Cogent Communications Holdings, Inc. Q4 FY2022 Earnings Call

Good morning, and welcome to the Cogent Communications Holdings Fourth Quarter 2022 and Full Year 2022 Earnings conference call. As a reminder, this conference call is being recorded, and it will be available for replay at www.cogentco.com. A transcript of this conference call will be posted on Cogent's website when it becomes available. Cogent's summary of financial and operational results attached to its press release can be downloaded from the Cogent website. I would now like to turn the call over to Mr. Dave Schaeffer, Chairman and Chief Executive Officer of Cogent Communications Holdings. Please go ahead, sir.

Thank you, and good morning, everyone. Welcome to our fourth quarter 2022 and Full Year 2022 Earnings conference call. I'm Dave Schaeffer, Cogent's CEO and with me on this morning's call is Thad Weed, our Chief Financial Officer. Hopefully, you've had a chance to review our earnings press release. Our press release includes a number of historical quarterly metrics and we present these in a consistent way each quarter. Now for a couple of comments on our results. Our revenue growth accelerated this quarter and our corporate revenues increased by 0.3%. Our revenue for the quarter increased sequentially by 1.3% to $152 million, an increase of 3.2% year-over-year. Our revenue increased by 1.7% for the full year to $599.6 million. On a constant currency basis, our quarterly revenue grew year-over-year by 5.5%. And for the full year, on a constant currency basis, our revenue growth was 3.9%. On a constant currency basis and also adjusting for the negative impact of universal service fee revenues on our revenue quarter-over-quarter or year-over-year on a quarterly basis, our revenues grew by 5.7% and for the full year 2022 grew by 4.4%. Our corporate business continues to be influenced by real estate activity in the central business district. Key statistics that we look at are building entries as measured by security card swipes in buildings and leasing activities that year-to-date indicate that while the central business district markets have seen some improvement, this improvement continues at a slow pace and is still not at pre-pandemic levels. We continue to remain cautious about the outlook for our corporate revenue growth given the uncertain economic environment and other challenges as a result of emerging from the pandemic. Our NetCentric business continues to benefit from accelerating continued growth in video, traffic, and streaming. For the quarter, our network traffic was up 11% sequentially and traffic growth accelerated on a full-year basis to 25% year-over-year. For the full year, our network traffic growth continues to improve. On a U.S. GAAP basis, our NetCentric revenue growth grew sequentially by 2.6% and grew by 9.6% year-over-year. On a constant currency basis, our NetCentric revenue growth for the quarter increased by 15.2% from the fourth quarter of 2021 and for the full year 2022 grew by 16.7%. Our sales force rep productivity was impacted by the rapid expansion of our sales force. Our rep productivity was 3.8 orders installed per full-time equivalent per month. We increased the number of sales reps this quarter by 26 or 5%. We ended the quarter with 548 reps and 503 full-time equivalents. This is an increase of 8.2% in full-time equivalent sales reps. For full year 2022, we increased our sales force by a net number of 58 million or 11.8%. Our average rep tenure declined due to this rapid pace of sales force expansion and did have a negative impact on full-time equivalent sales rep productivity. Now I'd like to take a moment and talk about our progress in closing our pending merger for the Sprint T-Mobile wireline business. We are diligently working to close our acquisition of the Sprint Wireline business with T-Mobile. We have received all material approvals from regulatory agencies globally, including in the U.S., the FCC and the California Public Service Commission. We anticipate this acquisition will close in the second quarter of 2023. We did incur approximately $200,000 in professional fees associated with this acquisition in the quarter and for full year 2022, approximately $2.2 million in fees associated with the acquisition. We anticipate Sprint wireline revenues at year-end 2021 were $560 million. We also anticipate that the run rate at the time of closing will be an annual number of approximately $450 million. We ultimately expect the number of Sprint wireline products to be reduced dramatically from approximately 30 unique products or SKUs today down to four. We have begun to sell the wavelength service this quarter to our customers through a commercial resale agreement with T-Mobile. Over the next three years, we expect to continue to achieve significant cost savings through the integration of our networks, approximately reducing the network expense of the Sprint wireline network in North America by $180 million, achieving approximately $25 million in savings in the rest of the world on the Sprint wireline network; and then finally, achieving $15 million of savings for Cogent in its operation and maintenance expense for its North American network. We anticipate also additional SG&A savings and other cost and revenue synergies over the next several years. With all of the regulatory approvals now in hand, we do feel the transaction is imminent to close in the second quarter of 2023. Now for a comment on our dividend and buyback strategy. During the quarter, we returned $44 million to our shareholders through our regular quarterly dividend. And for the full year 2022, we returned $169.9 million to shareholders through our regular quarterly dividend payments. Our Board of Directors, which reflected on our strong cash-generating capabilities, investment opportunities and the pending merger with Sprint decided to increase our quarterly dividend yet again by $0.01 sequentially this quarter, raising our quarterly dividend from $0.915 per share to $0.925 per share. This represents the 42nd consecutive sequential quarter of growing our regular dividend and the annualized growth rate in this dividend is now approximately 4.4%. Now for a couple of expectations about our long-term guidance. Once we have combined with the Sprint wireline businesses, we anticipate an annual revenue growth rate of between 5% and 7% and an annual EBITDA margin expansion on the combined business of approximately 100 basis points per year. Our revenue and EBITDA guidance targets are intended to be multiyear goals and are not intended to be used as quarterly or even annual guidance.

Thank you, Dave, and good morning, everyone. This earnings call includes forward-looking statements that are based on our current intentions, beliefs, and expectations. These statements, along with any others made during this call that are not historical facts, are subject to risks and uncertainties, and actual results may differ materially. Please refer to our SEC filings for more information on the factors that could cause actual results to differ. Cogent has no obligation to update or revise our forward-looking statements. If we discuss non-GAAP financial measures during this call, they will be reconciled to the corresponding GAAP measurement in our earnings releases, which are available on our website. Like many companies, we are still impacted by the COVID-19 pandemic, and our risks related to COVID-19 and other factors are detailed in our upcoming annual report for 2022 and in our quarterly reports. Regarding corporate NetCentric revenue and customer connections, we analyze revenues based on network connection type and customer type. We classify our customers into NetCentric and corporate customers. Corporate customers purchase bandwidth from us in large multi-tenant office buildings or carrier-neutral data centers, typically consisting of professional services, financial services, and educational institutions. Our NetCentric customers, including streaming companies and content distribution providers, buy significant bandwidth from us in carrier-neutral data centers. This quarter, corporate business accounted for 56.4% of our revenues and 57.1% for the year 2022. Corporate revenue decreased year-over-year by 1.2% to $85.8 million compared to the fourth quarter of 2021 but saw a sequential increase of 0.3%. For the full year, corporate revenues declined by 4.4% to $342.6 million. We reported 44,844 corporate customer connections at year-end, reflecting a 0.7% sequential decrease and a 1.3% year-over-year decline. Our total revenues, as well as corporate revenues, were negatively influenced by changes in USF tax rates. The impact of USF on our revenues for the quarter was flat, but there was a negative year-over-year impact of approximately $300,000 from last year’s fourth quarter, with a total negative impact of $3.1 million for the full year. On a constant currency basis, adjusting for the decline in USF revenues, our total revenues for the full year increased by 4.4%, compared to 2.4% growth from 2020 to 2021. Our NetCentric business, contributing 43.6% of revenues this quarter, grew by 2.6% to $66.2 million sequentially and 9.6% year-over-year. For the full year, NetCentric revenue increased by 11.1% to $257 million. The foreign exchange volatility predominantly affects our NetCentric revenues, which on a constant currency basis, decreased year-over-year by 15.2% and sequentially by 2.5%, but for the full year, NetCentric on a currency-adjusted basis grew by 16.7%. At the end of the quarter, we had 51,670 NetCentric customer connections, an increase of 1% sequentially and 7% year-over-year. Our on-net revenue was $114.9 million for the quarter, a sequential rise of 1.5% and a year-over-year increase of 3.8%. Full-year on-net revenue grew by 2.2% to $452.8 million, with 82,620 on-net customer connections at year-end, served across 3,155 total on-net locations. We are successfully selling larger 100 gigabit and 400 gigabit connections, boosting our on-net ARPU. Our off-net revenue was $36.9 million for the quarter, showing a sequential increase of 0.7% and an annual increase of 1.6%. The decline in the average price per megabit for our installed base was 2.8% sequentially and 19.6% year-over-year. The average price per megabit for new contracts was $0.13, with declines of 13.5% sequentially and 24% year-over-year. The shift towards larger connections is influencing our connection mix, impacting average price more heavily than ARPU changes. Our on-net ARPU increased sequentially by 1.3% to $464, while off-net ARPU, primarily from corporate customers, declined to $914, a sequential decrease of 0.7%. We maintained stable churn rates, with on-net churn at 1% and off-net at 1.1% for the quarter. To reduce turnover, we utilize a dedicated sales team to retain customers at risk of terminating their service and may offer pricing discounts to encourage them to stay. During the quarter, some NetCentric customers signed long-term contracts for over 2,150 connections, increasing their revenue commitment to us by over $20 million. Regarding EBITDA, we reconcile it to cash flow from operations in our quarterly earnings reports. Seasonal factors that typically affect our EBITDA, particularly SG&A expenses, include payroll tax resets, cost of living increases, seasonal vacations, year-end bonuses, audit and tax service timing, Sprint acquisition costs, and annual benefit plan cost increases. For the full year, we incurred $2.2 million in Sprint acquisition costs and approximately $200,000 in the fourth quarter. Our quarterly EBITDA, including these costs, declined sequentially by $0.7 million and by $0.3 million year-over-year. However, for the full year, EBITDA increased sequentially by 1.2% to $230.6 million. Factors affecting our EBITDA results included an increase in sales representative headcount, higher circuit costs from international expansion, and a one-time $1.5 million bonus granted to employees to counter inflation. Our quarterly EBITDA margin, inclusive of Sprint costs, decreased to 37.6%, down 100 basis points sequentially and 140 year-over-year. For the full year, the EBITDA margin, including Sprint costs, decreased to 38.5%. Excluding Sprint costs, our EBITDA increased by 2.2%, while the margin for the quarter decreased sequentially by 220 basis points to 37.8% and by 120 basis points year-over-year. For the full year 2022, the EBITDA margin, excluding Sprint, increased by 20 basis points to 38.8%. On foreign currency, approximately 25% of our quarterly revenues came from outside the United States, with 16% from Europe and 9% from other international operations. Current exchange rates indicate a potential positive sequential impact of $1.4 million but a negative year-over-year impact of the same amount. Our revenue concentration is low, with our top 25 customers representing about 6% of our quarterly revenues. Our CapEx expenditures decreased sequentially to $19.6 million, totaling $79 million for the full year. Supply chain uncertainties and anticipatory purchases related to the Sprint acquisition adjusted our procurement schedule for network equipment to ensure adequate inventory for growth, including new wavelength offerings and interconnection of networks. Regarding our balance sheet, our finance lease IRU obligations, typically long-term leases, totaled $304.2 million at year-end. Our cash and cash equivalents stood at $275.9 million, which includes $52.1 million of restricted cash related to our interest rate swap agreement. Our cash flow from operations was $36.3 million for the quarter and $173.7 million for the year. Total gross debt at par, including IRU obligations, was $1.3 billion, with a net debt of $978.3 million. At quarter-end, our gross debt to trailing 12 months EBITDA ratio, adjusted for Sprint costs, was 5.39x, while our net debt was 4.2 million. Our consolidated leverage ratio calculated under our indentures was slightly different at 5.32%, with a secured leverage ratio of 3.41%, and a fixed coverage ratio of 3.49%. We entered into an interest rate swap agreement to convert a fixed obligation to a variable one, with fair value reflecting market interest changes. At quarter-end, the fair value of the swap decreased by $2.6 million. We maintain a restricted cash balance equal to our liability under this agreement, with settlement payments occurring in November and May. Finally, our bad debt expense was only 0.5% of revenues for the quarter and 0.4% for the year, an improvement from the previous year. Our days sales outstanding remained excellent at 22 days at year-end, thanks to our billing and collections team for their continuous efforts. I will now turn the call back over to Dave.

Thanks, Tad. I'd like to highlight a couple of strengths of our network, our customer base and our sales force. We've achieved excellent revenue growth in our NetCentric business. We are a direct beneficiary of increased over-the-top video and streaming services, particularly in international markets. At quarter's end, we are on net in 1,458 carrier-neutral data centers. We also have 54 Cogent-owned data centers, bringing our total on-net data center footprint to 1,512 data centers, more than any other carrier globally as measured by independent third-party research. The breadth of this coverage enables our NetCentric customers to better optimize their networks and reduce latency. We expect to continue to widen this lead over the market as we project to connect an additional 100 carrier-neutral data centers per year to our network over the next several years, and we will be also adding approximately 45 Sprint data centers to our network footprint. At quarter end, we directly connect to 7,792 unique networks. The collection of ISPs, telephone companies, cable systems, mobile operators, and other carriers give us direct access to the vast majority of the world's broadband subscribers and mobile phone users. At quarter end, we had a sales force of 216 professionals who focus solely on this NetCentric market. We believe this group of professionals is the largest and most sophisticated sales team focused on this market segment. This sales force will also primarily be responsible for the sale of our wavelength products as we close the Sprint transaction. Now for a couple of comments on the corporate market segment. We are seeing positive trends in our corporate business as the work-from-home environment becomes established as part of how people work in an office. We believe our corporate customers will continue to look to upgrade their Internet access infrastructure to larger capacity connections and more diverse connections. Our corporate customers are aggressively integrating some of the new applications that became part of the work-from-home environment, such as video conferencing. This usage will require high bandwidth connections both inside and outside of the premise. Our aggressive push to lower bandwidth costs and provide greater coverage has begun to boost corporate demand for our robust bidirectional symmetric 1 gig and 10 gigabit ports. Corporate customers are increasingly buying connections in carrier-neutral data centers for redundancy and to support the ad hoc VPNs that their hybrid workforce requires. We remain focused on our sales force productivity and sales force growth and continue to improve our sales training programs and manage out underperforming sales reps. On a sequential basis, our total sales rep headcount increased by a net of 26 or 5% sequentially and by 11.8% year-over-year to 548 quota-bearing reps. Our number of full-time equivalent reps, those reps that have been in the seat for more than three months increased to 503 from 465. Our sales force turnover rate improved again to 4.7% per month from last quarter when that number was 5.5% per month. This is down substantially from a peak during the pandemic of 8.7% turnover per month and that was, in fact, the worst sales force turnover rate in our corporate history. But we are now operating below historic averages, with sales force turnover rate averages over the past 17 years, averaging 5.7%. So in summary, we remain optimistic about our unique position in serving small and medium-sized businesses located in the central business districts of major North American cities. We have 1,837 large multi-tenant office buildings on net representing over one billion square feet of rentable space, an average building size of approximately 550,000 square feet. We are excited about adding large enterprise customers to our customer base through the Sprint acquisition and are most excited about our ability to sell wavelengths or optical transport networking services, a new product to Cogent that we are adding to our portfolio and expanding into all of our U.S. data center footprint. Currently, key indicators of office activity, including workplace reentry and leasing activities are improving but do remain below pre-pandemic levels. We are encouraged that many tenants have indicated that they are returning to office, they are leasing new space, and they are now settling on what their new network architectures will work like. As certain corporations elect to downsize their office space requirements, multi-tenant office buildings, this increases the opportunity for new tenants to move into those buildings and allows us to increase our total addressable market and on-net corporate buildings. Under our indenture agreements, including the $250 million general restricted payments basket carve-out, we have a cumulative amount available for dividends and stock buybacks that actually exceeds the $275 million of cash that we have on hand. We are diligently working to close the Sprint wireline transaction. We are excited and very optimistic about the cash flow generating capabilities of the combined company with our expanded product set and footprint. We expect the transaction to close in the second quarter of 2023. With that, I'd like to now open the floor for questions.

Thank you. We'll take our first question from Frank Louthan at Raymond James.

Great. Just to clarify, do you anticipate that this will close closer to the beginning of the quarter or towards the end? Also, can you provide any updates on how the revenue and EBITDA will be recognized based on any updates from the accountants?

Yes. Thanks for both questions, Frank. So, while we have all regulatory approvals in hand, and we have been working diligently with over 30 working groups on the integration, the longest lead item remaining is the successful completion of our transition services agreement. While we have the framework of that agreement and we'll have identified all of the requisite items to be included, the negotiation of that is continuing. I would say probably mid-quarter is the most likely closing date. But we feel highly confident it will close during the quarter. It is probably unlikely that we will close at the very beginning of the quarter, just based on the fact we only have five weeks to complete that transition service negotiation. And there are over 200 discrete items that have to be addressed and agreed to by both parties. While I would characterize the negotiations as very constructive, and both parties are determined to get this done, there is just a lot of detail work that needs to occur. Now to the question about the transit services agreement that we entered into with T-Mobile as part of this transaction. In consultation with Ernst & Young, our auditors, we feel comfortable that we will be able to recognize the vast majority of that $700 million stream of revenue as revenue for Cogent and therefore, will be included in our reporting and in our financials.

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

Just a follow-up on that. You mentioned the revenue run rate at closing, which is around $450. Do you see that as the lowest point and expect it to improve from there, or do you anticipate further revenue decline after the deal closes?

Yes. Thanks for your question, Jim. So, two points. We do project the stable revenue run rate of the acquired enterprise customer base to be approximately $450 million because the transaction is probably going to close sooner than we had initially anticipated, it is our expectation that at closing in the second quarter, the annualized run rate will actually be above $450 million. There was a requirement in most of the acquired customers' contracts to give them adequate notice about the end-of-life product sets. So we will probably see some of those end-of-life products at closing being acquired, and then over the next several months, we'll probably get down to that $450 million run rate. We've also identified a team of individuals, primarily Sprint sales professionals, but also some Cogent sales professionals will be focused on that customer base of about 1,200 large enterprises with the goal of improving the quality of their service, modernizing their networks and preserving that revenue initially, and the ultimate expectation is that we will be able to grow the enterprise revenue base at about 1% to 2% per year. We think we're probably a year or two away from that, but we do expect to plateau at about $450 million of large enterprise revenue.

Great. And then just on the cash flow side, around the transaction, the first payment that you get from T-Mobile, would that come right as the deal closes? Or is there a lag period between when the deal closed and when you receive that payment?

It comes at the end of the month, first month after closing, and that payment is slightly over $29 million a month for the first 12 months and then reduces down to approximately $9 million per month for the next 42 months. Because this is going to be able to be treated as recognized revenue, we will probably end up having to straight-line that contract value. So, we will have some initial deferred revenue that will then burn off over the life of the lease.

We'll move next to Brandon Nispel at KeyBanc Capital Markets.

It's Evan Young on for Brandon. Did you talk about your SG&A expenses for the quarter and your sales rep hiring going forward, do you think that that's going to continue impacting expenses?

Yes. So, for the quarter, Evan, we, first of all, elected to do a $1.5 million bonus to our global workforce as in lieu of a larger CPI increase, we felt that much of the impact of inflation was a one-time shot. While there is a continuing component, we did not want to be locked into that continuing higher expense. So, that was an extraordinary expense in the quarter and was roughly 100 basis points of margin hit through that bonus. Second of all, to address the headcount number, we were very clear that pre-pandemic, we had about 600 total reps, about 550 quota-bearing FTEs. And the pandemic was particularly hard on our headcount. We actually experienced headcount decline of approximately 18%. As we have emerged from the pandemic, we have accelerated our hiring to the fastest pace in the company's history in order to regain that lost brand. We anticipate getting back to that pre-pandemic level of about 600 quota-bearing reps, about 550 full-time equivalent quota-bearing reps. These numbers are not inclusive of the roughly 60 quota-bearing individuals that will come from the Sprint transaction. So, they are additive. So off of that base of about 660 total reps, and with the Sprint team, almost the entire team is an FTE since our average tenure is quite long. We expect to grow the sales force headcount on a going forward basis at a much more normalized rate of about 7% per year. So, once we get to that roughly 660 headcount, we expect our hiring pace to decelerate. We hope that our turnover rate will continue to be below historic averages. We think the training programs that we put in place have helped quite a bit. And with that, the total sales force should grow at about 7% per year or about 45 net reps, a lower pace than we're growing today.

We'll move next to Nick Del Deo at SVB MoffatNathanson.

This is Michael Srour on for Nick. On your last earnings call, you shared how you had received expressions of interest for wavelengths and dark fiber from customers you spoke with at NANOG shortly after the deal was announced. And that was obviously very encouraging. And now you've mentioned that you began to sell wavelengths through a commercial resale agreement with T-Mobile. I was wondering if you could give us an update on the degree to which that's continued. And what it might mean for the speed with which you might start recognizing wavelengths and dark fiber revenue after closing?

Yes. Sure, Michael. Thanks for the questions. So, we needed to negotiate that commercial agreement that would allow us to sell wavelengths pre-closing. That agreement was put in place last month. Actually, in February, we began training our sales engineers and training staff at the beginning of the month to sell wavelengths and then actually just last week, formally launched the product to the sales team through a series of teach-ins around the world. In the initial period, the process is actually more complicated than it would be after the deal closes in the next couple of months. It does require us to enter into arm's length agreements to then resell. We also have a much smaller number of locations that we can sell in today than what we ultimately anticipate. So just to refresh your memory, Sprint had trialed this product and was only able to sell it in about 25 locations. By integrating the networks, we have added an additional 50 locations immediately. So today, wavelengths are available at Cogent or Sprint can resell our metro footprint in about 80 locations. Over the next 60 days we expect that number to go from 80 to approximately 200. It will probably take another year before the wavelength product is standard across our 800 locations. We have already booked initial orders. We do have a great deal of interest. I would say the initial order count at this point from the Cogent side is probably about a dozen wavelengths. I think that's pretty good for less than a week of selling. I think the Sprint side using our metro network since they could formally do that also beginning last week is also in that same kind of roughly 10 or so. So, these are pretty encouraging discussions. In terms of dark fiber sales, we have a memorandum of understanding with one counterparty on one route. We have been cautious in trying to sell dark fiber in advance of the transaction closing for three reasons: one, it's not a product that Sprint has ever sold. Two, there's not a lot of precedent for arm's length pricing. So if we were reselling it, we would need to demonstrate arm's length prior to the deal closing, and we didn't want to take any regulatory risk. And then finally, we wanted to make sure that the value of monies received from a sale would accrue to Cogent and not T-Mobile, so we would not want a dark fiber sale to be completed until after the transaction is closed. I think realistically, we're going to focus on wavelengths first, dark fiber second, and I think by the time we report the third quarter, we should be in a position to give much more color on the dark fiber market opportunity.

Got it. And one more, if I can. Can you talk about the cadence of corporate sales activity from Q3 into Q4 and how it's trended into the first quarter and the overall state of the funnel?

Yes. I would say that it has continued to improve. And in fact, the sequential 0.3% growth is much better than what we had been experiencing during the pandemic. But I also want to temper that by saying we're far from the pre-pandemic kind of 2% sequential growth rate. So, we have averaged. And when we look at all of the leading indicators, whether it be proposals, orders, ARPU, all of those are giving us encouraging indications that there'll be continued sequential corporate improvement. But as I've tried to counsel on the last couple of earnings calls, we anticipated a more rapid return to our normalized corporate growth rate. And now we're really taking a wait and see. Again, we're encouraged, but I think it's going to still be more than a year before we're back to that kind of 2% sequential corporate growth rate.

We'll go next to David Barden at BofA Securities.

You got John Crawford on for Dave. Just curious about leverage. So it's continuing to grow over the past few quarters above that 2.5% to 3.5% range. Can you tell me what direction you're expecting that to head in 2023? And if your mindset about discouraging excess cash has changed since last you spoke on it?

So thanks for the question and a correction on Dave's name, it's Dave Barden. But I think we've had a consistent policy of utilizing our balance sheet and slowly discouraging cash through a systematic increase in the dividend. I think we are going to remain committed to that strategy. It is true that our leverage did increase in the quarter primarily due to our elevated capital lease principal payments and our increase in CapEx. Much of this is an effort to tie the networks together. As I indicated earlier, we've got an additional 50 data centers connected to the Sprint network. Each of those had a capital expenditure associated with that. We have another 150 in process where we have also already spent that capital. I think we will see our leverage on a net basis, maybe go up a little from here, but not much and then start to come down. And remember, the recognition of the $700 million in cash payments from T-Mobile as revenue flows directly through to EBITDA because our transient product is a 100% gross margin product. The locations that we have reserved port capacity for T-Mobile to utilize that service if they choose to would have no incremental OpEx cost for Cogent. So, with that additional revenue stream and margin, we will rapidly delever.

Got it. And then another, if I can. So, I know you mentioned that the NetCentric team would be working on wavelength sales. I was wondering if your philosophy on headcount rationalization from that transaction has changed since you announced it, like the number of people you're hoping to bring over?

It really has not. Remember, the Sprint organization at the day we signed the deal, had 1,320 employees and only 60 of those were in sales. All of those salespeople remain; there have been a number of voluntary reductions in force that have occurred under T-Mobile stewardship and T-Mobile has been very generous in the severance packages that they have offered employees. That headcount today is slightly below 1,100 employees. We expect that number to continue to decline through the use of these severance packages. All of the employees are expected to become Cogent employees, and we have realigned our organizational structure with both senior Sprint people and management roles as well as Cogent, but following the Cogent organizational structure. My guess is there will be some additional fallout post-closing as some individuals choose not to take those new assignments. We negotiated an additional $25 million pool of capital above and beyond the $700 million with T-Mobile. So, we can mirror their severance program post-closing. I would say that we anticipate at closing to be somewhere around 1,000 employees, additional to the about 1,075 that Cogent has. So just over 2,000, and we would expect some additional operational attrition post-closing, offset in part by our growth in the sales force. So, we anticipate the combined company's headcount to be between 2,000 and 2,100.

We'll go next to Bora Lee at RBC Capital Markets.

So, can you remind us where you're expanding internationally and how you're thinking about those efforts and CapEx while integrating the Sprint wireline assets?

Yes. Thanks for the question, Bora. So Cogent today operates in 51 countries around the world. Our most recent expansion was Peru, and we are constructing through IRU acquisition, multiple rings in Lima as our initial market. So, when we go into an international market, we typically buy dark fiber, intercity and then also dark fiber intracity and look to connect to all significant data centers in those markets. We are in the process of several other countries that Cogent is organically adding. And in addition to that, we will be acquiring three new markets where Cogent had intended to go but has had significant regulatory lags that being India, Thailand and Malaysia. Sprint already has the requisite licenses in each of those markets. So, they will come on net quickly. So, I would suspect in 2023, we will add those three markets and probably about another five or six additional countries that Cogent organically is working on. That is factored into our expectation of adding approximately 100 carrier neutral data centers a year. And just to remind investors, the Sprint network outside of the U.S., inclusive of Canada and Mexico was entirely based on lit rented capacity. The Cogent network is entirely based on owned dark fiber. So, we are in the process of converting over those Sprint intercity networks and data center connectivity onto the Cogent-owned network. About 15 of those international data centers have already been brought on-net. We have about another 30 in process, and then we'll continue that effort until we have completely eliminated the least Sprint international network, and that is anticipated to result in about a $25 million a year annual savings rate.

And one more if I could. So, you somewhat alluded to this in your answer to my question. But in terms of the long-term guidance of 5% to 7% revenue growth and more specifically, 100 basis points margin expansion per year. How are you thinking of the sources of that margin expansion between COGS and SG&A?

Yes. Historically, Cogent has gotten its 200 basis points of margin expansion almost equally from COGS efficiency and SG&A efficiency, roughly a 50-50 split if you go back and look at the past 17 years of public filings. I think going forward, we would expect about the same. When we looked at our EBITDA performance, we actually had the very best gross margin quarter, I think, in the company's history, it's 62.5% in the fourth quarter, continued improvement. The smaller rate of EBITDA margin expansion year-over-year was really impacted by SG&A and it was really this rapid rehiring of our sales force post-pandemic. I think we'll be through that by midyear this year, and we should get back to kind of a cadence of about half of our EBITDA margin improvement coming from gross margin pickup and about half of it coming from SG&A efficiency.

And we'll take our next question from Michael Rollins at Citi.

I'm curious if you think about the process of taking share in each of the Corporate and the NetCentric segment, how does the practice of taking share generally compare with your expectations of that pace when you create the multiyear business plan and update that every so often. And what are the factors that you find cause the variance, whether it's a positive variance or negative variance? And is there something to learn from that as we try to consider the opportunities to take share from these products that you're going to expand with this upcoming acquisition?

Yes. Sure, Mike. Thanks for the question. And I think going forward, there'll actually be a third category, which is large enterprise, which really is a very different customer base than our traditional corporate customer, which tends to be a small or medium-sized business. I think when we look at the corporate segment, our gain in market share is as a result of more footprint and better penetration in that footprint. I think the variance came as a direct result of the pandemic. The increase in vacancy rate in that footprint and the, I think, paralysis of decision-making by the remaining tenants with the uncertainty around their office configurations. As we are emerging from that and companies are kind of settling on a work plan, whether it be hybrid or entirely in the office, we're seeing an improvement, but because there is still elevated vacancy and those work plan implementations are not universally adopted yet, I think we've seen a slower-than-expected rebound in the corporate business. It is rebounding. It is clearly much better than it was during the pandemic, but it is not back to kind of the pre-pandemic consistency that we had. And I think we're trying to be realistic about that. We look at all of the leading indicators as well as actual performance and then modify our plan. And that's why earlier in the call, I said we're probably at least a year away from that 2% sequential growth just based on the data we have. Moving over to the NetCentric business. There, we look at actual traffic patterns on a daily basis. We look at the need of our customers to have multiple connections to reduce latency in their applications, and we look at aggregate traffic growth. And there, we were probably too conservative in our modeling around the NetCentric business. I think we're very accurate in a rate of price decline, and I think that will continue, and we've been very consistent in that messaging. But in terms of the rate of traffic growth and the gain in market share, we were probably too conservative. We continue to grow our NetCentric business on a year-over-year basis, constant currency at about 60% faster than our long-term average. We expect that outpaced growth to continue for longer than we expected, but we also think that it will, at some point, begin to moderate. The final piece, which we have less empirical data on other than the diligence materials we've reviewed inside of Sprint is that large enterprise customer base. And the fact that they've been with Sprint for 30 years is extremely encouraging. They've been through multiple technology refresh cycles, and we can improve margins and quality by bringing those services on net. We think that we're being pretty conservative in our belief that we will see the $560 million run rate decline of $450 million that will plateau for a year or two and then only grow at a modest 1% to 2%. So that's kind of the thinking that goes into each of those assumptions, and they are literally revisited every quarter, and we adjust our projections to the Board each and every quarter based on real-world inputs in each of these categories.

And next, we'll go to Walter Piecyk at LightShed.

Dave, this is Joe on for Walt. Apologies if I missed this earlier, but I guess, capital intensity remains elevated. So, is there a way to think about kind of what's maintenance versus growth and what's kind of baked in from an acquisition standpoint going forward? Because this year, I guess, at least in Q4, there was the large capital lease principal repayment?

Yes. So, we have been very clear about the three categories. The first being the maintenance number for Cogent as kind of its organic business. That number needs to be about $35 million a year, and we anticipate that number continuing. What we would be spending above and beyond that would be expansion. We have added about 120 buildings at an average cost of about $125,000 a building, or about $15 million in building connectivity, and then we also add additional metro fiber, either in new markets that we expand into or densifying markets that we are in. Above and beyond that, we have said that we will spend $50 million one-time for the integration of the Sprint and Cogent networks. Some, actually a significant part of that capital lease was to get the fiber to begin to integrate those networks. The Sprint locations were oftentimes not in central business districts and needed to be connected in, and we've spent a significant portion. There is still some more to spend on that one-time integration expense. And then the third bucket of capital intensity, which will be higher for the combined company is that the Sprint business that we are acquiring, inclusive of the capital needed to support the wavelength product, is about $30 million a year. So, the way to think about the business once this one-time integration effort is complete is $35 million for the Cogent business, $30 million for Sprint, so $65 million of recurring CapEx, and then whatever footprint expansion we're going to do. And that's probably right now running in the neighborhood of about $30 million to $35 million, with about $15 million of that being buildings. And as we said, we'll be adding about 100 carrier neutrals and that the necessary either intercity or metro fiber in these new markets.

And that does conclude our question-and-answer session. I would like to turn the conference back over to Dave Schaeffer for any closing remarks.

I'd like to thank everyone. Hopefully, we've been complete in answering all of their investors' questions. And most importantly, we look forward to closing the Sprint transaction quickly, making sure that we have a fully negotiated transition services agreement that protects both parties, and we feel quite confident that will happen in the second quarter. Take care, everyone, and we'll talk soon. Bye-bye.

And this concludes today's conference call. Again, thank you for your participation. You may now disconnect.