American Tower Corp (AMT) — Q4 2018 Earnings Call Transcript

Thank you. Good morning and welcome to CoreSite's fourth quarter 2018 earnings conference call. I am joined here today by Paul Szurek, President and CEO; Steve Smith, Chief Revenue Officer; and Jeff Finnin, Chief Financial Officer. Before we begin, I would like to remind everyone that our remarks on today's call may include forward-looking statements as defined by Federal Securities Laws, including statements addressing projections, plans or future expectations. These statements are subject to a number of risks and uncertainties that could cause actual results or facts to differ materially from such statements for a variety of reasons. We assume no obligation to update these forward-looking statements and can give no assurance that the expectations will be obtained. Detailed information about these risks is included in our filings with the SEC. Also, on this conference call, we refer to certain non-GAAP financial measures, such as funds from operations. Reconciliations of these non-GAAP financial measures are available in the supplemental information that is part of the full earnings release, which can be accessed on the Investor Relations pages of our website at CoreSite.com. And now, I'll turn the call over to Paul.

Good morning and thank you for joining us. Today, I will share our key accomplishments for 2018, provide an annual overview of our business and service models, and share our 2019 priorities. Steve will then cover our sales results and initiatives to strengthen our market position, and Jeff will take you through our financial results and our 2019 financial guidance. Our 2018 annual financial accomplishments showed another solid year of growth, including revenue growth of 13%, FFO per share of $5.06, reflecting an increase of 14.2% or 11.9% growth after adjusting for the 2017 one-time non-cash charge, and declared dividends of $4.14 per share, reflecting 15.6% growth. We achieved one of our primary sales goals of substantially growing no logo acquisitions. Annualized gap rent sold to new logos in 2018 increased 81% over 2017, and we added many important new companies to our customer communities. We also made significant progress on restoring our development pipeline in 2018. We placed in service nearly 172,000 net rentable square feet, including data centers completed in Reston and Washington DC. We expect two new buildings under construction to come online in mid-2019 in Santa Clara and Northern Virginia, two of our biggest markets, and we proactively pursued further development in Santa Clara. In January, we entered into a contract for approximately 4-acre property adjacent to our existing campus to prepare for continued growth in that market. On the operational front, we delivered an approximate 7% improvement in power utilization efficiency on a same-store basis, compared to 2017, due to wise investments in new infrastructure and an ongoing program of operational improvements. We also achieved our highest uptime as a company, what we call seven nines of reliability, thereby improving customer experience and reducing our operating costs. We also rolled out significant product improvements this year. We upgraded our open cloud exchange to an SDN architecture, which makes it significantly easier for customers to provision cloud interconnection and enables us to add additional features in a more agile manner. We also piloted and put customers in our first inter-market connectivity product and expanded our blended IP product. We also experienced some challenges in 2018. Our overall sales fell short of our goals. We underestimated the challenges associated with selling when our occupancies were high and, therefore, capacity constrained, leading us to miss out on a number of good opportunities due to lack of contiguous space. We expect this challenge will be mostly remedied as our SB8 and our VA3 Phase 1B buildings come online this year. We also expect to complete our LA1 expansion midyear, which will provide further opportunities. Construction commencement at our LA3 data center project in Los Angeles is still on hold due to delays at the Los Angeles Department of Water and Power, and designing the power fleet for the new facility, but we expect that design to be completed soon. On balance, 2018 was a solid year with much heavy lifting across our company to position us well for future years. Looking forward to 2019 and beyond, I would like to remind investors of the key components of CoreSite's business model. Simply stated, we concentrated our activities in eight large metropolitan edge markets that are rich in consumers and enterprises, and where our extensive customer communities provide an outstanding platform for digital activities and transformation. Our campuses are characterized by major interconnection nodes and numerous cloud on-ramps, which anchor large, energy-efficient, highly flexible data centers where the widest possible range of customer deployment sizes and densities can be accommodated, and where customers can interconnect with each other via short run, dark fiber or virtual connections providing the maximum performance, bandwidth, reliability, flexibility, and security. Our primary target customers include enterprises in these markets who are moving to co-location to reduce their data service costs while increasing uptime and performance, and interconnecting to a powerful hybrid cloud and network ecosystem. Second, companies with edge needs in these markets; third, companies exchanging data traffic at these key internet points; and finally, companies who value deployments in these markets as part of a larger performance and cost strategy built around wide-area network realignment. The flexibility of our model is shown by the range of leases we can accommodate. Our smallest new and expansion sales in 2018 was 1 critical kilowatt and our largest was 3 megawatts, and we have added significantly larger leases in the past when we have more capacity to sell. We believe this speaks to the diversity of how we can drive revenue as we add new capacity. We leverage our service model to focus our leasing efforts on higher value deployments where there is a greater need for the components we offer compared to less differentiated offerings in the marketplace. Our 2019 priorities include; first, to translate new construction into more abundant sales, including larger scale deployments in our edge markets; second, to continue to focus on acquiring additional new logos; third, to continue to bring new connectivity and customer service products online to increase sales from new and existing customers; and fourth, to continue to deliver great customer experience and ongoing operational efficiencies. We believe these priorities and our other operating objectives will continue to drive value for our customers, employees, and shareholders. With that, I'll turn the call over to Steve.

Thanks, Paul. I would like to start with a summary of our quarterly sales and leasing results and then spend some time on product enhancements. Turning to our sales results. For the quarter, we had sales of $4.2 million of annualized revenue, which included 16,000 rentable square feet at an average GAAP rate of $259 per square foot, comprised entirely of our core retail co-location sales. While our sales in scale and hyperscale leasing were below our goals for the quarter, our performance included excellent growth in new logo additions, solid pricing, and strong renewals. So, let me walk you through a few of these in more detail. We have discussed in prior calls the goal of substantially growing new logo acquisitions in order to achieve greater diversity in our base, enhance our ecosystem, and provide a platform for future organic growth. And we’re seeing some solid traction in that regard. Here are a few highlights. The annual GAAP rent for new logo additions in 2018 increased 81% over 2017. In addition, the average weighted monthly lease term for new logos in 2018 increased by about six months from new logos in 2017, and the majority of these 2018 logos were driven by processes with ongoing logo growth for cloud and network providers. Pricing was also an important focus for us. For the quarter, pricing on new and expansion leases was consistent with a trailing 12-month average on a per kilowatt basis, driven by continued execution in the core retail co-location space, which typically supports better pricing and profitability. Moving to scale leasing as Paul mentioned, our capacity constraints impacted our sales opportunities primarily in scale co-location, which we believe will be addressed by our ongoing property development and our presale activity for projects nearing completion. We also have several markets with available inventory where we look to execute better in driving scale leasing. Renewals are another key aspect of the health of our business. During the fourth quarter, customer renewals were strong, with annualized GAAP rent of $22.5 million, rent growth of 3% on a cash basis, and 7% on a GAAP basis, and rental churn of 1.9% for the quarter, which is better than expected. So the fundamentals of our business remain strong, with increasing growth from new customer logos signing with CoreSite, existing customers including key strategic accounts choosing to extend their term and expand their space with us, and core retail co-location pricing levels remaining firm. Additional focus going forward will be on executing well in the scale segment as those opportunities align with our value proposition and available capacity. In my role as Chief Revenue Officer, we're evaluating opportunities to evolve our offerings in order to deepen our customer value or provide additional forms of revenue to the company. We have been working directly and through partners with the goal of making it easier for customers to continue their digital transformation journey. As such, there are several steps we have delivered this year and look to expand as we go forward. In 2018, we upgraded our open cloud exchange with SDN architecture; enabling easier onboarding to cloud and SaaS providers, as well as paving the way for future development; added Microsoft Azure Express Route private connectivity in our Silicon Valley and Northern Virginia markets; enabled enterprises with direct private connectivity between VMware and AWS in four markets including Boston, Denver, New York, and Northern Virginia; introduced AWS' logical redundancy offering, a second native point of connectivity to AWS Direct Connect at our Silicon Valley campus; provided dedicated access to the Oracle Cloud infrastructure in Northern Virginia and Washington DC campuses; enabled inter-market connectivity; expanded the flexibility of our blended IP product and formalized the launch of our solution partner program to extend our reach and value through our ecosystem to help customers along their journey from IT assessment to transition services to ongoing support and management. I hope this gives you a sense of our ever-evolving customer-centric initiatives in our data centers. We’re focused on the collective value of our solutions to customers as we enable them to execute against their business plans and grow the relationships among each other. Technology is driving higher expectations in all companies to make it easier and faster to do business with them. We believe we are a key resource for enterprises upgrading their technology to better support their business in the future. Enterprises that can drive a better digital experience for their employees, customers, and suppliers should win in their respective industries. We expect to be the cornerstone of how they successfully deliver digital transformation. With that, I will hand the call over to Jeff.

Thanks, Steve, and hello everyone. Today, I would like to share some highlights of our 2018 financial performance, review our detailed fourth quarter financial results, update you on our property operations and development, and provide you our financial guidance for 2019. Looking at our financial results. Our full year 2018 results included 13% revenue growth, 14.2% FFO growth per share, or 11.9%, after excluding a 2017 non-cash charge related to our preferred stock redemption, and adjusted EBITDA growth of 12.5%, while maintaining an adjusted EBITDA margin of 54.4%, and declared dividends of $4.14 per share or 15.6% growth. In 2018, we successfully accessed the capital markets and amended and expanded our credit facility, which provided us $250 million of additional liquidity while extending our debt maturities at attractive rates. We were also an early adopter of the new lease and revenue accounting standards. So we've had that in our rearview mirror since the beginning of 2018. Moving to our fourth quarter financial results. Our total operating revenues were $139.1 million for the fourth quarter, which was in line with the third quarter and reflected a 10.5% increase year-over-year. Operating revenues consisted of $118.3 million of rental power and related revenue, $18 million of interconnection revenue, and $2.8 million of office-wide industrial and other revenue. Interconnection revenue increased 1.8% sequentially and 10.9% year-over-year. FFO was $1.26 per diluted share and unit, in line sequentially, and an increase of 15.6% year-over-year, or 6.8% growth after excluding the 2017 non-cash charge related to our preferred stock redemption. Adjusted EBITDA of $74.6 million increased 1.1% sequentially and 8.5% year-over-year. Adjusted EBITDA margin was 53.6%, up 57 basis points from the prior quarter and down 100 basis points year-over-year, primarily due to the new lease accounting requirements that we implemented in 2018, higher power rates, and increased property taxes. Sales and marketing expense totaled $5.4 million for the quarter or 3.9% of total operating revenue. For the year, sales and marketing expense was $21 million or 3.9% of operating revenues, in line with 2017. General and administrative expenses totaled $10.5 million for the quarter or 7.6% of total operating revenues. For the year, general and administrative expenses were $14.1 million, essentially in line with our guidance. These expenses represented 7.4% of total operating revenues for 2018, compared to 7.8% in 2017. In the fourth quarter, we commenced 23,000 of net rentable square feet of new and expansion leases at an annualized GAAP rent of $192 per square foot, which represented $4.4 million of annualized GAAP rent. Moving to backlog, as of year-end, projected annualized GAAP rent from signed but not yet commenced leases was $9.9 million and $14.3 million on a cash basis. We expect substantially all of the GAAP backlog to commence during the first half of 2019. Turning to our property operations and development. Fourth quarter same-store monthly recurring revenue per cabinet equivalent was $1,537, reflecting a 1.6% sequential increase and a 6.3% increase year-over-year. Q4 same-store turnkey data center occupancy was 90.7%, an increase of 60 basis points sequentially. We ended the quarter with our stabilized data center occupancy at 92.8%, an increase of 40 basis points sequentially. During the fourth quarter, we completed construction and placed into the pre-stabilization pool 25,000 net rentable square feet for our DC2 data center. We have a total of 271,000 square feet of data center capacity in various stages of development across the portfolio. This includes ground-up construction in VA3 Phase 1B and SB8 Phase 1, which together total 108,000 square feet with $118 million incurred to date of an estimated $246 million. Data center expansions at LA1 and LA2, which together total 45,000 square feet with $9 million incurred to date of an estimated $34 million. And preconstruction of LA3 and CH2, which together represent 118,000 square feet with $39 million incurred to date of an estimated $250 million to complete the first phases of these projects. Collectively, these projects total $166.4 million invested as of the end of the fourth quarter of the estimated $530.2 million required to complete the projects. For more details on our development projects, please see page 19 of the supplemental information. Turning to capitalized interest. The percentage capitalized in the fourth quarter was 17.9%. The full-year percentage was 13.6%. For 2019, we expect the percentage of interest capitalized to be in the range of 20% to 24%, which is elevated compared to 2018 based on our development pipeline. Turning to our balance sheet. Our ratio of net principal debt to Q4 annualized adjusted EBITDA was 3.8 times. As of the end of the fourth quarter, we had $236.2 million of total liquidity, consisting of $233.6 million of available capacity on our revolving credit facility and $2.6 million in cash. As a reminder, we announced last quarter that we expect to access the capital markets for $350 million to $400 million in the form of additional debt to term out the outstanding balances of our credit facility. We expect the majority of the financing to be completed in the first half of 2019. As I stated last quarter, we are comfortable with modestly increasing our targeted debt to adjusted EBITDA ratio to 4.5 times. I would now like to address our 2019 guidance. For context, it's important to note that our guidance reflects our view of supply and demand dynamics in our markets as well as the health of the broader economy. We do not factor in changes in our portfolio resulting from acquisitions, dispositions, or capital markets activity other than what we have discussed today. And bear in mind that our guidance reflects coming into the year with constrained capacity due to high occupancy and significant new capacity expected to be available mid-year and beyond. As detailed on page 23 of our fourth quarter supplemental information, our guidance for 2019 is as follows: total operating revenue is estimated to be $580 million to $590 million, based on the midpoint of guidance, which represents a 7.5% year-over-year revenue growth, which reflects the timing of our development pipeline and the current level of capacity entering the year. Connection revenue is estimated to be $74 million to $77 million. This reflects 8.3% growth at the midpoint. General and administrative expenses are estimated to be $42 million to $44 million, representing 7.4% of total operating revenue at the midpoint. Net income is estimated to be $104 million to $109 million, representing $2.15 to $2.25 per share of net income to common shares, or $2.20 at the midpoint. Adjusted EBITDA is estimated to be $316 million to $321 million, and at the midpoint, reflects a 54.4% adjusted EBITDA margin and 7.6% year-over-year growth. FFO is estimated to be $5.21 to $5.31 per share in operating units. At the midpoint, this reflects 4% growth. Other guidance includes rental churn, which is estimated to be 6% to 8% for the full year. Cash rent growth on data center renewals is estimated at 2% to 4%. Capital expenditures are estimated at $400 million to $450 million, including data center expansion of $380 million to $415 million, non-recurring investments of $5 million to $10 million, tenant improvements of $5 million to $10 million, and recurring capital expenditures of $10 million to $15 million. That concludes our prepared remarks. Operator, we would now like to open the call for questions.

So I was interested in Steve's comments about better execution on scale leases going forward. And if you could maybe elaborate a little bit about what that looks like apart from having inventory to sell, what is it on the execution side that you think can drive better success? And then on the product side, you mentioned the OCX, the Open Cloud Exchange SDN offer. Is that something that you're offering because you can in anticipation of potential customer demand, or are you seeing meaningful takeaways already that you've kind of ramped that product and active customer interest, or is that more on the come? Thank you.

Great. Hey, Jonathan. Thanks for the question. As it relates to better execution, I think you'll hear me say that consistently regardless of our results. But I do think we can and should do better, especially in the scale leasing and really across the board. As it relates to scale leasing with the inventory that we have today, there are a few pockets where we have availability to sell scale deals today albeit a few. And I just think there are always areas where we can execute better as far as how we are aligning to the market, making sure that the market is aware of our availability out there, and ensuring that our value is driven throughout the organization in making decisions so that we position ourselves in the most favorable light to win. So that's really just how do we ensure that those aspects come together, and we continue to just try to get better in that regard. As it relates to your question around OCX and SDN, it really just sets the platform up for us to be able to provide those capabilities and automate the provisioning of logical circuits across our OCX. We have seen some initial interest, and we are encouraged by the future of what that can provide us, not only for intra-market, but also inter-market and some additional enhanced services as we go forward. So it's really just kind of setting the stage for the future.

Yes. So if I could just follow up on that. Are there any sort of use cases that you could maybe illustrate in a generic fashion, that would be driving that demand? And then a last one, perhaps for Jeff. You mentioned the PUE improvement. How would you translate that into kind of margin lift? So on a same-capital basis without the benefit of cross-connect growth, what kind of margin lift would you have seen the PUE improvement that you referenced in the script? Thanks.

Sure. So just to give you a couple of use cases, a great use case that we've already seen some demand for is where customers may be deployed in one market, but need access to a native cloud on-ramp in another market, where they can access our Open Cloud Exchange, logically provision that circuit to gain access within our Open Cloud Exchange on the other end in different markets, and get that native connectivity. So that's a pretty common request, I would say. Another typical use case would be where our customers are deployed in more than one market, where they may have an active-active type of environment or perhaps want to just load share their requirements between different markets. They can share those from a connectivity standpoint and be able to turn those up and down based on the amount of demand and amount of traffic that they need. So just a couple of examples for you. Now I'll let Jeff to answer your other question there.

Jonathan, I would say that in general, the PUE uplift contributed about 60 to 70 basis points to our adjusted EBITDA margin during the year. These are some of the operational improvements Paul has mentioned as we consider our strategy for 2018 and that our team is continuing to focus on as we move into 2019.

So Paul, in your opening remarks, you mentioned that the slower sales will likely be mostly remedied, I think, as it relates to additional capacity. And I'm just curious, is there something else beyond the additional capacity that you're focusing on to increase the sales effort?

Really just two things. One is the continuous improvement that we strive for that Steve mentioned, and they do a great job here, along with the rest of the team, diving into every opportunity from the previous quarter, lessons learned, how do we change our approach for future opportunities. And we've seen a lot of success from that continuous improvement. The other thing is what you guys have seen happen in this space since it was created is that you have different waves and cycles of user demand. One year, we see a lot of scale sales in our markets, and then those get filled up. Subsequent years, there's less of that. As those fill up, the cycle comes back. So I think having the capacity is a prima facie requirement for winning those deals. I feel good about what we see out there in the industry in general and pipelines.

Is there anything you could add vis-a-vis the cycle, where you think we are in terms of what types of customers have been slower to lease, which may be coming around in 2019?

So every side of the edge cycles are different than undifferentiated or hyperscale cycles. We had a significant edge cycle in 2015 and 2016. It started tapering off in 2017 and 2018, but we're seeing good signs going forward for that. That's not a promise of anything, but I do feel good about where we are in the edge cycle.

And then along those lines, I was curious regarding the interconnection revenue growth assumption or guidance of 8.3%. Is it obvious moderation from what you did in '17, but also from the pace in Q4 of 10.9% growth, down 8.3 for the year? Is there anything else embedded in terms of consolidation or churn that you could point to?

Yes. Jordan. I would say that all of those items you mentioned, whether it's some of the consolidation or just some of it associated with some churn, are all embedded into those estimates. But I think if you look at the past two to three years, you can see that it has decelerated by 200 to 300 basis points per year, largely driven off of volume growth. On the positive side, what we continue to see is customers that migrate from a certain volume to a higher volume product, which obviously increases price. So while you may not see volumes increasing from a per unit basis, overall pricing continues to go up as they migrate to a higher-priced product. We continue to see that, but all of that has been embedded into those estimates. As you point out, the midpoint is about 8.3%, with a range somewhere around 6% to 11% and we'll just see how we perform as we work through the year.

Okay, thank you. Hey, Jeff, just while I have you there. The commencements, I noticed you didn't provide any specific commentary for the year. Is there any number that's embedded at the midpoint here that you could point to beyond the backlog you talked about?

Yes. I mean the backlog, as I mentioned, you've got about $9.9 million, call it $10 million that we've got visibility into for next year. The remaining of those commencements are going to be largely dependent upon our leasing results, and some of that is dependent upon our construction deliveries. We’ll continue to update as we work through the year. In terms of trying to estimate that, I would just point to whatever our high-level revenue guidance is and our churn estimates and make some estimates on timing of leasing and commencements which should get you there based on the high-level revenue guidance we've provided. We'll continue to update you as we progress during the year.

First one is for Steve. You noted the revenue booked from new logos in 2018 was up 81% versus the prior year, which is a pretty substantial change. Can you talk about some of the specific initiatives you undertook to help drive those results? And were there particular selling points you found to be particularly effective in getting new enterprises on board?

Sure. Well, thanks, Nick. Thanks for the question. I mean, yes, there was one area that we did see some good traction on, I think, is an important aspect. We've mentioned it on prior calls as a key focus area for us even more than a year ago, where we wanted to really provide more diversity across our base, be less reliant upon our existing customer base for future growth, and just enhance the ecosystem. That's been the real impetus for this whole initiative. There has been a concerted effort from our sales, marketing, product teams, our channel team really across the board to ensure that we drive the right messaging, the outreach that we align our incentive plans, all of those kinds of things to motivate our teams toward rolling toward those new logos. So, I'd say it's really just a combination of all those things.

Okay, that's helpful. Then maybe one on Santa Clara specifically. I note that's historically been a very good market for you guys in part, because it's hard to bring new capacity online. It seems like there has recently been a spike in projects under development there with a lot of capacity in various stages of development. How do you see that influencing that market over the next couple of years? And are you confident you can get your desired returns on SV8 and the facility you plan on building on the parcel you just acquired?

This is Paul, Nick. We feel good about the Santa Clara market. There is always room for more participants in that market. But we like our campus ecosystem, the value that it drives, and the performance that it drives. That's why we are, despite those new participants in the market, continuing to look to expand that campus as I mentioned in my prepared remarks.

Paul, I think you mentioned earlier that LA3 construction has been delayed a little bit more. How does that impact your previous outline for 2020 growth as it relates to revenue, EBITDA, and FFO? Do you still feel comfortable returning to the double-digits there?

I do. And I don't expect this to be a significant delay, although it's a bit of a disappointment. And let me just be clear, I have a lot of sympathy for these power companies that have to manage these grids, and I'm guessing there's more sensitivity around this today than there was in the past. So I get it; they want to get it right, and we want them to get it right, and we're hopefully very close to resolving that and getting construction started in the first half of this year. So it shouldn't meaningfully move from where we were expecting previously, but it could be a quarter or two, maybe slightly more from what we had before.

Okay. And then, as far as you mentioned the positives from new logo growth, what are you seeing from existing customers? How are they growing their footprints and do you still feel confident about that side of the business?

It was a little bit lighter in 2018. I would refer you back to my comments earlier about cycles of demand in various areas. I think we'll see that be better in 2019.

Hi. I just want to get a little bit more color. I guess you have completed projects, pretty stabilized but completed in Virginia, New York, and D.C. It seems like they've been pre-leased, but the commencements haven't kind of come in. Can you give us a little more color on when they will start commencing and when we can see revenue coming from those projects? And I have a follow-up.

Some projects have already started generating revenue, while others will begin this quarter and the upcoming quarters. To provide more detail, VA3, 1A is a co-location asset and doesn't support scalability, which means it takes time to fully lease up, but it is currently leasing very well. DC2 was completed late in December and secured its first tenant earlier this month, making it a very new asset. New co-location assets usually experience a more consistent leasing process, whereas scalable assets tend to see more fluctuations in their leasing timeline.

And giving the development table for 2020 and knowing it's a little early. If you look at the megawatt numbers, they're relatively similar to the amount you're doing in 2019, should we expect there might be a little bit drop off in CapEx? But it really seems like the CapEx level for 2020 going into 2020 should probably be equal to 2019. Would that make sense?

Yes, just broadly speaking, if you look at the megawatts for 2019, we're looking at delivering about 19 megawatts, and for 2020, it's only 12. I think keep that in mind. I do believe 2019 would be elevated compared to where we think 2020 will be, at least as we sit here today. That may change based upon ultimate sales and absorption trends as we kind of work our way through 2019. Obviously, we'll provide some detailed color around that as we get closer to the end of this year. But I wouldn't expect it to be higher than what we're expecting for this year. If anything, similar or slightly down based on where we sit today.

Yes, thanks for taking the questions. Just a kind of back-to-the-scale location, that obviously continues to be challenged, and finished the year with roughly half the amount that you signed last year. Given your guidance for 2019 and the capacity expansion plans you've talked about, would you say the 2018 was the trough and we'll see scale colo higher in 2019? Here, we're just trying to understand your comments about new leasing opportunities given a lot of the construction and development plans you guys have. Then I have a follow-up.

So yes, obviously our guidance implies very strongly that we expect 2018 to be a trough. The requirement of increased sales is you have to have capacity to sell, and it has to be the right capacity in the right markets. Our trademark has been, as I mentioned in our comments, that we've been able to accommodate deployments from as small as a cabinet or sometimes even less all the way up to substantially larger requirements of 8 megawatts, 9 megawatts, and even occasionally a powered shell. We just didn't have anything like that kind of capacity to accommodate those types of opportunities in 2018, but we will in 2019 and continuing further into 2020.

And then maybe in terms of your new development and construction projects, at this stage, how conservative do you think your development completion timeline is? The one chart on 19 there seems that VA3 did slip to Q2. What were the factors there, and where do you see the major hurdles to hit your overall timeline on that chart? Thank you.

We strive to be straightforward based on the documentation we have. During the permitting phase, we rely on our advisors' estimates for how long it takes to obtain permits in specific jurisdictions. Moving forward, we'll categorize this as pre-construction and won't provide as much detail about permit timelines, as predicting them in current markets is very challenging. For construction, we base our expectations on the delivery dates outlined in the construction contract. However, these dates can change due to weather conditions, and we account for some allowances in our contracts. Additionally, unforeseen conditions at the site can cause delays. This is mainly why VA3 has shifted from late in the third quarter to around mid-second quarter; it's due to a mix of site and weather conditions in Virginia.

When we met with you guys at NAREIT in, I think, it was early November, it seemed like you guys thought pretty good about the pipeline for a scale deal at VA3 and that you're tracking various opportunities, and obviously, you didn't sign anything during the quarter. Curious if those opportunities are still there, they've gone elsewhere? And if the delayed opening had anything to do with it? And then also, as you look to get that anchor scale deal with VA3, how are you thinking about pricing? I know in the past, even though the pricing has come down, are you willing to do a deal at what you perceive now as the market clearing price or are you going to kind of hold in there and take the price that you want? Thanks.

Hey, Colby. This is Steve. I'll take a shot at this, and then Paul can chime in on anything that I might miss here. But I would just tell you, as far as the overall pipeline is concerned for pre-leasing, whether it be in Virginia or in the Bay Area, activity still remains strong. We are having multiple conversations, and we're still positive with the outlook. As far as any individual deals are concerned, not sure exactly which individual deals we're in discussion. I would tell you that even customer requirements, especially of that size, move all over the place based on their dates when they want them or whether or not they remain a real opportunity. So the overall market in the pipeline still remains strong, and we're optimistic about where we can go with it. As it relates to pricing, we feel like we've underwritten the asset to where we can still execute against it and meet the returns that we underwrote. The market clearing price, I would say, there's not necessarily a market clearing price. I think there's a lot of variables that go into what prices in any given market, but maybe especially in Virginia, as to what customers value, whether or not they're willing to pay for it, and how much scale they're looking to bring on board. So we feel like based on the value, the ecosystem, the scale and redundancy that we've built into the asset that we'll be able to meet the market.

Great. And then maybe just one for Jeff. You've taken your leverage target up from what has historically been sub-4 and this year it's 4.5, and in response to, I think, Richard's question, you noted that you expect CapEx next year to be down, but maybe similar. Once model would assume then that your leverage continues to go up as you go into 2020. Do you see that you just can't be taking up your leverage target further, or are there other actions that you're likely to take?

Yes, good morning, Colby. Those are items, it's going to be very dependent upon ultimately based upon our EBITDA growth in terms of where that leverage ultimately gets to. But more specifically, to your question, our leverage at 4.5, which we've messaged to, we continue to have those conversations with our Board when we meet and assess where we want to take that leverage, but that is one option that we would look at. There are other options that we need to consider based upon many factors in terms of the economics out in the marketplace. It's one lever we can pull, not the only lever. We just got to evaluate those with our Board as we work our way through the year and see how the lease-up is coming and ultimately the EBITDA growth associated with it.

Going forward, are we aiming for increased growth in the sales team, or is it primarily about having more inventory available to boost sales? Additionally, could you explain the recent increase in rent per square foot this quarter? What was the nature of the lease mix that contributed to this sequential uptick? Thank you.

Hey, Frank. This is Steve. As it relates to the sales headcount, we had a couple of open positions, a few open positions, I would say, coming out of 2018, that we now have filled in a couple of markets. For example, Chicago, where we have obviously new buildings coming on board, we expect to have better traction there and be prepared for that building coming online. We’re just adjusting a few heads here and there. Overall, I wouldn't expect any significant changes as far as total number of heads or dollars associated with them. As it relates to the rent and per square footage piece of it, I'll just give you my perspective, and then hand it off to Jeff. The uplift was primarily driven just from increased density that we saw in the fourth quarter, as we saw some of those smaller deals that were also more dense that drove the dollar per square foot up in that quarter.

Frank, the only thing I'd add to that is that Steve did make a reference in his prepared remarks that our pricing on a per kilowatt basis was consistent with the trailing 12-month average coming into the fourth quarter, and so that gives you a better sense for that density driving some of that increase on a per square foot basis.

Yes, hey guys. Maybe for Steve and tying on some comments that Paul had made earlier as well, but when you look at your completed pre-stabilized projects, there's nearly 150,000 square feet, it sounds like a large portion of that is not really colo scale available. So I was kind of curious what the percentage of that portfolio is leasable for scale colo? And then compare and contrast that to what's under construction of the full development pipeline, is all of that colo scale available? And then the third part of that question, just to be more confusing, is the 5,000 square feet that you use for the bottom of the scale colos portion of your business, is that too small? Is the scale colo today for you needing to be substantially larger than the kind of low end of the bar?

Yes, I'll give you a little bit of color on the scale versus retail. We don't necessarily break it out specifically as far as which sites are suited one for the other. I would say in general, Paul kind of mapped this out earlier in his comments around our initial phase at VA3, which is primarily geared toward colo; the DC2 facility, which recently came online, is more colo-oriented. Obviously, the LA1 15th floor that we'll be building out this year will be more colo-oriented. As you look at our newer builds, it's probably easier to look at them as being able to accommodate either retail or scale or wholesale, just in the modular fashion where you have a fresh floor plate that you can grow from. So, I think if you look at the next phase of VA3, for example, SV8, LA3, and CH2, those all will be able to accommodate both retail and scale or hyperscale.

Dave, this is Paul. I just would point you to page 13 of the supplemental. You can kind of get the picture that it's a combination of tight occupancies and the market spread of where we had capacity. Like Santa Clara, 97.3% occupied, I think you can easily interpret what that means for the ability to handle scale there. NY2, 77% occupied, which means more capacity to handle scale, but a less strong market, and I think that illustrates the dynamic we've been working through in 2018 and will be solving in 2019.

How do you guys think about land holdings in the long term? Obviously, your product is dependent on having good locations, and a lot of the markets you are in are tighter markets in terms of available supply. So how should we think about how you're positioning the land bank for the longer term? And maybe are there any new markets that are potentially on the table?

The new parcel we secured in Santa Clara demonstrates our strategy. As we have mentioned in previous quarters, we need to take a more proactive approach in acquiring, entitling, and preparing land for development in key markets. We quickly positioned the SV8, CH2, and VA3 lands for development, and we want to avoid falling short like that again. We are actively searching for additional opportunities to expand in all our major markets. The availability of land in these areas can be unpredictable, so we must consistently engage in the market, seek off-market deals when possible, and ensure they meet connectivity, power, and proximity requirements. The new land we have under contract in Santa Clara serves as a solid example of this. I apologize for forgetting the second part of your question.

Yes, are there any new markets that could be on the table? And maybe just your thoughts on the U.S. like are you always going to be only in the U.S., or are you open to expanding internationally?

So let me try to answer this one. When we sit down with our major customers, really all the way down to the top 20 or 30 customers, the primary thing they value about us is that we keep expanding to these infill edge markets, where we exist, because they have tremendous needs, and they foresee significant needs in the future, and they realize it is hard to proactively expand in those markets. That doesn't preclude us from going into other markets when we can find the right conditions. Believe it or not, we do consistently look at opportunities, but we just haven't found the right ones. I think that would be more of a domestic thing if and when it happens; international may happen in the future; I wouldn't rule it out, but it's not something that we're focused on at this time.

I have a few questions for you. First, you mentioned the different waves and cycles in the industry. I'm interested in how you plan to position CoreSite strategically based on your market outlook. The second part of my question is about your thoughts on scale demand. How does your perspective on this threat compare to where you were at the beginning of 2018? Do you feel more positive or negative about scale demand now? Lastly, could you provide any comments on build costs and what trends you are observing in that area? Thank you.

We are just as optimistic about long-term, intermediate-term, and even short-term demand now as we were at the start of 2018. Using a hockey analogy, we want to position ourselves strategically because we know that’s where the opportunities are. Having sufficient capacity in key urban markets is essential for low latency and high-performance applications for both consumers and businesses, along with Internet peering points, and it forms a key part of our strategy, which has been effective so far, and we remain confident in it. Regarding build costs, they have indeed increased more than general inflation over the past couple of years, with significant variability from market to market. Labor costs are the most unpredictable factor. Our design team regularly updates our designs to enhance efficiency and reduce costs to help manage these expenses. Our construction management team works diligently to keep costs under control, and our equipment procurement team continually navigates our supply chain to mitigate some of these increases, but we are still experiencing some inflation in construction costs.

Given the comments that leasing was held back by a lack of contiguous space, can you provide a little more color on frictional vacancy for the current portfolio?

I'm sorry, on what vacancy?

Is there a natural state for frictional vacancy in the low '90s right now? It's hard to exceed that due to the lack of contiguous space.

I mean, really, again, I'd go back to my response to an earlier comment. It really varies by market, but yes, there are markets where we are so tight that. I think when you get above 91%, 92%, you see this a lot more where you just don't have those contiguous spaces, where you can handle 1 megawatt deployments, much less 2 megawatts, 3 megawatts, or 4 megawatts.

The only thing I would add to that is that I think our model does lend itself, in some cases, to tighten that up on occasion because we do sell smaller deployments that, in the right floor plate, will allow us to fill that in tighter with smaller deployments, whether they are networks, smaller enterprises, or otherwise. Our model, as opposed to being just focused on hyperscale, also focuses on trying to bring in more logos and get more diversity across our install base, so that we can fill in some of those other pockets.

So coming into Q4, you'd expected I think 100 basis points of churn, extraordinary churn from the customer that churned about 70 basis points in the prior quarter. Net-net, the turnover came in lower sequentially, and I'm looking at guidance which is 6% to 8% for 2019. I think you've previously set 2% to 2.5% per quarter through '19, which would be 8% to 10%. So what's the color there on how your perspective on churn has changed? Secondly, if you could provide the number for new logos, which I think we have in prior quarters. And I have one follow-up.

I saw your note this morning, so I appreciate you raising the question. Just to give you some color on the fourth quarter, first of all. As Steve alluded to, our churn in the fourth quarter was better than anticipated. We had a good outcome ultimately with the customer we were renewing in the fourth quarter. While they did churn some level of their space, it wasn't as high as what we anticipated. I think overall, that was good news. In terms of 2019, the commentary I provided last quarter, which is consistent with where we sit today, is we would expect elevated churn from 2% to 2.5% in each of the two first quarters, so just the first half of the year. The second half of the year, we would expect that to return to normal levels. Just factor that in; that should explain your question related to the guidance for the year.

Thanks. The number of new logos is encouraging. Can you provide information about the new logos?

What was the question around the number?

I think we've had a number, it was 27, 28 in the prior quarters.

Yes, sure. So it was 32 this last quarter.

Great. With DC2 launching at the end of the fourth quarter, rental expenses remained fairly stable compared to the previous quarter. Should we anticipate an increase in the first quarter, and what is your outlook on EBITDA margin for the year?

Yes, great point Robert. The DC2 did come online very late in the fourth quarter. So you will see an uptick here in the first quarter associated with that rent for that particular data center. In terms of the adjusted EBITDA margins, if you look at our guidance, you can see that it's very flat compared to where we ended 2018. I think the implied EBITDA margin is at 54.4%, right in line with where we ended 2018.

Just a quick follow-up on the revenue growth guidance of 7.5%, Jeff, the total operating revenues. Do you have a breakdown of rental revenue versus power revenue growth?

I don't, George. What I would utilize, though, you can see where those percentages came out for the full year of 2018. I would use that same pro-rata allocation as you will look at your estimates for 2019. I don't think those are going to be moving meaningfully.

Okay. So a similar split, I think, rental revenues were up 11% compared to 2017 for power, something like that to a similar spread?

Yes.

Okay. And then I was just coming back to the interconnection question. It sounds like maybe that cycle of equipment upgrades may be slowing a little bit. I'm not sure if I read that correctly in your commentary, but what does longer-term growth look like in terms of interconnection revenue in your mind sort of three to five years?

Yes, I think just maybe to clarify as it relates to some of the items that were impacting churn for interconnection growth, I'm sorry in 2018. We alluded to some of the consolidation taking place inside the portfolio as one aspect of it, and the second being some of the migration from 10 gig to 100 gig. On the second, the migration, there’s only a minority of customers that really is applicable to. It’s hard for us to ultimately forecast when it makes sense from their perspective based on the volume of traffic they are moving through those switches. It is something that we watch and try to make estimates on. I don't know if it's going to moderate for next year or not, but it's something we continue to watch. But just keep in mind, I think it really is only applicable to a minority of our customers that we haven’t based today. I missed the last part of the question, the second part.

I am considering what organic growth will look like in the interconnection business over the next three years. Are we expecting to moderate by 200 basis points per year or will we stabilize around 8%?

Yes, I’m not sure I can provide a specific number right now; we need to assess how the interconnection revenues will develop. What I can say is that it should be closely linked to the volume of kilowatts we anticipate, which should positively influence pricing. We need to grasp this better to generate accurate estimates.

The only thing I'd tell you, Jonathan, is if you look at the broader industry and technology as a whole, interconnection is not reducing; it's accelerating, if anything. The amount of interconnected things out there just continues to grow. I think that provides a good opportunity for the industry and for us as well and how we monetize that. We have physical versus logical interconnection, I think, is playing out as we speak and that's why we've made some of the investments around the OCX and providing the additional capabilities there. To answer maybe an embedded question you had around SDN eroding any pricing around physical cross-connects, we have not seen that as of yet.

First, thank you all for being on the call. I appreciate the interest and the good questions. One of the analyst reports that came out overnight mentioned they were glad to have 2018 behind us for CoreSite. I certainly appreciate that sentiment because it has been a year of transition from utilizing our native capacity since the IPO to entering a new development phase for the company. I genuinely feel very good about what the team accomplished in 2018. I'm not referring to myself; I probably have a longer list than any of you of things I could have done better. However, the team achieved remarkable progress in 2018 in building blocks for future growth. We made significant construction progress, entitled new land, and began permitting and development processes while achieving operational excellence and enhanced efficiency, particularly in power usage. There are many other smaller advancements in IT and other areas that will collectively be crucial as we move forward. Overall, it was a year of hard work and dedication, but I feel good about what the team achieved, and I'm excited as we move forward into 2019 and beyond. Thank you very much for your interest, and we'll talk to you next quarter.