is a key driver of our ability to continue to grow those on-net revenues.

Our sales force productivity for the quarter remained at 5 orders per rep per month. When converted to full-time equivalents -- those reps that have achieved their ramp and are at 100% of their quota expectations -- productivity remained at 5.4 units per rep per month.

We believe that as we as additional resources, we will be able to continue to drive both additional unit connections as well as additional revenue growth on our network and, in addition to that, continue to see the trends in traffic growth that we have been able to demonstrate historically.

We continue to expand our footprint. We have over 9,600 miles of metro-fiber; over 23,000 miles of intercity fiber. We remain directly connected to over 1,980 networks globally. Roughly 380-plus of these networks are settlement free peers; and approximately 1,600 networks globally are Cogent customers.

We believe that there is substantial excess capacity in our network. We have no material constraints. We will be able to continue to add traffic to our network without any substantial increases on our capital expenditure.

Our business is entirely based on the Internet and Internet traffic growth. This is not true of many of our competitors. We continue to benefit from the growth in Internet traffic and continued trends as social networks and video continue to migrate to the Internet.

This is different than many of our competitors, who derive only a small portion of their revenues from Internet traffic; and in fact, are faced with product sets that are under tremendous pricing pressure and perhaps even experiencing declining revenue streams.

So in summary, our traffic continues to increase. We continue to gain market share, because those increases are substantially faster than those of our competitors.

Our on-net revenue -- this is our growth business -- continues to increase as a percentage of our revenue and, most importantly, continues to increase at approximately 10% quarter-over-quarter.

We see our gross margins expanding. The unprecedented expansion of 450 basis points in the last quarter was very encouraging to management. Our EBITDA margins continue to expand, demonstrating the heavy operating leverage of our business.

If you look at the guidance the Tad has given, at the midpoint of that guidance for next year, $47.5 million, you can see over 100% growth in our EBITDA from where we will end 2006.

Probably most importantly, we remain encouraged that the Company's operating cash flows will continue to translate into the ability to become true free cash flow under the definition that Tad laid out for Q1 2007. Again, we are extremely encouraged by the results of this last quarter and the trends that we see in the business going forward. Now, I would like to open the floor up for questions.

(OPERATOR INSTRUCTIONS) Jonathan Atkin with RBC Capital Markets.

Two questions. First of all, were there any onetime or out of run rate items that affected either on-net revenues or any of the expense items during the quarter?

Secondly, I wonder if you can elaborate a little bit on your on-net guidance for '07. Talk about the type of pipeline you're seeing that leads you to guide towards 35 to 40% growth year-on-year. Thank you.

Jonathan, it's Tad. There were no material nonrecurring items in the third quarter of '06 that impacted EBITDA or revenue. There were those two items I mentioned that impact EPS, but those were below the line.

Now with regard to pipeline, we continue to monitor our sales-force productivity as well as the size of our funnel. Each of our reps is expected to build a funnel of 400 legitimate opportunities that they will be working. Those opportunities are located in on-net buildings and are potential purchasers of our service over the next 12 months.

As we look at the funnel that is in place, we have an unprecedentedly large funnel that would represent about $16 million in monthly recurring revenue. Looking at historical fallout rates from that funnel, we remain encouraged that the approximately 10% quarter-over-quarter sequential growth in on-net revenue can be maintained throughout 2007.

As Tad indicated, we expect to see that guidance -- or those on-net revenues grow at approximately 35 to 40% year-over-year. We continue to gain share, and we see our value proposition to customers probably continuing to increase.

Anything about Internet bandwidth usage demand that leads you to perhaps a different outlook for the NetCentric business?

We have two different types of customers. Our corporate customers generally experience about 120% annual growth in traffic. But we derive no additional revenues from those customers.

We also monitor our NetCentric customers, where there is a much closer linkage between traffic growth and revenue. In that customer base, we continue to see traffic volumes increase at about 75% per year, which again encourages us to the revenue growth numbers that we have laid out on a weighted average basis.

Thanks very much.

Tim Horan with CIBC World Markets.

Great quarter, guys. Dave, just elaborate on the point a little bit. Have you done any more detailed studies of your on-net buildings? Maybe looking at your penetration on either a square foot or a customer basis. Where you're at now, and where ultimately you think you might level out; maybe some experience, some buildings. Maybe you have kind of peaked out, do you think, on penetration of the customer base in those buildings? Thanks.

We constantly monitor our customer base within the buildings that we serve. We have two different types of buildings that are connected to our network. In those corporate buildings, there is just about 50 potential opportunities per building, and we remain at just under five customers per building. So we see a significant additional opportunity in those buildings.

We have not really seen any buildings that have reached maturity. There is a great deal of variance in that corporate footprint, probably with our largest and most significant building, with over 70 customers in the building, and obviously some buildings that are relatively new to the network, with only one or two customers. But across the board, we think we have a significant opportunity to continue to add and probably close to triple the number of customers that we have in those corporate buildings.

Now switching gears to the NetCentric buildings, these are the carrier-neutral data centers. On average, these centers have about 200 opportunities per center. We remain just under 6% penetrated in those facilities.

We have the ability to grow revenue in those NetCentric buildings really three ways. One is to continue to gain customers within those facilities. Secondly, derive additional revenues -- to Jonathan's question -- as traffic volumes increase from those customers in the facility.

Finally, different than our opportunity in our corporate buildings, we have the unique advantage of being able to bring customers to these facilities as a place to connect to Cogent, and continue to grow both the addressable market and also our penetration within those facilities by us bringing those customers. We have a historical pattern of being able to do so.

Thanks. One question maybe on the off-net side. Have you explored ways to maybe lower your cost there on the special access side, to stabilize or start to grow that business again? Are you looking at any new technologies or new [cell] arrangements?

You know, we constantly monitor our off-net circuit purchases. Those purchases are done in probably somewhere around 70 different jurisdictions, but the 50 states here in the U.S. as well as internationally.

We have elected to buy those special access facilities under tariff. We are not treated as a CLEC and use no regulatory mechanisms to lower that cost. We do enter into master purchase agreements or master service agreements with providers to continue to drive down that cost, as well as continue to [groom] the network.

We, however, believe that costs for those circuits have stabilized over the past several years. We will remain competitive in that market, but the real strategic value in Cogent is to be able to drive growth in our on-net business.

Thanks so much.

Jurgan Usman with Wachovia.

Can you talk about a little bit more color on your on-net business for 2006? It looks like initial -- earlier this year you were giving guidance of, let's say, 40 to 45% revenue growth on-net. It looks like you're going to be slightly lower than that in this year. Kind of maybe provide color whether that is due to you signing up more and more smaller customers; or is it because of some other things? Thanks.

I think there's actually two factors that weigh into that growth. Again, we are very pleased with the growth of a little bit over 35% year-over-year. As you can see, we have seen some acceleration in that growth in the last couple of quarters over the earlier part of the year and the end of last year.

That has been driven really by two factors. One being our ramp in sales force. While we today are actually ahead of our plan in terms of sales force additions, we were slightly behind earlier in the year. We are seeing some of the impact of that, in that we have less reps that are at full-time equivalency.

Secondly, as we have forecasted, as we added additional reps to our NetCentric organization, we would be focusing on smaller customers, as you pointed out. That has resulted in a decline in the average revenue per user in those NetCentric customers.

We expect, however, that looking into 2007, we have pretty much experienced the majority of that ARPU decline. At 4.7% sequentially quarter-over-quarter, that is a much steeper decline than what we are forecasting for 2007, where we see on-net ARPUs declining for the full year at about 5%.

So again, we feel very comfortable with our total on-net revenue growth. That will be driven both by our achievement of sales-force productivity, sales force staffing size, and finally the mitigation of ARPU decline as our mix remains approximately 50-50 between NetCentric and corporate.

Okay. then on the salespeople again, you ended up -- I guess you ended November -- sorry. Right now you're at 132 already. Do you plan to add more beyond that? I know that you have a target of 180 by the end of 2007. Has that changed?

The 2007 target remains identical at 180. What we are doing is adding additional people beyond the 130 target by the end of this year to allow us to do two things. One, close the gap between full-time equivalency and number of reps. Again, as I pointed out earlier in the call, there was a gap and we are trying to close that, so we can get as close to 130 full-time equivalents as possible.

Then secondly, you know, we experience turnover in our sales force in large part due to our very disciplined approach of managing out underperformers. Just to touch on that, for each of our sales reps what we expect is that once a rep has achieved full-time equivalency they must maintain at least 75% of quota on a three-month trailing rolling average.

That discipline has forced us to terminate a number of reps. We want to stay ahead of that curve. It is encouraging, however, that approximately 91% of the sales force at full-time equivalency is either at or above quota.

Okay. Can you give me the churn numbers for your on-net and off-net businesses?

Sure. For -- talk about the second quarter and the third quarter, we had a slight improvement in churn, but not that material. So for on-net, continued to be at approximately 2%, slightly lower than 2% for the third quarter and is about 2% for the second quarter. Then off-net was slightly improved, but basically 4% for each quarter.

Great, thank you very much.

Kevin Moore with K.M. Moore & Co.

Congratulations on the strong profitability results. I just wanted to understand how you guys were thinking about profitability versus growth; and whether or not you have had to [make] sort of any trade-offs in order to achieve these sort of consensus-beating profitability numbers.

We have been very clear as a management team that we have really two goals. First and foremost is to increase cash flow. I think we have demonstrated that and feel very comfortable that the Company will be generating true free cash flow in the first quarter of next year.

Beyond that, we do expect to see profitability continue to increase. But at the same time, we remain committed to gaining market share, because we view the volume of traffic that we carry, as a percentage of the world's Internet traffic, as a key strategic asset of the Company. Our ability to continue to be a price leader is driven in large part by those traffic volumes.

Fortunately, these two goals are closely correlated to one another. We have not really had to face any significant choices that would sacrifice one of the goals over the other. You know, we do remain focused on our on-net business as opposed to off-net and non-core, which really don't drive either profitability or traffic volume.

Thanks.

Daniel Berninger with Tier 1 Research.

The question I wanted to get to was again on a theme of growth. I know you are only giving guidance for 2007, but I just wanted to see if you had any thoughts on this tree growing and growing and growing, and when you might have to incrementally pick up CapEx. Because right now, the CapEx spend is really not even that tied to the growth, because it is going to a few new buildings.

Then related, whether there was something you can do on the marketing side to improve sales productivity.

I think a couple of things. Our network remains substantially underutilized. Our goal is to increase traffic volumes, on-net network, and the associated revenues.

While we have not given capital guidance beyond the end of 2007, we do not see any significant requirements for additional capital. You know, our capital expenditures of about $20 million a year are bifurcated into two groups. About $12 million is necessary for all of the maintenance to keep the network running. We don't see that number fluctuating substantially either up or down over the next several years.

We also have an expansion of our addressable market, which is our building expansion program. You know, our model works in a finite number of buildings. We have identified probably another 250 to 300 buildings globally that make sense for Cogent to add to its network. We think the measured pace that we continue at -- that being about one building a week -- is the right pace to continue to expand the network.

The third place we could spend capital would be significant augmentations to the network infrastructure, whether it be the metro backbone or the longhaul backbone. There, while we do have a number of ongoing projects, these projects are able to be accomplished within our capital budget and usually out of equipment that we redeploy across the footprint.

We look at traffic volume growth in the Internet in its [entirety] as well as our NetCentric customer base, which pretty closely mirrors that market, at about 75% per year. We do not see any significant requirement for additional capital.

When we look at that 75% per year volume increase on the Internet, what we see is the drivers of traffic over time continuing to shift, whether it be the increase in social sites, file-sharing sites, and probably most recently the continued trend of the migration of video to the Internet.

We don't think that these traffic trends necessarily result in an acceleration in traffic volume; but rather, allow us to continue to grow at this kind of 705% annum rate, off of an increasingly large base and as Cogent becomes an increasingly larger portion of the Internet. I think that was demonstrated in our 20% sequential quarter-over-quarter growth, both the growth within the installed base as well as our gain in market share.

The last part of your question, which is marketing, we have elected to use a strategic sales force that is targeted at our addressable customers, as opposed to a general branding campaign. We think that continues to be the right strategy with our very targeted approach to the market.

Okay, thanks.

Nick Netchvolodoff with Lehman Brothers.

I had a couple of questions on ARPU. I am glad to hear that you think that the on-net ARPU declines are going to start to abate a little bit. But could you comment on the popularity of your Fiber 500 service and 2-meg and what impact that is having? Sort of how many customers you have got there, and whether that is gaining popularity.

Then secondly, comment on your price leadership versus your competitors; and how you have seen that change over the last three or four quarters; and what you expect over the next few quarters. Thanks.

Okay, first of all, with regard to ARPU decline, the primary driver of ARPU decline has been basically the addition of salespeople to focus on the NetCentric organizations and our ability to go after smaller organizations. So we have seen a significant decline in our NetCentric ARPU.

On the corporate side, the Fiber 500 as well as the 2-meg product remain relatively insignificant as a percentage of our revenue, accounting probably for less than 1% of our revenue for each of those products. You know, our corporate ARPU remains pretty constant going forward. It has really been the decline in the NetCentric side.

Now to the point of our competitors, on our corporate footprint we typically compete with companies that are offering T1 or T3 connectivity to our corporate customers. There we remain significantly lower on a per-megabit basis and probably equivalent on a per-connection basis, with a vastly superior product.

At the core of the network, where we compete for NetCentric companies, who buy bandwidth on a per-megabit basis, what we have seen is most of our competitors mitigating their rate of price decline. Our prices remain constant at $10 a megabit. We have seen the market average drop to about $45 a megabit, probably from a year ago maybe $65 a megabit.

There has been, obviously, continued place price pressure on our competitors as we remain at about 20% of their price points. We believe that, in listening to our competitors and what we hear in the market, they are slowing the rate at which they are dropping prices.

We are also hearing that their networks are at or near capacity. Therefore, they are faced with spending significant amounts of capital to add capacity. Most of our competitors have elected to put their capital elsewhere, not at the core of the network.

So you would say [though] the corporate customers are spending over $950, $950 to say $1,000 a month, is that a way to look at that?

I think it is actually a little less than that, Nick. Again, I want to remind you that our 100 megabits for $1,000 is a month-to-month offering. We offer that same product on a one-year contract for $900; or on a two-year contract for $800.

So if you look at the blend of customers who take longer-term contracts as well as the kind of 1% in each of those smaller products, I would tend to say that the corporate ARPU is probably a little bit below what you have just laid out.

Okay, thanks.

James Breen with Thomas Weisel Partners.

Just a question about sales-force productivity in relation to going after some of the smaller businesses. Do you expect that productivity could start to increase a little bit as you're kind of going into this newer market that is arguably a little bit more price sensitive than some of your larger customers? Thanks.

I think we have always been encouraged when we have seen productivity above five units per rep per month. We think that is a reasonable goal based on just the amount of activity it takes to close that business. I think we may see some increase in productivity on a gross basis as the sales force gets more seasoned; i.e., the number of new hires as a percentage of the sales force decline.

I think secondly that as Cogent continues to become probably better known in the market, with a more stable balance sheet and a more credible offering for customers, we should see some slight improvement in sales-force productivity. But we remain encouraged that at the 5.4 units per full-time equivalent we're still substantially ahead of the plan that we have laid out.

Great, thank you.

Scott Goldman with Bear Stearns.

A few questions if I could. First, I wonder if you could talk a little bit about the economics of selling higher capacity off-net circuits? Talk about T3 and gigabit ethernets, you know on an off-net basis. Do they have better economics than just acquiring a T1 off-net?

Second, wondering on churn, when we look at the NetCentric customers churn tends to be a bit higher than the corporate churn. I wonder if any of that contributes to the decline in on-net ARPU, whether there are some dynamics there of higher bandwidth customers churning off.

Finally, Dave, I would love to hear your comments on M&A, particularly in light of Level 3's acquisition of Broadwing.

Sure. First of all, for our off-net products, no matter what the size of the product the gross margins remain at about 50%. We are encouraged that a number of local service providers are rolling out point to point ethernet services, therefore making off-net solutions a lot more practical at higher bandwidths greater than T1.

Virtually all T3 and ethernet services are deployed over fiber facilities. As fiber gets pushed closer to the end-user, whether it be Verizon's FiOS, or Project Lightspeed with AT&T, or some of the European initiatives such as the one France Telecom and Deutsche Telecom have recently been mandated to undertake by the regulators. All of this bodes well as saying that we can leverage that footprint of their fiber deployment for higher bandwidth off-net services.

The profitability remains about the same. We really strive very diligently to match the cost of the loop to the service and maintain that 50% gross margin.

With regard to the NetCentric ARPU decline, let me let Tad take that one.

I think your observation is accurate, that there's two things. There is a greater churn percentage on the NetCentric on-net customer base as opposed to the corporate customer base. The corporate churn remains slightly less than 1%, and the NetCentric churn is approximately 3%.

In that, if you look at the adds, since the ARPU for the adds of new NetCentric customers -- we are, as we have mentioned, typically going or have the ability now to go after smaller NetCentric customers. It is resulting in the ARPU for the adds is actually less than the ARPU for the declines, which is what your observation is. So that does contribute to the decline in NetCentric ARPU quarter-over-quarter.

Scott, let me take a stab at the M&A question. You know, our industry will continue to undergo consolidation. You know, we have been a very active participant in that consolidation as we built out our network.

We have not done any M&A over the roughly past year and a half as we feel very content with the network footprint that we currently have. We really would be looking to add customers and have a very disciplined approach to what we will pay for those customers and the gross margin that they will produce.

Other companies who have very different revenue streams driven by legacy products, that is, circuit switch voice, are continuing to expand through M&A. We think there will be continued consolidation in the industry around us. We will continue to evaluate whether or not we wish to participate in that consolidation. But we remain committed to being very disciplined about what we will pay.

I think the final point is that we remain very content with our product line, and the business that we have, and the operating margins that we derive from that business. We do not have a need to buy businesses that are tangentially related to ours. We remain committed to being a low-cost and low-price provider of network services for the fastest-growing segment of the market, that being the Internet. We do not wish to migrate into the application (inaudible).

Great. One quick question. Can you provide a breakdown of the connections between corporate and NetCentric, where they stand today?

It remains at approximately 70%, 30%, in terms of the unit mix, corporate versus NetCentric. Then on the revenue side, it remains at approximately 50-50.

Thank you.

Jonathan Schildkraut with Jefferies & Company.

Actually most of my questions have been asked and answered. But I was wondering if you might give us a little more color on kind of the competitive dynamic. Who do you run into the most when you're out there bidding for projects? Do you see any development in that, with some competitors kind of emerging more head to head with you as the consolidation moves forward? Thank you.

On the corporate side, our largest competitors are AT&T and Verizon. Occasionally we will see Qwest in their own service territory, but rarely see them outside of territory. We rarely see CLECs in our corporate footprint, because of the types of buildings that we focus on; being these very large, urban, multitenant office buildings.

Switching gears over to the carrier-neutral colocation facilities, where we sell to NetCentric organizations, whether they be other distribution or access service providers, or content rich companies, we generally compete with companies like Global Crossing and Level 3. Occasionally, we will see the MCI UUNET division of Verizon or AT&T in that space. We always compete with the PTTs in Europe, generally in their home markets, but not outside of their markets.

You know, most of the consolidation that has occurred in our industry has been among voice providers and traditional CLECs, who for the most part have focused on smaller businesses in suburban environments. We have not seen any significant impact from that type of consolidation. Obviously, the consolidation that occurred earlier in the year between Verizon and MCI and AT&T and SBC was directly related to the trends that we continue to see.

Thank you.

Ariane Mahler with Credit Suisse.

Two questions for Dave. I wanted to follow up on some comments you made. Internet traffic growth, you mentioned video over the Internet, the key driver; and I guess we would all agree. But there is a perception out there that Cogent doesn't really benefit from that, unless you are assuming people at work do download video all the time. So that is one question; I wanted to see if you had any comments on the composition of your traffic growth. What was driving it?

Secondly, on M&A, do you have any comments on perhaps prices of foreign networks versus what you have seen in the U.S.?

First of all, with regard to traffic growth, you know, we observe the traffic at a macro level. We actually look at every customer on a daily basis, and look at their traffic trends both in terms of average, peak, and total bite transfer of traffic.

We do not inspect the packets at Layer 7, going in and looking at the application layer. But we can get a pretty good sense from the customer on what that traffic is. If you look at some of the larger video sites, we carry the vast majority of the traffic, for example, for YouTube which has experienced tremendous volume growth. In fact, in the 17 months that YouTube has been a customer of Cogent, we have seen their traffic volumes increase by about 290-fold.

That is a very extreme example, but we have hundreds, if not thousands, of service providers who run video across our network. We think that this will continue to drive traffic growth off of a larger base.

We do not, however, believe that even with this huge amount of video traffic that we are going to see any significant inflections either up or down in total traffic growth. We think that the Internet has reached a level of maturity and the 75% annual total volume traffic growth off of these larger bases will continue.

Just to give you a sense of scale, on a given day, Cogent transfers the equivalent of about 2.9 million two-hour full-motion, full-color movies across its network. I don't want to leave you with the impression that we know for a fact that there's 2.9 million movies traveling across our network. But what we do know is how much traffic we carry on a daily basis; and we do think the percentage of that traffic that is coming from video continues to increase.

Now to your second question of M&A, you know, we continue to look at networks outside of the U.S. as well as inside to help us serve additional markets. We remain connected to locations that either generate or transit virtually the vast majority of the world's traffic. Probably somewhere between 70 and 90% of the world's traffic either goes in or out of locations that are already directly connected to Cogent's network.

We have looked at the opportunity to purchase networks outside of the U.S., and again have remained very disciplined about what we are willing to pay for those, and do not want those purchases to be dilutive. But we have looked at a number of opportunities, whether they be in Eastern Europe, Central Europe, Asia, to expand our footprint. We have not seen anything that made economic sense.

Okay, thanks very much.

Donna Jaegers with Janco Partners.

Congratulations on a good quarter. Most of my questions have been asked and answered. But just one more quick one, since you mentioned that YouTube, that you carry the vast majority of traffic for YouTube. Obviously, one of your competitors made a big splash about getting some of their backbone traffic. I was just curious if you are seeing any change in your volume from YouTube.

Donna, we actually have not. We have continued to see an increase in traffic volume from YouTube. I think the competitor that you're mentioning on their earnings call specifically stated that the traffic that they were carrying for YouTube was not counted in their IP total; and therefore they had a relatively anemic growth in their IP business. In fact, what they are providing to YouTube is point-to-point connectivity to allow them to mirror servers from the East Coast and the West Coast.

But we remain YouTube's primary provider of Internet connectivity and the distribution of that traffic. We do not sell point-to-point services such as wavelengths or dedicated access services which our competitor made a press release related to.

Finally, it has generally been the policy of Cogent not to tout any specific customer win, but rather to look at these macro trends, which I think are much more important in helping our shareholders understand our ability to gain share.

Great, thanks Dave.

Colby Synesael with Merriman.

My question has to do with the competitive environment as well. You stated earlier that you think the average price out there is probably about $45 per net per meg. I think some of your competitors out there would argue that, because they offer a more complex solution -- such as point-to-point, or dedicated services, or PSTN services -- when you bundle that together, that price point is actually lower than that $45. It is actually much closer to a price point that is being provided by Cogent.

How would you argue against that? From that standpoint, if in fact solutions going forward become more complex, how do you think you're going to be able to compete?

A couple of points. First of all, at the core of the Internet, where customers buy on a per-megabit basis, these customers tend to be pretty sophisticated and are generally not buying managed solutions or more complex products. They are, in fact, buying large amounts of Internet connectivity on a per-megabit basis.

Remember, Internet connectivity really has three constituent components to it. That being the interface to the customer, the transport of that bit, and the routing of that bit across the Internet, connected to other service providers through peering agreements.

You know, some of our competitors may elect to cross-subsidize different products within their portfolios. You know, we remain the price leader; and I think that has been demonstrated by our growth in IP traffic vis-a-vis our competitors.

Most of our competitors derive a de minimis portion of their aggregate revenue stream from the Internet. Yet the Internet dominates the number of bits that they are actually carrying.

For Cogent, all of our revenues derive from Internet services. In fact, all of the traffic that we carry is Internet. We look at prices in the market. That $45 a meg average that I spoke about is just that -- an average. What we have seen are a huge banding. Some of the higher-price providers are $80 or $90 a megabit. Some of the more aggressive providers are probably averaging $25 to $30 a megabit and, I will tell you, for some very strategic accounts will get down to price points that are generally within 20 or 30% of Cogent's.

We have had a pretty standing offer in the marketplace to have any customer provide us an invoice and we will undercut that invoice by 50%. All they have to do is show us another Tier 1 provider's invoice for a comparable service, and Cogent will sell that service at 50%.

We have not dropped our price, we are still at $10 a meg, and have not had many people show us invoices simply because it is cheaper to buy our list price product than take that 50% discount.

Thank you.

There are no other questions in our queue. Mr. Schaeffer, I will turn it back over to you for any additional or closing comments.

Well, great. Thank you all very much for taking the time to be with us today. You know, again, I just want to summarize that this was a very strong quarter for the Company. The trends we see going forward remain intact. Traffic volumes continue to increase. Most importantly, we are able to better utilize our network and increase the earnings potential of that network.

We see revenue growth next year and, most importantly, EBITDA growth, getting us to that midpoint of 47.5 next year, seeing over a doubling in our earnings before taxes, depreciation, and interest expense that are significant indication of the operating leverage of Cogent's business.

So again, thank you all for your support. Take care and we will talk on the next quarter. Bye-bye.

This does conclude today's conference. Thank you for your participation. You may disconnect at this time.