Sie sind auf Seite 1von 118
!
!

Muddy!Waters,!LLC

www.muddywatersr e search.com info@muddywatersresearch

Director/of/Research://Carson/C./Block,/Esq.

Use of Muddy Waters reports is limited by the Terms of Service on its website, which are as follows. To be authorized to access such reports, you must agree to these terms, regardless of whether you have downloaded its reports directly from this website or someone else has supplied the report to you without authorization from Muddy Waters.

/

Terms of Service:

By downloading from, or viewing material on, this website you agree to the following Terms of Service. You agree that use of Muddy Waters LLC’s research is at your own risk. In no event will you hold Muddy Waters LLC or any affiliated party liable for any direct or indirect trading losses caused by any information on this site. You further agree to do your own research and due diligence before making any investment decision with respect to securities covered herein. You represent to Muddy Waters that you have sufficient investment sophistication to critically assess the information, analysis and opinion on this site. You further agree that you will not communicate the contents of this report to any other person unless that person has agreed to be bound by these same terms of service. If you download or receive the contents of this report as an agent for any other person, you are binding your principal to these same Terms of Service.

You should assume that as of the publication date of our reports and research, Muddy Waters, LLC (possibly along with or through our members, partners, affiliates, employees, and/or consultants) along with our clients and/or investors and/or their clients and/or investors has a short position in all stocks (and/or options, swaps, and other derivatives related to the stock) and bonds covered herein, and therefore stands to realize significant gains in the event that the price of either declines. We intend to continue transacting in the securities of issuers covered on this site for an indefinite period after our first report, and we may be long, short, or neutral at any time hereafter regardless of our initial recommendation.

This is not an offer to sell or a solicitation of an offer to buy any security, nor shall Muddy Waters offer, sell or buy any security to or from any person through this site or reports on this site. Muddy Waters, LLC is not registered as an investment advisor in any jurisdiction.

If you are in the United Kingdom, you confirm that you are accessing research and materials as or on behalf of: (a) an investment professional falling within Article 19 of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005 (the "FPO"); or (b) high net worth entity falling within Article 49 of the FPO.

Our research and reports express our opinions, which we have based upon generally available information, field research, inferences and deductions through our due diligence and analytical process. To the best of our ability and belief, all information contained herein is accurate and reliable, and has been obtained from public sources we believe to be accurate and reliable, and who are not insiders or connected persons of the stock covered herein or who may otherwise owe any fiduciary duty or duty of confidentiality to the issuer. However, such information is presented “as is,” without warranty of any kind, whether express or implied. Muddy Waters, LLC makes no representation, express or implied, as to the accuracy, timeliness, or completeness of any such information or with regard to the results to be obtained from its use. Further, any report on this site contains a very large measure of analysis and opinion. All expressions of opinion are subject to change without notice, and Muddy Waters, LLC does not undertake to update or supplement any reports or any of the information, analysis and opinion contained in them.

You agree that the information on this website is copyrighted, and you therefore agree not to distribute this information (whether the downloaded file, copies / images / reproductions, or the link to these files) in any manner other than by providing the following link:

http://www.muddywatersresearch.com/research/. If you have obtained Muddy Waters research in any manner other than by download from that link, you may not read such research without going to that link and agreeing to the Terms of Service. You further agree that any dispute arising from your use of this report and / or the Muddy Waters Research website or viewing the material hereon shall be governed by the laws of the State of California, without regard to any conflict of law provisions. You knowingly and independently agree to submit to the personal and exclusive jurisdiction of the superior courts located within the State of California and waive your right to any other jurisdiction or applicable law, given that Muddy Waters, LLC has offices in California. The failure of Muddy Waters, LLC to exercise or enforce any right or provision of these Terms of Service shall not constitute a waiver of this right or provision. If any provision of these Terms of Service is found by a court of competent jurisdiction to be invalid, the parties nevertheless agree that the court should endeavor to give effect to the parties’ intentions as reflected in the provision and rule that the other provisions of these Terms of Service remain in full force and effect, in particular as to this governing law and jurisdiction provision. You agree that regardless of any statute or law to the contrary, any claim or cause of action arising out of or related to use of this website or the material herein must be filed within one (1) year after such claim or cause of action arose or be forever barred.

/
/

Headquarters*of*the* US$585.4*m illion *SiteSharing*valuation*firm,*Brazil *

Company: ! American/Tower/

We/rate/American/Tower/Corp./(NYSE:/AMT)/a/

Corp ! !

Strong/Sell/and/value/it/at/$44.57/per/share,/

representing/downside/of/ 40 %.//AMT/has/serious/ challenges/domestically /and/internationally /that/have/ not/been/factored/into/the/stock/price.//It/has/engaged/ in/a/value /destroying/investment/binge/overseas,/and/ we/have/identified/a/significant/material/ misstatement/in/the/Company’s/accounts/that/could/ amount/to/fraud. /

Ticker : ! AMT !

Industry: /Wireless/Telecom/

Services///REIT /

Thesis : ! Strong/Sell /

There/is/an/approximately/US$250/million/

Report!Date: ! July/17,/2013 /

discrepancy/between/what/AMT/claims/to/have/paid/ for/the/acquis ition/of/towers/in/Brazil,/and/the/actual/

Target! Price: ! $44.57 !

selling/price.//AMT/claimed/to/have/paid/US$585.4/

.

million/for/the/towers,/but/the/real/price/was/close/to/

Stock!Price:! $74.71 !

US$300/million./If/AMT/is/aware/of/this/discrepancy,/

Market!Cap: ! $ 29,546.7 / m illion /

it/would/amount/to/fraud.//We/have/provided/our/ researc h/to/the/SEC . /

CEO/James /Taiclet’s/consistent/sales/of/approximately/

Float: ! 99.9% / / /

90%/of/the/shares/he/receives/from/option/exercises/

suggests/a/lack/of/faith/in/the/sustainability/of/AMT’s/ business/ – /ironically,/this/is/a/view/we/share/with/ him. /

Avg!Volume: ! 4.2 /million/shares !

AMT’s/international/bus iness/is/in/part/a/de/facto/ lending/business/that/artificially/inflates/revenue,/ EBITDA,/and/AFFO./It/is/also/part/carry/trade/and/part/ levered/directional/currency/bet./This/bet/has/ resoundingly/gone/south/this/year.// /

AMT’s/international/IRRs/are/general ly/poor/and/far/ below/cost/of/capital. /

While/wireless/data/usage/will/grow,/much/of/this/ growth/w ill/circumvent/cellular/towers./ Wi ] Fi/and/ more/recently/introduced/technologies/are/making/ towers/the/data/delivery/option/of/last/resort. /

AMT’s/REIT ] focused/i nvestors/will/be/disappointed/ by/their/inability/to/access/AMT’s/overseas/cash/ flows.// /

how/to/mode l/it. /

Wall/Street/is/setting/investors/up/for/a/fall/in/AMT./ We/have/reviewed/five/analyst/models,/and/it/is/clear/ that/they/do/not/understand/the/company/or/know/ ‘Page 2 of 118 /

TABLE OF CONTENTS

INTRODUCTION

5 /

S UMMARY

6 /

US$250!MILLION!DISCR EPANCY!IN!AMT’S!US$5 85.4!MILLION!TOWER!A CQUISITION!IN! BRAZIL .!!THE!MATERIAL!MISS TATEMENT!COULD!AMOUN

8 /

T RANSACTION/ B ACKGROUND

   

8 /

T HE/ P ROBLEM

10 /

DODGY VALUATION FIRM

14 /

S OMETHING/TO/ K EEP/IN/ M IND

16 /

IS!UNETHICAL!(EVEN!I

LLEGAL)!BEHAVIOR!IN! AMT’S!CORPORATE!DNA?

!!AT!THE!

LEAST,!WE!S EE!A!CORPORATE!CULTU RE!THAT!REWARDS!UNMA

 

16 /

AMT’ S/ P OOR/ C ORPORATE/ G OVERNANCE/ DNA

17 /

AMT’ S/ C OMPENSATION/ I NCENTIVIZES/ V ALUE/ D ESTRUCTION ; / CCI /IS/ M ORE/ P RUDENT

20 /

AMT’S!INTERNATIONAL! BUSINESS!IS!A!DE!FAC TO!LENDING!BUSINESS, !AND ! ARTIFICIALLY!BOOSTS !AMT’S!REVENUE,!EBIT DA,!AND!PROFIT,!WHIL E!SHARPLY! CURTAILING

22 /

AMT’ S/ D E/ F ACTO/ L ENDING/ B USINESS/ E XPLAINED

 

22 /

H OW/ L ENDING/ D ISTORTS/ AMT’ S/ I NCOME/ S TATEMENT

23 /

L ENDING/ M ODEL/IS/ I NHERENTLY/ L OWER/ G ROWTH

24 /

H OW/ P REVALENT/IS/ L ENDING ?

26 /

L ENDING/ B USINESS S/ E FFECT/ON/ EBITDA

27 /

P OTENTIALLY/ A DVERSE/ T AX/ T REATMEN T/DUE/TO/ B OOKING/ P RINCIPAL/ R EPAYMENTS/AS/ R EVENUE . 27 /

C ONCLUSION

28 /

AMT’S!INTERNATIONAL! BUSINESS!IS!PART!CAR RY!TRADE!AND!PART!LE VERED! DIRECTIONAL!CU

28 /

POOR!INTERNATIONAL!I RR!AND!NECESSARY!IMP AIRMENTS

31 /

I NDIA

32 /

B RAZIL

39 /

G HANA

43 /

G ERMANY

45 /

CELLULAR!TOWERS!ARE! SOON!TO!BECOME!THE!M OST!EXPENSIVE!OPTION !OF!LAST! RESORT;!THE !HALCYON!DAYS!ARE!OV

47 /

IMPEDIMENTS!SPECIFIC !TO!US!GROWTH

60 /

I NCREASING/ G R OUND/ R ENTS/AND/THE/ “T RI S TAR/ E FFECT

61 /

THE!REIT!THAT!CAN’T! DO!IT!RIGHT! – !INTERN ATIONAL!AFFO!IS!INAC CESSIBLE!TO! AMT!SHAR EHOLDERS

62 /

AMERICAN!JAMES!AND!T HE!TOWER!OF!DOOM!(AN D!OTHER!TALES!OF!VAL UE! DEST RUCTION!IN!EMERGING!

64 /

M EXICO : / A N/ A MIGO/IN/ N EED/IS/AN/ A MIGO/ I NDEED/ / S ALE/AND/ L EASEBACKS/WITH/ D ISTRESSED/ T ELECOMS

67 /

S OVEREIGN/ E XTORTION

71 /

C ORRUPTION/ S TILL/ R E MAINS/A/ P ART/OF/ L OCAL/ C ULTURE

71 /

/

‘Page 3 of 118 /

F ROTHY/ M ARKETS/ M AY/ P ROMOTE/ N EFARIOUS/ C OMPETITION/AND/ I NTERNAL/ D EFECTIONS

72 /

M ACRO ] E CONOMIC/ R ISKS : // C OSTS , / C OSTS , /AND/ M ORE/ R ISING/ C OSTS

73 /

Q:!HOW!DO!YOU!KNOW!W

HEN!WALL!STREET!IS!S ETTING!INVESTORS!UP! FOR!A!FALL?!

A:!WHEN! ANALYSTS’!MODELS!DIS AGREE!ON!EVERYTHING, !EXCEPT!TARGET!PRICE

74 /

C

AN T/ M ODEL/

6% /TO/ 8% / R EVENUE/ G ROWTH/IN/THE/ US / G OING/ F ORWARD ? // J UST/ T AKE/THE/

C OMP ANY S/ W ORD/FOR/IT

75 /

APPENDIX!A.1:!VALUAT ION

77 /

EXHIBIT!A.2:!CONSOLI DATED!HISTORICAL!AND !PROJECTED!FINANCIAL !STATEMENTS 86 /

EXHIBIT!A.3:!COMBINE

91

/

EBITA

EXHIBIT!A.4:!DETAILE D!REGIONAL!EBITDA!HI STORICAL!AND!PROJECT ED!FINANCIAL!

95

/

/

108

112 /

DATA

APPENDIX!A.5:!COST!O

APPENDIX!A.6:!AFFO

/

D!AND!REGIONAL!HISTO RICAL !AND!PROJECTED!SEGME NT!

F!CAPITAL

/

‘Page 4 of 118 /

INTRODUCTION

Investors seeking to own a REIT that is going to rapidly grow its AFFO and dividend for as far as the eye can see should not own AMT.

American Tower Corp. (NYSE: AMT) is a great story. However, it is not a great company. Investors think that because wireless data usage in the US is growing rapidly, AMT will continue to deliver strong domestic growth year after year. The growth story is enhanced by AMT’s rapid expansion into emerging and frontier markets, supposedly positioned to capitalize on the oft-heralded newly moneyed middle classes of India, South America, and Africa. REIT investors see a company with plenty of headroom to grow its dividend.

The reality is disappointing.

The best days of the tower business are now behind it. Unlike traditional real estate, this model is susceptible to disruptive technologies. These technologies are not theoretical – in fact, the single largest replacement for the cellular tower is a technology we have all used for years: Wi-Fi. Mark our words, in five years you will be downloading data to your PDA or tablet in far greater quantities than today; however, the antennae to which you are connecting will seldom be on a cellular tower. Although tower operators once drove efficiency, they have now become rent-seekers. Carriers cannot afford this model, and towers will soon be the most expensive option of last resort.

AMT could have conserved its cash and prepared for a slow growth future with compressing margins. But that is not what it chose to do. It has embarked on an asset binge, seemingly scouring the globe for the most challenging operating environments possible. By borrowing money in US dollars and investing in high yielding countries, it is not running a business so much as a hedge fund with substantial execution risks. Despite making highly illiquid investments, our models show that its investments usually underperform local government bonds. Germany is an exception – it outperforms German government bonds by 200 bps – 4% versus 2%.

AMT pulls the wool over investors’ eyes by operating a de facto lending business overseas, often entering into transactions where it will overpay for towers in exchange for carriers overpaying rent. AMT books the principal repayments as rental revenue. This lending artificially boosts EBITDA margins, making the business look more profitable than it really is. It also ensures that the promised growth will never come – at least not organically. The REIT-focused investors also do not appreciate how expensive it is to repatriate money from many of these countries, thereby making AMT’s international cash flow largely inaccessible.

/

‘Page 5 of 118 /

There is a potentially unseemly underbelly to AMT’s overseas activities. As we show in this report, there is approximately US$250 million missing from a transaction in which AMT bought towers in Brazil. There are therefore material misstatements in AMT’s financials. (We are providing our research on this matter to the SEC.) We also note indications that there could have been similar problems with transactions years earlier. Regardless of what the explanations for these issues ultimately turn out to be, there certainly are reasons to question the ethics of senior management. At the least, we note that chairman and CEO James Taiclet is an avid seller of AMT’s shares, disposing of approximately 90% of the stock he receives from option exercises. He appears to be leaving investors holding the bag.

Summary '

We rate American Tower Corp. (NYSE: AMT) a Strong Sell and value it at $44.57 per share, representing a downside of 40.3%. AMT has serious challenges domestically and internationally that have not been factored into the stock price. It has engaged in a value destroying investment binge overseas and we have identified a significant material misstatement in the Company’s accounts that could amount to fraud.

There is an approximately US$250 million discrepancy between what AMT claims to have paid for the acquisition of towers in Brazil, and the actual selling price. AMT claimed to have paid US$585.4 million for the towers, but the real price was close to US$300 million. Our analysis is based on financial statements of the company AMT acquired, Central Bank of Brazil records, and six industry sources, including two who had direct knowledge of the closed transaction. If AMT is aware of this discrepancy, it would amount to fraud. We have provided our research to the SEC.

There are reasons to question the ethics of AMT’s management. In addition to Chairman and CEO James Taiclet having led the company during some of its admitted option backdating, other senior management shows a disturbing propensity to push the ethical envelope. In addition, AMT’s audit committee chair is also the audit committee chair of NII Holdings, Inc. (NASDAQ: NIHD), which has been an AMT counterparty in the past. There is a disturbing discrepancy in the historical accounts for a series of tower sales and leasebacks between the two companies.

AMT’s compensation incentivizes value destructive acquisitions by compensating management in part on tower growth, regardless of the quality of acquisitions. While management’s performance benchmarks are adjusted for currency fluctuations, shareholders have no such protection. Mr. Taiclet’s consistent sale of approximately 90% of the shares he receives from option exercises suggests a lack of faith in the sustainability of AMT’s business – ironically, this is a view we share with him.

AMT’s international business is in part a de facto lending business that artificially inflates revenue, EBITDA, and AFFO. AMT effectively lends money to tower operators, and then books their principal repayments as rental revenue. This line of business also makes it substantially harder for AMT to grow organically. The only way to produce

/

‘Page 6 of 118 /

growth is to “feed the beast” and make more and more (often value destructive) acquisitions. This model of “growth” is impossible to sustain.

AMT’s international business is also part carry trade and part levered directional currency bet. AMT’s model is to fund in US dollars and invest in high yielding countries, where high local interest rates cause rent yields to be higher than in the US. However, AMT often underperforms local government bonds. Considering that AMT has substantial execution risk layered on top of the high local rates, and it is invested in illiquid assets, investors are faring poorly. AMT does not hedge its exposure to the myriad emerging and frontier market currencies to which it is exposed, making it also a levered directional currency bet. This bet has resoundingly gone south this year.

AMT’s international IRRs are generally poor, and far below the cost of capital.

obtained granular data on all of AMT’s operations and found that the international businesses have a much higher risk profile than the US operations. We analyzed the IRRs for four countries: India, Brazil, Ghana, and Germany. In the case of the first three,

AMT’s IRRs are no greater than those of the local currency government bonds. AMT is outperforming German government bonds, but is only generating a 4% IRR.

We

While wireless data usage will grow, much of this growth will circumvent cellular towers. Tower operators have gone from drivers of efficiency to rent-seekers, and carriers, with compressing margins, cannot afford the status quo. Going forward we expect domestic growth to slow and international markets to remain volatile with increasing headwinds to growth. Wi-Fi and more recently introduced technologies are making towers the most expensive option of last resort. AMT faces rising ground rents in the US due to the ground rent market becoming more efficient due to the internet and the ubiquity of cellular phones.

US tower industry growth has peaked, and we expect limited revenue per tenant growth and stagnant tenancy ratios. Domestic sales growth has been driven by contractual escalators, and excess charges (“amendments”) that carriers pay to add network equipment. Carriers do not like the pricing structure and its substance. By using Wi-Fi and new generation antennae that combine multiple generations into smaller equipment and enable carriers to share network equipment (the logical evolution from sharing towers), tower companies will be marginal players in US carriers’ future network growth. History does not repeat itself, but it rhymes. We have previously seen tower suffer due to technological changes (i.e., when paging companies became irrelevant many tower companies were forced into bankruptcy). While we do not expect AMT to go bankrupt, we do not expect it to continue to post strong US growth numbers beginning in the near future.

AMT’s REIT-focused investors will be disappointed by their inability to access AMT’s overseas cash flows. Of all of the international operations, only Mexico is in the REIT structure. All other regions are subject to corporate taxation. Investors seem not to appreciate how expensive and difficult it would be for AMT to repatriate its overseas

/

‘Page 7 of 118 /

cash flow. Crown Castle, Inc. (NYSE: CCI) notably states that the REIT structure is incompatible with emerging market tower operations.

AMT is operating in challenging markets, and without sufficient risk management. Poor quality due diligence (i.e., only inspecting 10% of towers before closing purchases) has likely led to model-busting capital expenditures. More stringent tower regulations are also expected to increase costs and negatively impact margins. Capricious governments are always a wild card that AMT faces in emerging and frontier markets.

Wall Street is setting investors up for a fall in AMT. We have reviewed five analyst models, and it is clear that they do not understand the company or know how to model it. They use widely varying assumptions, but (surprise!) seem to largely agree on the value. Investors should be particularly wary given that we are in a REIT bubble, as evidenced by investors’ day-to-day fortunes being tied to the words of Federal Reserve chairman Ben Bernanke.

US$250 MILLION DISCREPANCY IN AMT’S US$585.4 MILLION TOWER ACQUISITION IN BRAZIL. THE MATERIAL MISSTATEMENT COULD AMOUNT TO FRAUD.

There is a US$250 million material misstatement in AMT’s accounting for a purportedly US$585.4 million tower acquisition in Brazil. It is difficult to see how the Company would be unaware of this misstatement; and, if AMT is in fact aware of the discrepancy, it would amount to fraud. AMT disclosed paying US$585.4 million for 666 towers in Brazil in 2011. It is highly probable that the real consideration for the towers was only approximately US$335 million. However, AMT’s accounts reflect the US$585.4 million purchase consideration. We have provided our findings to the SEC, and we call upon AMT’s board to thoroughly and transparently investigate this transaction.

This purchase was the only transaction that we scrutinized in detail, and given the cultural and control problems we believe are endemic to AMT (see infra), along with its substantial emerging and frontier market investments, we would not be surprised to learn of similar problems in other transactions. As we discuss infra, we note a potentially similar, albeit smaller, discrepancy in a series of transactions with NII Holdings, Inc. (NASDAQ: NIHD). AMT and NIHD have the same audit committee chair.

Transaction'Background'

AMT was evasive about this transaction from the beginning, including its failure to (even to this day) disclose the identity of the seller. 1 The transaction size rapidly upsized from

////////////////////////////////////////////////////////

1 AMT disclosed the identity of a non-carrier seller of ground rights only months later. See AMT 8-K, filed September 7, 2011, announcing the acquisition of a portfolio of properties held by various limited liability

/

‘Page 8 of 118 /

US$420 million to US$585.4 million, purportedly starting with 565 sites and ultimately becoming 666 sites.

On February 28, 2011, AMT announced that it was acquiring 100% of the shares of a company that owned 565 sites in Brazil for approximately US$420 million. 2 The acquisition was to close on March 1, 2011 (the day after this filing), and achieve financial close on June 1, 2011. 3 The per tower price would have been US$743,000, which also would have been by far the most AMT had ever paid for towers overseas.

Barely two months later, AMT announced that on March 1, 2011 – literally the day after it filed its 2010 10-K that cited the aforementioned deal terms – AMT had closed the acquisition, but for a larger amount of towers and money. In the new filing, AMT disclosed that on March 1, 2011, it had acquired 627 sites for approximately US$553 million, which was a per tower valuation of US$882,000. 4 On a conference call held the same day, AMT CEO and chairman James Taiclet explained that AMT upsized the deal by 62 towers after completing its due diligence, and that “There may be up to an additional 40 sites under construction that [AMT] will acquire over the next several months”. 5 When AMT filed its 2011 10-K in 2012, we learned that the transaction had been purportedly upsized again to 666 towers for US$585.4 million. 6 This works out to a per tower price of US$879,000.

Muddy Waters became interested in this transaction at first because the per tower valuation was so high. The price of US$879,000 is high even by US tower price standards, which is approximately US$656,000 in recent years. The disparity is even greater when compared to other international transactions, which often have purchase prices inflated by AMT’s de facto lending business. Below is a graph setting out AMT’s per tower purchase prices for international acquisitions. Many of these transactions reflect the de facto lending business’s purchase price inflation effects. 7 We would not be surprised if there are similar problems with other transactions – we researched in depth only one transaction and quickly saw a major problem that amounts to fraud, unless AMT can provide some exonerating explanation.

/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////

companies owned by Unison Holdings, LLC and Unison Site Management II, LLC for approximately US$500 million.

2 See AMT 2010 10-K, filed February 28, 2011, p. F-52.

3 Id.

4 AMT 8-K, filed May 3, 2011, p. 3.

5 AMT May 3, 2011, conference call.

6 AMT 2011 10-K, filed February 29, 2012, p. F-29.

7 We believe that this includes the data for all international transactions in which AMT disclosed the purchase price and number of sites.

/

‘Page 9 of 118 /

1000 International Acquisitions 900 800 700 600 500 400 300 200 100 0 1998 2000
1000
International Acquisitions
900
800
700
600
500
400
300
200
100
0
1998
2000
2002
2004
2006
2008
2010
2012
2014
Announcement Date
Avg. $/Site ($ '000)

To identify the seller, we searched AMT’s Brazilian corporate documents through an online government portal, and saw that on March 3, 2011, AMT resolved to buy the shares of Site Sharing NE S.A. 8

We then obtained financial statements for Site Sharing NE and the parent from which it was spun out, Site Sharing do Brasil Empreendimentos em Telecomunicações S.A. (“Site Sharing Brazil”, and together with Site Sharing NE, “Site Sharing”). It is common practice for privately-held companies in Brazil to file their financial statements publicly. 9

The'Problem '

Site Sharing’s financial statements and Central Bank of Brazil (“BCB”) records make clear that the consideration for Site Sharing NE was only approximately US$335 million, approximately US$250 million less than disclosed. AMT’s initial PP&E allocation confirms the accuracy of Site Sharing’s financial statements. Conversations with six industry sources, including people who had been shopped the transaction, and at least two people with direct knowledge of the closed transaction, all confirmed that the purchase price was in the US$300 million range. 10

////////////////////////////////////////////////////////

8 American Tower do Brasil LTDA registry references the company’s resolution authorizing the Site Sharing purchase. The actual resolution is not available online. The resolution is Document 094.913/11-0, resolved on March 3, 2011 (www.cade.gov.br) 9 AMT does not follow this practice. 10 As per our usual practice, we do not reveal information that could compromise our sources and subject them to reprisal.

/

‘Page 10 of 118 /

AMT discloses (to this day) that it paid 12x to 13x EBITDA for Site Sharing. Giving AMT every benefit of the doubt, 11 the EBITDA of Site Sharing should have been at least BRL 70 million, yet in the year ended December 31, 2010 it was no more than BRL 45.5 million. 12 We are left wondering what Mr. Taiclet knew and did not know. He made the following statements about the transaction on the May 3, 2011 conference call:

James Taiclet: “The sites we acquired in Brazil are high performing towers and the multiple pay was in the range of 12x to 13x adjusted EBITDA.”

/

Philip Cusick, Macquarie: “This is Richard Choe for Phil. Going into the Brazil deal, I guess it was upsized a little bit and you said you paid 12x to 13x cash flow. Was that because the incremental sites were higher or was this kind of looking ”

/

James Taiclet, Executive Chairman, Chief Executive Officer and President:

back and kind of re-upping the price for the sites?

“Richard, it's Jim Taiclet. Basically, what happened with our Brazil transaction is the purchase prices adjusted for the actual cash flows that were occurring up the towers new and existing in the deal. But the bulk of it was driven by the new towers that were completed and ready to transfer. So those both contributed…”

The EBITDA AMT was implying for Site Sharing at that time was at least 53.5% higher than Site Sharing’s actual EBITDA. We combined the income statements of Site Sharing NE with those of Site Sharing Brazil to analyze whether their EBITDA jibed with AMT’s disclosures. Site Sharing NE was a subsidiary of Site Sharing Brazil. Site Sharing NE SA was incorporated on March 3, 2008, likely in preparation for a sale of the towers. Site Sharing Brazil transferred all assets into Site Sharing NE by September 10, 2010. The transferred assets included up to 687 rooftop and tower sites. 13 We combined the two sets of financials in order to calculate the maximum EBITDA that Site Sharing NE could have generated.

////////////////////////////////////////////////////////

11 Mr. Taiclet said that the additional 40 sites AMT might acquire were under construction, so the implied EBITDA is based on the size of the transaction as it stood the day he discussed the valuation, and using an EBITDA multiple of 13x in order to give the Company the maximum benefit of the doubt.

12 Based on the March 1, 2011 (deal closing date) rate 1.665 BRL. Using the average rate of BRL 1.760 during 2010 would have implied an EBITDA of BRL 74.9 million. Using the rate of BRL 1.59 on May 3, 2011, the implied EBITDA would have been BRL 67.6 million.

13 There was a schedule of assets transferred, which gave the highest numbered tower as 687; however, fewer than 678 towers were actually listed in the schedule.

/

‘Page 11 of 118 /

AMT ’s initial purchase price allocation for the acquisition confirms Site Sharing’s combined financial statements.

AMT’s initial purchase price allocation for the acquisition confirms Site Sharing’s combined financial statements. AMT initially allocated US$33.2 million to PP&E. 14 Per the Brazilian financial statements, the book value of Site Sharing’s combined PP&E as of December 31, 2010 was US$32.3 million (BRL 53.6 million). It appears as though AMT initially allocated to PP&E based on Site Sharing’s PP&E carrying value, with the slight US$ difference due to currency fluctuations. Thus, AMT’s disclosure confirms the Brazilian financial statements of Site Sharing Brazil and Site Sharing NE. (Note that the EBITDA margin is so high because of Site Sharing’s abnormally high rents, which were apparently up to 3.0x market on some towers.)

At 12.0x to 13.0x EBITDA, AMT would have paid US$327.6 million to US$354.9 million, as shown below.

paid US$327.6 million to US$354.9 million, as shown below. Central Bank of Brazil (“BCB”) records make

Central Bank of Brazil (“BCB”) records make clear that the total consideration paid was only approximately US$335 million. The Brazilian Real (BRL) is a controlled currency, and by law all exchanges into BRL must be registered with the BCB within 30 days of closing. The BCB has made available records of all such forex transactions through December 2011 on its website. AMT injected US$335.0 million into its Brazil subs in May 2011. It had not injected any money in almost two years. The last time it had injected money, it transferred US$46.0 million to fund a US$50.5 million tower acquisition. 15,16 Previously, the largest amount AMT had ever injected into Brazil in one

////////////////////////////////////////////////////////

14 AMT Q1 2011 10-Q, filed May 5, 2011, p. 18. 15 We define “significant” as in excess of US$7.0 million.

/

‘Page 12 of 118 /

month was only US$7.0 million, which was when its Brazil business was much younger and presumably free cash flow negative. 17

The graph below shows AMT’s USD injections to its Brazil subsidiaries by quarter between Q1 2006 and Q2 2011, and the deltas to acquisition consideration.

600,000,000 500,000,000 400,000,000 300,000,000 200,000,000 100,000,000 - Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
600,000,000
500,000,000
400,000,000
300,000,000
200,000,000
100,000,000
-
Q1
Q2 Q3 Q4 Q1
Q2 Q3 Q4 Q1
Q2 Q3 Q4 Q1
Q2 Q3 Q4 Q1
Q2 Q3 Q4 Q1
Q2
06
06
06
06
07
07
07
07
08
08
08
08
09
09
09
09
10
10
10
10
11
11
USD Injections
Consideration Delta

AMT’s disclosures regarding the amounts and timing of money it purports to have paid make clear that its accounts contain material misstatements. AMT purports to have (partially) paid the selling shareholders US$419.2 million in March 2011. In its Q1 10- Q, AMT stated:

“On March 1, 2011, the Company acquired 100% of the outstanding shares of a company that owned 627 towers in Brazil for $553.2 million, of which $419.2 million was paid using cash on hand in March 2011.” 18

Putting aside that AMT overstated the purchase price, it almost certainly would not have injected US$335 million into Brazil in May 2011 if the preceding statement were true. It would have only owed needed to pay US$124.1 million to $166.1 million to complete the

/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////

16 http://www.bcb.gov.br/rex/registrosCE/2011/rce052011.pdf

17 AMT injected $7.0 million into Brazil in November 2008.

18 AMT Q1 2011 10-Q, filed May 3, 2011, p. 18.

/

‘Page 13 of 118 /

transaction. 19 AMT made no other tower purchases in Brazil in 2011. This amount is enormously more than AMT would have needed for its operations, particularly considering it appears to have been free cash flow positive in Brazil for at least two years, judging by the absence of other injections.

The Q2 2011 disclosure is more egregious:

“On March 1, 2011, the Company acquired 100% of the outstanding shares of a company that owned 650 communications sites in Brazil for $583.7 million, of which $543.3 million was paid using cash on hand during the six months ended June 30, 2011.” 20

Regarding the extremely remote chance that AMT paid an additional US$250 million to the selling shareholders offshore (and the Site Sharing financials are wrong, the six sources are all wrong, and the Q1 2011 US$419.2 million disclosure never happened), AMT would have been taking a significant risk. Even though the parties could attempt to call such a payment something other than tower purchase consideration (e.g., consulting fees), AMT would be at risk of abetting the selling shareholders, who are Brazilian, in committing tax fraud. This is not something that multinationals with approximately US$1 billion invested in Brazil would usually do.

Going further into Fantasyland, if the reason AMT wanted US$335 million cash in Brazil in May 2011 had nothing to do with the Site Sharing NE purchase, AMT would have lost substantial money by converting the currency in May 2011 and leaving it on deposit. AMT executed its next Brazil transaction (i.e., its next major use of cash) on March 30, 2012. 21 In May 2011, the average exchange rate was 1.614, which was close to its all time high. On March 30, 2012, the rate was 1.823, which would have been a loss of 11.5%. That alone would have been a US$41.0 million drawdown, which would be a firing offense in most companies that could afford to lose that kind of money without going out of business.

Dodgy Valuation Firm

Every now and then, we at Muddy Waters reminisce about projects past – especially when a current project bears a striking similarity to a previous one. AMT gave us the opportunity to stroll down Memory Lane when we looked up information on the accounting firm AMT used to appraise Site Sharing NE. This brought to mind the photos of the sham Sino-Forest customers that were purportedly purchasing US$400 million per year of product, while operating out of run down apartments in remote parts of China.

////////////////////////////////////////////////////////

19 This range results from subtracting $419.2 million from both the $585.3 million AMT claims to have been obligated to pay, as well as from the $543.3 million AMT claims to have paid by the end of Q2 2011.

20 AMT Q2 2011 10-Q, filed August 4, 2011, p. 19.

21 AMT Q1 2012 10-Q, filed May 3, 2012, p. 21.

/

‘Page 14 of 118 /

We obtained the name of AMT’s appraisal firm, Econap Organizaçao Contabil Ltda., from minutes of a shareholder meeting of AMT Brazil. 22 The purpose of the meeting was to formalize the Site Sharing NE acquisition. The minutes list the accounting firm’s address as Rua Dom Joaquim de Oliveira 126, São Paulo, SP. 23 Numerous online directories confirm this address.

The appraiser on a purportedly US$585.4 million transaction is located in the house to the left, next to the wall with the bullet holes.

house to the left, next to the wall with the bullet holes. Source: Google Maps Street

Source: Google Maps Street View

This neighborhood appears not to be a central business district.

neighborhood appears not to be a central business district. Source: Google Maps Street View

Source: Google Maps Street View

////////////////////////////////////////////////////////

22 The minutes were actually from an April 30, 2011 meeting for the AMT Brazil entity American Tower do Brasil Cessão de Infra-Estruturas Ltda., published on May 27, 2011 in the "Diário Oficial Empresarial", a public diary of corporate activities. 23 Id.

/

‘Page 15 of 118 /

Something to Keep in Mind

One bit of information that caught our attention, and might be relevant to the discrepancy are forum postings by someone calling himself Enki Ea. Without further evidence, the following should be discounted, but kept in mind. The posting, appearing in the comment section of articles or videos on numerous financial websites, including Zero Hedge, Max Keiser, and Bloomberg, is a posting titled “CRIME/TERROR/WEAPONS

Bank in South America

in the center of money laundering for narco-traffickers, corrupt Brazilian politicians, and even someone linked to Hezbollah. The post then names seven individuals whom it alleges are the “main culprits”. One of the individuals named is Flavio Guimaraes. 24

The post discusses a small Brazilian bank that purportedly is

Flávio Ognibene Guimarães was the majority shareholder of Site Sharing Brazil via his company Ltda Sestri Empreendimentos e Participações Ltda. Mr. Guimarães also owned shares in his own name in Site Sharing NE when it was sold to AMT. 25 We contacted and spoke with the author of these posts to query whether he is referring to Mr. Guimarães of Site Sharing. The poster, whose real identity we do not know, stated that he is familiar with Site Sharing Brazil, and that Site Sharing’s Mr. Guimarães is indeed the same individual to whom he is referring.

We want to be clear that we have been unable to document these allegations by "Enki Ea" and have no indication at all that AMT has any involvement in any activities of Mr. Guimarães other than purchasing a company of which Mr. Guimarães was the controlling shareholder. However, given the irregularities surrounding the purchase of this Brazilian subsidiary, we are of the definite view that any independent investigation or interested regulator should examine the involvement of Mr. Guimarães in the transaction.

IS UNETHICAL (EVEN ILLEGAL) BEHAVIOR IN AMT’S CORPORATE DNA? AT THE LEAST, WE SEE A CORPORATE CULTURE THAT REWARDS UNMANAGED RISKS.

We perceive that AMT has at the least an unethical streak in its culture, with two or more current senior executives having had some role in illegal activity at AMT in the past. While investors might recall AMT’s option backdating scandal, the highly questionable (albeit not illegal) behavior of AMT’s EVP of International Operations / president of Latin America is surprising – particularly because he, a former CPA, was serving as AMT’s general counsel at the time of the backdating and unrelated behavior that netted him over US$10 million in backdoor compensation. We also have questions surrounding the present dual role played by the chairwoman of AMT’s audit committee, Carolyn Katz, who also chairs the audit committee at NII Holdings, Inc. (NASDAQ: NIHD). NIHD is a potentially significant AMT counterparty, and there have been some troubling

////////////////////////////////////////////////////////

24 See comment made July 1, 2013 at 06:30 at http://www.zerohedge.com/contributed/2013-06-30/ten- things-watch-week-ahead 25 SiteSharing Brasil S.A. Extraordinary Shareholders Minutes, September 1, 2010.

/

‘Page 16 of 118 /

discrepancies in the past between the way the two companies disclose their sides of past transactions with each other. Industry sources speculate that AMT and NIHD are close to completing another transaction, which would likely range from US$500 million to US$1 billion in value.

AMT’s Poor Corporate Governance DNA

In 2006, AMT settled a stock option backdating class action lawsuit for $14 million after admitting that it did in fact illegally backdate options. 26 According to a 2006 Center for Financial Research and Analysis report, AMT was one of the more egregious backdater of options. 27 Several senior AMT executives and board members departed in the wake of the scandal; others did not.

The plaintiff with whom AMT settled alleged that the backdating continued into 2005. 28 Jim Taiclet became president of the Company in September, 2001; CEO in October, 2003; and, chairman in February, 2004. In addition to holding positions of authority during the time that AMT was backdating options, Mr. Taiclet was alleged in the class action complaint to have received backdated options. 29 Yet, he remained at the Company. We wonder why there was not a more thorough housecleaning; and, we wonder whether those who escaped relatively unscathed took away the wrong lessons from the episode.

However, our concern is more acute with William “Hal” Hess, who had been AMT’s general counsel from 2002 until 2007. This covers much of the period when the backdating occurred. In addition to being an attorney, he previously practiced as a CPA. It would be hard to find somebody who would seem to know more about proper corporate governance than he. Mr. Hess is now AMT’s executive vice president of international operations and president of Latin America, and is one of Mr. Taiclet’s five direct reports. 30

What really concerns us about Mr. Hess is that while he was general counsel, he earned profits of approximately US$11.3 million exercising options for stakes in two of AMT’s Latin American subsidiaries: ATC Mexico and ATC South America. Fortune Magazine called this practice “backdoor options filtered through a subsidiary”. 31 The gist of the episode was that Mr. Hess received options to purchase equity stakes in the two (non- public) subsidiaries, and the Company was obligated to buy the shares back from him should he exercise and certain other conditions were met. Those conditions were met,

////////////////////////////////////////////////////////

26 http://money.cnn.com/magazines/fortune/fortune_archive/2006/10/30/8391767/

27 In Re American Tower Corp. Securities Litigation, Case No. 06-CV-10933, U.S.D.C, D. Mass. (2006)

28 In re American Tower Corp. Securities Litigation, Consolidated Amended Class Action Complaint.

29 Id., para. 40, 44, 45. The Amended Complaint further alleges that one instance of the option backdating in which Taiclet received 125,000 options was timed to occur the day before the AMT purchase of approximately 535 towers from NII Holdings (NIHD) on December 10, 2002. Id., para. 45-46. We discuss inconsistencies in AMT's and NIHD's accounting for later sales of towers below in this report.

30 http://www.americantower.com/corporateus/company/leadership-team/index.htm

31 http://money.cnn.com/magazines/fortune/fortune_archive/2006/10/30/8391767/

/

‘Page 17 of 118 /

and AMT bought the shares back from him after very short holding periods, and at valuations far greater than what he had paid.

Mr. Hess’s transactions were disclosed in footnotes to the proxy statements. However, this type of arrangement had the effect of keeping Mr. Hess’s full compensation out of the compensation table portion of the proxy statement, which possibly caused a large number of shareholders to be unaware of his total compensation. While all corporate executives owe a fiduciary duty to the shareholders, we find it particularly troubling when a general counsel engages in such aggressive practices. General counsels are (theoretically) part of the controls that protect shareholders.

We also have concerns regarding the overlapping board memberships of Carolyn Katz, who presently chairs the audit committees for both AMT and NIHD. Ms. Katz joined AMT’s board and audit committee in 2004. 32 She became chairwoman of AMT’s audit committee in 2007. 33 Ms. Katz was also a defendant in the backdated options class action lawsuit discussed above, although the Amended Complaint appears to allege only that she colluded in the practice and not that she personally received backdated options. She has been a member of NIHD’s board since 2002, 34 and has chaired NIHD’s audit committee since as early as 2004. 35 The problem is that both NIHD and AMT carry on their balance sheets towers than AMT purchased in 2003 through 2008. 36 Moreover, the two companies do not agree on the number of towers purchased, and in possible foreshadowing to the Site Sharing US$250 million discrepancy, AMT has consistently reported purchasing 34 more towers than NIHD reports selling. This differential has translated into AMT booking consideration payments to NIHD that are higher than what NIHD has recognized by US$6 million to US$7 million. 37

AMT states that as of December 31, 2003, it had acquired 665 towers from NIHD for US$112.4 million. During this same period NIHD reports it sold only 631 towers for US$106.4 million. 38 Note that AMT claims to have acquired 34 more towers than NIHD acknowledges selling, and AMT claimed to pay US$6.0 million more than NIHD claims to have received. 39 While US$6.0 million and 34 towers normally would not be a material amount for a company of AMT’s size, we strongly believe that in light of the US$250 million discrepancy in the Site Sharing transaction, the transactions with NIHD should be closely examined for possible impropriety – particularly to see whether there is a relationship to the Site Sharing consideration and possibly that of other transactions.

////////////////////////////////////////////////////////

32 AMT 2013 Proxy Statement, p.8.

33 Id.

34 NIHD 2004-2013 Proxy Statements.

35 Id.

36 NIHD stepped disclosing related information after 2008.

37 The fluctuations between US$6 million and US$7 million likely reflect currency fluctuations. The discrepancy is approximately US$176,000 to US$205,000 per tower. The purchase price in effect at the time the 34 tower discrepancy appeared was US$187,000 per tower.

38 AMT, 2003 10-K p. 7, NII 2004 10-K p. F-37

39 This 34 tower and approximately US$6.0 million discrepancy between AMT and NII’s account is traceable throughout all years of disclosures of this sales and leaseback, and peaked at US$7.2 million.

/

‘Page 18 of 118 /

Also strangely, both companies are presently carrying these towers on their balance sheets. We believe that NIHD’s accounting is erroneous in this instance, rather than AMT’s; however, Ms. Katz has been on both boards and audit committees around the time of the transactions and during the companies’ ongoing reporting. We do not know how to interpret these facts, but we have yet to think of an explanation that gives us comfort.

NIHD’s annual disclosures on this transaction repeatedly report this sale and leaseback as a financing transaction wherein NIHD maintains these tower assets on its balance sheet. The below excerpt is from NIHD’s 2004 10-K:

sheet. The below excerpt is from NIHD’s 2004 10- K: To summarize the above disclosure: •
sheet. The below excerpt is from NIHD’s 2004 10- K: To summarize the above disclosure: •

To summarize the above disclosure:

NII accounts for the sale as a financing arrangement and did not recognize a gain or loss from the sale, 40

The proceeds from the sale are treated as financing obligations repaid through monthly rents for 15 years, 41

Tower rent payments are recognized as interest expenses and ground rent obligations would be represented as operational expenses, 42

The tower assets remain on NII’s balance sheet 43

NIHD recognizes co-location rents as other operating revenues a portion of which is recognized as “interest expenses”. 44

////////////////////////////////////////////////////////

40 NII, 2002 10-K, p. 44

41 NII, 2002 10-K, p. 44, NII, 2003 10-K, p. 4

42 NII, 2003 10-K, p. 4

43 NII, 2002 10-K, p. 44

/

‘Page 19 of 118 /

NII has been disclosing its retention of the towers on its balance sheet every year since 2002, and in subsequent years NII has expanded the level of detail and increased the thoroughness of the description. 45

If AMT has not yet finalized its expected purchase of approximately 4,500 towers in Brazil and Mexico from NIHD, 46 we encourage the independent, disinterested directors to take a critical look at this transaction. 47 Our industry contacts in Brazil informed us that NIHD has been informally shopping this portfolio as a sale and lease-back in an attempt to entice buyers to join the formal which bidding began in January. 48 NIHD is in weak financial shape. 49 Apparently many potential bidders are reluctant to take the credit risk particularly given that there is a real chance NIHD will be acquired by another carrier and eventually consolidate infrastructure. This tower sale would likely be in the range of US$500 million to US1.0 billion. Not only should investors keep in mind the aforementioned overlapping board and reporting issues, but they should note that AMT’s compensation structure (not unlike pre-financial crisis banks) incentivizes AMT to make acquisitions, regardless of whether they create or destroy value.

AMT’s Compensation Incentivizes Value Destruction; CCI is More Prudent

Regardless of controls and culture, AMT’s compensation structure goes a long way toward explaining why it is destroying value through poor quality and risky international acquisitions. Both Mr. Taiclet and Mr. Hess were compensated in part for growing the tower base through acquisitions. 50 We understand that other AMT international managers are also incentivized to buy towers. As we detail throughout this report, AMT places much more emphasis on quantity than quality.

It is inappropriate that AMT adjusts management’s financial performance goals for currency fluctuations. 51 As we detail infra, AMT’s international business model is part levered currency bet because it does not hedge its FX exposure. If investors do not receive the benefit of hedging, neither should management. This structure insulates management from the substantial risks AMT takes in currencies alone.

/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////

44 NII, 2003 10-K, p. 4

45 Although we do expect public companies to provide transparency, especially in the case of large

transactions, do not mistake our praise for this relatively more transparent disclosure by NII as expressing an opinion on the quality of its accounting.

46 An industry source in Brazil who is not affiliated with AMT believes that the transaction either has, or soon will, close.

47 Raymond Dolan was a NIHD director until May 2012.

48 Industry source in South America.

49 http://www.reuters.com/article/2012/11/07/us-niiholdings-brazil-strategy-idUSBRE8A61RE20121107 ,

http://www.telegeography.com/products/commsupdate/articles/2013/02/06/nii-results-fall-short-of-

expectations-tower-sale-planned-in-brazil-

mexico/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+telegeography%2Fcom

msupdate+%28TeleGeography+CommsUpdate%29

50 AMT, 2013 Proxy Statement, p. 2.

51 AMT, 2013, Proxy Statement, p.27.

/

‘Page 20 of 118 /

We believe that the financial targets should however be adjusted to exclude the impact of AMT’s de facto lending business, in which it books loan repayments as revenue, thereby artificially boosting EBITDA margins and other operating metrics (see AMT’s De Facto Lending Business Explained).

We are inclined to take a skeptical view of any company whose CEO sells substantially all of the shares he acquires from option exercises. CEOs ideally think about the share price years down the road, and manage accordingly. CEOs who are big sellers of shares tend to be overly-focused on the short-term. In today’s world of lackluster growth, high leverage, and stretched valuations, trading the future for the present has more deleterious consequences than ever. We have the impression that many of AMT’s emerging and frontier markets investments are driven more by a desire to have a “story” than a sustainable business. (We have a similar opinion of AMT’s decision to convert to a REIT, which we see as being incompatible with its geographic scope). The tower operators’ halcyon days are at an end, and AMT would have been far better served investing its cash in other ways.

We see the specific causes of AMT’s overseas value destruction as being CEO and senior management compensation tied to international tower acquisitions, including sale and leaseback lending transactions, 52 and CEO compensation tied to stock price performance. 53 To exacerbate this, AMT CEO Jim Taiclet sells approximately 90% of the shares he receives from option exercises, as shown in the table below. Substantially all buys are from option exercises.

below. Substantially all buys are from option exercises. We make comparisons to Crown Castle International Corp.

We make comparisons to Crown Castle International Corp. (NYSE: CCI) throughout this report because it is managed in a much more sustainable and prudent fashion than is AMT. We note the following superior aspects of CCI’s corporate governance and compensation:

CCI separates the chairman and CEO positions.

CCI does not compensate senior management for tower acquisitions.

////////////////////////////////////////////////////////

52 AMT 2013 Proxy Statement, p. 28. 53 AMT 2013 Proxy Statement, p. 27.

/

‘Page 21 of 118 /

AMT’S INTERNATIONAL BUSINESS IS A DE FACTO LENDING BUSINESS, AND ARTIFICIALLY BOOSTS AMT’S REVENUE, EBITDA, AND PROFIT, WHILE SHARPLY CURTAILING GROWTH PROSPECTS.

AMT’s international business is part tower operator and part lender to its anchor tenants. This presents several significant issues. Because AMT books these loan repayments (i.e. return legs of financial round trips) as rent revenue, the repayments artificially inflate AMT’s revenue, gross profit, EBITDA, and AFFO. AMT guides investors to forecast annual international revenue growth of 10% to 12%; but, because rents from anchor tenants are often partly loan repayments, adding tenants to the towers generates relatively little growth. Thus, AMT is forced to spend more money and make more loans in order to meet these expectations – in other words, it has to keep feeding the beast. The biggest growth issue though is that when the initial leaseback terms end, the anchor tenants’ rents usually reset much lower, and growth becomes significantly negative. AMT’s counterparties can be risky credits, and these de facto loans significantly increase counterparty risk. These loans are part of the reason why AMT often generates IRRs lower than government bonds in three of the four markets we analyzed, and these loans serve only to exacerbate the value destruction. There might be adverse tax effects of this business that further depress IRRs.

AMT’s De Facto Lending Business Explained

When AMT purchases a tower from a carrier and leases it back, both parties have the opportunity to structure the transaction in a way that improves the optics of their financials. If AMT pays an amount above market for the tower in exchange for the carrier paying an above market rent, it confers the following (investor unfriendly) benefits on the parties:

The carrier shows more cash on its balance sheet.

The carrier’s liability is off balance sheet.

AMT shows more assets.

AMT engineers more revenue, and at a 100% pretax profit margin, thereby making its income metrics more impressive in the short-term.

The typical monthly market rent in emerging markets is now about US$1,000 per tenant per tower. (Industry sources state that carrier rents can be as low as US$300 per month where there is a single antenna and no ground space required.) Market rents are what carriers pay to co-locate on a tower, or in the case of a new tower, to have a tower operator build to suit. In sale and leasebacks, as well as build to suits, lease terms are usually 10 years. The build to suit price of a tower in EM depends on several factors, but is often between US$100,000 to US$150,000. In this price range, carriers are willing to build their own towers; or, they might prefer to have a tower operator build the tower, and lease it from the carrier for approximately US$1,000 per month.

/

‘Page 22 of 118 /

If AMT is buying a tower from a carrier, two other major factors come into play: the age of the tower, and its construction quality. Both of these will determine whether AMT needs to spend money, and how much, in order to co-locate tenants on the tower.

Assume for the sake of simplicity that Carrier A owns a single tenant tower – in other words, it is the only tenant. After making allowances for any maintenance cap ex, AMT determines that it is willing to pay US$100,000 for the tower with a market rent back of US$1,000 per month with CPI escalators. If Carrier A realizes that AMT is a deep- pocketed aggressive company with low funding costs, it might tell AMT that it wants $150,000 for the tower. AMT might very well pay that inflated price, but in exchange it would receive inflated rent.

How Lending Distorts AMT’s Income Statement

Using the simple example above, assume that AMT and Carrier A agree that there are three options:

US$100,000 market purchase, 10-year market leaseback at US$1,000 per month.

US$125,000 purchase, 10-year leaseback at US$1,250 per month.

US$150,000 purchase, 10-year leaseback at US$1,500 per month.

We will use a 3% annual rent and expense escalator, and will assume that monthly cash operating expenses are US$400. As shown below, by increasing the price it pays for the tower, AMT increases its EBITDA margin from 60.0% to 73.3%.

AMT increases its EBITDA margin from 60.0% to 73.3%. As shown in the table below, the

As shown in the table below, the IRR drops as AMT pays above market price, assuming that there is a linear relationship between overpayments and over-rents. We explain how we calculate these IRRs below. (We analyze AMT’s IRRs in four overseas markets infra, concluding that investments in all four should be significantly impaired, and that the investments return less than government bonds.)

and that the investments return less than government bonds.) AMT can easily end up generating lower

AMT can easily end up generating lower transaction IRRs when it lends money to the carriers. The hypothetical above assumes:

/

‘Page 23 of 118 /

AMT is able to depreciate or amortize the entire amount of the asset purchase over 10 years. This ensures that we are assuming these transactions are tax neutral. In reality, AMT might be paying additional local tax due to these transactions (see Potentially Adverse Tax Treatment Due to Booking Principal Repayments As Revenue infra).

We assumed a local corporate tax rate of 25%.

Assumes no co-locations – we are only modeling the anchor tenant economics.

We assumed no maintenance cap ex or expenses other than US$400 per month, and we have allocated no corporate SG&A.

At the termination of the 10-year leases, Carrier A’s rent reverts to market, which we have assumed increased by 3% per year from the commencement of the lease.

To calculate terminal value, we assumed a WACC of 9.5% (which we believe is conservative, as we explain in Appendix A.5: Cost of Capital) and a long-term growth rate of 3%. Neither of these assumptions affect whether IRR decreases through lending.

One of the issues for investors trying to understand this business model is that rents do not necessarily move in a linear fashion with above market multiples. There are qualitative factors that affect tower operators’ decision making process with respect to these negotiations with carriers, so it is very difficult for investors to understand what IRRs AMT generates on these transactions. The Company has stated internal IRR hurdles, but as we detail infra in this report, AMT’s stated IRR hurdle rates are too low for the markets in which they operate, and our models show that AMT will likely not achieve its IRR hurdles for most of these markets.

Lending Model is Inherently Lower Growth

When AMT lends money to a carrier as part of a sale and leaseback, it creates two impediments to future growth. First, co-locations do not move the needle as much as AMT would like investors to believe. Second, the carrier’s rent will most likely reset to market rates (i.e. significantly lower) at the end of the initial leaseback period.

The table below shows the revenue CAGRs over 10 years with co-locating 0.1 tenants (at market) for the full year, beginning at the commencement of Year 2.

10-Year Revenue CAGR

Market

10.6%

1.25x

Market

9.4%

1.5x

Market

8.5%

The leaseback agreements are generally 10 years. When the lease ends, and the anchor tenant’s rent resets to market, the drag on growth is substantial. The table below shows

/

‘Page 24 of 118 /

the YOY changes in Year 11 revenue and EBITDA, assuming in Year 11, the tenants per tower ratio grows by 0.1 (to 2.0).

Year 11, the tenants per tower ratio grows by 0.1 (to 2.0). AMT will likely respond

AMT will likely respond to the foregoing by pointing out that its leasebacks provide for renewal rates at the above market rates. We understand that it is rare that a carrier renews above market. The carriers understand that they have repaid the loan, and are reluctant to pay more than market.

Overseas, the leverage is firmly on the carriers’ side. This is an important point for US investors to understand not only in the context of the de facto lending business, but in all aspects of the international business. The third party tower operator came into being in the US right at the transition from analog to digital, and before the cell phone became ubiquitous. The transition to independent operators happened quickly, and in their haste, the US carriers ceded too much leverage to the tower operators. As we discuss infra, this pendulum is now finally swinging back the other way. However, we believe that foreign carriers have learned valuable lessons at the expense of the US carriers about how to work with the tower operators and not lose the advantage. Jay Brown of Crown Castle, Inc. (NYSE: CCI) hit the nail on the head when he talked about overseas moats:

“…as we look at emerging markets, we just have not gotten comfortable that the moats around the business are anywhere close to as good as the U.S. and the credit quality of the tenants who are willing to co-locate or to sell their assets in those markets, at least up until this point has not been, to us, has not had the right risk reward ratio there.” 54

In practice when AMT’s overseas above market leases have come up for renewal, the carriers have usually negotiated the rents downward. AMT might get some build to suit business in return for ignoring the specified renewal rate. AMT cannot muscle the overseas carriers around – it has too few towers in these markets, and is dependent on a limited number of customers. That factor is particularly acute with Telefónica and American Movil, which are its customers in multiple markets. If push really came to shove, we have no doubt that the carriers would win, and we believe AMT knows that. Cellular phone service is a utility, and particularly given that most carriers in AMT’s markets are struggling, consumers and governments recognize that if costs are high,

////////////////////////////////////////////////////////

54 Jay Brown, CFO, SVP Crown Castle International at Deutsche Bank Securities Inc. Media and Telecom Conference, Palm Beach Apr 11, 2011.

/

‘Page 25 of 118 /

consumers will end up paying out of their pockets. As a foreign company with little political clout, AMT has to be very careful about threatening to take carrier equipment off the towers – few of its host governments would tolerate disruption of a public utility for the sake of AMT’s profits.

How Prevalent is Lending?

To illustrate how much of AMT’s international business consists of financing revenue (i.e. return of principal and interest), the below graphs by region of carrier sale and leaseback transactions make several points clear.

LatAm Ex. SS Acquisitons

$350 Telef Axtel $300 Algar $250 Telef Vivo $200 Telef Telef Telef Telef Telef $150
$350
Telef
Axtel
$300
Algar
$250
Telef
Vivo
$200
Telef Telef
Telef
Telef
Telef
$150
Movil
Movil
VTR
$100
$50
$0
Avg.$/Site ($'000)

3Q09 2Q10 3Q10 4Q10 1Q11 3Q11 4Q11 4Q11 1Q12 3Q12 4Q12 1Q13 1Q13

2Q10 3Q10 4Q10 1Q11 3Q11 4Q11 4Q11 1Q12 3Q12 4Q12 1Q13 1Q13 PP&E $ / Tower

PP&E $ / Tower

Network Location $ / Tower

Cust. Relationship $ / Tower3Q10 4Q10 1Q11 3Q11 4Q11 4Q11 1Q12 3Q12 4Q12 1Q13 1Q13 PP&E $ / Tower Network

Goodwill $ / Tower1Q11 3Q11 4Q11 4Q11 1Q12 3Q12 4Q12 1Q13 1Q13 PP&E $ / Tower Network Location $

Note: This graph excludes the Site Sharing acquisition.

In Latin America, it appears as though the tower steel is generally worth approximately US$100,000 in more recent transactions. Previously that number was about $50,000. It appears as though at least five to six of the 12 to 13 carrier sale and leasebacks are financing deals, 55 based on the premium over tower steel book values. This observation is based on the assumptions that co-location rent revenue is not significant due to a combination of low tenancy ratios on the acquired towers and monthly rent rates that are much lower than those of the US. Thus, we assume that transactions in which intangible assets approach (or exceed) 100% of the PP&E value are lending transactions. Depending on how diligently AMT allocates its intangible assets, significant goodwill and customer relationship intangibles might also be signs of lending transactions.

////////////////////////////////////////////////////////

55 We do not know whether the undisclosed seller was a carrier.

/

‘Page 26 of 118 /

Africa Acquisitions

$300 MTN $250 Cell C MTN $200 $150 $100 $50 $0 Avg.$/Site ($'000)
$300
MTN
$250
Cell C
MTN
$200
$150
$100
$50
$0
Avg.$/Site ($'000)
C MTN $200 $150 $100 $50 $0 Avg.$/Site ($'000) 4Q10 PP&E $ / Tower Network Location

4Q10

PP&E $ / Tower Network Location $ / Tower

4Q10

Date Announced

4Q11

Cust. Relationship $ / Tower$0 Avg.$/Site ($'000) 4Q10 PP&E $ / Tower Network Location $ / Tower 4Q10 Date Announced

Goodwill $ / Tower($'000) 4Q10 PP&E $ / Tower Network Location $ / Tower 4Q10 Date Announced 4Q11 Cust.

It appears as at least one of AMT’s three purchases in Africa could be a lending transaction, based on intangible assets to PP&E. This transaction would be the Cell C sale and leaseback announced in November 2010.

Lending Business’s Effect on EBITDA

We conservatively estimate that a total of US$162 million of EBITDA is coming from such over market contracts (on average). The present value of above market contracts is US$728 million based on our projections of tenant churn and utilizing a 10-year remaining life on the respective contracts. Our analysis is subject to change based on changes in assumptions on churn, and length of contracts, among other factors.

Potentially Adverse Tax Treatment due to Booking Principal Repayments as Revenue

AMT can experience adverse tax consequences because it books the loan repayments as rental revenue. (In contrast, banks only book interest payments as revenue.) Avoiding net negative tax consequences requires a degree of care in structuring these transactions that we do not believe is generally present in AMT’s international acquisitions. AMT’s use of a slightly rosy assumption about co-location lease ups could cause AMT to ignore the tax implications of treating principal repayment as revenue.

AMT will owe income taxes on the principal repayments, if it is not able to depreciate or amortize an asset over a period of time no greater than that of the leaseback agreement.

/

‘Page 27 of 118 /

In Brazil, the entire tower asset would likely be recorded as PP&E (unless AMT is acquiring a concession, rather than fee simple ownership of the tower). A new tower is depreciated over 25 years, and a used tower is depreciated either over half of its useful new life (i.e. 12.5 years) or over the residual of its useful life. Brazil therefore could provide for a serious mismatch between revenue and offsetting expenses. Income is taxed at 25% for the corporate income tax and an additional 9% for the social contribution tax.

In Mexico, AMT may depreciate its tangible assets over 10 years, meaning that it is likely able to fully offset the income tax effect. The corporate income tax in Mexico is currently 30%, but will be 29% in 2014 and 28% in 2015.

Conclusion

AMT’s de facto international lending business serves to mislead investors about AMT’s

overseas profitability, while incentivizing AMT to enter into even more such transactions

in order to feed the beast. As we detail infra, AMT’s international business is often

destroying value, and in some cases earning returns lower than those of local government bonds and usually below the required return for the market. The de facto lending

business exacerbates these problems.

AMT’S INTERNATIONAL BUSINESS IS PART CARRY TRADE AND PART LEVERED DIRECTIONAL CURRENCY BET.

AMT’s international business is also effectively part emerging / frontier market carry trade, and part levered directional currency bet. AMT is largely funding in the US by borrowing US$8.85 billion, which is up from US$3.5 billion in 2002. It invests this

capital in streams of fixed cash flows in high interest rate countries. 56 The high local interest rates are the primary determinants of AMT’s local IRRs. (Problematically, AMT

is generally underperforming local government bonds.)

A standard carry trade is dependent on the foreign currency not depreciating (whether or

not due to hedging). As we discuss next, AMT is un-hedged and is earning its carry in volatile (and generally greatly depreciated) currencies. 57 Beyond the currency risk, AMT’s business model layers on top significant risk, including counterparty risk and execution risk (i.e., inability to match performance to its models). AMT’s international returns certainly do not compensate investors for the risks of these carry trades in emerging and frontier markets. As we note supra, AMT’s executive compensation targets are adjusted for currency fluctuations.

////////////////////////////////////////////////////////

56 Germany is an exception to this, but we estimate AMT’s IRR in Germany at only 4.0%. 57 AMT has US$427 million of local currency debt, which is approximately only 10% of its international balance long-lived assets. It announced a multi-currency credit facility on July 1, 2013; however, because the rates are based on LIBOR rates, the facility

/

‘Page 28 of 118 /

It is crucial for investors to understand that AMT does not hedge its currency exposure, which puts its investments and cash flows at risk. AMT states that it does not hedge its currency exposure because hedges for these currencies are expensive. We estimate that AMT has exposure to emerging and frontier market currencies of approximately US$4.2 billion, or 36% of its long-lived asset base. The chart below shows the performance since 2010 of four of the EM currencies to which AMT has directional exposure, the Brazilian Real, Indian Rupee, Ghana Cedis, and the South African Rand – a hedge fund using leverage for this trade would have been unhappy.

fund using leverage for this trade would have been unhappy. Source: Bloomberg None of the five

Source: Bloomberg

None of the five sell-side analyst models we have reviewed have FX forecasts. 58 Yet, in two of AMT’s major currencies, Indian Rupees and Brazilian Reals, exchange rates have already moved substantially from prior year numbers.

The table below compares the ask prices for one-year non-deliverable forwards on AMT’s currencies versus the USD.

FX rates as of July 12, 2013

2015 AMT Analysts Est

current

one year NDF

BRL

2.00

2.27

2.44

CLP

472

504

527

COP

1821

1904

1979

GHS

1.95

2.045

2.07

INR

54.35

59.63

63.48

MXN

12.29

12.81

12.82

PEN

2.61

2.77

2.868

ZAR

9.08

9.98

10.55

UGX

2619.00

2590

n/a

EUR

1.30

1.3

1.3

////////////////////////////////////////////////////////

58 One analyst projected the next two quarters currency exchange rates, and held currency constant thereafter.

/

‘Page 29 of 118 /

It is clear from the table above that it is expensive to hedge all the currencies in which AMT does business, except for the EUR. However, the hedging costs are functions of currency volatility. The more expensive it is to hedge a given currency, the more likely it is that currency movements will impact asset valuations. This is one more reason why the perception that AMT’s cash flow is stable is incorrect.

AMT has a small amount of local currency debt that is a very partial offset to its exposures. It has US$144 million in Colombia, out of US$351.8 million invested; and, US$90 million in South Africa, out of US$249.5 million invested. On July 1, 2013, AMT announced that it had entered into a multi-currency credit facility with rates linked to LIBOR rates. This facility appears not to hedge AMT’s currency exposures, save for that to EUR. There are no LIBOR rates for any of AMT’s other currencies (beside dollars).

AMT’s use of inflation-linked escalators in its overseas businesses is a poor hedge against currency fluctuations. Further, we believe that AMT’s business in its largest overseas market by towers, India, operates with fixed escalators, rather than those linked to CPI. Even though market practice in Brazil is to use inflation-linked escalators, AMT’s average rents have been dropping year-over-year since 2008, despite the country generally experiencing inflation approximating 6%.

CPI escalators are poor substitutes for currency hedges because the movements are not highly inversely correlated. This year, Brazil’s official CPI has been 6.2%. However, the Brazilian Real has weakened 9.8%. South Africa’s CPI has been 5.8% YTD, yet the Rand has weakened to 16.3%. In India, the relationship has actually been more direct, with CPI at 9.3% and the Rupee only weakening 9.1% YTD.

Currencies do not necessarily move in an inverse correlation to inflation. Currency movements are relative to one another, while CPI measures inflation in one market. (Albeit currency movements affect inflation and vice-versa.) The factors determining the relative demand for currencies are highly complex, and beyond the scope of this report. Suffice to say, AMT largely funds in US dollars, AMT’s shareholders buy and sell AMT’s shares in US dollars, and receive dividends from AMT in US dollars. CPI escalators overseas do not hedge these risks.

AMT is likely not using inflation-linked escalators in India, which is its largest international market by towers and accounts for 32.9% of AMT’s international tower base. 59 According to a competitor of AMT and an India-based analyst who covers another competitor, Bharti Infratel (BHIN:IN), it would be unusual for a tower operator to have inflation-linked escalators; rather, contracts usually provide fixed escalators of approximately 3%. India’s CPI rose at 8.9% and 9.3% in 2011 and 2012, respectively. To the best of these individuals’ knowledge, AMT utilizes fixed escalators. Our calculations below confirm our thesis that AMT is not using inflation-linked escalators in India.

////////////////////////////////////////////////////////

59 Based on FY2012 tower numbers shown in AMT FY2012 10-K, p. 30.

/

‘Page 30 of 118 /

Even though AMT should be using inflation-linked escalators in Brazil, which accounts for 13.8% of AMT’s international tower base, its average rents are declining in local currency terms. Our calculations below illustrate the paradox.

terms. Our calculations below illustrate the paradox. P OOR I NTERNATIONAL IRR AND N ECESSARY I

POOR INTERNATIONAL IRR AND NECESSARY IMPAIRMENTS

Overstated International Growth and Value Destruction

Overs tated International Growth and Value Destruction Much of the value that is associated with AMT

Much of the value that is associated with AMT arises from the excitement over ATL’s growth compared to other REITs. With growth in the US tower market slowing, the market is expecting the double digit growth to be coming from ATL’s international subsidiaries. We have investigated a number of ATL’s international operations and instead of finding double digit growth we found disappointing markets with significant headwinds, risks and overpriced acquisitions leading to substantial value destruction. This section will specifically address the issues in India, Brazil, Germany, and Ghana and demonstrate that AMT has destroyed at least US$1 billion of shareholder value.

/

‘Page 31 of 118 /

We estimate that international entities generate a consolidated segment EBITDA of US $482 million as

We estimate that international entities generate a consolidated segment EBITDA of US$482 million as of the latest twelve month period ending March 31, 2013. The market multiple for the Company as a whole is 21x EBITDA. Our analysis shows the market trends, revenue growth, profitability, return on investment, and risk factors do not support anywhere near a 21x EBITDA multiple for the international business. The cash flow generating capacity is just not there, and we caution “don’t believe everything you hear from management.”

India '

“Relative to India, you have to be careful about what you read out of all the press and all the things that are going on in India. But stepping back, we view, as we've always viewed India, is a very, very important market, a market that fits nicely within a diversified portfolio in American Tower.” / Tom Bartlett, ATL, CFO, EVP American Tower Corp at REITWeek®: NAREIT's Investor Forum Chicago June 6, 2013 /

“Only US investors would buy [Bharti Infratel, an Indian tower company going public] at this price.” Indian Telecom Analyst

Mr. Bartlett is referring to the 2011 crisis and its aftermath in the Indian telecom industry. The revelations of the 2G scandal in 2011 brought deep uncertainty to the market, which was further rocked by the Supreme Court quashing 122 of the spectrum licenses. Vodafone’s bitter tax dispute with the Indian government also highlights the regulatory risk in India. Adding fuel to the fire, our sources have told us that the Indian government realized that it could raise more money by selling licenses to a large number of carriers, which has brought about intense competition to the telecom market. The pressure has been passed down to the tower companies, where AMT is now the only foreign player left in India.

Nevertheless, despite all this news, India is seen by both management and analysts as a high growth market for ATL. As we discuss in the report, AMT is valued at such a high multiple due to the growth prospects of the international portfolio as the US growth is limited.

/

‘Page 32 of 118 /

Company and Street estimates for the international portfolio are far too optimistic and it appears there is a significant disconnect between how these international markets are viewed from the US and how they are seen by locals, India being a prime example of this. This disconnect can even be seen by the stock price performance of Bharti Infratel (BHIN:IN), a pure play Indian tower company, since its IPO in December 2012. BHIN IPO’d at INR220 but it is now trading at INR157.05, a fall of 28.6%. Local analysts strongly believed at the time that the stock was well overpriced because US investors pushed the IPO value up on the assumption that the Indian tower market would replicate the US tower market’s performance. As is the case with AMT, US investors mistakenly assumed the tower model can flourish elsewhere without understanding the local issues.

We also fundamentally disagree with Mr. Bartlett’s assertion that shareholders “have to be careful about what you read out of all the press and all the things that are going on in India”. We have performed deep due diligence in India and have found a market with significant head winds and regulatory uncertainty that will have a substantial impact on ATL’s future operations on the market. Shareholders should not take managements’ word, but should instead read and evaluate all the information in the local press and come to their own conclusions.

Marten Pieters, MD & CEO of Vodafone India, said recently in an interview with The Times of India regarding the telecom market in India – “The worst is over for this industry. That doesn't mean that we are there and the future is rosy”. 60 Mr. Pieters was referring to the carriers, however, based on the issues below it appears that the problems for the smaller tower companies are just starting.

Revenue Growth is Not What It Seems and Is Well Below CPI

ATL’s problems in India arose from its very first steps in the country. The contract escalators it negotiated with the telecom carriers were set at 3%, compared to CPI of 9%. At the time, AMT had hoped that the perceived substantial growth in the market would offset the loss from inflation, however, this growth never materialized.

The revenue growth is also not what it seems. Approximately 35% of total India revenue is reimbursements from carriers for energy costs, this has been grossed up in the AMT 10-Ks. AMT’s local accounts only show net revenue. This revenue has no impact on the bottom line at all, however, it does assist AMT in masking its revenue growth because these reimbursements are growing at CPI well above the rent escalators of 3%. While shareholders might be seeing a significant growth in revenue, little of it is hitting the bottom line.

India revenue increased from US$171 million in 2011 to US$182 million in 2012, an increase of only 6.5% which is far below the double digit growth that analysts were expecting. As per the above points we believe that some of this growth was in fact a growth in pass through revenue and did not hit the bottom line.

////////////////////////////////////////////////////////

60 http://economictimes.indiatimes.com/opinion/interviews/indian-telecom-sector-did-not-take-lessons-

from-other-countries-marten-pieters-vodafone-india/articleshow/21060585.cms

/

‘Page 33 of 118 /

The insubstantial rent escalators are also causing ATL’s margins to be squeezed because they are growing at CPI or higher, for example we understand from local sources that wages are growing at 15% per annum and there are significant security costs for many of the greenfield sites which are rising with every new tower build.

Currency risk

The below graph depicts the performance of the Indian Rupee against the US Dollar since AMT started investing in India. As can be seen the INR has deprecated significantly against the US Dollar and highlights the risk of investing in foreign jurisdictions without hedging the currency exchange which AMT has not.

without hedging the currency exchange which AMT has not. Whilst AMT has not taken an impairment

Whilst AMT has not taken an impairment of its intangible assets, there was nevertheless a decrease in long-lived assets in its Indian subsidiaries from US$807 million in 2010 to US$671 million in 2011, 61 which may have been caused by currency fluctuations.

As AMT has not hedged its currency risk we believe that this could cause further devaluation of revenue and assets in the future and will most likely lead to AMT performing below expectations.

Competitive Pressures Limit Growth

Indus Towers, the largest tower company in India, is a joint venture between India’s three largest carriers Bharti Airtel, Vodafone and Idea Cellular; and, is 42% owned by Bharti Infratel. Indus has first option on building new towers for each of these carriers. We believe that Indus will take up this option unless the tower is economically unviable, and this will limit AMT’s growth in new tower builds, or at least AMT will find that its

////////////////////////////////////////////////////////

61 AMT 2011 10-K, p. F-65.

/

‘Page 34 of 118 /

anchor tenants will be the smaller carriers with higher counterparty risk. With Indus and Bharti Infratel having substantial access to the three largest carriers (62% market share) it means that smaller tower companies, such as AMT, may be more reliant on smaller operators that have greater counterparty risk.

We understand from local industry sources that there is talk the tower companies may also face increased competition from power transmission operators, such as Power Grid Corporation of India Limited, which might let its existing infrastructure be utilized by wireless telecommunications service providers for installation of their active telecommunications equipment.

Another issue facing smaller tower companies is the lack of stringent zoning laws in India. Tower companies are free to build towers close to a competitor’s tower, which has led to a large number of towers being situated in close proximity to each other. As a result, the carrier has stronger pricing power as it can move towers without a material effect on its network. Additionally, one can see a number of towers that have no tenants. We understand that some of these are AMT towers.

There are also significant rent abatements in India, that come into effect when an additional tenant co-locates to a tower. We understand that a carrier can save up to 15% if there is one other tenant on the tower and this discount will increase with any additional tenants. This means that even when additional tenants co-locate to an AMT tower, there is a heavy discount on the rent, and revenue growth is not linear with tenant growth.

We understand from Bharti Infratel’s management that it offers additional discounts on top of this rent abatement in order to attract more tenants. We do not believe that ATL, with only 10,000 towers, can compete with Bharti and Indus, which own approximately 150,000 towers between them.

Regulatory issues

The Indian Department of Telecommunications is expected to impose an annual 8% of revenue license fee on tower operators. 62 In our discussion with local lawyers we understand that the fee is currently in abeyance and a date for its introduction is yet to be set. Nevertheless, this license fee is another cost that tower companies are going to be facing which they cannot pass on to customers.

Tower companies including American Tower have recently been given notice by the National Green Tribunal that its towers may have been constructed in violation of stipulated guidelines. 63 We understand from local lawyers that due to the gravity of the allegations and potential health hazard, the Tribunal has directed that there should be no construction of towers without following the mandatory provisions of law, and without

////////////////////////////////////////////////////////

62 http://www.business-standard.com/article/economy-policy/govt-considering-8-licence-fee-on-telecom-

tower-firms-113021300776_1.html

63 http://www.indianexpress.com/news/green-tribunal-issues-notice-to-american-tower-corp-on-mobile-

towers/1050137/

/

‘Page 35 of 118 /

obtaining necessary permission from the competent authority. The hearings are currently in process and we are not able to assess the potential impact, however, this could be significant if the Tribunal require towers to be pulled down.

The Indian government has recently approved the Telecom Regulatory Authority of India’s recent suggestions to reduce the carbon footprint of the telecom sector by mandating the reduction of dependence on diesel. 64 We understand from local sources that Indus is expecting that it will cost INR1 million per 5KW tower to come in line with government expectations. Based on this and assuming that AMT have a 50:50 split between rural and urban towers, we believe that the India operation will have additional capital expenditures of approximately $128m over the next ten years. It should be noted that this expenditure will not improve ATL’s margins as energy costs are all passed through to carriers who will solely benefit from this expenditure.

Carrier Effects on Growth and Pricing

Carriers are suffering from margin squeezes and have been looking to decrease the rents they pay on towers. We have heard from a number of local industry experts that the rents in some of the contracts coming up for renegotiation are as low as half of what they use to be pre the 2011 crisis. This has not come through in ATL’s revenue yet as leases have not come up for renewal yet but once they do they will find significant pressures on rental prices received.

Our local sources also insisted that cost saving pressures coming from carriers is one of the main threats facing tower companies in India. From 2007 to 2012, ARPUs in India have fallen at a 13% CAGR and there has been intense competition in a market that is oversaturated with carriers. These sources have informed us, that to combat falling margins carriers are starting to pressure tower companies to cut the costs that are passed through to them, fuel being the main component. We understand that carriers are willing to move equipment from their current towers to any tower that offers the greatest cost savings. In order to retain their tenants, tower companies will be forced to invest significant capital expenditure to achieve the required cost savings. Unfortunately for shareholders, these large expenditures will have little if any benefit on ATL’s margins.

The smaller tower companies, including ATL, had been hoping to benefit from Reliance Industries’ 4G roll out on their towers, however, Reliance has recently made a deal with Reliance Communications to collocate on their towers only to the exclusion of any other tower company. 65

////////////////////////////////////////////////////////

64 http://www.igovernment.in/site/telecom-towers-be-powered-renewable-energy

65 http://in.reuters.com/article/2013/06/07/reliance-industries-reliancecomm-towers-

idINDEE95603P20130607

/

‘Page 36 of 118 /

We also understand that Reliance will be paying INR15k per month compared to AMT and Bharti Infratel’s INR30-35k per month, which demonstrates the effect of pricing pressure. 66

It does not appear that the 3G roll out has had any substantial impact on revenues as has been the case with Bharti Infratel’s revenue per tenant per tower and as such we believe that the 3G roll out is unlikely to provide any significant benefit to ATL. 67 The high cost of using 3G data has seen a slower uptake than expected and this could also impact the 4G roll out. 68 If carriers decide to lower rates in order to increase the 3G take up then this could lead them to further squeeze tower companies. Some analysts also believe that the slowdown in the subscriber numbers and maturity of the 2G segment have put a question mark on some of the growth assumptions. 69

The Indian government sold license to too many carriers for the market size which has led to fierce competition with a number of carriers being loss making. As such it is expected that there will be further consolidation of carriers in the market, which will impact the number of tenants and could have a disproportionate effect on the smaller tower companies such as ATL.

Turn for the Worst

The tower market in Indian is in the worst shape in years. The market peaked in late 2010 / early 2011 and since then the market for tower companies has collapsed. We understand from India media sources that a local tower operator called TowerVision, which has a comparable portfolio to ATL, has been up for sale for some time with only AMT appearing to be a potential buyer. 70 Based on a recent article, AMT value the transaction at US$50,000 to US$58,000 per tower, 71 which compares to tower values of US$100,000 to US$120,000 at the peak of the market before 2011. This shows how far the market has fallen and indicates that ATL’s portfolio is also not as valuable as it once expected to become. Our industry contacts advise us that few tower operators are willing to pay up for Indian towers given the poor market conditions and increasing operating risks.

////////////////////////////////////////////////////////

66 http://www.business-standard.com/article/opinion/bharti-infratel-reducing-expectations-

113062701109_1.html

67 http://www.business-standard.com/article/opinion/bharti-infratel-reducing-expectations-

113062701109_1.html

68 http://www.reuters.com/article/2013/06/23/india-telecoms-4g-idUSL3N0EP0PJ20130623

69 http://www.business-standard.com/article/opinion/bharti-infratel-reducing-expectations-

113062701109_1.html

70 http://timesofindia.indiatimes.com/business/india-business/USs-ATC-redials-TowerVision-

deal/articleshow/19671359.cms

71 http://articles.economictimes.indiatimes.com/2013-07-02/news/40329310_1_tower-vision-american-

tower-corporation-atc

/

‘Page 37 of 118 /

Threat of Low-Cost Towers

Low cost towers may change the dynamics of the industry even further. Reliance Industries are said to have developed a tower that would cost merely US$2,000 and Chinese companies are said to have developed a tower to be used in India for US$10,000. 72 This may put further pressure on tower operators to reduce prices and costs.

Value Destruction

As a result of the above, we performed a calculation to assess ATL’s investments in the country. We estimate that AMT has made acquisitions totaling US$763 million and those investments are now worth US$399 million. We believe the market conditions and the outlook for ATL’s business prospects will remain very weak and that AMT should be forced to impair its intangible assets and/or goodwill in India by US$364 million, or 48% of the combined acquisition prices . We project that ATL’s existing Indian investments will only generate an IRR of 7% which is equivalent to the Indian government bonds yield at 7%, and are rated BBB with negative outlook.

We use a WACC of 15.5% India, and an exit multiple of 10x EBITDA. Such discount rates might sound high to US investors, but, again, the Indian government 10-year bond

yield is 7%, and inflation is at 9%. The volatility of the India equity market is also high

at 20%.

movement in the stock market index (up and down) and is somewhat of a proxy for the myriad risks a Western company takes when entering an emerging market.

Equity market volatility is the potential percentage change in the price

The table below summarizes the results of our analysis.

The table below summarizes the results of our analysis. Our View ATL’s Operations in India There

Our View ATL’s Operations in India

of our analysis. Our View ATL’s Operations in India There appears to be a disconnect between
of our analysis. Our View ATL’s Operations in India There appears to be a disconnect between

There appears to be a disconnect between how US investors are seeing growth in India and how local analysts view this and this is demonstrated by the Bharti Infratel stock performance. In light of all the reasons stated above, we believe that AMT face significant obstacles to its growth in revenue and EBITDA and a cash drain from further capital expenditure which will provide no benefit to the margins. We believe that at best, AMT will achieve a 6% CAGR in EBITDA which is hardly the growth driver that the

////////////////////////////////////////////////////////

72 http://www.indianexpress.com/news/reliance-india-plans-lowcost-telecom-towers/1003341/

/

‘Page 38 of 118 /

market is looking for in India and this includes the benefits of the additional capital expenditures that are unlikely to provide a meaningful return to shareholders. As per Mr. Martens quote above, the worst is over for Indian carriers, however, based on research, we have found that expectations on the ground in India are far below the lofty expectations in Boston.

Brazil '

"The strategy among our leadership team at the moment is to sort of consolidate and integrate and heavy up in the countries that we are at today. And so in Brazil, for example, Hal and his team accomplished 600 plus Tower acquisition. Grew the portfolio by over 1/3 in a country that we are already in this year" Jim Taiclet, JPMorgan Technology, Media and Telecom Conference 16-May-11

We believe that AMT has placed such an extreme emphasis on growth that they went for quantity over quality and this is evident in acquisitions such as Vivo in 2012. AMT are also guilty of using flawed models to justify other acquisitions, as in the case of the Site Sharing acquisition in 2011. These mistakes have led AMT to overstating growth and destroying shareholder value in Brazil. ATL’s Brazil is really just a story of two acquisitions.

Site Sharing

Aside from the mystery of "the missing $250 million" discussed supra, we believe this acquisition is plagued with further issues, which will hamper further growth. From our discussions with local industry sources with firsthand knowledge of this transaction, AMT were far from happy with the outcome of this transaction but attempted to only show the positives to investors.

Growth Based on Unsustainable Cash Flows

The Site Sharing sites were leased to Oi and TIM at well above market rates, we have been informed by local sources that these were as high as 2.0x to 3.0x over market rents. It is unclear why these towers were at such a premium. However we understand that the deal was made between Site Sharing and Oi and TIM to build these towers on which they would be the only tenants and in return Site Sharing would be given disproportionately high rents. This most certainly benefited Site Sharing management when they sold it on to AMT for 12 to 13x cash flow. We believe this higher than market rents are contributing US$66 million to EBITDA.

Unfortunately for shareholders, these cash flows are unsustainable. We understand that on the acquisition of these sites there was only an average three to five years left on many of the leases. We understand from sources close to the transaction that these carriers are already strongly pushing to renegotiate their rent even prior to the termination of the contract. Oi and Tim are using their network size as leverage to bring down the cost of the Site Sharing portfolio in line with market rent.

/

‘Page 39 of 118 /

Market rent abatements will mean that any additional tenants on the Site Sharing towers may not add any discernible revenue growth. Firstly, a new tenant would most definitely not be paying the over market rent that Oi or TIM are, otherwise they would find a different tower. A working example would be, if the anchor tenant was paying US$5,000 and a new tenant was added paying US$2,000, then the anchor tenants rent would be reduced to US$3,500, so the additional tenant is adding little growth. We should also note that it would take at least five years for a new tenant to be added on to every tower.

Based on the problematic Site Sharing portfolio, we believe that the above market rent that AMT is currently receiving, which amounts to 43% of operating revenue, is unsustainable and any average rent metrics based off this will be overstated.

Infrastructure Sharing Is a Threat To Growth

Earlier this year, four of Brazil's largest carriers announced infrastructure sharing agreements, which could severely impact the future growth of tower companies in Brazil 73 .

Oi and TIM announced their agreement to share infrastructure (towers and equipment) for LTE networks in the 2.5 GHz spectrum band deployments. The move was approved earlier this year, on April 19, 2013, by ANATEL, The Brazilian Telecommunications Authority. ANATEL voted to allow Oi and TIM to share towers, transmission equipment and broadcast stations. ANATEL further stressed that its intention is to reduce costs for carriers so that benefits can feed through to the consumer. 74

Vivo and Claro announced in March of this year that they had signed a memorandum of understanding to share telecom infrastructure, such as cell sites, backbone and backhaul and even their 3G networks. CADE, Brazil antitrust regulator, has already approved this agreement 75 .

Infrastructure sharing is expected to become an increasing trend and focal point for carriers across Latin America which are looking to deal with falling average revenues per user 76 (“ARPU”) by cutting network costs.

We have discussed this issue with multiple local industry sources, including ATL’s competitors and local tower infrastructure contractors and we are confident that this trend poses a major threat to tower companies in Brazil. Infrastructure sharing will hinder expected growth and hamper any expectation to benefit from the rollup of new networks.

////////////////////////////////////////////////////////

73 / http://www.rcrwireless.com/americas/20130307/carriers/brazils-carriers-announce-agreements- infrastructure-sharing-lte/

74 http://online.wsj.com/article/BT-CO-20130418-717923.html

75 http://www.rcrwireless.com/americas/20130307/carriers/brazils-carriers-announce-agreements-

infrastructure-sharing-lte/

76 AMT Semi-Annual International Market Overview March 31, 2013

/

‘Page 40 of 118 /

With the new technology enabling antennas to handle multiple carriers signals this threat is now a reality.

Poor Acquisitions Mean That Growth Will Require Significant Additional CapEx

We understand from local sources that the poor condition and quality of the 1,692 towers acquired in the 2012 Vivo acquisitions, will require AMT to invest substantial capital expenditure to strengthen the towers so that they can add more tenants. From our due diligence of this acquisition, we suspect that these towers were largely older towers that were constructed in the early stages of Brazil’s tower deployment, with a large percentage of these towers not having a permit (we note this a wider spread problem in Brazil), and many not even having site plans, which would have made it extremely difficult for AMT to determine whether the towers are able to support any additional tenants and plan their CapEx accordingly. Without this most basic information, AMT is having hard time signing new contracts with collocating tenants, without knowing which towers can actually support colocation. Thus, we believe this supports what heard from sources we deem reliable that AMT is not satisfied with this acquisition, having modeled it based on estimations that are not materializing. The post-acquisition work on each tower to assess it structural strength will also delay tenancy growth.

Based on local knowledge we gathered, we estimate that the redevelopment of these sites will cost US$30,000 to US$40,000 per tower. If only 50% of the Vivo towers are ultimately redeveloped to handle more than one tenant per tower, the result would be an additional capital expenditure of US$31 million. Our expectation is that a larger number of the Brazilian towers will need to be redeveloped in order to add more tenants to towers and that the total cost will be approximately US$100 million. If alternative energy regulations are implemented and/or carriers push more of the total cost of operating a tower onto tower companies, additional capital expenditures would be needed to remain competitive.

Currency Risk

The Brazilian Real has depreciated significantly against the US Dollar since AMT accelerated its investments in Brazil. The BRL has deprecated by over 30% in the last 2 years, and 12% in the last 3 months. By not hedging the currency, AMT have taken a substantial currency risk on shareholders’ returns, and if indeed AMT benefited from the carry trade when it first entered the Brazilian market, it is no longer the case and we expect these devalued revenues to have an impact on ATL’s revenues going forward.

/

‘Page 41 of 118 /

Value Destruction As a result of the above, we performed a calculation to assess ATL’s

Value Destruction

As a result of the above, we performed a calculation to assess ATL’s investments in the country. We estimate that AMT has made acquisitions totaling US$988 million and those investments are now worth US$713 million. The material overpayment for the Brazil assets relative to their profitability and risks lead us to believe that AMT should be forced to impair its intangible assets and/or goodwill by US$274 million, or 28% of the combined amounts paid for acquisitions. We project that ATL’s existing Brazil investments will only generate an IRR of 11%. Brazil government bonds yield 11%, and are rated BBB with a negative outlook.

We use a WACC of 17.5% for Brazil, and an exit multiple of 8.0x EBITDA. Such discount rates might sound high to US investors, but the Brazilian government 10-year bond yield is 11%, and inflation is at 6% . The volatility of the Brazilian equity market is one of the highest in the world at 20%. Equity market volatility is the potential percentage change in the price movement in the stock market index (up and down) and is somewhat of a proxy for the myriad risks a Western company takes when entering an emerging market.

somewhat of a proxy for the myriad risks a Western company takes when entering an emerging

/

somewhat of a proxy for the myriad risks a Western company takes when entering an emerging
somewhat of a proxy for the myriad risks a Western company takes when entering an emerging

‘Page 42 of 118 /

Our View of ATL’s Operations in Brazil

AMT have shown little regard for shareholder value when investing in the Brazilian market, its two main investments appear to be classic cases of knee jerk acquisitions with no regard to potential risks and minimal due diligence. We are not surprised to hear from local sources that AMT are looking to potentially acquire a Brazil and Mexico tower portfolio from NII Holdings (NASDAQ: NIHD). Our industry contacts have advised us that few tower operators are willing to take credit risk on NIHD at anywhere near the current asking price. With substantial head winds from infrastructure sharing and carriers looking to cut costs, we do not believe that AMT should be gambling any more money in the market place based on their poor track record.

Ghana '

Growth Ceiling

We understand both from sources operating in Ghana and from an AT Kearny article on Tower Xchange 77 that the tower companies in Ghana and Africa in general should only expect to achieve a maximum of two tenants per tower.

This contention is based on the following:

a. In Ghana there are three main carriers, AMT have a deal with the largest, MTN whilst the other two have deals with Eaton Towers and Helios Towers, which means there is less chance of collocation.

b. We also understand that new tower companies are entering the Ghana market which is increasing competition and will lead to further dilution of tenants and potentially reducing prices of any additional tenants.

c. We have also heard that all future expansion in Ghana is likely to be rural which will require extensive capital expenditure and will require the operation of linking towers in order to increase coverage and few tenants per tower.

As a sale and leaseback transaction, the rent paid by MTN to AMT is above the market rent that AMT would receive from non-anchor tenants and so any additional tenants will be paying lower rent.

Loss Making

At this point in time, the Ghana joint venture is loss making. We understand from the MTN conference call that this loss is below the EBITDA line, nevertheless, after

////////////////////////////////////////////////////////

77 http://www.towerxchange.com/why-tenancy-ratios-above-two-will-be-difficult-to-achieve-in-africa/

/

‘Page 43 of 118 /

investing in a US$517.7 million joint venture it is surprising that it is loss making.

Cash Drain

We understand from industry experts in Ghana that a significant amount of capital expenditure is required in Ghana. Whilst, they did not have direct information regarding ATL/MTN joint venture in Ghana, they have looked at newer tower portfolios and have found that at least 25% of towers built by local carriers were only strong enough for one carrier, another 15% needed refurbishment and a further 10% needed to be fully replaced. We believe that as the MTN’s tower portfolio is older, it is likely to minimally require this amount of capital expenditure.

This raises questions over the level of due diligence that AMT have put into assessing their acquisitions, at a significant cost to shareholders. This cash drain also raises questions over the value of the transaction itself.

Purchase Price Anomalies

On December 6, 2010 AMT had entered into a definitive agreement with MTN in Ghana to form a joint venture to buy 1,876 towers for approximately $430m. Once the final tranche of towers were transferred on December 23, 2011, the joint venture paid $518m for 1,856 towers. No reason has been provided for the increase in purchase price.

In MTN’s 2011 accounts, it seems that ATL’s share of the cost remained at US$219 million but this only 42% of the final purchase price, it should be noted that the amount AMT paid does not appear in MTN’s 2012 accounts.

This anomaly raises a number of questions, not least as to what is the true value of the assets? Why would MTN sell AMT 51% of the assets for just 42% of the value, it does not appear that they were desperate to do so as there would have been other bidders? Could this mean that the assets are overstated?

Value Destruction

As a result of the above, we performed a calculation to assess ATL’s investments in the country. We estimate that AMT has made acquisitions totaling US$518 million and those investments are now worth US$246 million. We believe the market conditions and the outlook for ATL’s business prospects in Ghana are uncertain given the extremely high inflation in the country, high government bond yields, high business risk and questionable joint venture structure. Given the decline in value relative to the amount paid for acquisitions, we believe that AMT should be forced to impair its intangible assets and/or goodwill by US$272 million, or by 53% of the original acquisition price.

We project that ATL’s existing Ghanaian investment will only generate an IRR of 11%. Ghanaian government bonds yield 17 %, and are rated B. We use a WACC of 25% in Ghana, and an exit multiple of 6.0x EBITDA. Such discount rates might sound high to

/

‘Page 44 of 118 /

US investors, but, again, the Ghana, again, the government 10-yeaer bond yield is 17 %, and inflation is at 11%. The Ghana market is also very illiquid and the capital markets are still emerging.

very illiquid and the capital markets are still emerging. Our View ATL’s operations in Ghana The

Our View ATL’s operations in Ghana

are still emerging. Our View ATL’s operations in Ghana The Ghanaian investment appears to be overvalued
are still emerging. Our View ATL’s operations in Ghana The Ghanaian investment appears to be overvalued

The Ghanaian investment appears to be overvalued particularly when compared to their discounted cash flow model valuation and the unjustified increased purchase price. Unfortunately, the acquisition is another example of AMT not performing enough due diligence, at the outset it should have had a better handle on the purchase price and given guidance on the amount of capital expenditure that was require in order to increase its tenancy ratio.

Germany '

AMT bought 2,031 towers in the end of 2012 from KPN/E-Plus for $500m, KPN have since claimed that these towers are not core assets. KPN’s 2012 financial statements seem to indicate the parties have signed a 15 year lease. E-Plus, the tenant on the towers, is the third largest mobile operator in Germany with a market share of 21% and is owned by Dutch KPN, which is considered a player in the low cost end of the cellular market.

The German wireless market is saturated with a penetration rate of 140% for 3G and 47% for 4G. Germany has about 35,000 site locations with 70,000 base stations, ATL’s acquisition represents about 5.7% of the total German market.

The sale by KPN was the first tower sale in Germany. Historically all the towers were owned by Telekom and when the market opened up for competition in the 1990’s other operators started to build their own towers, however, a significant proportion of tower space is leased from Telekom. E-Plus, as discounter, chose to build its own towers to save on the rental payments and we understand that they might have prioritized cost over quality.

Low Growth Market

We also understand from local sources that there is little collocation demand for rural towers in Germany as all operators have complete coverage and E-Plus is not deploying LTE in rural areas. As such and based on our discussions with local experts, there is very

/

‘Page 45 of 118 /

limited growth for AMT in Germany.

Threats of Carrier Consolidation

A major risk for AMT is the possibility of a merger between E-Plus and Telefónica. This

merger has been rumored many times before and KPN has also tried to sell E-Plus to raise funds to fight a takeover attempt by Carlos Slim. 78 We note that there are other examples of KPN and Telefónica working together such as a network sharing pilot in the

Netherlands. 79

A merger might happen in order to save on costs 80 and would likely need to the

decommissioning of sites as both companies have coverage of all of Germany and both

companies use the same technology.

Value Destruction

As a result of the above, we performed a calculation to assess ATL’s investments in the country. We estimate that AMT has made acquisitions totaling US$526 million and those investments are now worth US$419 million. We believe the market conditions and the outlook for ATL’s business prospects in Germany are weak with increasing risks from carrier consolidation, a low quality tower portfolio, and stagnant new tenant growth prospects. Given the decline in value relative to the amount paid for acquisitions, we believe that AMT should be forced to impair its intangible assets and/or goodwill by US$107 million, or by 20% of the original acquisition price. We project that ATL’s existing German investments will only generate an IRR of 4%. German government bonds yield 2%, and are rated AAA with stable outlook. We use a WACC of 7% in Germany , and an exit multiple of 14.0x EBITDA.

of 7% in Germany , and an exit multiple of 14.0x EBITDA. /////////////////////////////////////////////////////// / 7
of 7% in Germany , and an exit multiple of 14.0x EBITDA. /////////////////////////////////////////////////////// / 7
of 7% in Germany , and an exit multiple of 14.0x EBITDA. /////////////////////////////////////////////////////// / 7

////////////////////////////////////////////////////////

78 http://www.fiercewireless.com/europe/story/kpn-ceo-pushes-network-sharing-o2-germany/2012-11-16

79 http://www.mobileworldlive.com/vodafone-and-kpn-partner-for-dutch-sharing-pilot

80 / http://www.mobileworldlive.com/telefonica-mulls-cost-saving-german-alliance-with-kpn /

/

‘Page 46 of 118 /

Our View of ATL’s Operations in Germany

ATL’s assets in rural Germany have low growth prospects but yet there is a risk of carrier consolidation. There does not appear to be any apparent reason for this acquisition and it appears to be expansion for expansion sake.

CELLULAR TOWERS ARE SOON TO BECOME THE MOST EXPENSIVE OPTION OF LAST RESORT; THE HALCYON DAYS ARE OVER.

Fiber and Cell Towers: Is History Repeating Itself?

In 1998, investors were bullish on fiber optic cable owners, because internet traffic was projected to grow 15x over the next five years. 81 At the time, investors were drawn to the logical conclusion that a major increase in traffic meant the need for an equal amount of new fiber optic cable.

Internet traffic met those expectations, but demand for fiber barely grew because improved laser transmission and copper-bonding technologies increased the data capacity

of a single strand of fiber to the point where it and existing distribution could absorb the

increase in demand without much need for additional fiber.

A similar phenomenon is happening right now in the tower industry. AMT is scouring the

globe, purchasing tower assets from other operators. Given the projected increase in global demand for mobile data, investors are lauding its efforts. 82 But are towers the only way to relieve the congestion of global data traffic. Just as in the past, there are other technologies out there that can increase capacity of the existing tower network, including

Wi-Fi, smart cells, sectorization, antenna tilting, and improvements in spectrum efficiency and spectrum sharing. Investors seeking to profit from data demand growth by owning AMT will be disappointed.

Wireless History and Development

The ideas used to develop the wireless network were adapted from traditional telephony. The equipment carriers developed included analog, mechanical devices meant to interface with the traditional analog public switched telephone network (PSTN). One of these strategies developed including overlaying a honeycomb shaped grid over a map of the region in which the carrier was implementing service. They placed a transmission tower at the center of each ‘cell’ in the grid, hence the term ‘cellular’. These towers relayed mobile user’s signals as the user travelled from one transmission area to another.

////////////////////////////////////////////////////////

81 raw.rutgers.edu/MiklosVasarhelyi/Resume%20Articles/CHAPTERS%20IN%20BOOKS/C11.%20network%20demand%20model.p

df

82 http://reboot.fcc.gov/blog?entryId=1257085

/

‘Page 47 of 118 /

Since carriers built and operated their own networks, radio frequency engineers had precise information about their network’s capabilities and performance. Because the carrier controlled all aspects of their private network, these systems were termed Homogeneous networks.

/ / / / Carriers building private networks used electromagnetic spectrum purchased from government auctions
/
/
/
/
Carriers building private networks used electromagnetic spectrum purchased from
government auctions to replace the copper wires of a fixed network. Once their private
networks were complete, carriers began offering consumers access by selling packages of
voice minutes, and eventually access to SMS and data access packages. The
Telecommunications Act of 1996 introduced a new regulatory scheme where entrants
could pay wholesale prices to access carrier networks. This legislation, in conjunction
with the FCC final ruling on Docket 16509 83 concerning private access to AT&T’s
network, eliminated the carrier’s competitive advantage of a private network. Carriers
responded by divesting themselves of assets, including through sale and leaseback of
towers, and marketing arrangements with third party companies such as tower marketing
companies. These transactions provided carriers with added liquidity and allowed them to
focus on their core business. Tower companies could market the towers to other carriers
looking to expand or improve communications at a lower cost than carriers would pay to
develop those assets themselves. Additional tenants greatly improved the economics for
tower companies, giving them strong rents and margins.
Carriers have since begun to invest in a next-generation network based on modern
technology. This new network is digital, uses open communication protocols (Internet
Protocol (IP)), and is designed for metro-based data consumption. These new
‘heterogeneous’ networks are ad-hoc and unplanned compared to the first generation, but
they succeed where it matters: they can meet consumer data demand.
/
////////////////////////////////////////////////////////
83 http://www.cybertelecom.org/industry/mci.htm
/
‘Page 48 of 118 /
/
/

/

/

‘Page 49 of 118 /

Growing Data Demand Will Largely Bypass AMT

The AMT bull case is that data demand is surging and AMT will benefit. However, the next generation data delivery network is here already, and makes towers optional and comparatively expensive. Wi-Fi data delivery already greatly exceeds cellular data delivery, and a new generation of equipment, network sharing, government regulation, and competitors will combine to marginalize tower operators, leaving their networks as the highest cost data delivery option of last resort.

Myopic AMT Management Does Not Grasp the Challenges Ahead

AMT CEO James Taiclet stated in 2012 that “95% of our towers are outside of urban and near-urban environments”. 84 In its June 2013 overview, AMT stated it is "well positioned to capture incremental demand from network densification as a result of the rapid growth in wireless data consumption." 85 However, only 10% of cell sites (including macro towers) handle almost 90% of cellular data traffic. 86 We question how AMT’s towers can be positioned to capture incremental growth when, according to their President, CEO, and Chairman, the vast majority of their towers are not located where the people generating the increased demand are.

Carrier Upgrade Completions Will Reduce AMT Amendment Revenue

AMT generates revenue from its tower network by charging carriers rents. AMT charges for the amount of equipment carriers place on the tower, and for changes to the equipment already on the tower, a fee known as “amendments”. In the US, the tower industry has already captured much of the growth it will see from high-speed data network build outs; therefore, amendments will generate much less revenue than they have in recent years. Verizon’s (VZ) CEO has said that Verizon is 95% done with its 4G upgrades. AT&T (T), the other large domestic carrier, is midway through a 3-year upgrade announced in 2012.

////////////////////////////////////////////////////////

84 http://www.morningstar.com/earnings/49066062-american-tower-amt-q4-2012.aspx

85 http://www.americantower.com/Assets/uploads/files/PDFs/investor- relations/AMT%201Q13%20Financial%20Operational%20Update.pdf, page 20

86 www.jdsu.com/ProductLiterature/smallcellhetnet_wp_nsd_tm_ae.pdf

/

‘Page 50 of 118 /

Technology and network sharing will reduce amendment revenue

Technology and network sharing will reduce amendment revenue Industry observers note that the development of wireless

Industry observers note that the development of wireless communications usually occurs in two phases. During the first “coverage” phase, carriers seek to establish a minimal amount of coverage in their service areas. This coverage is often described as “a mile wide and an inch deep”. During this phase of development, carriers are reluctant to share any network component, because coverage can be a product differentiator. During the second “capacity” phase, carriers focus on expanding the data delivery capacity of their networks and seek out opportunities to reduce costs by sharing some or all of their network.

AMT vs. The Cloud

The second “network densification” stage is focused on adding data delivery capacity and will largely bypass towers to a significant extent. Tower operators’ ability to generate amendment revenue by charging the same carrier for multiple generations of equipment on the same tower will decrease as new generation equipment with greater capabilities replaces older equipment, and carriers can remove old equipment from the towers. For example, equipment today is built with multiple chipsets to handle various voice and data technologies. This reduces the need for additional equipment, and allows carriers to consolidate by sharing network equipment, which is the next logical evolution of tower sharing. As software becomes more embedded in the network infrastructure, upgrades

/

‘Page 51 of 118 /

and configuration changes will be handled via the “cloud” at the software and firmware level, and carriers will not have to replace equipment as often. Technological advancements reduce AMT’s tower utility

JDS Uniphase estimates that network capacity can be increased by 1,000 times in the next ten years without adding much equipment to macro towers through a multiplier effect of adding spectrum, carrying voice over LTE (VoLTE), and relying on small cells. If carriers’ spectrum increases by three times, then LTE can multiply the spectral efficiency by another six times, and small cells can increase the spatial efficiency by 56x, creating a total gain of 1,000x. 87 This gain can be achieved without much reliance on towers.

We believe that Wi-Fi will increase the spectrum availability by more than three times and this combination of technologies greatly threatens rents of the towers themselves.

Wi-Fi is a Grave Threat to AMT’s Business Model

By nature of its ubiquity, improving Quality of Service (QoS), and very low total cost of ownership (TCO), Wi-Fi represents the single biggest threat to the tower industry. It is a mature technology with widespread adoption and as the Cisco VNI data meter shows, it is already used to deliver more data worldwide than the cellular network. Approximately 80% of data traffic is from indoor users. 88 As a communications technology, Wi-Fi has several major advantages over traditional cellular. First, and perhaps most important, Wi- Fi is near-free for the carriers and it is unlicensed, meaning carriers can use it without having to pay license fees in the billions of dollars, and there is no authority to grant, revoke permission or regulate its usage.

authority to grant, revoke permission or regulate its usage. /////////////////////////////////////////////////////// / 8

////////////////////////////////////////////////////////

87 www.jdsu.com/ProductLiterature/smallcellhetnet_wp_nsd_tm_ae.pdf

88 www.jdsu.com/ProductLiterature/smallcellhetnet_wp_nsd_tm_ae.pdf

/

‘Page 52 of 118 /

Wi-Fi presently delivers more smartphone data than cellular. We direct interested readers to the Cisco VNI data meter website for additional regional and tablet data consumption graphs: http://www.ciscovni.com/data-meter/index.html.

In addition to being free, Wi-Fi can deliver significantly more data than 4G. Depending on configuration, the recent version of Wi-Fi, “802.11ac” has a theoretical peak rate of 1.3 GBPS while the 4G theoretical peak rate is 1 GBPS. Realistically speaking, radio interference, distance to antennae, ground speed, backhaul constraints and other factors limit both 4G and Wi-Fi, resulting in 4G data delivery rates in the 10-16 MBPS range and Wi-Fi delivery in the 331 MBPS 89 , an advantage of twenty times.

Average LTE data speeds in 2012

20 15 10 5 0 Mbps
20
15
10
5
0
Mbps

AT&T

Sprint

Verizon

Download

Upload

Source: RootMetrics

It is difficult to overstate the threat Wi-Fi and small cell pose to AMT’s economic model. Wi-Fi gives carriers and consumers a low cost alternative to the current mobile economic model. Although carriers have millions of subscribers, their margins are thin and eroding (as we explain in Carriers Under Pressure infra). This erosion is accelerated when subscribers switch to smart phones, because smart phone users generate an enormous amount of data traffic on the carrier’s private network. The carrier absorbs the cost of supporting this traffic.

These costs provide a powerful incentive to the carrier to find alternative delivery networks, and they in turn encourage consumers to offload to a free network (i.e. Wi-Fi) for their data as soon as possible. Carriers have incentivized consumers by eliminating unlimited data plans and charging consumers for the data delivered. Consumers naturally prefer a free, faster, unmetered service more so than a paid, slower, metered service and respond to the incentive by getting off of the private carrier network and onto the free Wi-Fi network whenever possible.

Carriers such as SK, Orange, and T-Mobile facilitate the transition to the Wi-Fi network by implementing SIM-based authentication, a technique that retains user-device information after the initial login so that so that users are ‘auto-magically’ logged in

////////////////////////////////////////////////////////

89 http://www.zdnet.com/the-five-things-you-need-to-know-about-gigabit-wi-fi-802-11ac-7000017112/

/

‘Page 53 of 118 /

again when they are in range of the network. 90 91 Smart phones simplify the process by notifying customers whenever they are in range of a free or previously accessed network. Carriers that originally invested in towers to support their network are now investing in Wi-Fi and small cells. To that end, AT&T now offers over 20,000 hotspots to its mobile subscribers and there are no usage fees. 92 Boingo (WIFI) claims a global network of over 700,000 hotspots available to consumers for US$8 per month. 93 The next evolutionary step will be integration of Wi-Fi access and the core cellular networks.

Additional technical developments include remote tilting antennas to minimize interference which have “the most direct impact on coverage and interference parameters”. 94 A technique called “transrating” can reduce cellular network bandwidth requirements for streaming video by 30% to 50%. 95 Local content caching is enabling carriers to manage video traffic better by enabling them to store high demand video (e.g., viral YouTube videos) at central locations, or cache video at the small cell itself.

AMT would like investors to believe it is positioned to significantly benefit from the growth in data demand. If AMT investors expect 15-20% growth they will be disappointed. Wi-Fi traffic already exceeds cellular traffic. As technology evolves, cellular tower industry growth will slow considerably. We believe that tower companies’ halcyon days of 74%gross margins from US carriers are numbered. 96 We expect that few, if any, international markets will ever offer tower companies the economics that have been generated in the US. (As we explain supra, ATL’s international gross margins are artificially inflated by its de facto lending business.)

AMT is at a dangerous crossroads in its business: its customers’ margins are compressing, yet carriers have ample room on the balance sheet to invest in the high value, low cost alternatives being offered by existing vendors, as well as by newly- formed companies. The data delivered over these alternate delivery routes already exceeds the data delivered on the cellular network, and their continued development and growth will leave AMT as the highest cost option of last resort going forward.

Small Cells

First generation radios were analog, bulky and could typically perform only a single function. Radios are now embedded with integrated circuits, making them more similar to modern computers than their predecessors. LightRadio (pictured below) is part of a new generation of palm-sized Wi-Fi equipment manufactured by Alcatel-Lucent (NYSE:

ALU) that puts 2G, 3G, 4G and Wi-Fi functionality into a single antennae with its own base station equivalent hardware and software. It is inexpensive and small enough to

////////////////////////////////////////////////////////

90 http://www.ipass.com/press-releases/sk-telecom-2/

91 www.jdsu.com/ProductLiterature/smallcellhetnet_wp_nsd_tm_ae.pdf

92 http://www.att.com/shop/wireless/wifi.html#fbid=d_SfZbBu3Iz

93 http://www.boingo.com/wifi-plans/mobile/boingo-mobile/?lang=en

94 www.jdsu.com/ProductLiterature/smallcellhetnet_wp_nsd_tm_ae.pdf

95 www.jdsu.com/ProductLiterature/smallcellhetnet_wp_nsd_tm_ae.pdf

96 Bloomberg

/

‘Page 54 of 118 /

enable carriers to place them anywhere extended network coverage is needed, and can be managed in the cloud.

coverage is needed, and can be managed in the cloud. Source: Money.CNN.com These devices are power

Source: Money.CNN.com

These devices are power efficient and can handle multiple spectrums simultaneously, enabling carriers to share network equipment – the next step in an evolution that began with sharing towers. Small cells are designed for, and excel at providing inexpensive additional capacity in metro areas. In metro areas carriers have advantages not available on towers such as access to previously built public infrastructure, including electrical power and data backhaul on fixed line networks, often controlled by the carriers.

Multifunction small cell manufacturing is a growing industry segment. Others producing similar equipment include traditional manufacturers such as Ericsson, Nokia Siemens, Cisco, and Texas Instruments and young guns such as BelAir Networks, Ruckus Wireless, IP.Access, and SpiderCloud. Carriers deploying small cells include Verizon 97 and AT&T 98

In November 2012, the number of small cells deployed surpassed the number of macro cells deployed since inception, with an estimated 6.0 million small cells, compared to 5.9 million macro cells. 99 Small cells are cheap to buy, easy to deploy and offer carriers a high value mechanism to bring data access to data consumers in metro areas. Carriers and consumers are incentivized to use them because the data delivery cost and total cost of ownership (TCO) is lower than macro cells.

////////////////////////////////////////////////////////

97 http://www.fiercewireless.com/story/verizon-taps-alcatel-lucent-ericsson-lte-small-cells/2013-05-21

98 http://www.fiercetelecom.com/story/att-evolves-small-cell-volte-webrtc-drive-broadband-ubiquity/2013-

05-02

99 www.jdsu.com/ProductLiterature/smallcellhetnet_wp_nsd_tm_ae.pdf

/

‘Page 55 of 118 /

Network Sharing I s Caring 100 Carriers have long understood if they could share infrastructure,

Network Sharing Is Caring

100

Carriers have long understood if they could share infrastructure, they can reduce operating expense and increase profitability, which was the impetus for divesting tower assets in the first place. AMT may have sown the seeds of its own destruction with the initial success of its business model. Carriers are loath to allow tower companies to

continue to reap outsize margins when theirs are compressing. Technology now makes it possible for carriers to share as much of their network as they want, and is swinging the balance of power back to carriers. Foreign governments in some of AMTs markets, such

as Brazil and Colombia, are legislating network sharing because it maximizes scarce

resources, minimizes environmental degradation, saves fuel, and is more efficient economically.

Equipment sharing is a natural evolution from sharing towers.

 

Site

Mast

RAN

Network

Core

Core and

Sharing

Sharing

Sharing

Roaming

Sharing

Platform

Sharing

Property

Yes

 

Yes

 

Yes

 

Yes

 

Yes

 

Yes

 

Mast

No

Yes

Yes

Yes

Yes

Yes

Antenna

No

No

Yes

Yes

Yes

Yes

BTS/BSC

No

No

Yes

Yes

Yes

Yes

Core

No

No

No

Yes

   

Network

Yes

Yes

Value

No

No

No

No

No

Yes

Added

Services

A

combination of carrier need and technological advances are enabling carriers to greatly

consolidate their tower-based network equipment, share network equipment with one another (the next logical step after sharing towers), circumvent towers altogether through the use of small cells and Wi-Fi, and increase data capacity by one thousand times with minimal additional use of cell towers. / /

////////////////////////////////////////////////////////

100 www.senzafiliconsulting.com/Reports/TheeconomicsofsmallcellsandWiFioffload/Aboutthereport.aspx /

/

‘Page 56 of 118 /

If You Can’t Beat ‘em, Join ‘em

*

Carriers have begun to realize that they cannot service the data demands of their customers over cellular networks and have taken a “if you can’t beat ‘em, join ‘em” approach. They have already begun deploying and promoting Wi-Fi networks as an alternative. For example, AT&T and the NYC mayor’s office are collaborating to deliver free AT&T branded Wi-Fi to 26 parks around the five boroughs. Cities now offering free Wi-Fi stretch from Silicon Valley to Ponca City, Oklahoma to Ocean City, Maryland. 101

Carrier Checkbook Fatigue: Data Demand Creates Systemic Problems

Checkbook Fatigue: Data Demand Creates Systemic Problems Source: Unstring Insider: Feb, 2007 The problem facing

Source: Unstring Insider: Feb, 2007

The problem facing carriers is systemic: revenue growth is outpaced by the cost of delivering data and is increasing as smartphones and tablets penetrate the market. Competitive pressure and consumer contracts are revenue constraints that prevent carriers from raising rates to cover the increased cost of delivering data, thereby forcing the carrier to find alternative lower cost data delivery methods.

US Carriers: Declining Wireless Margins

 

2004

2005

2006

2007

2008

2009

2010

2011

2012

Sprint

N/A

29%

32%

25%

16%

18%

16%

14%

13%

AT&T

34%

19%

25%

45%

39%

33%

29%

21%

20%

Verizon

48%

39%

36%

33%

30%

24%

21%

18%

N/A

*Source: Bloomberg

The situation in India provides an example of this phenomenon. As India begins to roll out 3G, the subscriber base is increasing, while ARPU declines.

////////////////////////////////////////////////////////

101 / A/comprehensive/list/can/be/found/at/ http://www.openWi] Fispots.com/guide_free_Wi] Fi_wireless_hotspot ] zones.aspx/

/

‘Page 57 of 118 /

Carriers are responding by deleveraging and raising cash by selling underperforming assets , such as

Carriers are responding by deleveraging and raising cash by selling underperforming assets, such as cell towers at emerging market bubble prices to willing buyers, such as AMT. These deals usually take the form of sale and leaseback transactions wherein AMT purchases the towers from the carrier and the carrier agrees to pay rents to AMT for the length of the contract. The section of this report detailing AMT’s SiteSharing deal supra takes a look under the hood at such transactions.

IP Based Softphones Threaten Carriers and Towers

Softphones are software versions of phones. They have a telephone number, can make and receive calls, they have voicemail, and they can send and receive SMS messages. The main difference between a softphone and a standard phone is that a softphone is designed to run on VOIP (voice over internet protocol), a standard that is supported on any current device that can connect to the internet. Softphone pricing plans are a fraction of traditional mobile phone plans, and can be as low as $99 per year. Consumers living in cities with free Wi-Fi, or who have Wi-Fi at home and work, can skip the traditional cellular network altogether, circumventing the carriers (and AMT) in the process. As discussed infra in section Cable and Softphones Aligning to Offer Consumers a Cheaper Alternative, cable companies are creating alliances to promote Wi-Fi based softphone use.

The softphone industry is large enough to threaten AMT. Microsoft Corp (Nasdaq:

MSFT) spent $US 8.5 billion to acquire Skype in 2011 102 , and Vonage is a popular service. Google Voice is a free app offered by Google. A cursory look at the Google Play store reveals at least 45 other softphone applications.

Spectrum: Why Buy the Cow When You Can Get the Milk For Free

When carriers built the first generation of private wireless communication networks, they relied on governments to provide adequate spectrum to meet their needs. Governments

////////////////////////////////////////////////////////

102 http://news.cnet.com/8301-30685_3-20120355-264/microsoft-closes-$8.5-billion-skype-acquisition/

/

‘Page 58 of 118 /

met the need by conducting auctions. In the US, those auctions lasted as many as 276 rounds and resulted in fees as high as US 18.9 billion for a single band of spectrum. 103

This time around carriers have the luxury of being able to pick and choose from a variety

of technologies that reduce their dependence on paid spectrum, while simultaneously

increasing their ability to deliver increasing amounts of data to consumers.

In some countries, such as Peru carriers are not paying the governments for spectrum

they agreed to buy. 104 In India, the government appears to have difficulties deciding on

what spectrum allocation regime they want to implement after their recent allocation scandals, and in Brazil, as well as other countries including the US, the government mandates carriers provide coverage in unprofitable rural areas. 105

Carriers, sovereigns, and citizen consumers all desire ubiquitous, affordable high-speed mobile communications at the lowest cost possible. Wi-Fi enables this goal and the

proliferation of Wi-Fi will continue to expose towers as the high cost data delivery option

of last resort.

If you Can Beat ‘Em, Beat ‘Em

AMT’s margins are too attractive to go unnoticed. Aside from carriers looking to reduce their dependence on towers, cable companies have now set their sights on the tower industry and are leveraging their ‘Last mile’ advantage.

A recent June 2013 Jeffries report estimates that cable operators actually carry more

consumer wireless data on their networks than mobile carriers, yet cable generates $0 directly attributable revenue from wireless. Jeffries also notes that the CableWiFi confederation has tripled its hotspot count to more than 150,000, from 50,000 in the 12 months since the initiative launched, that technologies are being developed to facilitate

authentication between contiguous hotspots for mobile users, and that all network gear being deployed today is enabled with this technology including devices such as the Samsung Galaxy S4. Jefferies expects all smartphones going forward to have the technology. 106

Cable and Softphones Aligning to Offer Consumers a Cheaper Alternative

Five cable companies including Cablevision, Comcast, Time Warner Cable, Cox Communications and Bright House Network founded the CableWiFi Alliance in May 2012 to promote free outside-the-home Wi-Fi service to coalition customers. This will be accomplished by replacing existing home cable modems with a new free modem, which will broadcast two signals. One of those signals will be private and protected for the

////////////////////////////////////////////////////////

103 http://wireless.fcc.gov/auctions/default.htm?job=auctions_all#completed

104 http://www.congresistalescano.pe/web/prensa/557-fonica23

105 http://www.infodev.org/infodev-files/resource/InfodevDocuments_1128.pdf

106 http://www.marketwatch.com/story/cable-stocks-to-buy-as-they-eye-plans-to-offer-wireless-service-via-

wi-fi-2013-06-27

/

‘Page 59 of 118 /

specific use of the customer while the other signal will broadcast and open, unsecured guest signal to be used by passersby.

One of the primary intended uses of this free neighborhood network is support for a branded softphone application, 107 Voice2Go, that would enable its customers to make free web-based phone calls and send SMS messages without it counting against their wireless account. 108 Since the network is Wi-Fi based, consumers can use other VOIP softphone apps such as Skype (acquired by MSFT in 2011 for $USD 8.5 billion), Vonage, or Google Voice. This development could significantly reduce the need for cellular or broadband data transmitted between macro cells, particularly in densely populated areas.

VZ and SpectrumCo are awaiting FCC approval for Verizon’s purchase of 122 spectrum licenses for US 3.6 billion. Comcast (NASDAQ: CMCSA) CFO Michael Angelakis framed the move during the UBS Global Media and Communications Conference, “We don’t have to invest in building a wireless network. We aren’t going to acquire a wireless network. It’s quite a significant transaction. If you’re a Comcast customer, this means that you will eventually be able to purchase cable television, Internet service, home phone service, and cell phone service, all through your cable company." 109 If this sale is approved, investors may see a global wave of similar deals, all of which are bad for the tower industry and AMT investors.

Wi-Fi is a free public technology that empowers competitors who seek to develop low cost high availability networks. Wi-Fi is a global threat to the private cellular network that AMT provides infrastructure for. AMT’s customers are collaborating with AMT’s competitors and using new technology to develop data delivery networks that reduce their TCO. AMT is myopic if they think that carriers and cable companies are going to allow them to keep generating fat margins without a fight.

IMPEDIMENTS SPECIFIC TO US GROWTH

We do not have a sanguine view of AMT’s US growth prospects, despite acknowledging that Americans are going to be using increasing amounts of data in coming years. Carriers will be focusing their network densification in areas in which AMT has relatively few towers. AMT is also facing increasing cost pressure. The ground rents under many of its sites are likely to double or triple in coming years. AMT has been slow to react, and has instead directed its investments to often poorly-conceived emerging and frontier market investments.

////////////////////////////////////////////////////////

107 http://news.cnet.com/8301-1035_3-57439667-94/new-comcast-services-threaten-wireless-text-and-data-

revenue/

108 http://news.cnet.com/8301-1035_3-57588325-94/comcast-expands-wi-fi-network-with-new-

neighborhood-initiative/

109 http://www.digitaltrends.com/mobile/comcast-execs-verizon-deal-to-bring-the-quadruple-play

/

‘Page 60 of 118 /

Increasing Ground Rents and the “TriStar Effect”

AMT has not been forthcoming about its exposure to ground rent increases, which we expect to materially impact AMT’s margins. Most alarmingly, but in an opacity typical of AMT, it has not disclosed in its filings the litigation between it and TriStar Investors, Inc. (“TriStar”); nor, has AMT disclosed that because TriStar has gained control of the ground beneath over 650 of AMT’s higher yielding towers, AMT is likely faced with a choice to buy TriStar (likely for at least US$400 million to US$500 million) or lose annual revenue purported to be approximately US$77 million. 110 Even though AMT expects to spend approximately $100 million per year going forward, it still leaves itself exposed to the problem.

Many of the cellular towers have ground leases that are coming up for renewal in the coming years. Typically ground leases gave tower owners (or their assignees) rights of up to 20, 25, or 30 years.

When ground leases come up for renewal, landowners typically have significant amounts of leverage over AMT. AMT does not actually own the tower steel for a significant portion of its portfolio – namely the towers it leases from Verizon (NYSE: VZ) and AT&T (NYSE: T). The lease agreements for these portfolios, which cover at least 4,250 of its towers (approximately 20% of the US portfolio), 111 provide that if AMT is unable to renew a ground lease, then it loses it lease on the tower.

Even in situations in which AMT owns the tower, the typical landowner has a substantial amount of leverage over the Company. The cost to decommission and re-assemble a tower elsewhere generally runs 50% to 100% of the cost of a new tower, and must be incurred upfront. Zoning restrictions are significant barriers to moving towers in many locations. The zoning environment is far less permissive than it was when the towers were originally constructed – particularly due to concerns about landscape clutter and health concerns. (As we discuss infra, even developing countries such as Chile, Peru, and Brazil are imposing tight zoning restrictions and even forcing AMT to move towers for health protection purposes.) The other issue for AMT in trying to move towers is that it plays havoc with the carriers’ network coverage. Carriers create coverage maps that attempt to minimize dead zones in between towers, but without causing them to spend inefficiently.

AMT’s landlords are much more sophisticated than they used to be, and ever increasing numbers of them are seeking a greater share of the economic pie. Over the past 15 years, the total cash flow from a typical tower has grown by more than 400%, 112 while the typical landlord has seen its share of the tower cash flow fall from approximately 40% to less than 15%. 113 The market for ground leases is becoming far more efficient than it used to be, caused in part by the ubiquity of cellular communications and the ability of

////////////////////////////////////////////////////////

110 Investors familiar with TriStar.

111 AMT 2012 10-K, p. F-63.

112 TriStar Second Amended Complaint, para. 34.

113 TriStar Second Amended Complaint, para 2.

/

‘Page 61 of 118 /

the internet to deliver market information. There are numerous ground lease consulting firms and attorneys helping landlords renegotiate their leases on more favorable terms. Often these consultants receive a portion of the new economics, incentivizing them to be solid advocates for their clients when dealing with tower operators.

We therefore conservatively expect AMT’s US ground rents to increase 4.5% annually for the next ten years, despite AMT’s expected spending of US$100 million per year buying easements or long-term lease renewals.

THE REIT THAT CANT DO IT RIGHT – INTERNATIONAL AFFO IS INACCESSIBLE TO AMT SHAREHOLDERS

We do not believe AMT’s international business is compatible with its REIT structure. There are numerous tax barriers to AMT paying dividends to shareholders from its international cash flows. The myriad risks to these cash flows are incompatible with REIT investors’ desires for steady, predictable dividends. Our impression is that REIT investors generally expect the underlying assets to hold their value, if not appreciate; however, towers have limited useable lives. We view AMT’s conversion to a REIT as an attempt to ride a REIT bubble and produce a stock pop that will prove to be a short-term gain. We also wonder to what extend AMT could be out of compliance due to the site sharing discrepancy.

Because AMT is doing business in various overseas jurisdictions, its international cash flows will effectively be inaccessible to AMT investors. The main issue is that many of these jurisdictions impose substantial taxes on transfers of profits out of the country. There are of course ways to minimize the taxes, but taxes can still be substantial. We researched statutory taxes that could be issues for AMT in attempting to repatriate profits to the US. AMT’s structuring is extremely complex, and it is impossible for us to know the actual rates that will apply to it. As a case in point, AMT has a SOFOM in Mexico – SOFOMs are non-bank lending institutions. In any event, our research gives investors an idea of the issues associated with repatriating profits.

In general, AMT has three ways of expatriating profits from its overseas operations. AMT can lend money from its offshore parent to its onshore operation, and make principal and interest payments to the parent. AMT appears to have structured a number of parent loans between its Netherlands entities and its onshore operating companies. AMT can also charge its onshore companies for management services and intellectual property (i.e. trademarks) usage. Finally, AMT’s entities can pay dividends to their parents.

Interest payments on loans from parents can be deductible to the onshore subsidiary, but the interest charges are often subject to “transfer pricing” limitations, i.e., AMT cannot charge itself usurious interest rates. There is also usually a country-specified limitation on how much debt its operating entities may incur. Regardless, interest payments from

/

‘Page 62 of 118 /

the onshore subsidiary to the parent are often subject to “withholding” tax. 114 The following highlights some of the withholding taxes that could be applicable to interest and dividend payments to AMT’s parent companies from the operating entities.

In Mexico, withholding taxes on interest payments range from 4.9% to 40%. There are no withholding taxes on dividends.

In Colombia, withholding taxes on interest paid to foreign entities can range from 14% to 33%. Dividends to offshore shareholders are generally taxed at 25% to

33%.

In Ghana, withholding taxes on interest paid is 10%. The withholding tax on dividends to offshore shareholders is 15%.

In Germany, there is no withholding tax on interest. Dividends to US shareholders are generally taxed at a total rate of 26.375%. (We do not have the rate applicable to Dutch shareholders.)

In South Africa, there are no withholding taxes on interest. The withholding tax on dividends is 15%, but can be recoverable under certain circumstances.

In India, withholding taxes on interest payments are generally 20%. 115 There is no dividend withholding tax.

In Uganda, interest paid to foreign lenders is taxed at 15%. Dividends are also taxed at 15%.

At the risk of this report seeming to be a love letter to Jay Brown, we quote him regarding the incompatibility of the REIT structure with the risks of emerging and frontier market cash flows: 116