Sprint vs. Verizon Standoff on Network Reliability and Capex

Verizon wasn’t happy about Sprint stealing their “Can you hear me now?” spokesperson or Sprint’s efforts to suggest that Sprint’s network reliability performs within 1% of AT&T and Verizon‘s networks. Verizon issued their own press release full of pretty interesting facts that attempt to demonstrate that network reliability is important- but that Verizon’s network is far more extensive and they have devoted far more resources to their network performance. Verizon can’t overtly come out and say that they think that Sprint is gaming the third party network analyzers by focusing their limited capex on improving metrics in enough markets to make claims like they have, knowing that no third party analyzer can test the breadth of coverage and capacity that Verizon has built into their network. So instead- they provide the following:

Verizon points out that they have invested over $116B in the last 16 years- more than any other wireless company, which included $5B over the first 6 months of 2016.

Most importantly, we get to see some real numbers on the number of macrocells, small cells, and DAS systems deployed by Verizon in 2016. Here are what Verizon indicates that they have deployed in 2016. (On small cells, Verizon isn’t clear in their press release whether this is total or just in 2016. Personally, I would have thought it was high for just 2016, but low for a longer timeframe).

Small Cells=     2,100

DAS Systems:      350

Macrocells:     1,500

Site Mods:     1,700

I sincerely doubt that we won’t see any similar response from Sprint. For reference, here are our best guesses as to Sprint’s network efforts in 2016 to date. If you think otherwise, please don’t hesitate to let us know.

Small Cells=      200

DAS Systems: 50

Macrocells:      100

Site Mods:     2,000

One thing I find interesting is that Verizon indicates that their network “is engineered to provide, on average, 5Mbps to 12 Mbps while downloading and 2Mbps to 5Mbps while uploading.” The key here being that this is the AVERAGE of their entire network. This seems to me to be a veiled challenge to Sprint to respond with the design standard for Sprint’s entire network, not just the areas they focused their limited network capex on to improve their 3rd Party analyzed test results.

Whatever the case, I love seeing the actual site deployment numbers from Verizon. Unfortunately, given the lack of activity by AT&T and Sprint this year, we won’t see similar disclosures by them. T-Mobile is a wild card in terms of disclosure.

 

Crown Castle Small Cell and Tower Update- 2ndQ 2016

New replacement pole small cell
New replacement pole small cell

While the call itself was pretty understated as compared to even other CCI calls, it was in the Q&A where the call got interesting.  Here is what we took from the call.

TOWERS:  ($115M organic revenue growth in 2nd Quarter)

New Builds:  In regards to new builds, CCI is not building many new towers- only 50 of them in the last quarter.  They don’t say it, but we believe that the majority of those towers are replacements of their existing towers where the underlying landowner wasn’t willing to extend the underlying tower ground lease at a fair market value rent.   Crown doesn’t expect this slow pace to change, noting specifically that there are a number of new tower company entrants or established mid-tier tower companies that will do non-sensible build-to-suit deals in order to establish market share.

Collocations:  It appears that Crown had a limited number of new collocations and that the primary justification for revenue growth was from modification work on existing towers.   Crown expects that 2017 will be similar to 2016 but that 2018 may see some increase in terms of new colos as FirstNet and possible other holders (hoarders) of spectrum who haven’t built anything out may seek to start building.  (DISH)  Personally, I believe this is optimistic.

Ground Lease Buyouts/Extensions:  Crown Castle closed about $19M in land purchases during the 2nd quarter.   This represents about 75-100 leases that were purchased.  Crown now controls over 80% of their “Site rental gross margin” for more than 20 years.   We would be curious about how many actual leases as a percentage of ground leases that represents- as most of their single tenant towers have little to no margin.

AT&T and T-Mobile “Noise about Possible Relocation”:  Unsurprisingly, Crown was pretty vocal in response to concerns by analysts about AT&T’s RFP to move colos off expensive CCI or other tower company sites.   They suggest that the return on investment to the carriers to move (2.5% to 3%) isn’t justified.  I believe that CCI’s numbers are optimistic but that relocation isn’t justified except in rare circumstances.  (High colo rent in rural areas with low barriers to entry for new towers or relocation to another similar existing tower nearby that isn’t owned by the Big 3 tower companies)  We regularly do our own internal analysis of when relocation is justified for our clients who are considering lease extensions of CCI or AMT expiring ground leases and the results rarely justify relocation.   On a macro-economic basis, relocation of existing collocations really doesn’t make sense.  On a site specific basis, it can.   If you need help evaluating a threat to relocate your tower please reach out to us.

SMALL CELLS:  ($55M in organic growth in 2nd Quarter)

New Site/Collocation Ratio:  Crown is experiencing 30% collocation and 70% new system builds for their small cell buildouts.   The collocation number includes both additional nodes within existing small cell fiber routes and “laterals” that stretch from the existing routes.   We believe that there should be more visibility provided to what percentage of “collocations” require new laterals as opposed to those that just require addition of the specific node to an existing fiber route.

Location/Difficulty:  Crown clarified that 90% of their revenue is from the top 10-15 markets in the US.  Only 2% comes from valuable locations outside those markets- which consist primarily of individual venues.  The vast majority of these builds are capacity builds, not coverage although there are some coverage builds in hard to zone neighborhoods, such as the Denver example they held out earlier this year that we questioned was representative of their typical small cell build.  This is as to be expected given the location of the Sunesys fiber routes.  Crown alleges that it takes between 18-24 months to procure permits and authorization to build new small cells.   They don’t clarify if they mean locations where there is not existing Crown owned fiber or not.   We assume the latter.   This doesn’t match what Mobilitie is claiming- they are indicating that they are ahead of schedule with their permitting and acquisition of small cells with the exception of some difficult markets.  (We suspect otherwise.)   Crown directly addresses that other companies don’t have the RF design experience and/or operational experience necessary to compete with Crown’s depth.

Competitors in Small Cells:  Interestingly, one of the questions from the analysts was about rumors that the analysts had heard that some of the wireless carriers were in discussions directly with metro fiber providers to attach small cells directly to those provider’s fiber rings.   CCI claims that this hasn’t impact their returns yet, but that doesn’t mean anything given that it is still too early to see whether metro fiber providers and the above referenced carriers will choose competitive locations in the same area as Crown’s fiber routes or whether they will collocate on the same route.   I anticipate that the answer will be contingent upon pricing and available assets.   This is a material question that bears further investigation and clarification from Crown- especially since as they admit in their earnings call they see 20% margins on their third tenant on a fiber route.   Given that Sprint is solely focused on using Mobilitie for the time being, that means that Crown has to get all three remaining carriers (AT&T, VzW, and T-Mobile) in order to see the expected returns.

Crown anticipates that 50% of small cells currently are self-built by the carriers, while 50% are built by third party companies like Crown.  We don’t know whether Crown infers that Mobilitie/Sprint is self-built or third party built.  If the latter, Crown’s share of the 50% that are third party builds outside of Mobilitie/Sprint should be fairly high.

Capital Reimbursement:  Crown indicated in the call that they have paid approximately $3B in capex towards their small cell systems.  ($1B of which was for Sunesys) Of the remaining, $2B, it appears that Crown has collected $500M of that capex from the carriers in the form of prepaid rent.

 

AT&T Forecasts 6,700 New Macrocells from 2017-2022

In a presentation at the Cohen and Company’s 44th Annual Technology, Media, and Telecom conference, AT&T’s CTO and President of AT&T Labs Krish Prabhu indicated that AT&T believes that small cells will make up the substantial majority of their future cell tower and cell site development over the next 5 years.   In response to a question on how small cells will fit into their future 5G, fixed wireless, and IOT networks, Prabhu indicates that AT&T has approximately 90% of their macrocell network that they expect to be in place in the next five years already standing.   However, they have only deployed 5-10% of the total count of small cells that they expect to have at 2022.

Based upon our best information, AT&T currently has approximately 67,000 macrocell sites in the United States.   Assuming 10% growth, that means that AT&T will deploy approximately 6,700 new macrocells over the next 5 years.   We are unsure of the total count of small cells owned by AT&T in the US, but would estimate that number at 5,000 or less.  (If anyone has better numbers, please comment below).   If 5,000 represents 10% of their projected small cell builds, we would expect to see 50,000 new small cells over the next 5 year period.   Below is a chart showing a straight line estimate of new AT&T macrocells and small cells based upon 6,700 new macrocells and 50,000 new small cells over 5 years.

Graph of AT&T small cells and macrocells projected over 5 years

Projected AT&T Small Cell and Macrocell Growth- 2017-2022

A word of caution:  AT&T previously indicated that they would be building 40,000 AT&T small cells which never actually occurred.   In this case, though, we believe that AT&T is likely to build a significant number of small cells and should exceed the 40,000 previously anticipated.  As Prabhu indicates in the discussion, 5G wireless in the millimeter wave bands “gets really bad once you go several 100 meters”.   There are only so many towers and the cost of adding new macrocells is significantly more expensive than adding new small cells.

By way of comparison, AT&T has completed just over 10,000 new macrocells over the last 5 years based upon information provided by RBC Capital Markets.

5G and LTE-U Test Markets for Each Wireless Carrier

We were curious what markets we would start seeing modifications requests come in for new antennas on existing cell towers for 5G and LTE-U installations.   These trials in most cases will require new antennas with new frequency bands.   So we created the map that you can see below- or see the online version of the 5G and LTE-U Test Market map.  US 5G and LTE-U Test Markets by Wireless Carrier- 2016

US 5G and LTE-U Test Markets by Wireless Carrier- 2016

Verizon Small Cell for Comparison

As an addendum to our earlier article about Crown Castle and Mobilitie small cell proposed installations in Orlando, we thought it would be helpful to include a drawing of a Verizon proposed small cell from Massachusetts.   In this case, Verizon is installing a Cloud Radio Access Network (C-RAN) with two small cells to augment their coverage and capacity in the vicinity of an existing Verizon macrocell which is collocated on an existing SBA Communications tower.  Verizon is expanding the baseband unit (BBU) at their macrocell to support these two small cells.  The small cells utilize remote radio heads (RRH) which are connected via fiber to the C-RAN avoiding the need to deploy separate baseband units at each small cell location and shrinking the equipment that is necessary to place on the poles.   This allows Verizon to add capacity and coverage to their existing macrocell network on a much cheaper basis overall by pinpointing where they have service deficiencies.   See below for an image of their coverage map.  Note the item labeled Lowell 9- this is the SBA tower Verizon and the two blue dots are the small cells proposed on existing utility poles.

 

Verizon C-RAN Coverage Map
Verizon coverage plot for C-RAN -small cell

Below is a screenshot of the small cell installation on the existing pole.  Note the large antenna on the top of the pole (12″x28.7″) and the RRH located on the pole.

Verizon_Pole_Top_Antenna_CRAN

Crown Castle (CCI) Small Cell Initiatives and Reporting

Crown Castle DAS Node
Picture of Crown Castle DAS Node from FCC Presentation by Crown

So as a clear indication that Wall Street is very focused on small cell initiatives by the public tower companies, Crown Castle
started reporting their small cell financials separately from their tower financials in the Q1 2016 quarterly earnings and call.   They must have been receiving a significant number of questions from the analysts because the earnings call presentation is carefully crafted to show a rosy picture even though Crown hasn’t been completely transparent on their small cell financials.

SOME VISIBILITY- BUT QUESTIONS STILL REMAIN

In general, we are excited to see them Crown add this reporting, as we have been suggesting to the various analysts that retain us that it is difficult to measure how successful their small cell efforts are without this breakdown. Unfortunately, Crown still isn’t distinguishing between small cells and DAS in the breakdown preferring to treat all DAS nodes and small cells as if they are the same and have similar financial attributes.  Interestingly, an analyst from Bank of America specifically asked this same question in the Q&A without getting a substantive answer.

What we do know from the earnings call is that Crown’s small cell business still amounts to approximately 12% of their consolidated site rental revenue similar to what it was in late 2015.  Crown indicates that new small cell builds amount to 75% of their small cell systems’ incremental revenue – while 25% is additional collocation on existing fiber routes/DAS networks. They suggest that they have 16,500 miles of fiber, but don’t disclose how many miles are actually used for small cell nodes or DAS.   CCI says they are focused on the top 25 markets which isn’t surprising given the location of Sunesys fiber in these same cities. This suggests a few obvious questions for CCI that were partially addressed in this call and should be expanded upon in future calls:

1.  How do they expect to grow once those 25 markets are complete?

2.  Now that the world is fully aware of the value of dark fiber and surplus capacity, is it reasonable to expect another fiber company acquisition?

3.  How many nodes are in top 25 markets or Central Business Districts (CBD) as opposed to non urban core areas?

CROWN’S SAMPLE MARKETS – LAS VEGAS AND DENVER

In the earnings call and presentation, they point to two “small cell” networks that they suggest demonstrate the viability/profitability of their efforts.

The first is a Denver project where they suggest they are seeing a 20% yield. It is important to note that this is an Outdoor DAS, not small cells, and shouldn’t be held up as representative of the returns CCI will see for all small cell installations. For more information, see this description of Greenwood Village Crown Castle DAS.

The second example is in Las Vegas. Crown points to a 13% yield on this project, and notes that they are seeing the second tenant on this network. We assume the first is Verizon, but we cannot conclusively identify the second tenant.  Either way, this is a positive indication of future collocation on fiber based small cell initiatives albeit with limitations.

Both of these systems are what Crown refers to as “vintage systems” –  i.e., older more mature systems. Both are in primarily residential areas where the DAS or small cell installation is needed for coverage, not capacity. In its Q&A, CCI references the probability that these systems will become 3 tenant systems. We suspect that this is likely for the Denver DAS, as adding to an existing DAS provides true value to the wireless service provider. The incremental cost of connecting to additional nodes isn’t significant, and providing coverage in those areas is much more problematic, thereby increasing the probability that 3 tenants will use the system.  For Las Vegas, the 3rd tenant is also likely given that it would be difficult for a wireless service provider to build a tower in this residential area especially since the small cells/fiber has already been deployed and approved by the city.

But to hold out the two examples that Crown did as indicative of the future performance of their small cell networks is at best creative marketing.    Here is Crown CFO Jay Brown’s quote:  “We tried not to pick actually the best systems that we had, but we tried to pick some that we thought were somewhat representative as the markets develop.”   The problem is that these small cell networks aren’t representative of what Crown is currently building.   Most of the activity we see from Crown these days (and we readily accept that we aren’t seeing everything Crown does) is in large metropolitan areas.  Large metropolitan areas are not like residential areas where there is no deployed fiber nor are there conduits already in the ground that can be easily expanded.    The cost to deploy fiber is more expensive in urban areas.   However, the probability of greater density or “collocation” on existing networks is greater as well.

In other words, we question whether these two networks are “representative” of what Crown is actively deploying.   If these aren’t their best systems- we would like to see more information about those systems especially if they are more urban small cell deployments as opposed to DAS or small cells greenfield builds in residential areas.

CROWN’S OPPORTUNITIES AND HURDLES IN SMALL CELLS

For small cells, there are fewer barriers to entry for subsequent wireless service providers in an urban area.   While the ease of access may be attractive, the combination of upfront capex and long term continuing opex isn’t as desirable. Crown suggests that small cells are susceptible to a “land grab,” just as towers were. While true to an extent, in most cases, Crown’s acquisition of small cell rights is not a significant barrier to entry to other small cell providers or the wireless service providers who wish to establish similar and (mostly) equal rights.  As opposed to a “land grab,” we characterize it as more of a “first choice” type opportunity where CCI may procure some select locations, there are in most cases many other locations that will work just as good.   Unlike towers where there is often monopoly type restrictions placed on the building of new towers, small cell agreements with municipalities and utilities are very rarely exclusive such that they preclude another competitor from building a similar network.

Where we do agree with Crown is that they point out that competing established fiber companies don’t necessarily have an edge over Crown.   The wireless service providers are currently looking at fiber holistically and are trying to avoid lit fiber to the extent possible preferring their own dark fiber.   So to the extent that Crown can offer fiber and pole access concurrently, they are at an advantage to local or even national fiber companies.  Furthermore, to the extent that their acquisition of Sunesys included space fiber capacity on the existing fiber routes, Crown has a further advantage.

We don’t wish to seem overly pessimistic about Crown’s efforts in small cells.  To the extent that any company is doing it right, they are in our opinion.   There is limited value add to the wireless service provider of having access to poles.   However, fiber is and will continue to be a core requirement for future densification efforts.   And once T-Mobile actively starts deploying small cells (which they will), they will seek existing networks that are ready to use like those that CCI has deployed or is deploying because adding capacity rapidly is more important to T-Mobile than controlling costs.  Crown’s anchor tenant for most of their new small cell builds is Verizon and as such they are well insulated from future consolidation and well positioned for future densification modifications/expansion on existing small cells by Verizon who seems very focused on adding capacity to their networks with an emphasis on small cells.

As to competitors, Crown is well situated for now.  American Tower (AMT) is looking at small cells in the US- our clients have reported inquiries from American Tower regarding small cell opportunities.  SBA Communications (SBAC) seems content to sit it out for now.   So for now, Crown has the market primarily to itself with some competition from Extenet and Mobilitie and minimal competition from local and national fiber companies.

IMPACT ON TOWERS

If you have read our previous posts, you know that we aren’t that bullish on future growth on existing towers.  Small cells enable the wireless service providers to avoid costly and difficult densification using macro cells.   This isn’t just our supposition: Crown suggests in the earnings call that they are expecting one new tenant every 10 years on the average tower that they acquired from AT&T and T-Mobile.  Contrast this to the expectation in as late as 2006/2007 of one new tenant every four years on the average tower.   While Crown doesn’t attribute the reduced “lease-up expectations” to small cells and efforts by the wireless service providers to avoid future collocation on public tower company towers, it seems pretty clear that they are forecasting lower lease-up as a result.

CONTACT US

If you have been contacted by Crown Castle, Extenet, Mobilitie, or Verizon regarding a small cell master lease, please contact us.   We have worked with airports, subway systems, amusement parks, major and secondary US cities, and private landowners in regards to similar proposals.

T-Mobile’s Rampant Success

Just thought this was worth sharing- a great article on T-Mobile’s Tremendous Success in the last quarter of 2015.  Also of interest is that Sprint seems to be churning its good established customers while adding new customers presumably at 1/2 of the rate of the other wireless service providers.  That can’t be a sustainable business model.

Wireless Adds and Drops
T-Mobile’s rampant success comes at the expense of other carriers- not from new additions.

Pricing War in South Africa with Chinese Tower Manufacturers

In a random web search, I came across this document on tower steel pricing which is from a South African tower manufacturer describing the issues that they are having with Chinese manufacturers flooding the South African tower market with cheap towers.   I can’t comment on the accuracy of the article or the impact in the US- but I just wanted to share it because it is fascinating.

Please note that this is a link to a word document on tower steel pricing- so it may be blocked in your browser.

Tale of Two Small Cell Proposals – Crown Castle vs. Mobilitie

The City of Orlando, Florida recently received two separate applications for two separate sets of DAS nodes and/or small cells. The applications were submitted by Crown Castle and Telemobilitie, both of whom purport to request planning approval for new poles and/or the replacement of existing poles. We believe that Telemobilitie is a “Doing Business As” name for Mobilitie – the applicant who submitted the plans to the City on behalf of Telemobilitie has a LinkedIn page indicating she works for Mobilitie, a company which is purported to be assisting Sprint with the development of 70,000 “mini-macro” small cell sites. As previously mentioned Mobilitie has set up a number of LLCs with different names, for some unknown reason. An examination of the two applications illustrates the inherent issues that companies like Mobilitie will face when proposing new 70′ -100′ structures for the public rights-of-way, as opposed to the 35′ pole replacements being proposed by Crown Castle. Here are the two planning applications and summary details of each.

Crown Castle’s Proposed 35′ Poles

Crown Castle Small Cell Plan Staff Report – submitted for 8 locations in downtown Orlando. Six of the proposed poles would replace existing utility poles owned by the Orlando Utility Commission, and 2 new poles would be constructed (and owned) by Crown Castle. Crown Castle is proposing new or replacement poles that are all under 35′ in height, which therefore negates the necessity of a Conditional Use permit in the City of Orlando. Below is a snapshot of one of the drawings for a small cell. It is not particularly appealing as a light pole, but certainly fairly attractive as a small cell. Note the placement of an electric meter on the base of the pole.

Crown Castle Small Cell
Crown Castle Small Cell Proposed Pole – Orlando, Florida

Mobilitie’s Proposed 70′ Poles

Telemobilitie Small Cell Staff Review – submitted for 2 locations near downtown Orlando for the installation of two 70′ wood poles. Mobilitie submitted these two applications despite the Code’s explicit terms that a Conditional Use permit is required for all DAS node poles or small cell poles that are  taller than 35′. Furthermore, the Code requires that the poles be galvanized metal, yet Mobilitie proposed wood poles. If you examine the Mobilitie proposal, you can see that they are including two antennas on the pole: one at the top for standard cellular/PCS/AWS/2.5 use, and one down the pole for UE relay or LTE backhaul (whereby Sprint will be using their own spectrum for wireless backhaul).

Mobilitie Small Cell
Mobilitie Proposed Small Cell Pole – Orlando, Florida

One of the two poles proposed is to be constructed near an existing parking garage and is likely appropriate for the area. The second is to be built in a lower end residential area where the buildings are less than a few stories tall. It is hard to imagine that a 70′ pole will be approved in this location, especially when there is a 30′ pole just across the road that could have been replaced much less conspicuously. The equipment appears to be attached to the exterior of the pole instead of being built within the pole, a fact which the City of Orlando properly identified as an issue in their evaluation.

Small Cell in Residential Zone
Tele Mobilitie Small Cell Proposed in Residential Area

How are these applications being treated?

So in reviewing the planning staff recommendations in each of these Appearance Review Board reviews, you can see how these two similar applications will end up being treated differently. Because the Crown Castle poles are under 35′ in height, the City of Orlando Appearance Review Board has recommended approval (with conditions including relocation of some poles to nearby locations) for all 8 poles and has reviewed a specific design from Crown for future poles to make it easier and quicker for future approval. Conversely, the City of Orlando is requiring that both Mobilitie poles obtain a Conditional Use approval, which will entail an application and a formal review and hearing process. A Conditional Use review will require notice to nearby residents and a public hearing, which reduces the probability that a pole will be approved.

Another interesting component: if you examine the comments in the Mobilitie application for N. Rosalind Avenue, the staff notes that Crown Castle submitted their application first and therefore has location priority. Telemobilitie will either need to resubmit at a new location outside of separation distance, or co-locate their equipment with Crown Castle on a single pole (which will never happen).

So what does this tell us?

1. First come, first served. Multiple wireless service providers or neutral hosts like Crown Castle or Mobilitie may be interested in a single location – but the first one to submit will typically get it.  In this case, the City of Orlando has required a separation distance between poles such that Mobilitie will be required to either co-locate on the Crown Castle pole (which is next to impossible) or find another pole outside the area of separation.

2. Mobilitie and Sprint have an uphill battle in their efforts to get approval for 70′-100′ poles.  As Crown Castle’s plan illustrates, the poles do not have to be 70′-100′.  We suspect that Mobilitie and Sprint are interested in deploying them at that height to reduce the number they have to deploy and to augment Sprint coverage and capacity without collocating on public tower company towers. That may explain why they only submitted two pole applications as opposed to Crown Castle’s eight.  For more information on why Sprint may be avoiding new collocations, please see our article on Sprint’s plans to avoid future collocations on AMT, CCI, and SBAC.

3.  Is Mobilitie testing the City of Orlando?  We wonder if Mobilitie is testing the City of Orlando to see what they will approve or not and what the process will be. Perhaps they are intentionally submitting in order to establish standing to sue for not approving their poles in the public rights-of-way.

4. The City of Orlando is and should be working on their small cell policy.  Like many other municipalities, it appears that the City was unprepared for what they refer to as the “large number of applications for outdoor Distributed Antenna System [DAS] installations in the public rights-of-way, including many applications for antenna and other equipment on or about various City utility structures.”  As indicated in the staff reports, the City is working on a small cell policy which will undoubtedly make approval more difficult (or at least more clear cut) for future applicants.  I would be willing to bet that the new revision of the small cell policy will likely prohibit or restrict 70′-100′ poles.

How can Steel in the Air, Inc. help?   

If you are a municipality and need assistance with the review of proposed small cell master lease agreements or the drafting of intelligent small cell policies – we are available to help.   We have assisted other municipal and private entities to review and evaluate small cell proposals and draft appropriate policies and standards.   Please contact us for more details.

Key Takeaways from American Tower (AMT) Q4 2015 Earnings Call

American Tower Corporation (AMT) provided its 4th Quarter 2015 earnings call this past week.  In reviewing the transcript, here are our takeaways:

1. American Tower now has upwards of 140,000 towers worldwide – which represents a sizable increase in India of 42,000 towers over 2015 and the 11,500 sites from Verizon this past year.  AMT expects to be at 145,000 sites by the end of the year, which suggests that there aren’t any sizable acquisitions out there for 2016.

2. American Tower built approximately 3,000 towers and 50 new indoor DAS networks in 2015 (almost all of them outside the US).

3. AMT says they have been effective at containing churn due to “skilled master lease contracting.”   However, they don’t provide any details to support this conclusion.

4. Growth for AMT in the US was fairly limited at 6% on an organic growth basis as compared to 11% in international markets.  AMT expects US growth to remain steady in the mid 5% range while international will accelerate slightly to 12%.  AMT does expect the US growth to be straight lined throughout the year – which is surprising because 1Q so far hasn’t been very exciting from our vantage point.

5. Average tenancy per tower of the newly acquired towers in the US and internationally is 1.4.

6. American Tower makes a very clear effort in this earnings call to talk about communications real estate and property revenues as opposed to towers and tower revenue.  This is a clear nod to concern by the market that towers are no longer the only way to provide wireless service.  American Tower generates 3% of its revenue through DAS, and another 3% from rooftops.  Interestingly, American Tower is now coming around to rooftops again – after distancing themselves from their rooftop management history (Spectrasite/APEX) in the 1990s. They allege that they intend to grow these areas, but offer nothing concrete on how or when they will do so.

7. American Tower continues to point to 30%-50% data usage growth in the US and make general aspersions that such growth will correlate to more amendment revenue.  They also suggest without demonstration that the carriers will continue to build out their sites based upon carrier aggregation and the use of higher frequencies which will increase site density needs by all wireless carriers.  We don’t believe that any of the tower companies have sufficiently answered questions about the impact of small cells and C-RAN as they impact tower collocation uptake rates.

8.  Ironically, American Tower has now done an about face from proclaiming how many of their towers are in core urban environments/top 50 BTAs to emphasizing that 95% of their towers are in suburban, corridor, and rural environments.   They do this because they are trying to answer questions about small cell and densification efforts by the carriers that don’t use towers.  (See the Sprint mini-macro efforts – or Verizon’s substantial small cell efforts in suburban and urban areas.)

9.  American Tower claims that 95% of capex is being spent on macro networks, although they clarify that this may fall to 90% in the next few years.   I don’t know where they pull these numbers from – Verizon has indicated explicitly that a majority of their spending this year will be on small cells and DAS installations.  Sprint has limited plans for macro deployments – publicly indicating that they intend to use mini-macros (big small cells) for densification.  Not unexpectedly, American Tower declines to talk about master lease negotiations in place and whether the carriers are using alternative technologies to push back on lease rates and other terms.

10.  American Tower clearly doesn’t like small cells as a business opportunity, stating that they are focused where they think revenue opportunity is – which they define as towers and indoor small cells.   They keep posturing about their indoor small cell space (really DAS not small cells), pointing to their in-building DAS systems in the US as having 2.3 tenants per DAS.  They do point out that their outdoor DAS has not grown as fast as they expected.  I find it interesting that AMT states they have the leading multi-tenant indoor small cell portfolio. I doubt this is true – but don’t know who else has more indoor DAS facility leases in place.

11.  AMT completed lease extensions or lease buyouts on approximately 3,000 properties in the US, with 2/3 of those by lease extensions and the other 1/3 by lease buyouts.  They now control the land rights under their towers for 63-65% of their total portfolio in the US with an average of 22 years remaining.

Overall, it appears that American Tower is focused overseas and that while the US market still commands attention from the market analysts, American Tower isn’t that bullish on it. The growth opportunities are limited other than site amendment efforts and they don’t appear to be focused on building or buying new towers in the US, at least not for the foreseeable future.   American Tower gives lip service to small cell opportunities – but really doesn’t appear to be doing anything other than continuing to operate those indoor and outdoor systems that are already in place.