This is a recommendation to short SBA Communications (SBAC- $37), the third largest wireless tower company in the U.S. This is a longer term short and it could take a while to play out. One less aggressive way to implement the short is to short calls, though personally I don't think the implied volatility is high enough to make this implementation particularly attractive.

The company earns most of its money leasing space on its towers to wireless service providers like Verizon. SBA drove its growth in the past 5 years by acquiring or building towers. Its revenues went up about 150% from $260m in 2005 to probably $620m this year while its towers about tripled from around 3,066 at the end of 2004 to about 9,000 at the end of this year.

I once read this on VIC (I wish I had written down who wrote it, so I could give credit) and it stood out to me as good advice:  "Moreover, when the market is giving a high multiple to a rollup that is facing rising industry-acquisition multiples with steep competition for assets at top-tick margins, it's time to short." I've seen this many times as well. I think this is the case with SBAC.

The three largest public tower companies have rolled up a big chunk of the industry, going from 29k towers in their combined U.S. portfolio at the end of 2004 to 52k towers today (see table at end of report), driving up tower price multiples as a result. SBA in 2005 paid 10x forward tower cash flow, now they pay 17x. I estimate SBA's IRR was 16% on tower acquisitions in 2005, now I think it is below 10%. I don't see how SBAC's incremental EBITDA margins get any better from their 85% today.

To sum it up, I think SBA's future incremental return on capital has shrunk to mediocre levels. Yet the market has priced SBAC as if the company still has high incremental returns and its growth story is intact. As SBA's returns and growth decline due to the much higher price they are paying for towers today versus five years ago, I believe investors will re-rate the stock to a much lower valuation. I think $19 for the stock is more appropriate. If I am wrong, there does not appear to be a lot of risk to the short given the stock's already very high valuation. Key points:

• High valuation
• Declining and mediocre returns
• Heavy insider selling
• Declining quality of earnings/reported EBITDA
• High customer concentration is risky
• 2011 estimates likely not conservative like the Street expects
• High leverage and tight covenants
• No year in the last five with free cash flow (including acquisitions)

High valuation. The stock is very expensive trading at 29x free cash flow, 22x operating cash flow, 18x EBITDA (using 3Q10 annualized EBITDA), 11x EV/sales and 12x book value. Investors clearly expect strong growth in future cash flow. As you can read in the next section, I think that SBA's returns have declined to their cost of capital. Given that, I think capitalizing current run rate free cash flow at 15x, or a 7% free cash flow yield, is more than fair. If I am correct that the company no longer has investment opportunities above its cost of capital, then the company's valuation should get no credit for growth.

The company is currently generating about $50m per quarter in operating cash flow, or $200m per year*. Assuming a 40 year economic life on their towers, 8700 towers and a cost to build the towers of $250k (SBAC has recently been spending $400k), we get maintenance capex of $54m from this calculation: 8700 towers x ($250k per tower / 40 years). That results in $146m in free cash flow. Capitalizing that at 15x results in a per share value of $19. That assumes that the company never pays any material taxes, a conservative assumption from the perspective of the short. The company spends about $150m per year on interest.

Capital structure

Stock price $37

Shares out 115m

Market cap: $4.2B

Net Debt: $2.8B

EV: $7.0B

Target price

CFO: $200m (company spends about $150m per year on interest and pays an immaterial amount of tax if you want to start with EBITDA)

Maintenance capex: $54m

FCF: $146m

Multiple on FCF: 15x

Target price: $19

*I would use management's adjusted EBITDA, but because it appears inflated (see below) I use a rough average of the last two quarter's cash flow from operations.

Declining and mediocre returns. The price the company has paid to acquire towers or build new ones has increased over 70% since 2005, the year that SBA started its aggressive expansion strategy. This has cut the company's pre-tax IRR on these investments from 16% to under 10%. The company recently paid about 17x forward one-year tower cash flow versus about 10x in 2005 for tower acquisitions (see 8-K from 11/13/2006 for more details).

Why have multiples increased and returns delinked? The industry has been consolidating for some time and competition for towers has been increasing through the years. Purchase price multiples have therefore increased, and returns have commensurately declined toward the cost of capital. Another reason that might explain why prices went up is that landowners became increasingly more sophisticated and informed. Today there are attorneys and consultants that specialize in helping landowners get the best price from tower companies. Here's one example: http://www.celltowerattorney.com/

A future source of declining returns could be new technology. For example, Sprint is attempting to cut its number of towers by 30% from 65k to 45k. Newer technologies like DAS are helping to provide new capacity, reducing the need for new tower growth.

Tower companies and analysts have talked up potential growth from 4G. However, I don't think it is that simple. First, 3G was what was going to drive tower industry growth starting a couple years ago, but it did not lead to accelerating organic leasing growth for SBA, so it's not clear to me that 4G will either. Second, one source of growth that SBA touts is Clearwire, but not it is clear if they Clearwire is funded. Third, and related to the first point, 4G installations won't generate as much leasing revenue per Drew Caplan, the Chief Network Officer of LightSquared. According to a recent article, he "cautioned the industry to curb its enthusiasm regarding lease rates, stating that the equipment on their structures will be 'at a size and weight that will make my friends in the tower business awfully unhappy.'"

International could be a source of economic returns, but it is early and it is unclear if SBA will be successful in countries like Panama. I will monitor this risk, though international is a tiny part of SBA's business now.

New tower model

• IRR from above: 10% pre-tax
• Assumptions (I think conservative for the short, i.e. if anything the actual IRRs will be lower):
  • 3.5% price increases per year, every year, forever. My model indicates SBA hasn't achieved this level historically, though they say in their SEC filings this is what they expect.
  • 85% incremental EBITDA margin forever
  • no taxes
  • $1,000 per year of maintenance capex
• SBA's recent tower purchases have been in the $625k range excluding 100 towers it bought in Panama. This is up from the average of $355k in 2005. Plugging $355k into the model above results in an IRR of 16%.

Not included in SBA's non-discretionary capital spending of $1,000 per tower figure is the amount that the company is spending to extend ground leases, buy land and buy perpetual easement purchases. These amounts totaled $12.7m in the 9 months ended September 2010. Including these amounts in the company's "equity free cash flow" for the YTD period would reduce this figure by 8%. Of the 8,705 tower sites SBA owned as of September 30, 2010, approximately 30.3% were located on parcels of land that SBA owns, land subject to perpetual easements, or parcels of land in which SBA has a leasehold interest that extends beyond 50 years. So future payments to extend ground leases or buy the land are likely to be material.

SBA is very aggressive in its assumption of $1,000 of maintenance capex, which the company uses to calculate equity free cash flow and in turn, what analysts use to value the company. AMT uses a similar figure, so it isn't just SBA, but I still think it is not correct.

SBA claimed $2.2m in 3Q10 for maintenance capex for its 8,700 towers. It may be true that is what SBAC spent that quarter on maintenance capex. My gut says it is much higher. But putting that aside, from an economic point of view, economic depreciation must be far higher. That $2.2m in 3Q is about $1,000 per tower per year. Today SBAC is spending $400k to build a new tower. Dividing $1,000 into $400k implies a useful life of 400 years for a tower. Let's say it's $250k to build a tower. 250 years? Assuming a useful life of 250-400 years for a structure, even a concrete foundation, brick building and steel, just makes no sense.

Here's a little color on how SBA thinks about the maintenance capex figures it provides the Street. You can sum it up as: we won't be around when it matters. Three years ago I spoke with IR. I asked how much it cost to build a tower. IR said $225k. I asked how long a tower lasts. IR said towers will last 40 years, but that SBA depreciates towers over 10-15 years for tax purposes. I asked how much per year per tower that SBA spends on maintenance capex. IR said $1,000. I then asked that if a tower lasts 40 years, doesn't SBA have to replace a tower every 40 years and spend $250k, meaning $40k in lifetime maintenance capex ($1,000 per year x 40 years) is nowhere near enough? IR's reply, and I quote verbatim: "I won't be around in 40 years."

Heavy insider selling, including by one insider that appears to have excellent timing. Brian Carr sold all of his shares in March at $36; he had previously bought 20k shares at $12, near the stock's low in November 2008. He recently exercised almost all of his vested options and sold all the underlying shares from the exercises. The VP & Chief Accounting Officer has been heavily exercising options and selling as has the SVP - Property Management (the head of sales). Neither of these two executives own any actual shares (just options). The CEO sold substantial amount of shares recently.

Declining quality of earnings/reported EBITDA. Management's adjusted EBITDA historically has tracked EBITDA derived from cash from operations. But recently the two have diverged, with management's adjusted EBITDA significantly higher than EBITDA from CFO. This suggests a more aggressive / lower quality of management's adjusted EBITDA. Part of the reason for the divergence is that management is excluding higher acquisition expenses from adjusted EBITDA than historically, even though acquisition activity is not higher ($6m more in TTM versus same timeframe a year prior).

High customer concentration is risky. AT&T accounts for 29% of site leasing revenue, Sprint 24%, Verizon 15%, and T-Mobile 12%. Sprint is planning to reduce its number of cell sites by 20k or about 30% due to the use of new technology and consolidation of its networks. I think many are skeptical they can do it, but if they did, that sure doesn't seem priced into the price of SBAC. Wireless tower companies have touted Clearwire's rollout as an important growth driver, but Clearwire's funding is in doubt and Sprint could be on the hook for Clearwire's debt (see various Street research).

2011 estimates likely not conservative like the Street expects.

 "We expect numbers to steadily rise over the course of 2011, consistent with a strong leasing environment and SBA's penchant for historically conservative initial guidance setting." -- October 2010 report from BofA Merril:

Obviously the BofA has become accustomed to SBA beating and raising guidance. My reading of other Street research suggests a similar sentiment.

But I think 2011 estimates are high enough that the company is unlikely to beat them; beating 20011 estimates is probably required to keep the stock at current levels. SBA is calling for 9% organic site leasing growth in 2011, same as they called for in 2010, but I think organic growth in 2010 will wind up being about half, or 4-5%.

SBA doesn't provide organic / non-acquisition revenue that we can use to easily calculate the growth. But we can back into it. Site lease revenue growth is a function of 1) price changes per tenant lease, 2) tenants per tower and 3) number of towers.

For 2010 we can get a rough estimate of the combined effect of price changes and tenants per tower, or what I think SBA means by organic revenue growth:

Organic revenue growth = (1 + total site leasing revenue growth) / (1 + growth in number of towers) -1

Where total site leasing revenue growth = (1+tenant per tower growth) *(1 + price change per tenant)-1

For 2010: (1+.12) / (1 +.09) - 1 = 3%, a lot lower than the company's 9%.

Note, however, this is a rough estimate year to year (should smooth out over time) because it assumes the company adds towers evenly through the year. My quarterly model shows organic site leasing growth of 4-5% this year.

SBA will still beat their overall site leasing growth guidance for 2010 even though I think organic growth will come in below expectations. This is because SBA will buy a lot more towers than their original guidance, making up for the slower organic growth.

This could happen again in 2011, but it doesn't seem likely. The company is already highly levered and nearing its debt covenants, which means the company may not have enough capital to by buy/build towers in excess of guidance to make up for slower than expected organic growth. See the next section. To sum it up, accelerating organic leasing revenue growth from 4-5% in 2010 to 9% in 2011 and acquisitions above SBA's plan mean that guidance is probably not conservative like the Street expects.

High leverage and tight covenants limit SBA's growth and elevate risk. This report is getting long so I am going to refer you to this article: http://seekingalpha.com/article/228226-sba-communications-too-much-debt-not-enough-growth?source=qp_investment_views.

Management incentives. Management's cash bonuses are based on meeting EBITDA targets, except for Mr. Silberstein, as Senior Vice President-Property Management, whose cash bonus is based on achieving revenue levels. Since there is no capital charge, management can earn its maximum bonus even if it destroys shareholder capital by overpaying for tower acquisitions or pursuing uneconomic new tower builds. Equity comp is based on peer group analysis and an evaluation of the individual executive's responsibilities, contributions and performance in the prior year.

Risks:

Lease pricing goes up more than investors expect because of 4G.

Abililty to increase tenants per tower. SBAC's average tenant per tower has been dead flat at about 2.5 since 2006. Part of the reason SBA's ratio has been flat could be that their addition of new towers to their portfolio has masked an increase in the tenants per tower for seasoned towers. This assumes that new towers have fewer tenants than seasoned towers. In any case, I think a future increase in tenants per tower is built into the stock price.

International opportunity.