CyrusOne Inc.
Q4 2016 Earnings Call Transcript

Published: 24 Feb 2017

[Operator:]

Good day, and welcome to the CyrusOne Fourth Quarter 2016 Earnings Conference Call. All participants will be in listen-only mode. After today's presentation, there will be an opportunity to ask questions. Please also note that this event is being recorded. I would now like to turn the conference over to Mr. Michael Schafer. Please go ahead.

[Michael Schafer:]

Thank you, Andrea. Good morning, everyone, and welcome to CyrusOne's Fourth Quarter 2016 Earnings Call. Today, I am joined by Gary Wojtaszek, President and CEO; and Diane Morefield, CFO. Before we begin, I would like to remind you that our fourth quarter earnings release along with the fourth quarter financial tables are available on the Investor Relations section of our website at cyrusone.com. I would also like to remind you that comments made on today's call and some of the responses to your questions deal with forward-looking statements related to CyrusOne and are subject to risks and uncertainties. Factors that may cause our actual results to differ from expectations are detailed in the company's filings with the SEC, which you may access on the SEC's website or on cyrusone.com. We undertake no obligation to revise these statements following the date of this conference call, except as required by law. In addition, some of the company's remarks this morning contain non-GAAP financial measures. You can find reconciliations of those measures to the most comparable GAAP measures in the earnings release, which is posted on the Investors section of the company's website. I would now like to turn the call over to our President and CEO, Gary Wojtaszek.

[Gary J. Wojtaszek:]

Thanks, Schafer. Good morning, everyone, and welcome to CyrusOne's Fourth Quarter 2016 Earnings Call. 2016 was another record year for CyrusOne. As you can see in the earnings presentation on slide 21, our full year revenue of $529 million, adjusted EBITDA of $279 million, and normalized FFO per share of $2.66, each exceeded our 2016 guidance by 6% to 7%. Over the course of the year, we signed nearly 1,500 leases, totaling 642,000 square feet of raised floor and 92 megawatts of power, nearly doubling 2015's record pace and power totals. The deals we signed in 2016 will generate nearly $150 million in annualized GAAP revenue and represent more than $1.2 billion in total contract value. To put that in perspective, in 2015, we generated nearly $400 million in revenue. So the bookings in the past 12 months represent total revenue growth of almost 40% over the prior year. Further, we'd like to highlight that while we don't have all the fourth quarter leasing numbers yet, we're punching well above our weight in sales velocity compared to the rest of the industry. Through the first three quarters, we signed more than one-third of all publicly reported data center REIT leases on a revenue basis, which is significantly above our respective market share. This demonstrates the value of investing in a robust and experienced sales force that can drive growth across a very broad product offering. I am particularly excited about our recent announced acquisition of the Sentinel Data Centers, which has key strategic and financial benefits, and importantly for our shareholders is expected to be immediately accretive when closed. Beginning with fourth quarter performance on slide 4, revenue of $137.4 million was up 21% over the fourth quarter of 2015 and adjusted EBITDA of $73 million was also up 21%. Normalized FFO of $0.68 per share was up 11%. We leased 74,000 colocation square feet and 9 megawatts of power, totaling $19 million in annualized GAAP revenue, which is down 56% from the prior four quarters' average, but up 37% from the average of the eight quarters prior to that. Also, pricing this quarter increased 38% versus the trailing four-quarter average. We also added two more of the largest cloud companies as customers in the fourth quarter and now have nine of the 10 largest cloud companies in our portfolio, up from three just a few years ago. Our backlog stood at nearly $60 million as of the end of the quarter, representing nearly $470 million in total contract value. We're also announcing an 11% increase in the quarterly dividend for the first quarter to $0.42 a share. Lastly, as I mentioned before, we recently announced the acquisition of two data centers from Sentinel. I'll talk more about the benefits of this transaction in a minute. Slide 5 highlights our track record of strong growth with revenue, adjusted EBITDA, and normalized FFO per share growing at compound annual rates of 26% to 30%. In 2016, the rate of growth actually accelerated, with revenue growing at 33%, which is the fastest annual growth we have experienced since our IPO. Adjusted EBITDA also grew at 32% in 2016. As we look ahead to this year, we expect this trend to continue, which Diane will discuss later. As we have shown previously, with this growth, we're continuing to strengthen our cash flow streams by signing longer leases with escalators, further extending the lease duration of our portfolio. This past year, the average lease duration was almost nine years, which is indicative of the maturation of the industry as customers are willing to engage with us for longer term commitments. Slide 6 provides an overview of the size, duration, and type of bookings we have generated over the past few years. There are a few trends to point out here. First, the amount and size of the bookings have continued to increase over the past few years as the market matures and customers are willing to transact with us over a longer time period. Second, over 90% of the leases signed this year included escalators at an average of approximately 2%. As of year-end 2016, 58% of our portfolio now has built-in lease escalators, up from less than 10% at the time of the IPO. Also, while our recent success with the hyperscale cloud providers has been impressive, I would point out that retail customers make up 93% of our customer base. Lastly, we booked $8 million of annual interconnection revenue in 2016, which is a record volume of IH bookings for us, up approximately 45% compared to last year. We have always believed that the value we bring to our investors as demonstrated by our sales velocity and returns we generate is attributable to the ability to sell mixed data center solutions ranging from a single cabinet up to a 30 megawatt deployment, which we do very efficiently. We do not have an either/or strategy of selling to either retail or wholesale customers. As we have said, no customer has ever come to us and asked for our retail product or our wholesale solution. Rather, we sell data center solutions to close to 200 Fortune 1000 companies who have varying data center needs. Also, the investment we're making in our sales and marketing organization is very attractive when considering the growth rate of our company as our sales and marketing costs are really being made with an eye towards the company we are building. That said, the amount we spent on sales and marketing is significantly below what a traditional real estate company would pay in commissions to third party real estate brokers. As I mentioned earlier, this year, we booked $1.2 billion of total contracts. If we sold through traditional real estate broker channels, we would have incurred commissions of roughly $48 million, which would have been 65% more expense than the $17 million of sales and marketing expense we incurred for our staff and the $12 million of commissions we paid, which will all continue to generate significant benefits to us over time. Slide 7 lists the key benefits of the recently announced Sentinel acquisition, the first being the addition of a market in the Southeast and further diversification of the portfolio. As we have stated for a long time, we planned on establishing a presence in the Southeast as we believe that having a broad geographic product offering is becoming increasingly important for customers as their IT applications become more distributed. Approximately 65% of our revenue is generated from customers in multiple locations, and 76% of our MRR signed in 2016 came from existing customers. This has very relevant connection with our acquisition of the Sentinel assets, since more than two-thirds of their roster of nearly 30 customers are new to CyrusOne, including five new Fortune 1000 companies. Given our success in doing repeat business with our customer base and in multiple locations, we can capitalize growing our revenue with this new set of customer relationships, while selling to two new locations to already existing customer base. We not only acquired two great data center assets, but a very valuable set of customer relationships as well. This deal also increases our penetration in the healthcare vertical, which we believe has significant growth potential in the coming years, with the digitization of healthcare data, continued technological innovation and compliance with regulatory requirements. We are acquiring a portfolio of long-term leases with escalators with a weighted average remaining term of more than eight years and only 3% of rent due for renewal on the next three years. These are high credit quality customers with approximately 70% of the rent generated from companies with an investment grade rating. Another highlight of the transaction to point out is that the North Carolina location will represent the lowest power cost in our portfolio and one of the lowest in the country, which is the reason why Google, Apple and Facebook have all located data centers in that region of the country. Both locations will be interconnected with our other existing data centers. These data centers will be owned by CyrusOne, increasing the contribution of NOI from owned properties to the entire portfolio to nearly 80%. We expect this percentage to increase over time as we continue to build our development pipeline, which are all wholly-owned facilities. Lastly and very importantly, a strategic consideration in our evaluation of M&A opportunities, this transaction provides us with significant potential upside. Currently, there is about 35,000 square feet of raised floor and 8 megawatts of power capacity that is either available or can be brought online with minimal investment. Additionally, there is built-out shell capacity in North Carolina that will double the size of the facility and which can be fitted out in about 10 weeks. Also, a shell is currently being built at the New Jersey facility which will double the size of that facility. In total, we will have an additional 230,000 square feet of raised floor and 37 megawatts of power available to sell. Slides 8 and 9 show the revenue contribution by market pro forma for the acquisition. We have significantly diversified the portfolio geographically through successful organic expansion into new markets as well as acquisitions such as this one. As you can see, we have a balanced mix with our biggest market, Dallas, only accounting for 17% of revenue. As we deliver pre-lease capacity in the markets such as Chicago, San Antonio and Northern Virginia, the market mix will continue to diversify further. As I mentioned earlier, the contribution from the healthcare vertical will more than double as a result of this transaction. Moving to slide 10, the weighted average remaining lease term of more than eight years in the Sentinel Data Centers extend the remaining lease term of CyrusOne portfolio to nearly five years, which is twice the length of the remaining lease term at the time of the IPO. Only 3% of the rent in the Sentinel portfolio is due for renewal in the next three years, and 70% is under contract for more than seven years. Approximately 95% of contractual rent includes escalators with a weighted average rate of nearly 3%. This will increase the percentage of rent in the CyrusOne portfolio that includes an escalator to nearly 60%, up from less than 10% at the time of the IPO. Slide 11 shows the growth opportunity associated with this transaction. As we have shown previously, most recently in our Northern Virginia development, we generate mid-teens development yields on our facilities. In this case, we're acquiring a cash flow stream locked in for more than eight years, while gaining immediate access to a new market without incurring the typical land, entitlement and development risk that is required when entering a new market. As also noted, we are adding new customers and getting upside of the lease-up and expansion opportunities. The chart on this slide shows the yield progression over time at these locations. The underwritten going-in yield was approximately 7%, with lease-up of existing inventory in the additional capacity that can be developed relatively inexpensively. We expect the yields to increase to approximately 9%. Over the longer term, we expect the yields to increase to the 12% to 15% range as we develop and lease-up the nearly 40 megawatts of total additional power capacity at the two locations. This yield progression is something we see in all of our investments and, most recently, with last year's acquisition of the Chicago Mercantile Exchange data center in Chicago. Slide 12 shows the yield progression at this location in Aurora. This transaction had a very similar profile to Sentinel, including a near term opportunity to lease-up additional capacity that could be built out at a very low cost and a longer term opportunity to develop additional capacity on the adjacent 15 acres of land that was included in the acquisition. Here, the underwritten going-in yield was approximately 8%. Based on fourth quarter annualized NOI plus the impact of the fully pre-leased second phase within the building, the yields on this investment will increase to approximately 12%. So just one year after the transaction, we have leased-up almost all the additional capacity within the building and increased our going-in yield by almost 50%. With the development of the additional land, we expect to ultimately generate a yield in the 13% to 16% range, which is consistent with the yields we earn across our portfolio. In conclusion, I couldn't be more pleased with the position we are in as we begin 2017. The backlog will deliver nearly $60 million in new revenue. Our sales funnel remains very strong. We have capacity to expand in our markets, and we have a strong balance sheet. I believe our sales velocity combined with the high yields we earn on our capital will ultimately benefit our shareholders, and we are well positioned to capitalize on these investments and our existing development pipeline in 2016 – 2017. I will now turn the call over to Diane who will provide more color on our financial performance and cover our guidance for 2017. Thank you.

[Diane M. Morefield:]

Thanks, Gary. Good morning, everyone. As Gary mentioned, 2016 was a record year for CyrusOne and as slide 14 shows, the strong year-over-year financial performance continued in the fourth quarter. Revenue and adjusted EBITDA each grew 21%, driven by continued strong leasing across our markets, the impact of the acquisition of the CME data center in Chicago earlier in the year and our interconnection product lines. Churn for the quarter was 2.2%. For the year, net of company-initiated churn in the second and third quarters, churn was 7.4%, slightly lower than our expectation of approximately 8% mentioned on our last quarter call. As you know, we report churn in total for declines in rate, leases that don't renew and early lease terminations. Some companies only report term churn. So our churn numbers by definition may be higher or appear higher than some of our peer group depending on how they define this metric. There isn't really a necessarily right or wrong way to report churn. We're just highlighting that the comparisons might not be completely consistent. Turning to slide 15, NOI increased 24%, primarily driven by a 21% increase in revenue. Adjusted EBITDA was up 21% driven by NOI growth, partially offset by higher SG&A. As Gary mentioned, we have been adding experienced talent, particularly in our sales organization, and investing in critical systems to manage the business for our significant continued growth. We have a fully integrated internal sales team, and typically less than 10% of our leasing each quarter is generated through any indirect sales channel representing our customers. As Gary also mentioned, this is a much more cost efficient model compared to other types of REITs that are heavily weighted to tenant reps that charge significant commissions. Normalized FFO increased 28%, largely consistent with our revenue and adjusted EBITDA growth. However, normalized FFO per share grew 11% as a result of the impact of additional equity issued to fund our growth and manage our leverage metrics to maintain a strong balance sheet. The chart at the bottom of slide 15 shows the impact of the adjustments to normalized FFO to calculate AFFO for each quarter in 2016. In the fourth quarter, normalized FFO was up 2.9% sequentially, while AFFO was up 21.9% sequentially. The difference between normalized FFO and AFFO decreased significantly in the fourth quarter compared to the third quarter, driven primarily by a reduction in the impact of deferred revenue and straight line rent adjustments. This was primarily the result of the completion of brief ramp up periods associated with some of our larger leases to accommodate the phase-in of these customers' power requirements. These are generally long-term build-to-suit leases with rent escalators that are generating targeted mid-teens development yields and very attractive NPVs. As we said, we would expect there to be a lot of continued volatility in our AFFO, driven by the size and velocity of the larger deals we are doing for hyperscale customers. But as you can see, once the customers have completed their ramp up, the difference between AFFO and NFFO is more minimal. It's similar to an office development in which a large anchor tenant gets a free rent period, but once that period ends, the differential between the cash and straight line rent is a more normalized run rate. Moving to slide 16, we have significantly expanded our footprint over the last year to accommodate the demand we are seeing across our markets and our stabilized portfolio is 92% utilized. Our pre-stabilized data halls in Dallas, Houston and Austin are still in their lease-up phases and as we fill those data halls, they will supplement the growth coming from our pre-leased development projects. Slide 17 summarizes our development pipeline as of the end of 2016. We currently have development projects in Northern Virginia, San Antonio, Chicago and Phoenix that will deliver nearly 500,000 colocation square feet and 88 megawatts of power capacity. These are some of our strongest markets, where we are essentially out of inventory as of the end of the year. In addition, 72% of the pipeline is pre-leased. The trend toward pre-leasing in the last year has substantially de-risked our capital investment. In Northern Virginia, we have seen the most demand. As of year-end, we had a fully pre-leased building totaling 79,000 square feet of colocation and 15 megawatts that have since been completed and is now operational. We are developing an additional 27,000 colocation square feet and 9 megawatts at our Sterling IV location. This is also fully pre-leased and we expect it to deliver in the second quarter. As a result of all our success in this market, we are building another shell on the campus and expect to deliver the initial phase totaling 81,000 colocation square feet and 12 megawatts of power capacity in the second quarter. This building will have an additional 459,000 square feet of powered shell available for future development. Also as of year-end, we had a fully pre-leased building consisting of 132,000 colocation square feet and 24 megawatts in San Antonio that has since been completed and is now online. In Chicago, subsequent to the end of the fourth quarter, we completed the second phase of Aurora I and as mentioned on last quarter's call, this is fully pre-leased. Given the strong demand in this market, we have begun construction on a new building and expect to deliver the first phase totaling 77,000 colocation square feet and 10 megawatts of power capacity in the second quarter. This building will also have an additional 272,000 square feet of powered shell available for future development to be developed in phases. Lastly, in Phoenix, we are constructing two buildings, each expected to deliver in the second quarter. The first will deliver 73,000 colocation square feet and 12 megawatts and is fully pre-leased. We are also constructing a 185,000 square feet of powered shell based on the strong demand drivers we are seeing in this market. Once these facilities are completed, our portfolio, including the data centers acquired from Sentinel, will be 75% larger than it was just at the beginning of 2016. The land and powered shell we have available for future development will allow us to more than triple the size of our footprint to nearly 8 million colocation square feet across the portfolio. Our model of developing our data center campuses in phases, based on pre-leasing and our visibility from our sales funnel, significantly de-risks our capital spend and manages our capital outlay throughout the year. Moving to slide 18, our balance sheet remains very strong. Net debt to adjusted EBITDA as of the end of the fourth quarter was 4.3 times and adjusted for the impact of the Sentinel acquisition it is 4.7 times. In November, we expanded our unsecured credit facility to $1.55 billion, including a $1 billion revolving credit facility. We also extended maturity dates and reduced borrowing rates. We have a weighted average remaining debt term of 5.3 years and no near-term maturities. Available liquidity at year-end, adjusted for the acquisition, was nearly $500 million. Moving to slide 19, as Gary mentioned, we are increasing the dividend for the first quarter of 2017 to $0.42 per share, which is an 11% increase over the 2016 quarterly dividend rate. This represents an annualized yield of 3.5%, based on our closing stock price on February 21. The current dividend is up more than 160% compared to 2013. More than 85% of our 2016 payout was a return of capital, providing an additional after-tax yield for investors with taxable income. In considering how to maximize value to our shareholders, we target our dividend payout ratio, to retain a reasonable level of cash, to redeploy to our capital expenditure budget which fuels our continued growth. We balance that with paying an attractive cash dividend to return a portion of our free operating cash flow to investors as well. So we try to be very thoughtful in determining the optimal allocation of cash to create attractive shareholder returns. Turning to slide 20, our revenue backlog at the end of the year was $59 million, representing nearly $470 million of total contract value. As shown on the bottom chart, we expect approximately $47 million of the backlog to commence by the end of the first quarter and the remainder to commence in the following two quarters. This gives us good visibility into growth for this year. Slide 21 highlights our guidance for 2017. We're track to close the Sentinel Data Centers acquisition on February 28 and this timing is reflected in our outlook. The guidance mid-point for total revenue reflects a 27% year-over-year increase. Excluding the impact of the Sentinel transaction, revenue growth from our existing portfolio would be in the upper-teens range. We anticipate revenue churn towards the lower end of our assumed 6% to 8% annualized range. The adjusted EBITDA guidance mid-point represents a 30% year-over-year increase. The implied margin, assuming the mid-points of the revenue and EBITDA guidance, is 54.3%, approximately 1.7 percentage points higher than the 2016 margin. The estimated margin expansion is driven primarily by a decrease in our SG&A costs as a percentage of revenue as we continue to scale the business with less marginal new overhead. The guidance mid-point for normalized FFO per share reflects a 9% year-over-year increase, which is below the increase in our other financial metrics for 2017. We currently have a mix of 70%/30% floating versus fixed rate debt pro forma for the funding of the Sentinel acquisition with our $211 million equity forward which has been drawn and a $300 million borrowing on the revolver. We intend to issue fixed rate debt sometime in the near future to more permanently finance the debt portion of this acquisition and repay the revolver. Throughout the year, we have made assumptions for increases in interest rates which affects both floating and fixed rate assumptions and we plan to add additional fixed-rate debt over time, which flows directly through NFFO, but we believe is prudent balance sheet management. We have no near-term needs to issue equity beyond the forward. We're managing our balance sheet leverage in the four to five times range and comfortable periodically going over five times with plans to bring that leverage back in line to five times during the right windows. We do have an ATM that can be utilized to dribble out equity, but again, I want to emphasize that we feel good about our balance sheet position and leverage near term. In addition, we are evaluating alternate sources of capital recycling that also could help fund our ongoing development pipeline. We expect capital expenditures to be in the $550 million to $600 million range, which is similar to our actual CapEx spend in 2016. In closing, we are really thrilled with our 2016 results, including very strong financial and operating performance as well as a record year for leasing. As always, we remain laser-focused on providing excellent customer service in order to continue to expand with our customers to meet their demand for additional space. We will continue to grow the company, while prudently managing our balance sheet and risk profile in order to maximize shareholder value. As many of you know, my background is over 20 years in the REIT industry, with a variety of REIT product types. I have to say I've never seen the opportunity for growth at such compelling, unlevered returns on invested capital compared to what we are achieving and continue to project to achieve here at CyrusOne. It is a very exciting time to be a part of this company and the overall data center space. Thank you for listening, and this concludes our prepared remarks. Operator, please open the line for questions.

[Operator:]

We will now begin the question-and-answer session. Our first question comes from Colby Synesael of Cowen & Company. Please go ahead.

[Colby Synesael:]

Great, thank you. Just a few questions. One is, I think you may have said, Dianne, in your prepared remarks that you guys are considering recycling assets. And if that's correct, I was wondering if you can be a little bit more specific about what you're thinking there both in terms of assets as well as potential proceeds? And then also I know you said that you're feeling comfortable not having to raise equity. But in the company's long-term guidance, you guys have assumed you continue to be acquisitive. If the company was to do other acquisitions this year, do you think equity would have to be raised to fund those? Thank you.

[Diane M. Morefield:]

Okay. So let me take your first question, Colby. It's very preliminary, some of the ideas that we're looking at for recycling capital. Obviously, other REITs from time to time sell non-core assets or more mature assets or do joint ventures. There's certainly strategies that can raise capital. Again, too premature to give you anything definitive or an amount or anything like that. But suffice it to say, we're looking at other alternatives that would help us redeploy capital into our high-return development pipeline. The guidance we gave for 2017 does not include any acquisitions other than Sentinel. So to the extent there were future acquisitions, we would have to evaluate at that point how we would fund it with a combination of debt and equity.

[Gary J. Wojtaszek:]

Hey Colby, Gary here. Just to provide a little more color on the other sources of capital, what we've been seeing in the market over the last two to three quarters, in particular, is there's a significant amount of interest from long-term investors, particularly pension investors that are looking to deploy capital in the data center business. I think people have looked at this for several years now and are comfortable that this is an asset class within the real estate group that is here, it is growing and they want to participate in that. And I think what you'll see in a lot of the different M&A deals that are being announced is there is an influx of this longer term pension capital into some of these other assets that will be coming to market or be announced. The upside of that is two things, I think, for consideration. One is I think it just lends credence to the idea that this is a real estate asset and it will be around for a really long term. And the other thing that goes aligned with that is that those type of funds are looking for really long-term returns and they are chasing assets at really high valuations. And we've been approached by a number of different folks that have come to us with interest in working through us to take advantage of what we're able to do as we grow our business. So it provides a little more color to what Diane had mentioned, but the opportunities are pretty attractive because they're willing to pay rates that are generally above where we would buy an asset at. And so if we can partner together, it would be a win-win for both of us.

[Colby Synesael:]

So it sounds like you guys are kind of contemplating potentially doing some form of JV if the opportunity presents itself?

[Gary J. Wojtaszek:]

Yeah, yeah.

[Colby Synesael:]

Thank you.

[Operator:]

Our next question comes from Jonathan Schildkraut of Guggenheim. Please go ahead.

[Diane M. Morefield:]

I don't know whose name is harder to pronounce, yours or Gary's.

[Jonathan Schildkraut:]

Well, I'm two for two this morning with good pronunciations on my name, so.

[Diane M. Morefield:]

You should go by Jonathan S.

[Jonathan Schildkraut:]

I'll take that under advisement. So listen, I guess a couple of questions here. Gary, maybe for you, overall leasing velocity in the sector with you and DFT reporting this morning was down pretty substantially both sequentially and on a year-over-year basis. Admittedly relative to your base Cyrus had the strongest leasing result. And I was wondering if there was something in terms of the way that deals flow into the market, whether it was the pull forward – whether there was a pull forward early in the year or a push out of some late-year deals, election cycle, that would somehow explain some of the variability in leasing? And then if you could tie that back into your expectations for 2017, is the $20 million a quarter number still the right number? And then I have a follow-up question. Thanks.

[Gary J. Wojtaszek:]

Yeah. Hey, Jonathan. So a couple of things. So as we mentioned on the last quarter, we're trying to bring people down to realistic leasing assumptions, which is why we said people should assume basically $20 million a quarter for us. And that's basically what we did this quarter. And that for reference is below where we were this year given the record deals, but it's probably about 20% higher than the two years prior to that. So we definitely have a leasing trend that is up from where we were tracking, putting aside some of these really large, super hyperscale deals. That said, what I would say is that our funnel as of the end of the fourth quarter was about 30% larger than where we ended at the end of the third quarter, and was about 50% larger than where we were at that same point versus last year. And so we're looking at like an incredibly large funnel, and we feel really good about it. But I think what a lot of people I don't think fully appreciate it, and we tried sharing some of this on the last quarter's call, was that we have been effectively sold out in Northern Virginian, in San Antonio, in Chicago, in Phoenix, which is our biggest, strongest markets, for a while now. And so while we've brought on some additional capacity to meet that demand, we're still effectively sold out and won't have capacity in those markets until the third quarter of this year. So we're tracking a really large funnel. We feel just as good now as we did last year, but I think in terms of the assumptions about bookings a quarter of $20 million, that would realistic. And I would expect that there's going to be quarters above that when a lot of these kind of fall in place at the same time. But for modeling purposes, I think we're trying to make sure everyone got down to that $20 million per quarter number. So that's a more realistic expectation.

[Jonathan Schildkraut:]

Great. That's very helpful. And my other question has to do with some of the equipment revenue that you've called out in some of the prior quarters. You guys didn't mention it in the release or in the deck, and I'm just wondering if there was anything to highlight there? And then separately, in terms of those equipment revenues as we think about them, were these actually capital contributions from some of your hyperscale customers that went towards your build costs or were they something else? Thanks a lot.

[Gary J. Wojtaszek:]

Yeah. So in our guidance, we've assumed a couple of million dollars of this for next year, but nothing material relative to what we saw in 2016. And generally speaking, these are distribution- type capabilities that our customers are paying for directly. So we'll make a 20%, 25% margin on them, it's a one-time in nature basis, but that would be in lieu of us actually deploying that capital in the solution and charging them an appropriate amount of rent for that. So depending on what customers are trying to do in terms of managing their budget, sometimes they prefer we buy that and roll it in, and deploy it, and install it, and roll that into our overall lease rate. Other times, they want to own it outright, and they would pay for the separately, and that would just be a sale with a cost of goods implications for us.

[Jonathan Schildkraut:]

All right. Great. Thanks so much for taking the questions.

[Gary J. Wojtaszek:]

Sure.

[Operator:]

Our next question comes from Frank Louthan of Raymond James. Please go ahead.

[Frank Garreth Louthan:]

All right. Three for three on pronunciation.

[Gary J. Wojtaszek:]

She got mine right though, Frank.

[Diane M. Morefield:]

Michael further proof (37

[Frank Garreth Louthan:]

All right. So a couple of things, as you grow – this company is growing. Talk to us a little bit about the sales team build out? How do you plan to ramp that? What's sort of the hiring there? And then on the backlog that's come down from the recent highs, what are your thoughts there? It sounds like you're working through that fairly quickly. Where should we – what sort of a long term rate of backlog that we can expect from you guys with the sales funnel that you're seeing? Thanks.

[Gary J. Wojtaszek:]

Yeah. So with regard to your first question in terms of sales, we're hiring about a dozen new sales folks this year all around the country. And that's basically additive to John Gould, who we hired last quarter and also Brent Behrman, who is running a lot o these strategic sales for us. We've significantly beefed-up our sales talent at the executive level there, and now we're trying to bring in some additional support under that to accelerate our growth. If you look at our funnel, we're a little less than $60 million this quarter in funnel, most of that assuming we didn't get any sales will all be kind of deployed by the end of the second quarter, the majority in the first quarter. But in terms of where this can go, it's unclear. I mean, if we look at – the funnel right now, as we said, is larger than we were at this point last year. So if all of those deals close, I mean, that funnel should be increasing above where we currently sit now. But it's all based on the timing of it. Ideally, if I had to choose where I wanted that backlog, ideally, I'd like it to be zero, meaning that my capital was planned perfectly in line with my sales. I recognize that that's not going to – that's never going to be the case. So if you listen to Diane's commentary, she provided really good insight in terms of all the different projects that we are bringing online this year. So just to kind of give you an overview of that, we've got projects online in Virginia, where we've just completed the last phase of that first building that we just brought online in Virginia. We completed the second building that we just announced the acquisition of in July. And we've also now – and so that's – and those are both completely sold out effectively, except for 2 megawatts in that second building. And then we've got that other campus there, the Kincora campus, that we acquired last year. We will have capacity online there. Some of that is pre-sold. And that will be brought online in the third quarter. If you go down to San Antonio, it's a similar story here. We have been selling out that property really, really quickly. We just completed an 18 megawatt build there just a couple weeks ago. So now we're effectively sold out there. We started construction on the next building there which will also be delivered in the third quarter of this year. And in Phoenix, it's a similar story. We completed 4.5 megawatt deployment there at the end of last quarter. We've got two projects underway there now. One of those next projects is completely sold out. And then we're bringing on another building there for additional growth. Those are going to be brought online in the third quarter. In Chicago, it's a similar type of story. I mean, this has been a great success. I'd encourage folks to really look at our earnings presentation and my commentary to show just how nicely the CME transaction has gone. We went in with a 8% yield on that. We've leased out basically that entire capacity that they had built out. And then the second half of that building, our yields there are now roughly 12%, which is about a 50% increase from where we were nine months ago when we acquired this. And now we're bringing on another facility there, a couple hundred thousand square feet that will give us capacity there in the third quarter. So we're hoping that we continue to try to get out in front of where customers' needs are. But it's really hard to work much faster than we are. I mean, for reference, I think this last year, we probably delivered 100 megawatts of built-out capacity this past year and I think that may be a record in terms of our deployment in the industry.

[Frank Garreth Louthan:]

Okay. Great. Thank you very much.

[Operator:]

Our next question comes from Vincent Chao of Deutsche Bank. Please go ahead.

[Vincent Chao:]

Hi. Hey, good morning, everyone. Just want to go back to the commencements and backlogs here a little bit. Last quarter, I think you had $43.8 million of backlog expected to come online in the first quarter, and if I net out the fourth quarter bookings from the new 1Q 2017 backlog, it seems like that number goes down a little bit. So I was just curious some of the revenue and some of the upside that we saw here in the quarter, was that just driven by timing, earlier timing commencements? Is there anything else driving it?

[Gary J. Wojtaszek:]

Yes. No, that's it.

[Vincent Chao:]

That's it? Okay. Great and thanks. And then in terms of the discussion around the capital stack and the desire to keep the floating rate down, I thought I'd heard the numbers of what the mix would be on a pro forma basis, but what are you thinking about in terms of targeted mix of floating rate by year-end?

[Diane M. Morefield:]

I think, generally, we don't have a specific target to get to by year-end, and partly it will depend on how attractive the high yield markets are. We are not investment grade yet. We are always in front of the rating agencies, and our goal is to obtain investment grade ratings. So we do have to issue in the high yield market at this point. I think we are generally comfortable in the 50-50 range of fixed to floating, but would go more fixed if we can execute in attractive windows.

[Vincent Chao:]

Okay. And do you think 6% is a decent bogey for high yield today, 10-year?

[Diane M. Morefield:]

Yes. Sort of 5.5% to 6%, probably.

[Vincent Chao:]

5.5% to 6%. Okay. And then I know you talked about the initial yield on the Sentinel deal being 7%, and I think the disclosures have talked about EBITDA in the $34 million range on an annualized basis. I guess is that what's specifically embedded in the guidance for 2017 or is there some growth also added into the guidance here?

[Diane M. Morefield:]

Pretty much 10 months of the numbers that we provided. So you obviously have to just assume 10 months of it.

[Vincent Chao:]

Right.

[Gary J. Wojtaszek:]

Just on that, I know there was like a lot of confusion in terms of our performance this year. So if you back out the Sentinel acquisition, our EBITDA was a couple million dollars higher than consensus. So our organic business for this year still tracking really well. In Diane's commentary, she also talked about our churn number's going to be at the low end of what we had estimated. So the business organically is firing pretty consistently with what we had last year, and the delta between EBITDA and FFO impact is predominantly related to capital structuring decisions which we think is going to give us more security around interest rates.

[Vincent Chao:]

Right. Yep, that all makes sense, but okay. And then maybe just going back, so expectations are sort of crystal clear here. Just the $20 million that you've been talking about for a little while now per quarter in terms of signings, it sounds like that is what the plan is predicated on. So anything, any hyperscale deals or larger-scale deals like you saw in 2016 to the extent that they come through would be additive to the numbers we're seeing here. Is that the ideal...

[Gary J. Wojtaszek:]

No, no. Yeah, I think no, no and this goes back to the fact that we're basically sold out. So the answer would be, yeah, like if we sold a lot in a location where we have extra capacity where that sale can convert quickly to revenue, it will be upside. But the problem for us is that we're basically sold out everywhere and we're bringing on additional capacity in all of the four hottest markets in the country for us. So just factoring in just kind of the annualized impact of that, no one should expect that there's going to be a material impact for this year's results on that. I mean, it would be more impactful in 2018.

[Vincent Chao:]

Got it. Okay. That's helpful. Thank you.

[Operator:]

Our next question comes from Simon Flannery of Morgan Stanley. Please go ahead.

[Simon Flannery:]

Great. Thank you very much. Good morning. Gary, coming back to the leasing, what's going on with the hyperscale guys? Are you assuming that there's less of that sort of big lumpy stuff just generally in the industry in 2017 or are you trying to maybe have a focus this year where your mix is more back to traditional levels where you're more focused on maybe higher-margin enterprise? And then related to that, I think at your Analyst Day a year ago, you've made the projection that enterprises, the Fortune 1000, would put most of their workloads, I think it was 75% in colocation rather than putting it into the public cloud. I'm just wondering if you had any updated thoughts around what sort of role is public cloud playing in their decisions right now? Thanks.

[Gary J. Wojtaszek:]

Sure. Yeah, so to take your first question, no, I mean so we don't go in with any preconceived notion of specifically targeting hyperscale customers or retail-type customers. Our focus is that we built the model that we're completely indifferent and we can handle any of those customers coming in. So from an operational model, from a back office functions, we can handle any customers. We don't have any preconceived notion. So I wouldn't go down not do a hyperscale deal because I'm going to do a retail deal. Our focus is to do all the deals. And we can make really nice returns if the only deals we ever did was just hyperscale deals, but being able to do both allows us to drive our yields up beyond the mid-teens, what we would typically do in a general underwriting. With regard to the opportunity into this year, I mean, if I look at my funnel, I'm 50% larger now than I was last year at the end of the fourth quarter. So we're looking at a funnel that is significantly larger. And so this notion that's out there about there's pull-forward and cloud companies not wanting to do more, that is not what we're seeing at all in our business. The deals that we're looking at and conversations with customers are larger now than they were last year. And so we feel pretty good. But in terms of just providing kind of like baseline guidance for folks, that's what we've been trying to kind of guide people to that $20 million a quarter is like a good nice run rate. If you get deals above that, that's great. But that type of bookings velocity is something that everyone should expect. With regard to the mix shift between – enterprise is moving outside. So there's two things going on here. One is, there's a general focus of just everyone, all the enterprises, recognizing that it is not in their interest any longer to build and manage their own data center. So they've kind of overcome this reluctance that they had for the last couple of decades and what we've proven over the last couple of years is, is we've made a lot of success with inroads in convincing the enterprises to outsource to us. We expect that that is going to continue. What we have seen, though, in the last year is that there is much more willingness from enterprises to actually do more with their data to public cloud companies. And so those two things go together jointly. And what I can tell you is one of our big oil and gas company's customers that we've been talking with and we've had as a customer for almost a decade, they recently made the decision, so this is a company that has outsourced, right, previously and they made a decision that they're going to be outsourcing a lot of their IT infrastructure to the cloud companies. And what they're really focused on is outsourcing to a couple different cloud providers that can help them achieve what they're trying to do in terms of managing – efficiently managing their IT hardware stack as well as their applications. And they've chosen to go down the path with two different cloud providers. Our conversations with them are associated with how we can help them engage with those different cloud providers more efficiently, all on an outsourced manner. So if you think about our focus in the enterprise, we expect that that's going to continue to bear out well for us. They're going to continue to do more on an outsourced basis, with us directly in terms of some of their dedicated gear and then also on an outsourced basis to the cloud providers which we expect that they will do with us here as well. And I think the results that you've seen by our success this year in convincing the largest cloud companies to go with us, as I mentioned on the call, we have nine of the top 10 now, is recognition of the fact, the same success we've had with enterprise is convincing enterprises for the last decade that it's not in their interest to build and manage their own data centers. We're having the same type of success with cloud companies recognizing that we can do this more efficiently than they can and they're more willing to enter into contracts with us. And that's why when you look at our funnel, our funnel, broadly speaking, is up 50% versus last year. If you look within that mix, we're up probably around 20% or so, 25% in our base enterprise business from a funnel perspective and up even more with the cloud companies in terms of the bigger build-to-suit deployments that we're working on with them. So at least from what we're seeing in our funnel, we see continued growth on both sides of that equation.

[Simon Flannery:]

Great. Great color. Thank you.

[Operator:]

Our next question comes from Richard Choe of JPMorgan. Please go ahead.

[Richard Y. Choe:]

Great. Thank you. Wanted to go into Northern Virginia a little bit more detail. What's driving the strength there? I mean, there's some strong pre-leasing in both two large builds and you have a third large build coming. It's almost like you can't build fast enough there. And then if you can talk a little bit about, I guess, the acreage, the land bank that was left there, dropped from 40 acres to 16 acres. Are you going to be looking for more land to kind of develop? And can you give us a sense of how much of the third build might be pre-leased as we move through the year?

[Gary J. Wojtaszek:]

Yeah. Hey, Rich. They're all kind of related, right? I mean, so the demand that we're seeing in that market is really broad, robust, and it's a couple of different customers that are choosing us. And so we've been the beneficiaries of the demand in that market in general and our penetration in those cloud companies which were nonexistent a couple years ago has been really, really well received and we're just – you're seeing the growth of that there. We had acquired a 39-acre parcel of land in Virginia there. The reason why that number in terms of available land has decreased is because of the construction project that we have on there. That construction project, as we build in increments, right, will give us hopefully enough capacity for the next year. But if it doesn't and we sell out faster, then we still got the other 16 acres that we'll develop on there, and then we will also look to acquire some additional land. But as of right now, we have no – we're not focused on acquiring any additional land in that market at all.

[Richard Y. Choe:]

And to clarify on Sentinel, it sounded like in your prepared remarks that you could ramp that up pretty quickly, or there's space available to be built out relatively quickly. But I think you kind of alluded to that there isn't any real growth number put in there for guidance right now. So can you talk a little bit about how quickly you can integrate, build out and develop?

[Gary J. Wojtaszek:]

Yeah. So let me take your last point there first. So we expect that we're going to close this deal next week. We should have 90% of that company integrated into our operations by next week with a little bit more to go after that. So we've been working in tandem and expect that everything will be fully integrated, that's from operations, security, ticketing and then there'll be a little more back office integration, but that integration effort is going to go really quickly. Our focus just generally speaking on this as an aside is we're focused on scale. We're focused on how do we build a business that can get really large over the next couple of years in terms of anticipating all the data growth. So that's what we're building our platform for. With regard to the growth, so there's a couple different things that are driving here. So right now as of the assets that we'll pick up next week, we'll have about 35,000 extra square feet of space available in New Jersey and South Carolina that's immediately sellable. That's about eight megawatts of capacity. In the South Carolina market, it's about a 400,000 square foot facility, and there's about a 400,000 square foot facility available. Half of that is built out. We were going – we can bring on the additional capacity, the shell capacity there, or raised floor really, really quickly, within 10 weeks as much as they want. So we'll be able to double the capacity that we have in North Carolina. In North Carolina, just as a reference, it's the cheapest power rates in the country in our portfolio. So we're looking at power costs of around $0.04 per kilowatt, rivaling what you would get in the Pacific Northwest. And this is the reason why a lot of the big hyperscale companies, Google, Facebook, Apple, have located data centers there. So we could bring that capacity very quickly in Carolina within the existing shell in 10 weeks. In New Jersey, there's a couple thousand square feet available there, but there's also a new shell being developed which will be brought online in the next couple of months and then once that's completed, we'll be able to put up our additional megawatts of fit-outs in that market as well. This has been one of the things that we've been struggling with in terms of – since we acquired Cervalis about how we bring on additional inventory capacity into the metro area, particularly in New Jersey where we were essentially sold out in our Totowa facility there. This is going to give us that runway where we're going to be able to sell out that additional capacity in there. With regard to your comment about what's in guidance, what's not, as Dianne mentioned, we're modeling nothing in there, but we fully expect that we're going to get some additional customers in that facility and that'll be upside from what we had estimated.

[Richard Y. Choe:]

Great. Thank you.

[Operator:]

Our next question comes from Amir Rozwadowski of Barclays. Please go ahead.

[Amir Rozwadowski:]

Thanks very much, and good morning, folks.

[Gary J. Wojtaszek:]

She got yours right, Amir. Yours is a little more phonetic.

[Amir Rozwadowski:]

Oh, yeah, that's not the usual case. But going back to one of the earlier questions on sort of the cloud service providers, Gary, you've been talking about (58

[Gary J. Wojtaszek:]

Yes, I fully believe money always goes where it's most wanted. And no one buys a share of any of those incredible companies because they have any prowess in building expensive real estate assets. Every shareholder of Google buys that because of the dominant market share that they have in search of, what they've been able to do in YouTube and all these other really cool technologies. And historically speaking, the only way that they can provide those products to their customers which is search or YouTube videos or anything else, historically, was because there wasn't a really good alternative to building that factory themselves. What we're showing them is that we are better builders of digital factories than they are and that ultimately our shareholders like the nice, steady cash flows that we gain really nice returns, high visibility and they recognize that we're not the type of company that's going to grow at 500%, 600%, 700%. So our capital base and our investors like those type of returns with that type of profile. Over time, their customers and their investors are going to recognize that as well and I believe over time they will outsource. And this is – Google, when they separated and they created a bunch of different companies and they put a balance sheet attached to all the different lines of business that they have, when you attach a balance sheet to a P&L, that kind of creates an opportunity for more rational capital allocation decisions. And to my earlier point, money is always going to go where it's most wanted. So I believe that is absolutely going to happen and I expect that they will convert over to an outsource model no differently than all of the enterprise companies have done as well. I mean, to give you some perspective, the oil and gas companies that we do business with are arguably the most advanced builders of capital projects in the world. I mean, you're building a two-mile long oil rig in the Gulf of Mexico. That has some incredibly difficult engineering talent and those companies which could easily built the data center have recognized it's not in their interest to do so and I expect all the cloud companies will as well. With regard to the broad demand for data, look, these guys are in a difficult position. Their businesses are on fire. They are exploding at a rapid pace. There is nothing on horizon that is going to slow it down and they are struggling with how to manage their capacity to meet the products that they're selling to their customers. That kind of backdrop is ideal for us because we can come in and disrupt what's going on there because we provide real value to them. Our ability to deliver a data center in a couple of months at a price point that is not on – not touchable by other folks is incredibly attractive to them. And just like all the other enterprise customers, when they first went down this path and they were reluctant to outsource about giving up control, not owning it themselves, what we've shown over the last decade is that you're afraid of ghost. And eventually the best way to eliminate a ghost is to basically bring a spotlight to it and all the ghost go away and that's happening here. They will eventually outsource everything to a data center provider and our focus is and goal is to be that choice.

[Diane M. Morefield:]

And I think you've seen that other real estate migration of like large retailers that used to own their anchor stores and then realize they didn't need to tie up capital or large hotel companies that used to own the hotels and then when what was called asset-light and sold the real estate and did what they were good at, which was managing and operating the hotel. So very similar. This is just in an earlier stage for our product type.

[Amir Rozwadowski:]

And I certainly understand the industrial logic and sort of the capital efficiency allocation decisions. But are you seeing right now from a demand perspective that they're making those decisions, that they're in a place right where the decision making process has shifted from say a year, two years ago when they were entertaining the idea of building that right now? I mean, clearly your business with them has grown, so I would suspect so, but what is sort of the real time feedback that you're getting today outside of the bigger picture industrial logic?

[Gary J. Wojtaszek:]

Yes. So the biggest proof point is, if you just look at where we've come from, so like two years ago, right, we're just starting to have conversations with them, right? All of the sudden, they start recognizing that we can help and they start awarding us business. We just finished a record blowout year in 2016 and 60% of those bookings too were in big cloud companies, right? I mean, just to give you some perspective and color, just so there's no misunderstanding, so all the deals we did this year, 60% were in like these bigger cloud deals, but 40% of our bookings were still at our base traditional kind of enterprise business. But – so in spite of a record year, we're sitting on a funnel that is 50% larger than where we were last year and the vast majority of that increase in our funnel is all in these cloud companies. And so at least from what we're seeing in terms of the visibility to our funnel and the conversations we're having with our customers, it's just as big and I think as everyone just helps them work through that and more gets outsourced, I think just broadly speaking, the industry is going to be the beneficiary of that. And if we continue to do well, hopefully we'd take a bigger share of it.

[Amir Rozwadowski:]

Great. Thanks for the incremental color.

[Operator:]

Our next question comes from Jonathan Atkin of RBC Capital Markets. Please go ahead.

[Jonathan Atkin:]

Thanks. So I'm interested in the pipeline that you're seeing in Chicago, Phoenix and Dallas where you either have inventory or you have construction underway and how does the demand profile differ across those markets? And then secondly, on the West Coast, the Vantage assets got sold to a non-listed data center investor and interested in kind of how you would evaluate those sorts of situations. Obviously, apart from price, strategically, what is it about that type of situation versus say land that you own that you could still scrounge up elsewhere in the Northwest? And how do you sort of view the puts and takes around expanding to the West Coast? Thanks.

[Gary J. Wojtaszek:]

Yeah, hey, Jonathan. So with regard to the funnel that we're tracking, it is in the same four markets that we saw big demand this year. So Chicago, Virginia, San Antonio and Phoenix. So our funnel is larger and the vast majority of that is located in those same four markets. From probably mix shift, differences between them slightly, but in general, it's kind of more of the same. With regard to the Vantage business, look, that's a great company. They've built a phenomenal business in that market and I'm sure the buyers are going to do well on that. This kind of goes through my earlier point about capital partners, though, is what you'll see is that there is absolutely a different type of capital partner that was coming into this business that are looking at data centers as just another type of real estate asset that is providing long-term sustainable yields for pension funds. And that's a business that's going to do well and the buyers are going to make a lot of money of that and the investors are as well. And I think over time, as this industry matures, you're going to see that the cap rates that people are willing to pay in this industry kind of come down to be more in line with what you're seeing in the broad real estate market. And so that's why we feel really good about what we're able to do when we generate mid-teens result – returns. I mean, if I do an asset that generates a 15% cap rate and I'm selling at 7%, I could basically get out more than 100% of my investment in that and still retain a majority ownership position in a particular asset. So to Diane's point about capital recycling, it's an incredible opportunity to monetize the value creation associated with the distribution channels that we have, right. These things go hand in hand. The marketing and sales investment that we have been making in the business is extremely valuable because just making a return is only one part of the equation. It's the sales velocity combined with the high returns that ultimately create value and there's a lot of different capital partners that are approaching us to want to work with us recognizing that value creation we have.

[Jonathan Atkin:]

And then finally, there's the Design and Build Teach-In that you're holding, I guess, in several weeks. Any sort of preview as to what you're going to be talking about at that event?

[Gary J. Wojtaszek:]

Yeah, I was going to hold off until the end, but I can do it now. So on March 7, we're holding a panel discussion in Hollywood, Florida at the Margaritaville Hotel with several leading architectural, engineering and construction firms. The purpose of that meeting is to really discuss our unique approach to designing data centers as well as the supply chain that we've assembled over the last five years that enables us to deliver these data centers and the time to market that is just unmatched. There's a lot of confusion in the market that's been coming out over the last couple months on two things. One, our yield, the quality, the facilities that we build, and the speed at which we can build it by. There's a lot of confusion about how can you actually do this, and we recognize that, look, we're definitely on the leading edge of this and it's created a lot of concerns and negativity in the industry broadly and we're hoping to shed a spotlight to that. So the folks that we're going to have on stage are people that have probably built and managed 60 million square feet of data center space worldwide. So I think what you'll see is a really good representation of understanding what drives this industry. Our entire senior management team is going to be there. We're going to talk about these things. Specifically, Diane's going to talk about the yield, what goes in, each of these megawatts. I mean, we've pointed out numbers at the last call in Northern Virginia that we delivered a facility for $6.3 million per megawatt in six months. That is by far a record on both measures that people just can't get their arms around. And we're going to bring the engineers, the architects and the construction folks that were involved in that project to explain what it is that we do which is different than what everyone else does. And then we'll just have the regular team there open it up for a Q&A. So we're hoping for a really good discussion, hopefully demystify what it is that we're doing to show people that we actually do deliver these yields and we do deliver these data centers at the timeframe that we say we do.

[Jonathan Atkin:]

Thank you very much.

[Gary J. Wojtaszek:]

Hope you can come.

[Operator:]

Our next question comes from Matthew Heinz of Stifel. Please go ahead.

[Matthew Heinz:]

Thanks. Just one for Gary on Sentinel. If I just tie together your comments about the power costs and hyperscale interest in North Carolina paired with the robust enterprise installed base in both Jersey and Raleigh, it seems like there's a nicely balanced demand mix across the two markets that you'll be entering. But I'd like to just hear in, Gary, if you could sum up your rationale for the deal along the lines of strategic growth, immediacy of the timing and kind of the reasonably full valuation that you paid for the asset?

[Gary J. Wojtaszek:]

Sure. Yeah, yeah. So I think what we've always said is there's two things that we're looking at these acquisitions, by one and strategic one is financial, right? So from a strategic perspective, what we're looking at is can we get into a market where we're currently not and also do we pick up a customer base that would be added to what we currently have? And that's the lens by which that we're kind of looking at these things through strategically in terms of can we get some additional synergy value out of the combination of the two? We have been saying for quite some time that, look, we believe this is a scaled business and size matters and having a broad footprint is incredibly important. We specifically have been talking about the Southeast and the West Coast and so this gets us into that Southeast market. When we looked at this, it was great because from a customer perspective, about two-thirds of the customers are new to us. 70% of their customers are really large investment-grade type companies and particularly a big opportunity in healthcare with the presence that they have in New Jersey, the pharma market as well as some of the opportunities in Carolina is also with medical as well. So we really enhanced the vertical that we have in the healthcare space. Two years ago, we talked about this is an area that we've been targeting our marketing towards in terms of developing a bigger penetration because data associated with healthcare is going to be huge. It's going to be about 15% of the data created over the next decade is going to be in healthcare-related fields. And so you need to establish a presence in there now in recognition of the fact of the data growth that's going to come in there. So we feel really good about that opportunity. In terms of the financial play, so we're going to see this – with deals being announced, as I mentioned, there's capital sources that are coming into this that are looking at this differently through different lens. Looking at these is pure real estate type play. So cap rates that are down in the sixes and sevens is becoming – and it is going to be increasingly becoming the norm in this space. And if you look at our business and what we just acquired there, what we have is a building, two buildings that we're going to basically be able to double in size that we went in with like 7% type yield that we're going to be able to increase capacity at really attractive rate. So similar to what we saw we did in CME last year, we went in at an 8% yield and now it's at a 12% yield a year later. We're looking at a similar thing here as well. The difference, though, with this acquisition is there's 30 customers in there, so the opportunity to sell to those customers in those facilities is greater just because we have 30 of them. You're looking at an opportunity to acquire an average lease duration of a portfolio of 8.6 years. And so when you look at that business and if you PVed this back and you did like a replacement cost basis, or if you're looking at this on a cost per megawatt basis, a bunch of different metrics, like, if you brought this down and you looked at the present value of those existing customer contracts plus the additional amount of capital that we can grow, we're looking at an average cost per megawatt there of less than $5 million. So when we look at the upside for growth here, it's big. We think that we can absolutely sell more of our customers into this location. And that North Carolina market, at the power rates that are there, we believe more of the big scale cloud companies are going to go there and we believe that because they've currently gone there. So between Google, Facebook and Apple, they've spent a couple billion dollars of investment in data centers in that market because of the low power rates and we expect that we're going to be the beneficiaries of additional demand from those folks as well as the other enterprises that are there. So we're pretty excited about the opportunity there.

[Diane M. Morefield:]

I think if you draw the analogy similar to our sales philosophy that we sell to both bulk wholesale as well as retail, same with our capital allocation. We're still very focused on organic growth and ground-up development, but we supplement that with the right acquisitions where you have a going-in yield and take away the longer period of zero cash flow as you buy land, entitle it and build on it. And so the mix of doing both types of capital allocation enhances, we think, our overall cash flow and yield.

[Gary J. Wojtaszek:]

Just on that, I mean, this shouldn't be a surprise to anyone in terms of what we said last year we were going to do, right? What we said was, we provided at our Investor Day last year an overview to basically double the size of the company, right? And what we were saying is part of that was we're going to predominantly grow the company organically, but there was going to be roughly $300 million a year of acquisitions. And so if you look at the $100 million that we did last year and the $500 million that we did this year, we're tracking right on our plan. So the most important thing for us as a management team is trying to provide you as much insight and transparency into what we are doing and where we want to take the company as possible, because at the end of the day, I never want any investors questioning about well, we didn't know this or that. We've been completely clear about what we've been intending to do since we IPO-ed the company, and we're tracking to slightly above our IPO plans and this is five years later. So we're right on top of the plan that we laid out half a decade ago.

[Matthew Heinz:]

Really appreciate the detailed answer and I'll save the other questions for our follow-up call. Thanks.

[Gary J. Wojtaszek:]

Great. Thanks.

[Operator:]

Our next question comes from Jordan Sadler of KeyBanc Capital Markets. Please go ahead.

[Jordan Sadler:]

Thank you. Wanted to follow-up on the delta between adjusted EBITDA and FFO for 2017. So it sounds like there is some kind of capital transaction assumed. Did you give us any specifics around that, either a joint venture or an asset sale?

[Diane M. Morefield:]

We did not. We just again stressed that we're going to move some of our debt into more of a fixed-rate bucket, so it can affect models, if some people have it more back-ended and we have it more front-ended. Again, we have an increase in interest rates. We had – we also have additional stock-based compensation that, as you know, runs through FFO and then gets back up for AFFO. So that's just from additional head count and our 2017 awards. And we also have some non-real estate D&A and income taxes that affect FFO. So it's just a combination of things, but I think we – the bigger is probably moving to some fixed-rate and having interest rates rise this year.

[Jordan Sadler:]

So no asset sales, no equity as in guidance? Right?

[Diane M. Morefield:]

Yeah, we said the only acquisition that's in it is Sentinel and it doesn't include any other transactions or major capital assumptions.

[Jordan Sadler:]

Okay. That's great. And then, of the development pipeline, you've got 88 megawatts under construction. What is the leased versus available number today?

[Diane M. Morefield:]

Well, 72% is pre-leased.

[Gary J. Wojtaszek:]

On a CSF basis.

[Jordan Sadler:]

72% of the 88 megawatts?

[Gary J. Wojtaszek:]

No, on a CSF basis, Jordan. Colocation square feet.

[Diane M. Morefield:]

Of the almost 500,000 square feet, that's 72% pre-leased.

[Gary J. Wojtaszek:]

Correct.

[Jordan Sadler:]

72% of the colocation square feet. Got it. And then one other one is just on the sales people additions, I thought I heard you guys talking about, I think you said 12 additions across the country.

[Gary J. Wojtaszek:]

Yeah.

[Jordan Sadler:]

So I'm just curious, I know you're putting some capital out the door, but if you're sold out throughout much of 2017, how do you bridge the gap there? It doesn't seem like sales is your problem, given your funnel. It seems like supplier capacity is your problem.

[Gary J. Wojtaszek:]

Sales is always a problem.

[Diane M. Morefield:]

For a revenue (01

[Gary J. Wojtaszek:]

All companies that go bye-bye, all have problem with sales. Our focus here is, look, we have been very successful over the last couple of years selling. We do not believe that we are as successful as we could be and we are – we believe that having a more robust sales group can increase our sales trajectory go forward. So we generally get a really nice return on sales folks and so you want to basically establish those relationships with the customers recognizing that over time they will result in yields. I mean, I think there's – maybe it's just because of the success in the industry this year that there were so many deals that were out there at one, but our focus is still attacking that Fortune 1000 base, right? We still only have 20% market share in the Fortune 1000. So the opportunity for us is the other 80% that we don't have. We've been super successful at attacking the cloud companies and we have 90% market share with those ten companies, but the bigger opportunity is still with those enterprises and none of those companies make the decision overnight to outsource. It's a relationship selling process that is based on a really long sales cycle. It typically takes two to three years from the time we make first contact until we actually get a sale out of that. So when we're thinking about the sales investment that we're making, that sales investment is with the goal of where this company is going to be over the next several years. So it's a forward-looking investment, right? And so – and that's the reason why like last quarter when we hired John Gould and Brent Behrman, we brought those guys onboard with the full intention of developing a business for the long term in terms of how much larger this business is going to be over the next couple of years, because underlying all of this is we're very comfortable with the broad demand drivers for data growth. And so as long as we're comfortable with those underlying secular trends, we want to make sure we develop a sales and marketing capability that positions us to take advantage of the trends that will manifest in the next couple years.

[Diane M. Morefield:]

And again, even with the increase in the sales force that's budgeted, there are SG&A as a percent of revenue declines in 2017, and you'll continue to see that happen as our revenue grows and our marginal overhead does not have to increase.

[Gary J. Wojtaszek:]

Yeah.

[Jordan Sadler:]

That's helpful. Thank you.

[Gary J. Wojtaszek:]

Also, I mean, just like to add a high level, though, Jordan, like, if you're spending $500 million, $600 million a year in capital, to really manage that well, spending a couple million dollars for increased sales and marketing to make sure you could turn all that by the $600 million you're on, that is the easiest way to de-risk your investments.

[Operator:]

Our next question comes from Michael Rollins of Citi. Please go ahead.

[Michael I. Rollins:]

Hi. Good morning and thanks for taking the questions. Two, if I could. The first is a follow-up going all the way back to Jonathan's question on equipment sales. I didn't hear if you specified the amount of equipment sales for the fourth quarter, but was wondering if you could specify that or just remind me of what that number was. And then secondly, I was wondering if you could look at the straight line adjustments that you've been disclosing. And if you can unpack what happened in the third quarter between what was the actual straight line versus the rent adjustments for the ramping-in periods, how that's split up in the fourth quarter? And then maybe give us a sense of the outlook for that, especially as it relates to the timing of commencements that you've outlined in the deck? Thanks very much.

[Diane M. Morefield:]

First, equipment sales. If they had been material as we disclosed them in the fourth quarter, I think just generally if we have an unusual amount in any given quarter we highlight it. But as a percent of our total revenue, generally, it's not that material. So we don't necessarily call it out. Regarding – again, we can't stress enough, the difference between FFO and AFFO is going to be volatile. I think you saw in the third (01

[Gary J. Wojtaszek:]

Hey, Michael. Gary here. Just to provide a little more color for modeling purposes, right, if you just assume a 10-year lease with a 2% lease inflation factor in there or a lease escalator in there, right, that will give you some basis for what the total delta on a steady run rate basis will be between cash rents and GAAP rents over that 10-year period. So you could just come up with that estimate. What you see in terms of the delta between our FFO and our AFFO, which historically was fairly minor, right? I mean, it was only a couple million bucks difference between those two. That delta was historically, I mean, once you net out all the stuff with stock comp expense and leasing commissions and everything else, you're only looking at that – the delta associated with the straight line difference associated with the lease escalators. What you saw in the last couple of quarters with us was because we have been generating these massive deals and those customers are – have ramp-in period. So there's a couple months in there where your cash and your GAAP really diverge, but then once they're on that cycle and they fully lease-up and they're paying you that cash, the only delta between your FFO and AFFO on that is just purely attributable to your lease escalators. What you should know, though, is in those deals and in every deal that we underwrite is we don't focus on the accounting impact, right? We focus on cash and what our cash returns are going to be on all of the underwritings we do because we're focused on being great stewards of company capital. And we're underwriting those at the mid-teens type cash returns. And so I know, generally speaking, it's hard to kind of fathom, particularly in a real estate asset class when you have a high growth situation and you've got this delta between cash and non-cash. And typically and I understand even in telecom and some of the other companies is that's generally a tell, right? So if there's a lot of non-cash growth going on that's generally a tell to say, hey, there's something funky going on in those company' results. And what I can say here it's purely attributable to the leasing velocity that we are on relative to the size of our company, which is significant. And that's the biggest delta. And that's what we're trying to do with having those engineers and architects and other folks there so people can understand what we're delivering is nothing that different than everyone else is delivering. The way we deliver it and the cost we deliver it is what's unique.

[Diane M. Morefield:]

Yeah. I'm not sure why there's such a unusual focus on this distinction because you ultimately get the cash. It's just timing. And again, when you're in early – when you're in development and you're doing bigger deals, that, one, it's going to fluctuate and two, there's going to be bigger gaps. But at the end of the day, and we've run the NPVs on these 10-year deals with ramp up or more minimal ramp up, the NPV is virtually the same because it's just such a short period up front where you have the bigger delta between cash and GAAP. But you ultimately get it. I mean to have a weird – or example, you'd have zero gap between AFFO and NFFO if you did a 10-year lease with no escalators. But no one would suggest anybody should do that because that wouldn't be the right economic deal. So it's just accounting and timing in my opinion.

[Michael I. Rollins:]

And just one other follow-up on this. Is there a framework then to consider of the commencements that are coming on and that you've guided to over the next couple of quarters, is there a way to think about what percent are large deals that might be subject to ramping up periods so you can just help the buy side kind of think about what that impact AFFO? If you can't give it quantitatively to us, maybe qualitatively, just so we can all consider what those impacts might look like?

[Gary J. Wojtaszek:]

Yeah. It is really hard, like, because here's what's going on, on the other side. So what customers are trying to do here, and we're being responsive towards, is they are trying to de-time their expense and managing their P&L with when they are going to be selling their cloud products to their customers, right? So they're – they want to tie their expense, which is our lease payments that we will charge for them, in terms of the ramp ups that they're having and charging with their customers over this period, so that – what they're trying to avoid is that their P&L is not GAAP, because what would happen if they pay us that full burden of the lease charge up front. But their customers were only paying them on a ramp-up basis. Then, that expense delta will be fully burdened in their P&L. And so what they're trying to do with us and what we're trying to accommodate is to give them the opportunity to phase-in their lease payments with us over that same point in time and so that the whole downstream situation, basically everyone is matching what the ultimate customer is buying. So they're trying to focusing from a P&L perspective, us to a lesser degree, we're more focused on cash returns, but it's hard to predict how we're going to buy it and what those ramp ups are going to be.

[Diane M. Morefield:]

We're not trying to be frustrating on it. It's just really hard to estimate. And if you look at the four quarters in 2016, we had a low of $2 million of deferred revenue or straight line rent adjustment in one quarter and a high of $10.7 million. So I mean, that shows the magnitude of the fluctuation and we just – there's really no way to estimate it and again, we don't give AFFO guidance.

[Michael I. Rollins:]

Thank you for taking my questions.

[Diane M. Morefield:]

Sure, Michael.

[Gary J. Wojtaszek:]

Sure. Well, thanks folks. I just want to reiterate on March 7 we're holding that panel discussion, so if you have any questions at all about our building, our designs and our speed, more than happy to host you there answering those questions. And the last thing I'd just add, this was Ms. Morefield's first earnings call and she has been a welcome addition to the team. She's been here for a little over three months and we're excited to have her onboard and look forward to a great 2017. Thanks, everyone, and see you at the conferences.

[Operator:]

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.