Ladies and gentlemen, thank you for standing by, and welcome to the American Tower Third Quarter 2019 Earnings Call. At this time, all participants are in a listen only mode. They will conduct a question and answer session. Over to your host, Igor Kozlovskiy. Please go ahead.
Good morning, and thank you for joining American Tower's Q3 2019 earnings conference call. We've posted a presentation, which we will refer to throughout our prepared remarks under the Investor Relations tab of our website, www.americantower.com. Our agenda for this morning's call will be as follows. First, I'll quickly summarize our financial results for the quarter. Next, Jim Taglek, our Chairman, President and CEO will provide an update on some key technology trends in the U.
S, particularly around 5 gs. And finally, Tom Bartlett, our Executive Vice President and CFO, will discuss our Q3 results and revised 2019 outlook in more detail. After these comments, we will open up the call for your questions. Before I begin, I'll remind you that this call will contain forward looking statements that involve a number of risks and uncertainties. Examples of these statements include our expectations regarding industry trends as well as our future growth, including our 2019 outlook, capital allocation, pending acquisitions and future operating performance the pacing and magnitude of the Indian carrier consolidation process and its impacts on American Tower and any other statements regarding matters that are not historical facts.
You should be aware that certain factors may affect us in the future and could cause actual results to differ materially from those expressed in these forward looking statements. Such factors include the risk factors set forth in this morning's earnings press release, those set forth in our Form 10 ks for the year ended December 31, 2018 and in other filings we make with the SEC. We urge you to consider these factors and remind you that we undertake no obligation to update the information contained in this call to reflect subsequent events or circumstances. Now please turn to slide 4 of our presentation, which highlights our financial results for the Q3 of 2019. These results as well as our year over year growth rates were positively impacted by our new agreement with AT and T in the U.
S. And negatively impacted by Indian carrier consolidation driven churn. During the quarter, our property revenue grew 9.7 percent to $1,900,000,000 Our adjusted EBITDA grew over 12% to 1 point $2,000,000,000 and our consolidated AFFO and consolidated AFFO per share increased by 8.5% and 8.1% to $891,000,000 $2 respectively. Finally, net income attributable to American Tower Corporation common stockholders increased by nearly 36 percent to $499,000,000 or $1.12 per diluted common share. Additionally, like the last few quarters, many of our comments around 3rd quarter results and our updated 2019 outlook will be focused on growth rates normalized for carrier consolidation driven churn in India and the Tata settlement in Q4 2018.
We view these normalized results as important indicators of underlying trends in our business. Reconciliations of these normalized metrics to our GAAP results are included in the back of our earnings presentation, in our press release and in our supplemental package. And with that, let me turn the call over to Jim.
Thanks, Igor, and good morning to everyone
on the
call. Consistent with our past practice for Q3 reports, my remarks today will be focused on the evolution of mobile technology. I'll spend most of my time on the U. S. Where we generate roughly 2 thirds of our cash flows and where many exciting and potentially impactful developments, particularly on the 5 gs front, are beginning to emerge.
We continue to believe that the technology cycles driving our business in the U. S. Will be largely replicated in our international markets over time, thereby lengthening and strengthening our global growth trajectory. Longstanding trends in U. S.
Mobile continue unabated. Unlimited data plans, advanced devices and increasing mobile video consumption consistently result in 30% or more growth in annual mobile data consumption per year. To keep up, our mobile operator customers have added significant additional equipment to existing transmission sites, while driving forward with incremental spectrum deployments to support today's 4 gs networks. These factors have led to a sustained period of robust domestic organic tenants billings growth for American Tower. An average monthly U.
S. Smartphone usage now stands at roughly 10 gigabytes per month, nearly doubling in less than 3 years. And by 2028, 10 years roughly from now, industry estimates suggest that the total U. S. Mobile data usage will be around 6 times 2018 levels.
We recently concluded that the need to efficiently manage the network cost challenges that this continued explosion in mobile traffic growth creates will actually be the main driver for the deployment of 5 gs. Simply put, the cost per gigabyte delivered must continue to decline at roughly the same rate as the growth in aggregate traffic carried across the network that's needed to sustain our carrier tenants' margins. We've seen this pattern historically with each successive generation of technological evolution, 2 to 3 4 gs, with carriers using tools like more efficient radio technology, advanced network design, incremental spectrum deployments, and additional network density to drive down those operating costs. For 5 gs, we believe the equation will be similar, but with some new aspects. The technology is expected to enable the broad usage of newly developed techniques such as massive MIMO, dynamic spectrum sharing and self optimizing networks, as well as wider spectrum allocations.
All these will help enable wireless carriers to efficiently manage the cost of their exploding network demand. The deployment of new spectrum bands for 5 gs will be tailored to both coverage and capacity. And eventually, as nationwide coverage is achieved, we expect that 5 gs will pave the way towards a variety of interesting next generation products and services that may offer profitable opportunities not just for mobile operators, but also across a variety of industries. So from a cost, efficiency and practical perspective, 5 gs is the logical and necessary next step in network evolution. With that said, 4 gs is still carrying nearly all the usage in the United States today and will continue to be the primary network technology here for years to come.
And at the same time, with faster speeds, 5 gs is already being deployed in limited coverage areas with devices beginning to enter the market now. So for the rest of my comments this morning, I'll focus on 3 specific areas of that 5 gs migration. 1 is spectrum. The second is what 5 gs networks are likely to look at and the anticipated timeline of that topology and then thirdly, some of those advanced potential products and services that will be ultimately enabled by 5 gs technology. So on the first point, we expect 5 gs to be accompanied by significant deployments of new spectrum assets across the full range of low, mid and high bands.
We've already seen 600 megahertz low band spectrum selectively deployed on our towers. T Mobile has stated publicly that they expect 600 megahertz to serve as a significant component of their 5 gs coverage plan across the country. And our portfolio of nearly 41,000 U. S. Sites is positioned very well to support that deployment.
What may be even more impactful for us over time is the widespread rollout of mid band spectrum. These bands generally between 2.56 Gigahertz offer an intriguing blend of low band coverage benefits and high band capacity benefits. Both of these are expected to be critical attributes of 5 gs networks. So spectrum assets in this range include Sprint's 2.5 gigahertz spectrum, the 3.5 gigahertz CBRS band and the C band spectrum currently held by satellite operators between 3.7 and 4.2 gigahertz. Importantly, given that propagation characteristics of this spectrum aren't as favorable as low band, we expect that more transmission sites will be necessary to deliver an ubiquitous 5 gs level signal.
In turn, this should lead to incremental amendments and co locations across our nationwide portfolio, which today has ample capacity to support additional carrier equipment. Additionally, we continue to evaluate the potential of significantly expanding the addressable market for neutral host and private indoor systems by utilizing CBRS spectrum. And ATC was one of the earliest members of the CBRS Alliance to help facilitate that process. Finally, there has been considerable discussion around millimeter wave spectrum, small cells and their applicability to 5 gs. RV remains that millimeter wave spectrum and small cells will serve pedestrian hotspots and other predominantly fixed location applications in urban and dense urban areas.
Less than 1% of our U. S. Macro tower sites are located in areas with high enough population density to economically support such outdoor small cell arrays. Consequently, the impact of macro tower sites from 5 gs millimeter wave spectrum deployments are expected initially to be minimal, both from a risk and an opportunity perspective. But over time, there may be some macro related uses for 5 gs at these millimeter wave bands for products such as wireless backhaul, fixed wireless service to homes and enterprise and other applications.
As we previously discussed, we do not believe that fiber fed outdoor small cells in their current form offer sufficiently attractive economics in the U. S. For us to make material investments in that business. Instead, our aim throughout the 4 gs to 5 gs evolution is to maximize the value of our existing macro tower and indoor DAS real estate. We're also seeking to add to this core growth by selectively deploying capital through our innovation program towards complementary technologies and initiatives that offer similarly attractive returns as our tower business.
At the same time, we continue to look for ways to expand our tenant base and augment the value of our existing portfolio. In our view, 5 gs will likely have a number of layers, since any given geographic area will have specific topographic and population characteristics. In rural locations, low band spectrum and perhaps some mid band will likely be the main components. In suburban areas highway corridors where there are more people and more usage, we believe that mid band spectrum is likely to be an important component of 5 gs with low band coverage also broadly deployed. And finally, in dense urban areas, all three types of spectrum, including high band millimeter wave are likely to be deployed through a combination of rooftop antennas, indoor and outdoor systems, and other small cell solutions.
The net result is likely to be an even more complex radio access network architecture, requiring more density, considerably more compute power, and more intelligent design to deliver a consistent user experience to all of us. These deployments are likely to take a significant amount of time. Past technologies have lasted at least 15 to 20 years from inception to sunset. Given the scope and intricacies involved with 5 gs, we would at a minimum expect a similar timeline. Moreover, in the near term, 4 gs will continue to serve the vast majority of mobile usage across the country.
Currently, industry estimates suggest that 4 gs will still represent more than 50% of the embedded U. S. Device market share even through the year 2025. Consequently, investments by our tenants into augmenting their existing LTE networks are expected to continue well into the 2020s, with incremental 5 gs related spending progressively being added to the mix. This use further reinforced by currently projected timelines of when mid band spectrum assets will be available to be deployed.
While Sprint's 2.5 gigahertz spectrum is ready today, the CBRS option for license spectrum has yet to occur and the C band is likely years away from being made available for terrestrial use. Consequently, we at ATC anticipate a 5 to 10 year period ahead of us that will be driven by a combination of ever increasing data consumption and a long cycle evolution from 4 to 5 gs technology. Therefore, we are exploring through our innovation program opportunities to use our assets to support this transitory period, and in turn, to maximize the revenue from our macro tower sites and our in building systems capabilities. Edge compute solutions at tower sites are one example. As data demand increases on both wired and wireless networks, our macro tower sites have the potential to act as convergence points for wireless access networks, cloud services, the Internet of Things and enterprise networks, given the tower's positioning on the edge of today's mobile networks.
As 5 gs supported applications that require minimal latency develop, there could be a further opportunity for edge compute to play an even more important role, especially in the establishment of autonomous air and ground vehicle management and control systems, autonomous cars and drones. To better understand the potential of this and other industry developments, we're already moving forward with trial edge data centers at several of our U. S. Tower sites and also have acquired a midsized interconnect facility in Atlanta earlier this year. We're using this limited scope architecture to learn as much as possible about potential future business models and to work together with some prospective future tenants as they evolve their approaches to this type of distributed data storage and compute technology.
Drone applications are another promising area where we continue to focus efforts within our innovation program as well. We're already using drones internally, which promotes safety and enhances our site monitoring and maintenance functions. Similar to what we anticipate from our mobile operator tenants, we're first using this technology to reduce our unit cost of operations, as well as increase the quality and the cycle time of our site inspections and structural analysis. In the longer term, we continue to believe that our tower sites could play a significant role in a 5 gs enabled drone air traffic control system, and we're working towards that goal. Already, UPS, Google and some others have received regulatory clearance to begin drone deliveries for limited purposes in specific geographic areas, which we see as an indication to us that progress is being made.
Then there's augmented and virtual reality, autonomous long haul trucking, smart factories and buildings and a myriad of other next generation applications that are also in the 5 gs pipeline. Many of these will be deployed by existing mobile operators, industry verticals or even by the government. Some may ultimately fail to gain traction or scale, while there may be others that we aren't even thinking about today that will be introduced by creative entrepreneurs, just like the founders of Uber and Lyft did when a robust 4 gs platform became available to them. All these developments are important for American Tower as we expect our towers and indoor DAS assets to play an even more critical role in the 4 gs and 5 gs networks of both the near term and the longer term future. Given the spectrum bands that will be utilized and the much heavier data throughput of 5 gs networks, we anticipate that transmission site topologies will be denser and there will be more equipment for tower.
Further, we are striving to pursue a well organized and integrated innovation program that we hope will enable us to expand our tenant base and deliver additional value to our investors from our extensive asset portfolio. Importantly, this is a global expectation over a medium to long time horizon as we believe our properties outside the U. S. Will serve similarly critical roles and technology evolution across our 16 international markets. And our primary conclusion of my remarks today that 5 gs will become imperative to mobile operators for reasons of per gigabyte cost management should apply both in the U.
S. And internationally as well. And with that, I'll turn it over to Tom for more detail on the quarter and our revised full year expectations.
Thanks, Jim. Good morning, everyone. As Igor just mentioned, we again generated stronger than expected financial results in the quarter, signed a new master lease agreement with AT and T, built a record 1300 or so sites and acquired another 600. We also continued to opportunistically refinance and extend our debt maturities during the quarter, further strengthening our balance sheet, and we continue to make progress on a number of internal initiatives, particularly around innovation and process improvement. As a result, we're raising our expectations across our key metrics for the year.
With that, let's dive into the details of our Q3 results and updated full year outlook. If you please turn to Slide 6. During the quarter, we generated consolidated property revenue growth and organic tenant billings growth normalized for the impacts of India carrier consolidation driven churn of 11.9% and 7.4%, respectively. Volume growth from co locations and amendments, as we've seen all year, contributed about 6% of the organic tenant billings growth, driven by continued elevated levels of tenant investments in a number of markets. Our U.
S. Property revenue grew over 14% to $1,100,000,000 with organic tenant billings growth of 7.1%. Volume growth from co locations and amendments drove close to 6% of this growth and escalators contributed just over 3%. We also recorded approximately $87,000,000 in incremental straight line revenue during the quarter as a result of our new agreement with AT and T. This was roughly $20,000,000 more than our initial projections, resulting from the finalization of straight line calculations associated with the agreement.
These items were partially offset by churn of roughly 1.6%. Similar to the last few quarters, most of the activity we saw on the new business side was amendment driven with carriers actively investing to keep pace with a rapid growth in mobile data usage. Our reported international property revenue growth during the period was about 4% after factoring in the negative impact of approximately 4% from India carrier consolidation driven churn. Normalized international organic tenant billings growth, which adjusts for the impacts of that churn, was just under 8% in the quarter, including 7.5% in Latin America and around 8% in both EMEA and India. New business revenue from co locations and amendments drove nearly 7% of this growth, while escalators contributed another 4% or so.
Other run rate items added nearly 1% with normal course churn offsetting these items by about 3%. Importantly, gross new business commitments were up across the board in international markets, with around $1,800,000 of run rate added in the quarter coming from India, close to $700,000 in EMEA and roughly $1,400,000 in Latin America. These run rate additions of nearly $4,000,000 were roughly 50% higher than in the year ago period. And finally, as expected, the tenant billings impact of India carrier consolidation driven churn to organic growth was $8,000,000 lower sequentially, and we anticipate an even more significant drop in Q4. Similar to last quarter, we also recorded some higher than average positive non run rate revenue items, including roughly $18,000,000 in tenant settlement payments in India and around $13,000,000 in positive items in the U.
S, composed primarily of revenue reversals and some back billing. And lastly, the day 1 revenue associated with the over the course of the last year contributed more than 1% to our global tenant billings growth. Around 2 thirds of these sites were constructed internally, including nearly 1300 new sites in the 3rd quarter, almost exclusively in our international markets. Both the number of new builds and the over $1,000,000 in run rate revenue added from them were nearly double that of Q3 of last year. In average day 1 NOI yields on Q3 builds were around 13%, illustrating the attractive return profile of these sites, and we expect to sustain strong new build momentum into the 4th quarter.
Turning to Slide 7, we also generated solid adjusted EBITDA and consolidated AFFO growth during the quarter, fueled by the strong revenue growth I mentioned as well as our continued focus on operational efficiency. Adjusted EBITDA grew by over 12 percent with our adjusted EBITDA margin coming in at 62.9%, up 30 basis points sequentially. This reflects strong gross margin conversion as well as the benefits of the AT and T MLA, which helped to offset the margin impacts of India carrier consolidation driven churn. The adjusted EBITDA growth attained also factors in the negative impacts of around $16,000,000 in slow pay related bad debt reserves we booked during the quarter as well as roughly $10,000,000 in negative FX impacts. As a result of the strong cash adjusted EBITDA growth in the quarter, we also drove strong consolidated AFFO and AFFO per share growth with consolidated AFFO growing by 8.5% and consolidated AFFO per share up over 8% to $2 Meanwhile, AFFO attributable to common stockholders grew 10.4% or nearly 10% per share.
On a normalized basis, consolidated AFFO and consolidated AFFO per share grew by approximately 11%, continuing our long track record of delivering double digit growth in this key metric. Now moving to Slide 8, let's now take a look at our updated expectations for 2019. In the U. S, we are raising our full year organic tenant billings growth outlook to now more than 7%. This increase is based on a stronger than expected 3rd quarter and continuing expectations that growth will be around 6% in Q4 based on some tougher comps, slightly higher churn and some deceleration in new business activity relative to the record levels we have seen over the last 18 months.
This is also a reflection of 2019 being a front half loaded year in terms of new business, as roughly 2 thirds of our expected full year new business commencements were booked in the 1st two quarters. In aggregate, we're outpacing our initial revenue expectations due to much of the activity driving that outperformance being pulled forward. We're also raising RME organic tenant billings growth outlook to more than 7% for the year based on lower than expected churn in the second half of the year largely due to timing. Meanwhile, trends in Latin America and India are consistent with our prior assumptions, and we are reiterating organic tenant billings growth expectations of 7% to 8% in Latin America and 8% to 9 percent in India on a normalized basis for the year. In aggregate, we continue to expect normalized organic tenant billings growth for international to be roughly 100 basis points above that of the U.
S, driven by record new business volumes. And unlike in the U. S, we expect the year to be slightly second half weighted across our international markets in terms of new business run rate additions. Finally, and as we've said in the past, we continue to expect a return to positive organic tenant billings growth in our international business next quarter on a reported basis as the negative impacts of the carrier consolidation process wind down. Before moving on, I do want to briefly touch on last week's Supreme Court ruling in India on the definition of adjusted gross revenue for the wireless carriers and what it potentially means for our 2020 growth expectations in India.
Prior to the ruling, we anticipated that organic tenant billings growth rates in the market would approach historical rates at some point in 2020 after coming out of 2019 with solid gross new business trends and ICC related churn impacts fading. Today, due to the court decision and the lack of certainty with how and when the carriers and the government may respond, it's too early to determine whether those growth expectations are still reasonable. By the time we issue our 2020 outlook in February of next year, we anticipate having significantly more information available to provide a much more complete view. And having said that though, I'd also note that over the long term, we continue to believe that for ubiquitous 4 gs to become a reality in India, significant incremental network density and wireless capital investment is necessary, and we would expect to benefit from that given our comprehensive Indian footprint. Looking at Slide 9, we're raising our expectations for 2019 consolidated property revenue by $180,000,000 or 2.5%.
This reflects the $167,000,000 of straight line revenue benefits attributable to the AT and T MLA, the benefits of the non run rate items earlier and some organic outperformance in Q3 relative to our prior expectations as well as roughly $5,000,000 in additional FX neutral pass through revenue. These items are expected to be partially offset by around $32,000,000 from unfavorable FX trends, primarily in the 4th quarter. Additionally, I would note that as a result of continued strong demand for our real estate and carefully constructed master lease agreements, our book of non cancelable tenant revenue stood at approximately $46,000,000,000 at the end of Q3, up from $34,000,000,000 last quarter. This equates to more than 6 years of annual revenue for us, creating a strong foundation for the business for years to come. Turning to Slide 10, we're also raising our expectations for adjusted EBITDA by 4% or $180,000,000 This reflects the straight line revenue benefit from the AT and T MLA and the conversion of the other property revenue upside from Q3 as well as the benefits from the continuing direct cost outperformance, particularly in India.
This is being partially offset by the bad debt reserves around $6,000,000 in U. S. CIP write offs we expect in Q4 and around $16,000,000 in unfavorable FX translation impacts. Consistent with our prior assumptions, we would expect our adjusted EBITDA in Q4 to be down sequentially as the positive non rate items we saw in Q3 are not expected to recur. Finally, we are increasing our expectations for consolidated AFFO for the year by $5,000,000 or 0.1%.
This is primarily being driven by the cash adjusted EBITDA outperformance in Q3 as well as $5,000,000 in additional expected interest income, partially offset by around $5,000,000 in incremental cash taxes across our international operations and roughly $13,000,000 in FX headwinds. Additionally, and as you can see on the slide, our revised outlook includes roughly $14,000,000 in distributions to PGGM, our JV partner in Europe, which was not contemplated in our prior outlook, as well as the assumption that year to date underspend of maintenance CapEx reverses in the 4th quarter. Finally, and just to remind you, we recorded a sizable settlement with Tata in Q4 2018, so the non recurrence of that will further impact year over year growth rates for us next quarter. For the year, we continue to expect to once again deliver double digit AFFO per share growth on a normalized basis of roughly 11%. Turning to Slide 11.
Our capital allocation plans for the year remain consistent and are governed by the same investment discipline we've used historically. We expect our dividend in 2019 to grow by around 20% to a total of $1,700,000,000 subject, as always, to board discretion. In addition, at the midpoint of our outlook, we plan to deploy $1,100,000,000 of CapEx, around 85% of which is discretionary, focused on augmenting the capacity of existing macro sites, building new ones and buying land under our sites in the United States. This is up $50,000,000 as compared to our prior expectations, driven in part by a 750 site increase in our projected build program suit to $42.50 at the midpoint. In total, we anticipate that these sites will add more than $3,000,000 in monthly run rate revenue.
We also anticipate spending $15,000,000 more than our prior outlook on land purchases in the United States. We've had tremendous success with our land program over the last 5 years, leading to margin benefits of over 50 basis points for our U. S. Business. And as of the end of the quarter, 73% of our land in the U.
S. Was either owned or under a lease of at least 20 years. And further, we continue to add incremental scale through acquisitions with about $700,000,000 being spent on the nearly 1,000 sites acquired year to date and another $1,850,000,000 or so, including the assumption of existing debt committed to the Eaton Tower transaction in Africa, which we are targeting to close by year end. And as we noted in our press release this morning, we're now working with MTN on buying out their stakes in our joint ventures in Ghana and Uganda, which we would expect to happen in the first half of twenty twenty, assuming the closing of the Eaton transaction. And finally, we paid $426,000,000 for the first set of India puts in March and expect to pay an additional $350,000,000 for the 2nd set by the end of the year subject to regulatory approval.
So for the full year, we anticipate deploying about $6,000,000,000 of capital while keeping our leverage steady and without the need for equity financing. Looking forward, we expect to continue to evaluate a number of attractive portfolios to further diversify our portfolio and enhance our long term AFFO per share growth. As you can see in the middle and right charts on the slide, this same disciplined capital allocation process has enabled us to grow our consolidated AFFO per share at an annual average rate of 14%, while increasing return on invested capital by 140 basis points over the decade. Looking forward, we expect to prudently deploy capital towards growth initiatives globally as we enhance and further diversify our business. Although the business today is far larger than it was in 2,009, we continue to target double digit annual consolidated AFFO per share growth and an increasing return on invested capital.
Turning to Slide 12, we believe that our balance sheet is a key differentiator for American Tower and has been a critical component supporting our historical growth. Since becoming investment grade a decade ago, we've been focused on optimizing our capital structure from a cost of capital and investment funding perspective. This has included more than $25,000,000,000 in debt issuances composed of unsecured, secured, local currency denominated instruments as well as select issuances of mandatory convertible preferred stock supporting large strategic M and A transactions. Through this process and taking advantage of the market rate environment, we've reduced our weighted average cost of debt from roughly 5.1 percent as of the end of 2009 to around 3.5% today, including our latest issuance in early October. At the same time, we've moved the weighted average tenor of our debt out by roughly 1 year.
And this, along with the tremendous underlying cash generation capabilities of our business, has enabled us to deploy over $40,000,000,000 to a combination of dividends, CapEx, acquisitions and share repurchases over that time frame, while maintaining leverage levels solidly in the investment grade range. This balance sheet strength and flexibility has facilitated the expansion of our portfolio from around 27,000 sites in 2,009 to more than 171,000 today and a nearly fourfold increase in AFFO per share over the same period. Today, we are more focused than ever on maintaining and enhancing our balance sheet and continue to believe that it positions us well to drive continued portfolio expansion, make investments in growth, absorb and manage risk and achieve attractive total returns for our shareholders. And turning to Slide 13 and in summary, our 3rd quarter results were strong with demand trends across our global footprint supporting sustained leading organic growth. We were able to come to terms in a new MLA with a key tenant in the U.
S, augmenting our visibility into future growth and adding to our contracted revenue base, which now stands at over $46,000,000,000 Internationally, we continue to capitalize demand for denser, more ubiquitous mobile networks by constructing record levels of new towers for large multinational tenants at attractive day 1 yields. And finally, we continue to deliver strong consolidated AFFO per share growth, extended our long track record of meaningfully growing our dividend and opportunistically added around 400 towers to our foundational U. S. Footprint. As a result of our year to date performance and consistent expectations for Q4, we raised our outlook for all of our key measures and look forward to finishing the year strong and positioning the company for continued success.
With that, I'll turn the call over to the operator, so we can jump into some Q and A.
Thank you. Our first question is going to come from the line of Ric Prentiss from Raymond James. Please go ahead.
Thanks. Good morning, guys. Hey, Ric.
Hey. Couple of questions. First, some equipment companies and tower companies have noted the T Mobile, Sprint, DISH deal decision being delayed and uncertainty on the approval has caused maybe a temporary pause in the marketplace. And don't want you to talk to any specific customers, but in aggregate, have you guys seen any similar trends as far as maybe a pausing or slowing down as people trying to figure out which way the road is going to fork?
Rick, it's Jim and good morning. Over the years and you've been watching this as long as I have, the conventional way to deploy wireless networks in any market is a build out period, grooming period to follow and that results in more or less kind of a sine wave type of deployment schedule with some ebbs and flows. And that I think is all that's happening now in the U. S. It's very typical.
The carriers will whichever one will adjust that sine wave to their particular circumstances at the time. So we look at this as normal course.
Just to remind you, Rick, I mean, mean, as far as the years goes, much of our new business was pulled forward and we believe in the first half of the year, which has generated outperformance from a revenue perspective. But our expectations for this year from a new business perspective are second only to 18. And in international, we expect record levels. So it's been a tremendous year. And as Jim said, it's timing, there are ebbs and flows that carriers will experience when they're deploying their networks.
Makes sense. And then the other question, Jim, appreciate your 5 gs comments. There's obviously a lot of interest what's happening there. In particular, noting what you're talking about with indoor systems and the CBRS that auction scheduled right now for June of 2020. How do you see indoor systems?
What role would American Tower play? What's kind of the level of opportunity and the ability to put your balance sheet to work?
Sure, Rick. So we've got 400 indoor DAS, traditional indoor DAS systems up and running today. We've got rights for thousands of properties in the United States and we've got similar kind of portfolios in whether it's Brazil, India and other countries. So we've got what we think is a leading independent position in most of our markets on indoor installed systems. The goal is to expand the addressable market of the buildings that can actually economically support such a system.
When you're using traditional, even neutral host architecture that we have today under kind of a 4 gs regime with license spectrum, the carriers have to put a base station in the building, there has to be fiber wired throughout. It's a fairly expensive proposition to use the conventional technology. What CBRS will do and cost is increasingly important as I talked about earlier outdoors and indoors. It will drive the cost down of installing and operating these indoor networks once the 3.5 gigahertz phones are deployed and the transmissions are coming the mobile operator systems. And then the interior backhaul doesn't necessarily have to be fiber in that situation.
There's other ways to get it back to their base station. And so therefore, the costs should dramatically drop. The scaling, we hope, can ramp fairly rapidly as to the addressable number of buildings in the United States that you can get to with a CBRS and ultimately 5 gs kind of technology because the cost for building will be lower. Okay. And you guys
would be the neutral host kind of concept?
That's correct. So we'd work with the landlord, do the install, operate the shared part of the system and then there'd be different ways of connectivity back out to the cloud and back out to the mobile networks. Very good. Thanks, guys.
Thank you. Our next question then will come from the line of Jonathan Atkin from RBC. Please go ahead.
Thanks. So I wanted to ask about the innovation program and how much capital you would look to kind of deploy in terms of rough ranges on an annual basis? And maybe related to that, although I think it would be bracketed separately, but appetite for further fiber investments in Latin America or prospectively in Asia?
Hey, Jonathan. To date, spent probably $1,000,000,000 slightly over $1,000,000,000 largely on some of the investments that we've made in Latin America Atlanta. So I would say probably over $1,000,000,000 And the future will depend upon the return profiles of each of these initiatives.
And Rick, when it comes to John, when it comes to added fiber in Latin America and some of the other markets Tom referred to, again, it's going to be return based. But the real goal of our fiber deployments for ATC is to increase the lease up rate and the value of our macro tower assets in those markets, because unlike the U. S, there's not a vibrant fiber optic cable industry in most of these countries. And whether it's ourselves doing funding, bringing in partners to bring fiber to the tower, we've got to solve that issue to get a viable 4 gs signal back into the cloud. So we've got about 600 towers that we've literally wired with fiber in Mexico, 50 in India as we're just getting started and 25 in Brazil so far.
That's the real goal of the fiber investments. Now what comes along with that, which is pretty brilliant and Olivier Pouche runs our international business and has led this development, which is we can be a neutral host provider as we're running by neighborhoods, we're running by enterprises to get to our tower and we are now working with some of the mobile operators in those countries that want to do fiber or fixed wireless to the home that we're passing or the enterprise that we're passing. And we have nothing to do with that commercial arrangement between the enterprise, the consumer and the MNO, but they're leasing that fiber run from us that we're going to probably do anyway for our tower. So we're going to try to scale that. It's a great combination and it's really starting to work in Latin America first and we think in other places too.
Thank you. And then with the AT and T towers having recently traded about 1,000 or so, just wondering how you view the puts and takes around the purchase of carrier built towers in the U. S. Either related to that or prospectively other portfolios that still exist?
Yes, we can John, we continue to look at everything that might be available in the United States. And there are some smaller portfolios that are being put together. And just as part of our normal investment committee business development process, we'll a good hard look at them and see what kind of value can be created and whether that's a good use of our capital.
And there aren't any sort of game changing sized portfolios left in the United States anywhere, held by carriers or even smaller private companies. They're definitely worth going after. As Tom said, we accomplished one of those transactions already this year and really augmented our portfolio nicely and we'll continue to do that.
Finally, and you've mentioned this on other calls, but the operating efficiencies around fuel management in Africa and somewhat in India. And are you in the earlier or the later innings of realizing improvements on that side operationally?
John, I would say we're very much in the earlier innings of that. I mean, we started these initiatives, particularly in Africa with a lot of the newer base technology, gensets and things like that. And we've achieved some really terrific success, I think, in those markets, but we're really still at the early innings of being able to enjoy the future benefits of that.
Yes. And again, it all comes down to the customer and the viability and the performance of their site. We're very customer focused in that when we put investments into energy management, they are designed to increase the uptime of the site above and beyond any of our competitors in the country that you're speaking of. It's also designed to get the cost down for our customers. The less fuel that we use, the less we pass through to them.
So they're more efficient sites and therefore, our attractiveness for new tenants and stickiness of those existing goes up and that's two important reasons why we do it. And the third part is that it reduces the carbon footprint of the industry in that country. So, we've got 3 great reasons to pursue what Tom just described, and we use the same ROI criteria that we use with every other investment.
And just to size it, we probably spent roughly $30,000,000 $40,000,000 last year, dollars 50,000,000 $60,000,000 this year, And we're really excited about a lot of these solar battery enhancements that are coming.
Thank you.
Thank you. Our next question will come from the line of Phil Cusick from JPMorgan. Please go ahead.
Hi guys. Thanks. First, I guess not to get into guidance yet, but with a 6% exit run rate in 4Q, is there a reason that you can help us with that we should consider an acceleration in the U. S. In 2020?
Phil, I will talk through that in February. We'll the reason that we do wait until February is that we're really looking for a lot of the guidance that's being delivered by our customers. In the markets that we're spending, the markets that we're in, the carriers over the last couple of years have spent upwards of $40,000,000,000 to $50,000,000,000 And so we want to make sure we understand where that's actually going to be spending, who's going to be spending and what are they going to be spending on to really then come out to our investors to really get real clarity in terms of what we would expect for the year. So I would hold off that question for a few months and we'll clearly address it in February.
And maybe one more that's a I know a stretch. I know it's early, but the India tax ruling, I'm surprised there's no impact of that on the value of the put that you guys pay this quarter. Anything we should think about there?
Bill, as far as the put goes, that's a pre negotiated floor price and we're working with Tata and others to complete that transaction. We will go through with that. We're pleased that we're actually going to have upwards of 90 ownership of this asset now going forward and it's a rupee based payment. So the currency actually is a benefit to us given where it is today. So the put will be accomplished.
We'll get our ownership up where we want it, keep working with Macquarie in the market and grow the business.
Got it. Thanks guys.
Thank you. Our next question then will come from the line of David Barden from Bank of America. Please go ahead.
Hey, guys. Thanks.
A few questions if I could. Just the first one, following up on the India situation. I mean, I think that with or without the Supreme Court ruling and I think with the Supreme Court ruling, it gets tougher. Idea Vodafone, current course and speed, this doesn't have the capital to continue spending on the network through all of 2020. So when you kind of have a base case expectation that 2020 irrespective of the Supreme Court ruling would actually grow at some kind of normal rate.
Where do you get the comfort level that idea Vodafone is going to have the capital to spend? And how do you see that situation resolving itself? That'd be the first question would be helpful. And then the second question would be, Sprint's been doing some kind of innovative things with Altice in terms of trying to impact their capital intensity with the strand mount network architecture. And it's conceivable that, that could increase in terms of its use, if it could go to Comcast, it could go to Charter.
Could you kind of talk about how you see that approach to network deployment impacting macro tower demand? And then my last question, I guess, is for you, Jim, in particular, is kind of one of the finest things that's happened this year in the sector is that, big fund families have been changing their benchmarks to include towers. And from your perch at the NAREIT Board, do you sense that there's any movement there to think again about where towers sit in terms of the NAREIT benchmark hierarchy? Thanks guys.
Sure. So we'll tag you to pitch 1 at a time.
On Via
Vodafone, they've got a plan in place to spend $4,000,000,000 in CapEx through 2020. They did a $3,500,000,000 capital raise. IDEA and Vodafone are matched together. They've merged on purpose to be able to get the wherewithal to compete in 4 gs. And over the short to mid term period, we think they will continue to be competitive and work through their issues with the tax department and be a player.
There's an interest, I would imagine, in the Indian government, whose top level political regime has a program called Digital India, which is going to require at least 3 or 4 healthy mobile operators to pull that off. And this is something we expect to be worked through. And we also expect idea Vodafone with everything we know now to be part of Digital India going forward as they work through their issues. It will take some time to resolve and we'll continue to progress. Meanwhile, GEO and Airtel are moving apace and we'll continue to work with all 3, plus BSNL, which is a customer of ours.
So we'll go forward with that industry structure.
Before Jim takes the next pitch, Dave, I just want to remind, I mean even in the Q3, we generated 8% growth in the quarter on a normalized basis. But we've said that high single digit, kind of low double digit growth was what we expected. So we're already there in the quarter. So that gives you some sense of the level of spend that's being delivered in the market.
And then with essentially small cell type mountings and ancillary type of transmitters or repeaters, they've always been out there. There will always be more. The macro site is the foundation of all of these networks. It's going to continue to be. And when there's fill in capacity or a niche that's not getting great coverage, these kind of solutions should be employed.
But we don't think they'll materially affect our macro business. And then on fund benchmarking, the buy side is voting with their feet, we think, in the real estate industry and moving forward with including in their own benchmarks, as you know, Vanguard has been one of the leaders there, the most visible, but a number of the REIT type investor firms are moving towards technology real estate assets. As far as when the NAREIT index will include it or the RMZ, it's their decisions and I don't have access to those discussions.
Okay. Thanks, Jim.
Thank you. Our next question then will come from the line of Nick Del Deo from Moffett Nathanson. Please go ahead.
Hey, thanks for taking my questions. Sort of following up on India,
as you discussed, there are
a couple of financially distressed carriers there. They still own stakes in towers that might get shaken loose. The valuation described a big tower co there, Barti and Fortell, I think it's like 5x EBITDA. If you really believe in India's future, does this not present an opportunity for a well capitalized player like yourself to really make some big moves and create value over time? Or is there other considerations we
should bear in mind?
We absolutely think that's true. And we've moved very aggressively with ideas towers, Vodafone's towers and the 40 some percent of the joint venture with Tata that we haven't owned. And so all those assets we've brought in and we're happy with our current footprint at the moment and we're going to grow organically off of that as everything kind of gets back on track in India. We'll keep our eyes open as we go along. There's some very relatively small tower portfolios still around, and we look at those, of course.
But as far as something more strategic, we're going to operate what we have and integrate it all for the short term and get where exactly we want to be positioned to look at larger deals down the road.
And Nick, just in terms of the capital we're spending there, we increased our capital for new builds in the quarter for the year, and that's exclusively India. And so we're spending more capital there on increase our exposure and expand our footprint in the market.
Okay. That's helpful. And then Jim, in your prepared remarks, you noted that, I think you said less than 1% of your U. S. Sites are in locations with population densities sufficient to support outdoor small cells.
So I thought it was a really interesting statistic and it was pretty specific. What's the population density threshold that you're using to define what sites fall into that bucket? And understanding that's a small share of your business, have you actually observed any impact on new leasing due to small cells on those 1% of sites or rooftops you might manage in those areas?
The answer to the second part is no. There's been no impact on any of this sort of deployment on our macro site or rooftop management business. Remember that base layer in urban, suburban or rural needs to be there. It's not going to be taken away or tried to be chopped up. You just really shouldn't be taking that approach to network deployment.
But on the first instance, we use a metric of 10,000 people per square mile as the economic threshold with current technology and current costs. And we actually stress tested that. If you were able to cut the input costs in half, it ends up being about 5,000 people per square mile. So, even at the 5,000 person level, 85% of Americans live in places lower than 5,000 people per square mile. And so that's the network architecture in the U.
S. That is going to continue to be macro based just on that fundamental right there. There's no way to cover 320,000,000 people at under 5,000 people per square mile topologies with small cells economically or even technically.
Got it. Thank you, guys.
Thank you. Our next question will come from the line of Simon Flannery from Morgan Stanley. Please go ahead.
Great. Thank you. Good morning. On C band and 5 gs, I think you talked about having a 5 to 10 year runway of growth from that deployment. When do you think this can really kick off?
If you take the C band alliance proposals over the last few days, they're now looking to free up 100 megahertz within 18 months of a report in order. So that potentially gives you leasing in the 2021 timeframe. So you might get it, particularly if it goes to multiple carriers, some decent volume in the fairly near term. I just wanted to see if you think that is feasible. And then around that and around CBRS, what interest are you seeing beyond the big four carriers in terms of looking at potentially deploying that?
Thank you.
Yes. Thanks, Simon. And the next couple of years, early 2020s, we expect it to have some commercial activity and it will be a long term bill as you've even described a moment ago. We've got some trials, really interesting ones up and running now with some property owners actually. So without disclosing who they are, what specific business they're in, if you think about what venue owner could leverage 5 gs technology, you would think of far as an owner of a commercial piece of real estate, whether it's a far as an owner of a commercial piece of real estate, whether it's a warehouse, a factory, office building, just to manage again cost down and operating those buildings.
So those are the kinds of verticals that, so to speak, that we are working with on trials and other initiatives right now. And some of them are actually you may or may not know it, but they're up and running on 4 gs and licensed technology in certain trial sites already. So that's how we're going after this. Very methodical, ROIs on everything. CTO and he watches over the viability and sustainability in the business cases for these projects and Tom, of course, has got the financial side and we have Innovation Council that actually has to run through our investment committee when there's a sizable proposal that comes down.
So that's how we're going to go about it. We think there's real prospect there. It will take some years to play out.
And is a lot of this activity in CBRS?
Not necessarily right now, but again, we're trialing it with either licensed spectrum, private 4 gs LTE kind of setups and things like that at the moment, because when CBRS frees up and that spectrum, the general access is here. And then the priority access, as you described, will be auctioned. But once that becomes available, in both sides of it, especially if you have a commercial building that you can use the general access network to serve that because it's unlikely that the general access will be burdened within your building unless it's you that is controlling it. So that's available, we think, on a wide scale. And the last point I would say is that, again, it's going to take some time and the devices need to roll out as well and that's going to again take its own pace.
Great.
Thanks a lot. Thank you.
We have a question from the line of Michael Rollins from Citi. Please go ahead.
Thanks. Good morning. You mentioned as part of the change to 4Q domestic site leasing that there was a pickup in churn. Can you expand on what you're seeing on the churn front? And is there anything that's on the horizon just from past mergers or integration events that we should be thinking of?
Thanks.
Michael, I mean, it's up from a bit from Q3, but it's de minimis. It's below 2%. I think it gets up to 1.7% or something like that from the 1.6% or so that we had in Q3. So it's just an enormous across the footprint that we have in the United States.
It's just mostly timing, Rick, when certain contracts come due or certain we have other verticals that are consolidating and other things that are networks. And it's as Tom said, it's within the 1% to 2% we've had for 20 years. So really nothing different.
Thanks.
Thank you. And then our final question will come from the line of Colby Synesael from Cowen and Company. Please go ahead.
Great. Thank you. Sorry to go back to the India thing, but I just thought I'd ask one more question around that. Just given the local law in India, to the extent one of your customers was to declare for bankruptcy, would you expect to be paid through that process? And I guess, are all of your tenants in India currently, up to date, if you will, in terms of payment?
And then secondly, going back to the first question you got from Rick as it relates to T Mobile and your response being ebbs and flows. Just given how long we might be at a slowdown from summer of your U. S. Providers, would you anticipate that just because of that length, it could actually be material enough that we would see some moderation in your growth rate for some period of time measured in quarters, if you will, as a result of what you're seeing now when you think about the book to bill as we go into 2020? Thank you.
Okay. We're getting multiple pitches again Tom to finish up here. But Colby, we'll take them one at a time. There is a bankruptcy lot in India that was put in a couple of years ago, sort of nationwide. And the network is still going to be needed.
The subscribers are still going to have to be served. If there's a financial restructuring, as we've had in other selected other international markets over the years. The networks tend to operate and they need to pay the electric bill and the lease to keep that network going and that tends to be what happens. So that's our historical experience when there's a financial restructuring that goes on. Our customers by and large in India as far as their payment schedules and on time, if you will, is not materially different than it has been in the past in that market.
BSNL, which is I think our smallest customer, is a little bit behind what it's typically doing, but that's going to catch up, we expect. And then thirdly, cycling back to the U. S, with T Mobile, look, the data rate usage is still going up 30% a year. T Mobile, if it's got a bit of a delay to its merger or not, is trying to take share, trying to stay relevant and adding customers and data usage. So it's going to be up to them to answer that question, but we don't expect that there'll be a long cycle sort of hiatus from any of our customers going into 2020.
But as Tom said, we'll look for the CapEx guidance, we'll look for the public statements and we'll add up what we see in our field operation as far as
applications and interest in the sales force
with our sites. And
February.
Great. Thank you.
Good, Bobby.
All righty. Well, thank you everybody for joining this morning and have a great day.
Yes. Happy Halloween to everybody and a good weekend to all. Bye bye.
Thank you. Ladies and gentlemen, that does conclude our conference for today. Thank you for your participation for using AT and T Executive Teleconference. You may now disconnect.