Cogent Communications - Q1 2024

Transcript

Operator (participant)

Good morning and welcome to the Cogent Communications Holdings' first quarter 2024 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 Cogent's website. I would now like to turn the call over to Mr. Dave Schaeffer, Chairman and Chief Executive Officer of Cogent Communications Holdings. You may begin.

Dave Schaeffer (Chairman and CEO)

Thank you, and good morning to everyone. Welcome to our first quarter 2024 earnings conference call. I'm Dave Schaeffer, Cogent's Chief Executive Officer, and with me on this morning's call is Tad 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 metrics we present in a consistent manner each and every quarter. On May 2nd of this year we closed the issuance of our $206 million IPv4 securitization notes at 7.9%. These notes mature in five years but may be extended for up to a 30-year term. This securitization was the first ever of a securitization of IPv4 lease revenue. Cogent is the owner of approximately 38.8 million IPv4 addresses. We acquired 28.8 million of these addresses when we purchased PSINet and various other acquisitions early in our history.

We acquired an additional 9.9 million IPv4 addresses in May of 2023 with the acquisition of the Sprint network assets from T-Mobile. We are leasing approximately 12.2 million of these IPv4 addresses out for a monthly revenue run rate of approximately $3.4 million a month. We have securitized 3.1 million of that monthly leased revenue, and this represents revenue from 11.1 million leased addresses. We also included 1.4 million unleased addresses in the pool of this securitization. The IPv4 internet addresses are a finite resource. The price of these addresses has substantially increased over the past several years. Now for an overview of our results. Our combined Cogent business had a very good quarter. Our total revenues were $266.2 million in a quarter. This did represent a $5.9 million sequential decline. Our on-net revenues increased by 0.4% to $138.6 million.

Our revenue under the Commercial Services Agreement with T-Mobile declined sequentially by $5.8 million. Our non-core revenues declined by $1.2 million. Our wavelength service revenues increased sequentially by 7% to $3.3 million. All of the decline in our revenues was attributable to the decline in Commercial Services Agreements and non-core services as was expected. Our EBITDA as adjusted for the quarter was $115 million, an increase of $4.5 million sequentially or approximately 4.1%. Our EBITDA as adjusted margin for the quarter was 43.2%. This is up 260 basis points from the 40.6% we reported last quarter. We received three payments from T-Mobile for a total of $87.5 million in the quarter. Our Sprint costs are reported separately, and we're $9 million in a quarter compared to $17 million last quarter.

These costs include approximately $4.3 million of severance reimbursement in the quarter as compared to $16.2 million in severance reimbursements in the previous quarter. Despite the seasonally increased cost associated with SG&A in our first quarter, our SG&A did decrease by 6.4% from $74.9 million last quarter to $70.1 million this quarter. These SG&A numbers are net of that severance reimbursement that I mentioned earlier. Our SG&A as a percentage of revenues decreased to 26.3% for the quarter, down from 27.5% last quarter. Our cost of goods sold decreased by 3.2% from the previous quarter. Traffic on our network increased by 1% sequentially and was up 20% year-over-year. Our gross debt to trailing 12-month EBITDA as adjusted, and our net debt ratios both significantly improved in the quarter.

Our gross debt to trailing last 12 months EBITDA as adjusted was 3.57 in the quarter, and our net debt ratio was 3.17, substantially below the range we have set historically as a target. We are in the process of realizing cost savings and synergies over the next 3 years. We will continue to receive the impact of these savings and achieve an aggregate of $220 million in savings. We anticipate additional SG&A and other cost savings and revenue synergies as well over the next several years. Our recent progress in achieving these cost savings is very encouraging, and we intend to surpass our initial targeted savings goals. Our sales force performed well in the quarter. Our rep productivity in Q4 of 2023 was 3.3 installed orders per rep per month. This improved sequentially to 4 units installed per rep per month in the first quarter of 2024.

Our sales rep productivity results do also include the impact of enterprise sales reps that joined us from the acquired Sprint business. These new enterprise sales reps are continuing to receive training on Cogent sales processes and methods and have not yet fully reached their maximum level of productivity. Now for our total headcount: in connection with the Sprint acquisition, we hired 942 total employees. At quarter end, 718 of these employees remained with Cogent. During the quarter, our total sales rep count increased by 20%, or a 3% net sequential increase in our sales force. Now for our new wavelength and optical transport service business: in connection with the acquisition of Sprint, we have expanded our offerings to utilize the Sprint network to sell wavelength services or optical transport services across that network. We are selling these services to existing customers, to acquired customers, and to new customers.

These customers require dedicated optical connectivity without the capital cost and ongoing expenses associated with owning and operating their own transport network. We have connectivity and wavelength sales capabilities today in 419 locations. However, these locations do require longer than acceptable sales provisioning cycles. We have sold wavelengths to date in a total of 104 locations. By the end of this year, we will be able to offer wavelength services in over 800 locations across North America with much more rapid provisioning cycles. Our wavelength revenue in a quarter increased sequentially by 7% to $3.3 million for the quarter. Our Sprint acquisition materially expanded our network footprint.

To date, we have reconfigured 25 of the acquired Sprint facilities into Cogent data centers and added these data centers to our inventory of 1,586 third-party carrier-neutral data centers and 78 Cogent data centers, which today contain an operational 159 MW of power. We are in the process of converting an additional 23 of these facilities to Cogent data centers and optimizing our data center portfolio footprint. In a market where we had a former Sprint data facility that we converted to a data center and a legacy leased Cogent data center, we decommissioned one leased data center in the quarter. Now for a comment on our dividend and buyback strategy: in our first quarter, dividend was $45.8 million and was accrued at quarter end and paid on April 9th due to our expanding the period for our sales call.

Our board of directors, which reflected on the strong cash flow generating capability investment opportunities including the additional opportunities afforded us by the integration of the Sprint assets, decided to increase our quarterly dividend by yet another 1 cent a share, raising our quarterly dividend from $0.965 a share to $0.975 per share per quarter. This increase represents the 47th consecutive sequential increase in our regular quarterly dividend and a 4.3% annual growth rate in dividend. Now for a couple of comments on our long-term goals: now that Cogent is fully integrated and combined with the former Sprint network, we are anticipating long-term average revenue growth rates of between 5% and 7%, and EBITDA as adjusted margin expansion of approximately 100 basis points annually.

Our revenue and EBITDA as adjusted guidance targets are intended to be multi-year targets and are not intended to be used as quarterly or specific annual guidance. Our EBITDA as adjusted and leverage ratios are impacted by the $700 million IP Transit subsidy agreement that we received with T-Mobile in conjunction with the acquisition. Beginning in June of 2024, these payments monthly will be reduced from $29.2 million a month to $8.3 million a month, and then will continue for an additional 42 months. This reduction will impact our future EBITDA as adjusted or leverage ratios beginning in the second quarter of 2024, which are always measured on a trailing 12-month basis. We will also be looking to monetize other assets that were acquired in the acquisition.

This will include excess data center space and power, additional monetization of our IPv4 address unleased inventory, and dark fiber over the next several years. Now I'd like to turn the call over to Tad to read safe harbor language and give some additional operational performance metrics for the quarter. Following these remarks, we will open the floor for questions and answers. Tad.

Tad Weed (CFO)

Thank you, Dave, and good morning to everyone. This earnings conference call includes forward-looking statements. These forward-looking statements are based upon our current intent, belief, and expectations. These forward-looking statements and all other statements that may be made on this call that are not historical facts are subject to a number of risks and uncertainties, and actual results may differ materially. Please refer to our SEC filings for more information on the factors that could cause actual results to differ.

Cogent undertakes no obligation to update or revise our forward-looking statements. If we use non-GAAP financial measures during this call, you will find these reconciled to the corresponding GAAP measurement in our earnings releases that are posted on our website at cogentco.com. We analyze our revenues based upon network connection type, which is on-net, off-net, wavelength, and non-core, and we analyze our revenues based upon customer type. We classify all of our customers into three types: NetCentric, corporate, and enterprise customers. Our corporate business continues to be influenced by real estate activity in central business districts. We continue to remain cautious in our outlook for our corporate revenues given the uncertain economic environment and other challenges from the lingering effects of the pandemic.

Our corporate business represented 46.9% of our revenues for the quarter, and it decreased sequentially by 1.4% to $124.9 million due to the grooming of low-margin off-net connections and the elimination of non-core products. We had 51,821 corporate customer connections on our network at quarter end, and for the quarter, the sequential impact of USF on our corporate revenues was not significant. Our NetCentric business continues to benefit from continued growth in video traffic, streaming, and wavelength sales. Our NetCentric business represented 34.6% of our revenues for the quarter, and it declined sequentially by 1.3% to $92 million, and the decline was primarily due to the $5.4 million reduction in the Commercial Services Agreement provided at T-Mobile that Dave mentioned earlier. We had 61,599 NetCentric customer connections on our network at quarter end. Our enterprise business represented 18.5% of our revenues this quarter and was $49.3 million.

We had 19,463 enterprise customer connections at the end of the quarter, and our enterprise revenue decreased sequentially by 5.7%, primarily due to the elimination of non-core products and the grooming of low-margin off-net services. On revenue by network connection type: our on-net revenue was $138.6 million for the quarter, a sequential increase of 0.4%. Our on-net customer connections were 87,574 at quarter end. We serve our on-net customers in our 3,321 total on-net multi-tenant office and carrier-neutral data center buildings. We continue to succeed in selling larger 100 gigabit connections and 400 gigabit connections in carrier-neutral data centers and selling 10 gigabit connections in selected multi-tenant office buildings. Selling these larger connections has the impact of increasing our year-over-year on-net RPU. Our off-net revenue was $118.2 million for the quarter, a sequential decrease of 4.4%.

The sequential decline in our off-net revenue was partially impacted by our migration of certain off-net customers to on-net and the grooming of low-margin off-net contracts. Our off-net customer connections were 34,579 at the end of the quarter. Our wavelength revenue was $3.3 million for the quarter, which was a sequential increase of 7%, and that was 693 wavelength customer connections. Our non-core revenue was $6 million for the quarter, a sequential decrease of $1.2 million or 16.8% due to our decision to end-of-life these non-core products. Non-core customer connections were 10,037 at quarter end, a decline of 16.2%. Some comments on pricing: our average price per megabit for our installed base decreased sequentially by 5.8% to $0.26 but increased year-over-year by 5.9%. Our average price per megabit for our new customer contracts for the quarter was $0.11, a sequential increase of 5.1%.

On RPU: our on-net RPU increased sequentially, and off-net and wavelength RPUs slightly decreased. However, our year-over-year on-net and off-net RPUs increased primarily from the impact of the Sprint business. Our on-net RPU increased sequentially by 0.8% from $521-$525, and year-over-year our on-net RPU increased was 12.6%. Last year, it was $467. Our off-net RPU decreased sequentially by 1.3% from $1,120 to $1,106, and year-over-year it was an increase of 21.5%. Last year, it was $910. Our wavelength RPU was $1,638. Our sequential quarterly churn rate for our on-net and off-net connections for the combined business increased. Our on-net unit churn monthly rate was 1.4% compared to 1.2% last quarter, primarily due to the reduction in the T-Mobile CSA revenue and the associated connections.

Our off-net unit monthly churn rate was 2.1% compared to 1.3% last quarter, again from grooming low-margin off-net contracts and the T-Mobile commercial services contract changes. On EBITDA and EBITDA margin: we reconcile our EBITDA to our cash flow from operations in each of our quarterly earnings press releases. We incurred $9 million of Sprint non-capital acquisition costs this quarter compared to $17 million last quarter. Included in the $9 million of Sprint acquisition costs for the quarter are $4.3 million of severance costs. Included in the $17 million of Sprint acquisition costs last quarter were $16.2 million of severance costs. These severance costs are paid by us but are fully reimbursed by T-Mobile. We will incur some additional severance costs in Q2 2024 but none thereafter. Under U.S. GAAP, these severance costs need to be reported as a receivable at the closing date.

These severance costs are classified as post-acquisition costs and as a component of the bargain purchase gain. On EBITDA as adjusted and margin: our EBITDA as adjusted includes adjustments, again for the Sprint acquisition costs and cash payments received under our $700 million IP Transit services agreement with T-Mobile. We billed and collected $87.5 million under the IP Transit services agreement this quarter and last quarter. All amounts billed under the IP Transit services agreement have been paid to us on time. Our EBITDA as adjusted for Sprint acquisition costs and payments under the IP Transit agreement was at $115 million for the quarter. That was a 43.2% EBITDA as adjusted margin. That was a sequential increase of $4.5 million in EBITDA and a 260 basis points margin increase over last quarter.

Our first quarter has traditionally been a quarter when we experienced a decline in EBITDA margin due to cost of living salary increases, which again we had this year, the resetting of payroll taxes in the United States, and our audit fees. That did not occur this quarter. Our foreign currency impact: our revenue earned outside of the United States is reported in U.S. dollars and was about 17% of our revenues this quarter consistent with prior quarters. About 11% of our revenues this quarter were based in Europe and 6% related to Canada, Mexico, Oceania, South America, and African operations. Our average euro-to-dollar rate so far this quarter is $1.07 and the Canadian rate is $0.73.

Should these average foreign exchange rates remain at current levels for the remainder of this quarter, we estimate that the FX conversion impact on sequential quarterly revenues would be -$0.4 million and the impact year-over-year would be -$0.5 million. We believe that our revenue and customer base is not highly concentrated. Our top 25 customers represented about 18% of our revenues for the quarter. Our quarterly capital expenditures were $40.9 million this quarter, down 6.3% from last quarter. We are continuing our network integration of the former Sprint network and legacy Cogent network into one unified network and converting former Sprint switch sites into Cogent data centers. Our finance lease IRU obligations are for long-term dark fiber leases and typically have an initial term of 15-20 years or longer and often include multiple renewal options after the initial term.

Our IRU finance lease obligations were $517.5 million at quarter end. This is inclusive of an uneconomic finance lease that we acquired from Sprint. We have a very diverse set of IRU suppliers and we have IRU contracts with 328 different dark fiber suppliers across the world. At quarter end, our cash and cash equivalents and restricted cash totaled $163.3 million. Our $44.8 million of restricted cash is directly tied to the estimated fair value of our interest rate swap agreement. Our operating cash flow results are materially impacted by the timing and amount of our payments under our TSA agreement with T-Mobile for transition services and the presentation of the payments of our $700 million IP Transit agreement. Payments under the IP Transit agreement under U.S. GAAP are considered cash receipts from investing activities and not classified as operating activities.

Our operating cash flow was a +$19.2 million for the quarter compared to a -$48.7 million in the fourth quarter of last year. Payments under the IP Transit agreement, again, are reported as investing activities and were both $87.5 million this quarter and last quarter. Debt and debt ratios: our total gross debt at par, including our finance lease IRU obligations, was $1.5 billion at quarter end and our net debt was $1.3 billion. Our total gross debt to last 12 months EBITDA as adjusted and our net debt ratio both significantly improved this quarter. Our total gross debt to last 12 months EBITDA as adjusted ratio was $3.57 at quarter end and our net debt ratio was $3.17. This compares to gross debt of the last 12 months EBITDA ratio 4.07 last quarter end and a net ratio of 3.75.

Our consolidated leverage ratio, as calculated under our note indentures, was $3.51 and our secured leverage ratio was $2.33. Some comments on our swap agreement: we are party to an interest rate swap agreement that modifies our fixed interest rate obligation associated with our $500 million 2026 notes to a variable interest rate obligation based upon the Secured Overnight Financing Rate for the remaining term of our 2026 notes. We record the estimated fair value of the swap agreement at each reporting period and we incur corresponding non-cash gains and losses due to changes in market interest rates. The fair value of our swap agreement increased by $6.2 million from last quarter to a liability of $44.8 million. We are required to maintain a restricted balance with the counterparty equal to the liability. As of today, the value of our swap agreement is $35.5 million.

Lastly, some comments on bad debt and Days Sales Outstanding: our Days Sales Outstanding, or DSO, was significantly impacted at year-end by the conversion of all former Sprint customers to our billing system in November of 2023. Our DSO for worldwide accounts receivable significantly improved from year-end and is reverting to historical norms. Our DSO was 27 days at the end of the quarter versus 37 days at the end of last quarter, so a 10-day improvement. Our bad debt expense was $2.6 million and 1% of our revenues for the quarter and that's in line with historical performance. Again, I want to thank and recognize our worldwide billing and collection team members for a fantastic job in serving our Cogent customers. And with that, I will turn the call back over to Dave.

Dave Schaeffer (Chairman and CEO)

Hey. Thanks, Tad.

Now for a few highlights on the strength of our network, our customer base, and sales force. Our NetCentric business continued to experience significant traffic growth in our business from streaming and other customers. We are a direct beneficiary of the continued migration of video to over-the-top. At quarter's end, we were on net in 1,586 third-party carrier-neutral data centers and 78 of Cogent's owned data centers for a total of 1,664 data centers. This is more data centers connected by a network than any other carrier as measured by third-party research. The breadth of this coverage enables us to serve our NetCentric customers in a larger number of locations and helping them reduce latency on their network.

We continue to expand our footprint and anticipate adding approximately 100 carrier-neutral data centers to our network per year over the next several years, above and beyond the additional 23 data centers that Tad mentioned earlier that we are adding due to the conversion of the Sprint switch sites into data centers. We are continuing to experience extended provisioning cycles for wavelengths, but we now can offer wavelength services in 419 locations. By year end, we will have over 800 carrier-neutral locations connected to our network throughout North America with substantially reduced provisioning cycles that will mirror the provisioning times that we are able to achieve with our transit services. At quarter's end, we were directly connected to 8,098 networks. This collection of ISPs, telephone companies, cable companies, mobile operators, and other carriers, allows us to reach the vast majority of the world's broadband and mobile phone users.

Cogent remains the most interconnected network in the world. Our corporate customers are aggressively integrating new applications that become part of their working world, such as video conferencing. These usages will require higher-speed connections both inside and outside of their premises. Our enterprise customers continue to groom their networks and are focused on our core connectivity products of DIA and virtual private network services, including both VPLS and MPLS managed network services. Now for a highlight on our sales force: we remain focused on growing our sales force and increasing the productivity of those sales reps. We continue to expand and modify our training programs, and we routinely manage out underperforming sales reps.

Our sales rep turnover in the quarter was 5.5% per month, which was down from a peak of 8.7% per month at the peak of the pandemic and is in line with our historical averages, which have been at 5.6% of the sales force leaving the company each month. We continue to train new reps as well as provide supplemental training for the reps that joined us from the Sprint business. Our sales rep productivity increased sequentially 23% to 4 installed units per rep per month. At quarter's end, we had a sales force of 284 sales professionals globally focused on the NetCentric market, 379 sales reps focused on the corporate market, and 14 sales reps focused on our enterprise market segment.

In summary, we remain very optimistic about our unique position to be able to serve the connectivity needs of small, medium, and large businesses in major cities across North America in the central business districts. We have 1,861 on-net multi-tenant office buildings connected to our network with over 1 billion sq ft of office space. We remain excited and optimistic about our enterprise customer base and the ability to provide connectivity services to those companies globally and our opportunity to repurpose assets acquired from Sprint. The repurposing of the network to sell wavelength or optical transport services is progressing well, and the conversion of former Sprint switch sites into data centers is also progressing, increasing Cogent's data center footprint and available power. We have a significant backlog and funnel of over 2,400 wavelength opportunities.

However, due to these longer provisioning cycles, we are uncertain if all of these orders will remain with us through our ability to provision them. We will be able to reduce that provisioning cycle as we complete our network optimization and mirror the provisioning windows that we are able to achieve in our IP business. Key indicators show that office activity is improving. Workplace reentry and leasing activity remain substantially below pre-pandemic levels. However, many tenants are returning to offices, and leasing activity for commercial offices has begun to improve in many areas. We are working diligently to continue to reduce costs and integrate Sprint assets. We are optimistic about the cash flow capabilities of this combined business.

Over the next three years, we anticipate continuing to achieve annual cost savings and exceeding our initial $220 million of annual cost synergies that were projected at the signing of our transaction with T-Mobile. We look forward to monetizing many underutilized assets, whether it be excess data center space or IPv4 addresses or the substantial inventory of dark fiber that we have in our network. Our Sprint acquisition costs do not include separately identified integration costs related to the operating expenses associated with the integration. This has reduced our EBITDA and EBITDA margins as adjusted, but we felt that, it was more appropriate to just include these in our standard run rates. Finally, I want to take a moment to address a question that several shareholders have raised with me, and that is, my sale of a portion of my Cogent holdings.

Unfortunately, I have a large real estate portfolio outside of Cogent, and that portfolio has been under significant pressure. My stock sales are solely attributable to supporting that portfolio and have no reflection on my optimism of Cogent. In fact, I'm more optimistic today about Cogent's prospects to both grow its revenues and expand its profitabilities than I have been in Cogent's entire history. With that, let me open up the floor for questions. Operator.

Operator (participant)

Thank you. The floor is now open for questions. If you have dialed in and would like to ask a question, please press star one on your telephone keypad to raise your hand and join the queue. If you would like to withdraw your question, simply press star one again.

If you are called upon to ask your question and are listening via loudspeaker on your device, please pick up your handset and ensure that your phone is not on mute when speaking and when asking your question. Your first question comes from the line of Sebastiano Petti with JPMorgan. Your line is open.

Sebastiano Petti (Senior Research Analyst)

Hi. Thank you. Hi. Can you hear me? Dave, you continue to see issues with the provisioning on the wave side. I mean, what gives you comfort in hitting the 800 locations by year end? And how should we think about the trajectory or the ramp-up in the revenue portion of that as you progress over the course of the year and then perhaps on a 12- to 18-month basis thereafter? And then can you help us perhaps trying to get more comfortable with the decline in enterprise?

Maybe help us think about what you said a lot of the revenue decline was predicated on non-core and other legacy revenues kind of deteriorating. But help us think about enterprise specifically. I mean, what is the glide path there? Is that just a continued decline in revenue, or should that stabilize in the coming quarters, perhaps maybe before returning to growth? Thank you.

Dave Schaeffer (Chairman and CEO)

Options that were wave-enabled to 419 in the 12 months since the acquisition was closed. The wave-enablement contains four key work efforts. We need to put a transponder shelf in each of our carrier-neutral data centers to accept the wavelengths. We need to reconfigure the metropolitan networks to be optimized for wavelength delivery. We need to physically extend the Sprint network to intersect our metro networks. And then finally, we need to deploy a reconfigurable add-drop multiplexer at the intersection of the intercity and intracity networks.

All 4 of those efforts are underway to touch all 800 of the locations. This project has multiple subprojects and multiple constraints. Most of those constraints that were outside of Cogent's hands have been resolved, and now all of the additional work is work that is under our control. All of our markets are now physically interconnected. All of the markets have a clear project plan to complete them. And I think we'll see a continued increase in that number relatively linearly between 419 on May 9th to 800 by 12/31/2024. Now to maybe the more important question on wavelengths is provisioning. One of the key advantages that Cogent will have in this market is the standardization of our products and the ability to provision these services within a 2-week window, which is far better than the industry has historically been able to provision wavelength services.

25 years ago, when Cogent started selling high-capacity transit services, the provisioning windows in the industry for those services averaged over 90 days. Cogent transformed the industry by shrinking that to an average of 9 days to install, and that is a big part of the reason why Cogent has become the largest transit provider globally. Now, with wavelength services, most carriers provision those on a customized or bespoke basis. We have architected the network, so the provisioning is much more standardized. Each of the data centers will have the requisite transponder shelf, and the actual provisioning will require 3 key steps. The first step will be the plugging of a pluggable optic at each of the 2 endpoints. So that's actually 2 of the steps: 1 at the A end, 1 at the Z end, which does require a field technician coordination.

But since these are all in carrier-neutral data centers and we have nearly 500 people in our field services organization, we anticipate being able to do that within 2-3 days of order validation. The second piece is once those physical work efforts at the two endpoints are complete, the network operation center needs to build the physical path through all of the intermediate devices. All of that work has been automated and scripted. That work is done by the nearly 250 people in our customer care and NOC support group on a routine basis. We've been doing that literally for decades to support the Cogent IP network. It was invisible to customers, but now it'll be customer-facing. So we should be able to reduce those provisioning times substantially. Today, Cogent is averaging over 120 days to provision a wavelength.

So we're talking about a 10x improvement in that provisioning. But it's migrating from a custom one-off provisioning to a modularized, standardized provisioning program. And we feel very comfortable that we have the resources and the architecture to do that. So as these sites become fully wave-enabled, the amount of custom engineering declines materially. I'm now going to switch to your second question around enterprise. One of the challenges in the enterprise market globally, and it was a challenge that the Sprint enterprise business faced, is that the service providers to maximize revenue oftentimes sold products and services where they had little or no core competence. When we acquired the business from Sprint, there were tens of thousands of non-core services spread over 24 discrete product categories. We have been end-of-life-ing those services. We reduced our non-core revenue sequentially by another $1.2 million. This is by design.

These are products that have negative gross margin. Cogent adds no value in delivering those services. These tend to be security management services or land administration-type services, which were sold in conjunction with connectivity. Part of Cogent's differentiation from other service providers is the limited number of products that we support. We are a connectivity company. We basically sell 5 key products. We sell internet access, by far and away our largest product. Our second largest product is VPN services. These are connectivity services delivered over the internet as either an MPLS tunnel or a VPLS encapsulated mesh. We sell wavelength services, which are transport between point A and point C. We sell internet address space to customers, and then we sell space and power in our 78 data centers and 159 MW of capacity. That's it. That's what Cogent sells.

So in our enterprise business, we are going through two simultaneous grooming exercises. The first is the elimination of these non-core products. The second is analyzing the connectivity products on a circuit-by-circuit basis. We are aggressively migrating off-net to on-net. You saw that show up in our substantial margin improvement and reduction in Cost of Goods Sold. We are also grooming out circuits across delivery medium that are not supportable, whether it be wireless, coax, or even twisted copper, consolidating on an entire fiber-based platform that will give us the ability to scale. And we are withdrawing from countries and markets that are unprofitable and ones in which Sprint was offering services without the requisite regulatory infrastructure. So for all of these reasons, we are grooming less profitable business.

We think the enterprise business will stabilize over the next several quarters, and we actually have the opportunity to grow that business as many other global service providers are withdrawing from this market space.

Sebastiano Petti (Senior Research Analyst)

Thanks.

Operator (participant)

Thank you. Your next question comes from the line of Greg Williams with TD Cowen. Your line is open.

Greg Williams (Director of Equity Research)

Question. Maybe I can follow up on the Waves' question, just ask it differently. You had $3.3 million in Waves' revenue. Last fall, I think you said you were going to get a run rate about $20 million in revenue annualized. By around now, of course, you rescinded that goal. But when could we reach the $20 million or any other target you can give for Waves? Just help us figure out the cadence and the confidence in the Waves revenue increase. And then the second question is just on the integration costs for Sprint.

You noted $9 million in Sprint acquisition costs, but at the end of your scripted remarks, I think you were mentioning separate identified integration costs, and that's reducing your EBITDA. What is that number? How much of this one-time sort of integration cost to Sprint do you have, and where is that going throughout the course of the next year or two? Thanks.

Dave Schaeffer (Chairman and CEO)

Okay. Two very good questions, Greg. So first of all, on wavelengths, we continue to grow our funnel. We're actually trying to discourage customers until we can give them more realistic provisioning timelines. The demand for the routes that we have and the data centers that we are targeting is stronger than we had initially expected. As we kind of outlined, we anticipate the wave business to be about a $500 million run rate business five years after the acquisition.

We were hoping that we could grow that business linearly, meaning it would grow to a $100 million business a year post-closing, $200 million in the second year. Our hope was based on the fact that the limited number of locations that we initially served would actually be the locations that the market wanted. What we discovered as we began selling was that the wave demand was much more diffuse across a much larger number of data centers. It's not that the demand is not there; it's just spread out into facilities that we could not provision. We do still believe that there will be a $500 million business by run rate of May of 2028. I know that's a long time, and investors want much nearer targets. The reconfiguration of the network is progressing.

We are over halfway through bringing on the number of sites that we need. We are not, though, yet in a position to have the standardized delivery that I spoke about. That will be in place by year-end. We hope that that will allow us to eat into the backlog and to also accelerate new sales as we will be more comfortable in taking those orders with more standardized provisioning windows. So in terms of a specific target, burn once. I'm not going to get burned again and lay something out other than we anticipate a relatively linear approach to a $500 million business by the spring of 2028. And we will fully admit that the initial demand did not line up with the locations, but we will catch up as those sites come online. Now for your second question around integration costs.

The operational integration of the business has a real cost associated with it. Oftentimes, companies put a bucket of expenses into that and are kind of given a free pass by investors as these are viewed as transitory as opposed to permanent expenses. We already had enough accounting complexity associated with this transaction. It is highly unusual to have a transaction that has a bargain purchase gain, has an effective, someone gave you something, and it had the ability to take costs out. Normally, a company pays for a transaction and then integrates that over time. That additional complexity resulted in us ultimately going to the chief accountant of the Securities and Exchange Commission, we've mentioned this previously, to get a final ruling on how we should appropriately account for it. We included the transit service payment agreement from T-Mobile, which is the $700 million payment, in EBITDA as adjusted.

We did not want yet another extraordinary adjustment. But there are numerous operational costs associated with converting back-office systems and billing systems. We have not quantified those externally and don't want to report on them to yet have another metric that would be EBITDA as adjusted, as adjusted, as adjusted. I think that's just a little too confusing for people. But you should see continued operational improvement going forward that will include headcount rationalization, site optimization, network integration. And these integration costs will decline. They are baked into our overall guidance of at least 100 basis points a year of annual margin expansion. I think you saw the very rapid impact of that this quarter with a sequential improvement in EBITDA margins of 260 basis points in a quarter that would traditionally have a decline in EBITDA margins. Just add something to the Sprint costs that we do include.

So they're just professional fees directly associated with the acquisition, so non-recurring. And then because under the U.S. GAAP accounting, we need to record the severance amount as a receivable at the closing date. So you put that asset on the books, and it increases the gain. Even though they're reimbursed when we're actually paying those severance costs, they are also included as an expense. So that is obviously directly attributed to the acquisition, and those severance costs are included in the Sprint costs. I hope that helps.

Greg Williams (Director of Equity Research)

That is helpful. Thank you.

Dave Schaeffer (Chairman and CEO)

Great. Thanks, Greg.

Operator (participant)

Your next question will come from the line of David Barden with Bank of America. Please go ahead.

David Barden (Managing Director)

Hey, guys. Thanks for taking the questions. Just one kind of housekeeping item, Dave. You gave us the IPv4 contribution, $3.4 million per month.

Can you kind of give us a sense as to what that would look like sequentially and year-over-year from a growth standpoint as a contributor to presumably the corporate revenue line? And then just on a related point, you've discussed the reservoir of IPv4 that you have and the data center opportunity that's kind of midstream and then the dark fiber. Could you kind of give us a sense as to how we monetize that? Is the intention to do more kind of ABS-type financings to reinvest in the business, or have you had any inbounds for potential sales of these assets, that sort of thing? Thank you.

Dave Schaeffer (Chairman and CEO)

Okay. Couple of great questions, Dave.

So first of all, the growth rate in the IPv4 revenue stream has been running between 2% and 3% sequentially per month for the past year and a half, and we anticipate that growth to continue. As part of focusing on the IPv4 value, we did a few things. We modified our compensation structure for the sales force to increase their commission rates to make leasing of address space equivalent to that of selling bandwidth. And two, we have incrementally raised prices on addresses. Now, we did not raise the installed base, but for new sales, we have increased prices because of the pricing umbrella established by both Amazon through AWS and Microsoft through Azure. So I would expect this business to continue to grow at a similar rate.

We are also evaluating should we sell some of the addresses because it may take us too long at this growth rate to fully recognize value out of them by leasing them. That final decision has not been made. Today, we are leasing out about a third of our inventory, meaning two-thirds are unutilized. We'll continue to aggressively raise prices and increase sales, and then may also choose to pair our inventory by selling. Again, a final decision has not been made, and it's somewhat dependent on market conditions. I think the IPv4 purchase market was more robust six months ago. It slacked off a little bit. I think that will come back, and that may impact our decision on timing and exact magnitude of potential sales. Now, with regard to the data centers, there, we have a lot of foundational work.

Independent of everything that I've described on the network reconfiguration, there is a dedicated team of individuals working on clearing out these sites and then converting them into modern data centers. That work is well underway. We have converted 25 of the facilities. We have 23 more that are in the process of conversion. And even of the 25 that have been converted, there is still some additional remedial work. Within those facilities, we are basically dividing the facility into three areas. There is a small PoP room that is being established that will house our network equipment. Typically, that utilizes about a half a megawatt of power at about 1,000 sq ft of the facility. Secondly, we are establishing a retail Cogent data center in each of these locations.

Those will range from 5,000 sq ft-10,000 sq ft and will typically be between 0.5 MW-1 MW of power. They will look exactly like the inventory of data centers that Cogent had previously. The customer bases typically take 1-2 racks of space at a time with typically 2 kW-5 kW per rack. That leaves a substantial amount of inventory, approximately 1 million sq ft and about 100 MW of power today as surplus. We have begun the process of marketing that to third parties on a wholesale basis. That will either be direct users or other data center operators. We have offered those participants two different models. They can lease space from us on a per megawatt basis, or they can elect to purchase the facility along with the power. We've gone out to about 117 counterparties and have begun those negotiations.

We have three LOIs in hand after two weeks for three specific facilities. It is probably too early for us to bake that into our financial model, but probably over the next quarter or two, we'll be in a position to comment on the monetization. If they are leases, they'd be long-term streams of revenue that we would look to securitize if we wanted cash or just harvest. If they are sales, they would obviously reduce our leverage on our balance sheet, give us additional capital. Then on dark fiber, we have a substantial footprint of dark fiber. However, we have not been willing to sell that, not because of any philosophical reason, but because the team that's involved in provisioning that dark fiber is the exact team of people that are full speed ahead on delivering our wave network reconfiguration.

By year-end, that wave reconfiguration will be substantially complete, and then we can turn our attention to figuring out how deep and what is the right pricing model for the dark fiber in our network. So all of these assets are ones that we will ultimately create value out of, but we need to sequence this properly. Hopefully, that was helpful, Dave.

David Barden (Managing Director)

Super helpful, Dave. Thank you so much.

Operator (participant)

Your next question will come from the line of Frank Louthan with Raymond James. Please go ahead.

Frank Louthan (Managing Director)

Great. Thank you. Hey, Dave, can you comment on kind of where the non-core revenue bottoms out?

I assume it doesn't necessarily go to zero, but when you get rid of the negative gross margin and other products and so forth, how should we think about where that revenue line kind of bottoms out, and how long do you think it'll take to get there? And is there any part of it that you might consider putting in discontinued ops since you're clearly kind of pushing it out? First question. And secondly, can you comment on the usage of the circuits that T-Mobile is paying for with their cash payments? Are they using those? And is there potential for them to grow and do more business with you beyond that? Thanks.

Dave Schaeffer (Chairman and CEO)

Yeah. Two very different questions. So in the non-core business, prior to the Sprint acquisition, Cogent had about $100,000 a month of non-core revenue, about $300,000 a quarter.

That was services that were left over from acquisitions that were 18 years or more earlier. There is a very long tail on these services. We are actually still providing hosted email to PSINet customers 21 years after the acquisition. We kind of wish it goes away. It produces some margin, and we support it. The Sprint services were much more complex and probably much lower margin. Our goal is to get rid of them as quickly as possible. Our expectation is they will be substantially gone by the end of 2026, which is the longest dated contract that we're obligated to support. But there probably will be some residual tail beyond that. We are also looking at the enterprise customer base and realizing that some of the services they are getting for connectivity are not economically viable. The locations are very remote.

The services are delivered over inferior transmission off-net circuits, and the margins are very low. We're doing a combination of raising prices and migrating traffic. First and foremost, anything that's off-net that we can bring into on-net, we're doing that. And you see that in both our numbers in terms of on-net connections and also revenue improvement as well as margin improvement. We have no intention of putting these services into a separate bucket of discontinued operations. There is already too much accounting complexity here at Cogent.

Tad Weed (CFO)

Yeah. Just comparing the amounts related to this non-core revenue and the associated costs, it's really not material enough to group into a discontinued operations bucket.

Dave Schaeffer (Chairman and CEO)

All right. Thanks, Brian.

Operator (participant)

Thanks. Your next question will come from the line of Walter Piecyk with LightShed Ventures. Please go ahead.

Walter Piecyk (General Partner)

Thanks.

Dave, I just want to go back to Barden's question on the IPv4 stuff. So 2.5%, does that mean that a year ago, you had about $3 million of revenue? And when exactly did that revenue start for you in corporate, Dave?

Dave Schaeffer (Chairman and CEO)

Okay. So first of all, we began leasing IPv4s in 2015. One thing I did not correct in Dave's question was the breakdown between corporate, NetCentric, and enterprise. Let me give that. In IPv4 leasing, 85% of the revenues are NetCentric, 14% are corporate, and 1% are enterprise. Between 2015 and mid-year 2022, we would only lease addresses to companies that also purchased bandwidth from us at the same time. In the summer of 2022, we relaxed that restriction. At that point, the growth rate in this business materially accelerated.

The second thing that we did in the beginning of this year was to normalize the commission structure. So we would pay the sales force the same payout ratio if they leased addresses as opposed to just selling bandwidth. That had a positive impact on the sales rate. And then the third thing that we implemented, actually April 1st, is an increase in pricing. This was a relatively small part of Cogent's business and something that, quite honestly, we had not focused on initially. We were approached by a number of banks to potentially do an asset-backed securitization of our network. We reviewed that and concluded that would be impractical due to the fact that our network traverses 12,000 municipal, local, and national jurisdictions. And perfecting a security interest in that broad of a network is just not practical. Second, 60% of our network is based on IRU.

While there is precedent to securitize that, it's more challenging. And third, there was not a huge amount of customer diversity. As we looked at our aggregate balance sheet, we thought that it would make sense for us to reach out to ABS investors as a new group of investors and realized that our IPv4 revenues were an optimal candidate: widely diffused customer base, 8,000 customers, 12,000 unique agreements, a very sticky customer base. The churn rate in IPv4 leasing is 0.8% annually, almost 15 times better than the churn rate in our bandwidth business. And it is an extremely high-margin business with no real operating costs. So it was an ideal candidate for this market.

Now, what turned out to be challenging in doing this transaction were two things: the complexity involved in securitizing not only domestic but international revenue streams, and then secondly, educating investors on a new asset that they had never seen before. So it was a lengthy process. But our business for IPv4 leasing has some significant tailwinds. And the decision by Microsoft and Amazon to lease out at 12x Cogent's rates at the time in 2023 created quite a high umbrella for us in terms of pricing and giving us the confidence that we have the ability to move prices up. Hopefully, that was.

Walter Piecyk (General Partner)

Well, you mentioned that last quarter, Dave, about Amazon and Microsoft. And so it's been 90 days. It doesn't look like growth accelerated there, or you sold.

And then I think in answering Barden's question, you said, I guess, the market looked a little softer, I guess, for IPv4. So I mean, who knows what happens? Given what can happen with IPv6, it would seem like there would be some immediacy to either increase price with the existing leases, try and lease more aggressively, or sell it more aggressively in case that market disappears. So just comments of that. And also as it relates to corporate, I know this is more of a NetCentric business, but I think your buildings dropped for the first time. I look back at my model. I think I go back to like 2008. I don't think I've ever seen your multi-tenant buildings actually drop. What's going on with that? Because you usually add buildings, which obviously gives you incremental potential capacity for some corporate growth.

Dave Schaeffer (Chairman and CEO)

Okay.

I'll take those in reverse order. So post-pandemic, we have slowed the rate of multi-tenant building additions. We actually did add several buildings in a quarter, but we also had a half a dozen buildings converted to residential service from office. So the net reduction of 1 building from 1862 to 1861 was fairly insignificant. Actually, the square footage increased, meaning the new buildings that came on were larger than the ones that were taken offline because of the residential conversion. We do expect our multi-tenant footprint to continue to grow, but at a more moderate rate. It makes more sense for us to divert that capital into more data center connectivity as that is a more robust and growing market. Now I'll go to your IPv4 monetization question. IPv6 was introduced in 1998. It today accounts for a whopping 7% of internet traffic.

The U.S. federal government in 2010 put a mandate out for all agencies to be entirely on V6 within 18 months. Today, they're less than 2% converted. This has a very long tail. It is a very finite resource, and the expense of renumbering is not trivial. So even if you had V6 and it would work, the cost of leasing an address is so de minimis versus the benefit it gives, most companies will take a very long time to renumber. And no one wants a partial view of the internet. The primary reason sale prices have softened over the past 6 months is Amazon and Microsoft have been withdrawing from the purchase market as they accomplish their initial goals. Now, I think both of those companies will reenter the market. There is still a broad and active market.

The pricing for larger blocks has actually continued to go up. We will explore sales. We're very comfortable with our ability to continue to grow the leasing business, and we are going to do what it takes to create the maximum long-term value for shareholders out of this asset, our data center assets, and our fiber assets. No one should have any doubt that Cogent is sitting on valuable assets that can be monetized.

Walter Piecyk (General Partner)

Okay. That's helpful, Dave. And but in the absence of a sale, maybe you can just give us some sense of without having the IPv4s in order to secure that incremental financing. I mean, obviously, these TSA payments are going to drop from $87-$24. So that drop alone, I think, takes your EBITDA to a level that I'm not sure it covers CapEx and cash interest expense.

So when we think about incremental financing without IPv4 assets to lean on for incremental gross debt increases, what is that rate going to look like?

Dave Schaeffer (Chairman and CEO)

So if I remember.

Yeah. So first of all, if I remember well, I had a few fairly aggressive questions from you several quarters ago around our aggregate leverage. And as I pointed out, it peaked at a net leverage target of 4.6, which was, as you pointed out on that call, substantially above the 2.5-3.5 range that we had laid out. We have been able to reduce that leverage to 3.17, substantially below the high end of that range. And that number will come down even further in the next quarter as it is, as Tad pointed out, an LTM test.

We are going to look at our aggregate balance sheet and try to optimize it, whether it be through incremental high yield, whether it be through securitization, or through asset sales. We have many levers to pull. We have been very transparent around the reduction in payments from T-Mobile. They are going down. They will go down in June. However, we are achieving substantial cost savings ahead of what we laid out. And again, I think it was last quarter you had great doubts on our call about our ability to do that. And I think our improvement in EBITDA shows clearly our ability to reduce those costs, both SG&A and COGS, substantially. So I think it will be the combination of cost reductions and revenue growth. As the incremental wavelength business grows at a more normalized rate with very high contribution margins, our aggregate EBITDA will begin to grow.

As I laid out on the last call, we did $352 million in EBITDA in 2023, up from $233 million in 2022. While we're not giving exact guidance, it will be similar in 2024. But in 2025, the impact of wavelengths IPv4 monetization and data center monetization, coupled with growth in the core IP business, should grow our EBITDA and expand margins. As I said, it's a top-line 5%-7% growing business with a 100 basis point a year margin expansion. I'm also going to clarify a question that I didn't answer for Frank, which is T-Mobile's utilization. And that's two-part. One, the commercial services are typically connectivity services and colo. They're getting out of our facilities. They are stopping using us for backhaul. In many cases, we're buying it and reselling it to them. That's part of why off-net declined. They are very far along in that exit.

Their revenues decline sequentially at about $5 million. On the transit services, it was always meant to be a subsidy payment to Cogent. They are using a couple percent of the transit that we are providing them. They could use it all. It'd be fine with us. It's all provision. But I'm not sure there's a huge incremental opportunity with T-Mobile. Hopefully, this will.

Walter Piecyk (General Partner)

What do you think the incremental rate? Well, no. I mean, the question was, what's the incremental rate when you're not using IPv4 to back the cash you're going to need to pay the dividend? Because again, yeah, the leverage is going down, Dave. That's just math, right? You're getting TSA payments. But when those payments drop from $87 million to $24 million, so does the EBITDA. And then if the EBITDA can't cover CapEx or cash interest, I mean, it's just math, right?

The leverage is obviously going to go up. So the only question is, what is the rate? When you have to borrow more, what do you think the rate on that new debt is going to look like?

Dave Schaeffer (Chairman and CEO)

So a couple of things. One, you are correct. As the TSA payments go away, that will have a negative impact on EBITDA. However, growth in the business and the ability to monetize these assets will have a positive impact. They are roughly going to offset each other. So EBITDA for full year 2024 will be similar to full year 2023 and will grow in full year 2024. In terms of being a prognosticator on interest rates, that's difficult. I'm not Jay Powell, and I'm not in a position to answer that question totally.

But what I can tell you is that our current unsecured bonds, not IPv4 securitized, are trading at around 99, giving a yield to worst of about 7.5%. I think that would be indicative of about what our incremental cost of capital would be. We will explore a combination of additional securitizations and additional debt. We understand that as our EBITDA grows, we have a targeted range. We are committed to returning capital to shareholders. So your point that we're giving out more than 100% of our free cash flow is old news, Walt. It's been our policy since 2010.

Walter Piecyk (General Partner)

No, no, no. I didn't say I'm not talking about the dividends. I was saying the EBITDA is not covering CapEx or cash interest. That is new. That is new after these payments run because of the cash interest from the deal, right?

That's the new thing.

Dave Schaeffer (Chairman and CEO)

I don't want to. I'll disagree with your arithmetic. I think you're incorrect on that.

Tad Weed (CFO)

We'll just see how it plays out.

Walter Piecyk (General Partner)

Dave, just one last question. Just on the synergies, just ballpark, what have you realized so far and what's left in the bucket when we try and come up with EBITDA estimates? And we know what the targets are, so you don't need to review them. Just kind of percentage realized or dollars already realized. Thank you.

Dave Schaeffer (Chairman and CEO)

Yeah. We're probably 40% through the synergy realization.

Walter Piecyk (General Partner)

Awesome. Thank you very much.

Dave Schaeffer (Chairman and CEO)

Thanks, Walt.

Operator (participant)

We'll come from the line of Michael Rollins with Citi. Please go ahead.

Michael Rollins (Managing Director)

Hi, Dave. Good morning.

Dave Schaeffer (Chairman and CEO)

Hey, Mike.

Michael Rollins (Managing Director)

I'm curious. If we take a step back, can you simplify either how heritage growth rates in corporate and NetCentric are performing?

And if you'd rather do it pro forma given the integration of the business, but just some way to kind of appreciate the level of growth that you're achieving relative to what you're used to over the long term? And if you see things in those growth rates inflecting more positively or pulling back in terms of that rate of growth as you look at over the next few quarters and you look at the sales trends, the volumes, etc.?

Dave Schaeffer (Chairman and CEO)

Yeah, absolutely, Mike. Good questions. So Cogent has two market segments. It has a NetCentric segment that focuses on transit sales. That continued to grow. T-Mobile, as a customer, was a NetCentric customer. They're a service provider. If you just net out their decline in revenue, that business grew. Traffic grew sequentially 1%, roughly 20% year-over-year. And that business, on a heritage basis, is probably growing around 10%.

In the heritage business, that represented about 40% of Cogent's aggregate revenues. The other larger business, heritage, was Cogent's corporate customer base. That corporate customer business was actually declining during the pandemic far worse than its long-term average growth rate of 11%. Today is probably at a growth rate of around 3%-4% year-over-year, still far worse than what we had experienced for nearly 15 years between going public and the beginning of the pandemic. That business is slowly improving, but I've given up trying to predict when everybody's going to be back in the office at the same level of occupancy pre-pandemic. There is improvement. That business is improving sequentially and year-over-year, but it is a slow pace. We now have a singular integrated customer base.

We report now on three customer types, and we report on some additional products that were not material, non-core. We always reported on but was immaterial. Now it's more material. And we now have wavelengths as a service. So yes, Cogent became more complicated when we acquired Sprint. We got new customers and new products. But the trends in the underlying heritage business are reasonable. They're not at peak, but they're doing pretty well. I mean, a 10% growing NetCentric base and a kind of 3%-4% corporate is not terrible.

Michael Rollins (Managing Director)

And two follow-ups, if I could. Just first on the corporate side, just given where your share is in your buildings of unique customers, what do you see as the catalyst to try to improve share? Is there anything competitively that's shifting in that market that could help or hurt this performance?

Dave Schaeffer (Chairman and CEO)

It's been a gradual shift.

Everybody in our footprint is already using internet connectivity. And if they need a VPN service, they already have a VPN service. So they need to either relocate or they need to change their usage patterns. I think video conferencing was a huge tailwind to that as people became dependent on it. That maxed out their connections. However, many, many companies were reluctant to enter into new IP contracts until they had some clarity around their office real estate requirements. I think companies are now kind of settling into what that new real estate requirements footprint looks like. So I think we're seeing this gradual improvement, but there are still leases that have term left on them that people intend to exit or downsize from. And until they make that final real estate decision, they're not going to make a permanent bandwidth decision.

Bandwidth is a utility to support their office occupancy. I don't know if there's another killer application that's going to drive things, but I do think the dependence on video conferencing is a significant shift in the world from pre-pandemic to post-pandemic. And I think as companies figure out, "I'm going to stay in this office. This is how many sq ft," they're going to be much more interested in signing a long-term higher-cap bandwidth connection. It was interesting. 5 years ago, our average corporate on-net user was using about 18% of a 100-Mbps connection at peak. Today, our average corporate customer is using 13% of a 1-Gbps connection at peak. So they're using 8 times more bandwidth than they were 5 years ago. I think video conferencing is probably the one thing I'd point to. And remember, they're doing that with 40% less employee days in the office.

Hopefully, that was helpful, Mike.

Michael Rollins (Managing Director)

It was. Thanks, Dave.

Operator (participant)

Your next question will come from the line of Nick Del Deo with MoffettNathanson. Please go ahead.

Nick Del Deo (Managing Director)

Hey. Morning, Dave.

Dave Schaeffer (Chairman and CEO)

Hey. Hey.

Nick Del Deo (Managing Director)

How are you planning to use the proceeds from the securitization? Are you going to buy out that uneconomic dark fiber lease that you've seen from Sprint that you've talked about? And if that's the case, can you talk about the mechanics and the benefits?

Dave Schaeffer (Chairman and CEO)

Yeah, sure. So I'll start with the benefits and mechanics. So the lease has a provision that allows us to buy out at a 12% discount rate. This is a lease that is fiber that we don't need and would like to exit as quickly as possible. There's about a $130 million liability associated with that lease today. The payment stream on that is $4.2 million a month.

We would have to write a check for probably about $112-$113 million to buy out. I don't necessarily know we would use all of our cash to do that that we receive. We did $206 million. With an ABS securitization, there were both substantial costs and reserve accounts established. Our net proceeds were about $200 million, of which about $6 million were restricted in reserve accounts. So $194 million of net proceeds. I'm not sure I want to use $112 million of that in cash. We are looking at the points that Walt raised. Do we raise more money? I think what we're going to do is keep some cash on the balance sheet.

We will look to buy out of this lease and continue to be able to invest in the business at the appropriate rates, which include the conversion of the data centers, the wave enablement of the network, and our ability to demonstrate, I think, kind of all three legs of the value proposition that we were anticipating from Sprint. We've been able to, I think, validate the worth of our IP address inventory. I think we have to show the value of the wavelength and dark fiber assets and the colocation. And I think these proceeds will be used to help demonstrate all of that.

Nick Del Deo (Managing Director)

Okay. Okay. Great. And then two clarifications on EBITDA. So first, it looks like your unfavorable lease amortization went from $10.3 million in Q4 to $2.5 million in Q1.

So I think if we were to look at it on a cash basis, I think the sequential improvement would have been even stronger than you showed. Is that fair?

Tad Weed (CFO)

Yeah. The unfavorable lease liability had to be adjusted and increased for the extension in renewal terms that we had already recorded under the leases themselves. So the lease liabilities that are on the balance sheet kind of as a gross up, but the unfavorable lease needed to be matched to that. So that was one of the corrections that was made in the quarter that resulted in a net of a $5.5 million reduction in the gain on the $1.4 billion gain on the bargain purchase.

Dave Schaeffer (Chairman and CEO)

Yeah. So we would have been a little better without that accounting adjustment. And listen, there is so much complexity to Cogent's accounting for this transaction.

We are working very diligently to report everything in a consistent and non-confusing way for investors.

Tad Weed (CFO)

But going forward, in terms of changes to the purchase accounting, we'll have additional severance and perhaps an adjustment to the tax rate, which could be an adjustment to the deferred tax liability. That is all that is expected in the second quarter.

Dave Schaeffer (Chairman and CEO)

Yeah. We also have an anticipated substantial tax refund coming as a result of this. We had overpaid, based on Cogent's run rates going in, our federal estimated tax and have approximately $18 million of a pending tax refund.

Tad Weed (CFO)

Right. So that will be split. We'll get some prior to filing the tax return and the remainder when the tax return is filed. It's about $18 million.

Nick Del Deo (Managing Director)

Okay. Okay. That's great, color.

Just to be clear on the amortization point, the reduction versus Q4, you're saying it went into the gain rather than an expense reduction?

Tad Weed (CFO)

Yeah. Essentially, that's right.

Nick Del Deo (Managing Director)

Yeah. Okay. Okay. Cool. And then second, just a quick one. I know you call that FICA and vacation accruals in Q1. Tad, did you say that you did not have audit expenses in Q1? Because I thought you normally had those.

Tad Weed (CFO)

Oh, yeah. Absolutely did. So they were high.

Nick Del Deo (Managing Director)

Oh, okay.

Tad Weed (CFO)

Yeah. Those associated just with the Sprint acquisition, which was valuation services, are in the Sprint costs. But the traditional audit, we pay over $2 million. You can read it in the proxy on our regular audit. That's included in the first quarter costs. And the variation vacation, if you look at fourth quarter to first quarter, it's a couple million dollars as traditional.

Nick Del Deo (Managing Director)

Okay.

I must have misheard your comment then. Then did you have your sales meeting this quarter, or is that in Q2?

Dave Schaeffer (Chairman and CEO)

We're actually not going to have it this year because it would end up being just a large distraction. We've got so much integration work going on. Now, what we have done is ramped up our regional learning manager program, hired some additional resources, and are doing it on a more regionalized basis as opposed to a global meeting. We'll resume that next year. But we felt with all that was going on with the integration, that there just wasn't enough cycles to do that. We think there's great value in it, but we're kind of taking a less impactful strategy and doing it just through regional meetings.

Nick Del Deo (Managing Director)

Okay. Perfect. Thanks, guys.

Dave Schaeffer (Chairman and CEO)

All right. Great.

Operator (participant)

And so we have no further questions at this time.

I'll hand the call back to Dave Schaeffer for any closing remarks.

Dave Schaeffer (Chairman and CEO)

As always, our calls tend to be a little long. I want to thank everyone for their patience. I think hopefully, we've been clear in answering questions. We look forward to seeing each and every one of you as we get together at conferences. Take care soon. See you soon. Bye-bye.

Operator (participant)

That will conclude today's meeting. Thank you all for joining. You may now disconnect.