I recommend a short position in American Tower, (AMT).   They’re an SP500 component that owns and operates cellular phone towers.  It is the lowest yielding reit I’ve ever seen and an expensive stock which ever way you look at it.

Disclosure: I just went into the cell tower business so I am a competitor of AMT.  I had lousy cell coverage in my neighborhood so I put up my own mini cell tower.  It cost me $79 and I installed it in five minutes. Now I enjoy phone conversations that are almost as good as landline offers.   My cell station even has a GPS satellite antenna to make it 911 compliant.  My neighbor is probably talking on my tower right now (if you’re within 30 feet of me and use Verizon then you are on it, too).  Similar to my little cell tower, there’s a whole bunch of interesting gizmos of all sizes coming to market that route cell phone coverage over any broadband connection.  There’s a USB dongle that you pop into any networked computer for instant cell phone coverage.  There is nano particle spray-on antenna material that increases an iPhone’s range by 40%, or you can paint the side of a building to turn it into a massive antenna.  Lucent makes a Rubic’s cube-sized device that is capable of replacing much larger antennas in obscure places.  Qualcomm is testing a similar product.  They attach to existing street lamps, bus shelters, or traffic signal.  These advances in cellular transmission will show that existing cellular phone towers are not immune to technology’s creative destruction.   At the very least, they’re going to make tower locations less special.  Cell towers weren’t here 20 years ago and it is dangerous to bet that they’ll be here in the same form 20 years from now – or that they’ll command the same pricing power they do now. 

Value:  AMT doesn’t have any trademarks, brands, patents, competitive advantages or special expertise.  Their core revenue generating assets are simply 47,047cell phone towers spread all over the world.  The new construction cost of a tower is $150k.  Being very generous and assuming the land (or the long term lease) under the tower is worth on average $150k then that gives you a replacement cost for their assets of $14.1 billion.   Their enterprise value of AMT is over twice that, at $33 billion.  Another way to look at it is via their own balance sheet.  They carry the towers on their balance sheet at $5 billion.  This is not an old business and they depreciate slowly over 20 years so that $5 billion number is probably not too far off true value.   That being said, the best way to estimate replacement cost is just to look at their recent acquisitions.  Over the last year and half they’ve acquired 8,369 towers at an average price of $216k/tower.   To get a $33 billion valuation for the company you have to assume the land under each of their existing cell phone tower is worth $550k and the tower itself is worth $150k, over 3x more than what they’ve actually been paying.

To use actual cashflow to justify the $66 share price you’d need to assume 5 years of 30% annual growth in dividend, with a terminal value that assumes a perpetual 3% increase in dividend (8% discount rate).  There are 24 analysts covering AMT.  The average growth expectations over the next few years is only about half that, at 17%.  Yet, they are excited about it because it’s a “clean story,” it’s a reit with “growth profile,” and probably their strongest argument is that as a new reit it is underrepresented in reit indexes.  Reit investors look for yield.  Merrill Lynch predicts dividend will grow by 14.2% per year over the next 5 years.  That appears to be in line with other analysts.  It will take 17 years of 14.2% annual dividend boosts for the dividend to catch up to the average reit.   If you’re a “dividend” investor and you understand that reits payout substantially all their profits (thus dividend is all that matters) why would you put money in AMT when the average reit pays almost 3x as much?

Analyst estimates for growth over the next 5 years is too high because the easy growth is behind AMT.   In their May, 2012 investor presentation AMT indicates they can leverage their existing towers by putting multiple antennas on the same site, i.e. putting Verizon, AT&T, and Sprint all on the same pole.  Their ROI for 1 provider on a site is low (4% according to company presentation) but as they add multiple providers the ROI skyrockets (20% for 3 tenants).  Indeed, you were able to see this trend in their numbers going from 2002 to 2008.   Revenue per tower grew.  That trend is over and the towers are full. The last few years the revenue per tower has been flat domestically.  Most importantly, the revenue per tower is now over what they report as the maximum capacity for a tower ($60k).   Revenue per tower has been shrinking internationally.  They didn’t break out international vs domestic towers prior to 2005.

AMT’s own “U.S. Sample Tower Economics” indicates revenue per tower of $60k when it is fully occupied (3 tenants per domestic tower).  Since their existing domestic revenue per tower is $77k, you’d have to assume there no room left for increased revenue per tower and no more opportunities within their existing portfolio.   It looks like they have 4 cell providers on every pole already.  They don’t break out sample economics for foreign towers but the revenue per tower trend is down. 

This blurb on cell tower trends written by MD7, a company that specializes in negotiating tower leases, highlights the advantage of a distributed cell system and connects new technology:  “While one might think the Empire State Building may offer great coverage of Manhattan, it is actually not a useful cell site because a single site that high will not be able to handle the millions of calls made each day in New York. Obviously the best way to cover Manhattan is through several sites, much closer to the ground and spread throughout the city. Additionally, more cell phone users and more varied applications now require a greater bandwidth, which further increases the need for sites closer to the ground. Instead of one mountaintop site covering lots of users, carriers piece together lower elevation sites to accommodate greater bandwidth requirements necessary to meet technology demands. In other words, the average Rad Center is decreasing. As sites come closer to the ground and closer to each other, carriers are less particular about their location. This flexibility combined with an increasing ability to use non-typical cell sites (such as light poles) creates a competitive environment that drives cell site rents down.”

Lastly, there are numerous website like Steel In the Air dot com and MD7 that discuss alternatives for individual landowners who want to sell or lease their property.   The lease rates are becoming more transparent.  Glance at a few and make your own judgment about the “feel” of this market but to me it feels like the awareness levels of profit potential have risen to a level that will make the market more difficult for consolidators like CCI and AMT.

http://www.morningstar.com/invest/articles/5097-vertical-consultants-provides-free-review-of-cell-tower-lease-buyout-offers.html

Other issues:

AMT’s cash tax rate was too low.  The 10-year average of cash taxes paid for AMT was below average  (21% effective cash tax rate for AMT vs. 30% cash tax rate for the average publicly traded U.S. company).  Over the most recent 5-year period it was significantly below average (6.8% for AMT vs. 29% average).  In June 2011, AMT received a subpoena requesting documents from 2007 through 2011 relating to taxes.  AMT says the SEC investigation is relating to an employee complaint.  It is noteworthy that the IRS is not involved (how come they write me letters when I make tax mistakes?  Why is the SEC taking this one?).  The SEC has escalated it to a formal investigation. “SEC approves requests for formal orders when it finds that it is likely that a securities law violation has occurred… granting them the ability to issue subpoenas and to administer oaths.”   If they would’ve paid 30% cash tax rate over the last 10 years instead of 21%, that work out to be $81 million more than they actually paid (not including late fees, interest, penalties, SEC settlement, etc.).   This $81 million would not make a dent in their valuation but it would provide some headlines.

Emerging market meltdown - They’ve been on a major acquisition binge in emerging markets like Ghana, Africa, Columbia, Chile and Brazil.  Over the last 2 years they’ve acquired 8000+ towers (17% of their towers).   They’ve paid, on average, $216k per tower.  When you buy AMT stock you are paying $531k per tower (and assuming some debt).

Risks:  Passive Reit Funds have to buy AMT.  AMT’s market cap is large so if they were market cap weighted that would indicate they’d have to buy a lot of it.  Merrill’s June 5 report on pent up demand estimates passive funds will have to buy between $4 billion and $6.6 billion worth of AMT stock.   I agree that this will put upwards pressure on share price as passive reit index funds buy their quota.   Merrill does an excellent job of breaking down specifically which funds the demand might come from.  Interestingly, little has materialized so far this year.