Description

We recently wrote up GEO, the second largest private operator of prisons in the US.  Today we are proposing the idea of going long CXW, the market leader in this duopoly.  There are several characteristics we like about this business including high barriers to entry, recession resistance and a growing industry augmented by a penetration story.  The government can’t afford to build new prisons and for the first time they are selling prisons to the private sector.  There are essentially 2 buyers of these assets (CXW and GEO).  The value proposition is a better service at a lower cost.  Customer credit quality is good: they have AAA tenants that pay in a currency they print.  Their tenants don’t default. 

CXW is a more focused version of GEO i.e. they are a pure play on the adult corrections business, which is the business we like.  We like CXW’s more focused business strategy better but wrote up GEO previously because we felt its higher equity FCF yield (18% at $18) more than justified its flaws.  While we continue to like GEO, and it remains cheaper, we think CXW represents a very good risk reward, which has recently been skewed favorably as a result of a 13D filing by Corvex and Marcato.  We think the 13D filing, which highlights CXW’s undervaluation and prospects for a REIT conversion, represents an undervalued call option.  We feel stronger about our investments in both CXW and GEO and have added to our positions in both at these higher levels.

Our thesis is simple:

1. The REIT call option is potentially very valuable and costs very little.
2. Management is shareholder oriented and will do what is in the best interests of shareholders.
3. The REIT call option will increase in value as management responds to shareholder inquiries.

We are not real estate or REIT experts.  Our thesis is not that CXW will convert to a REIT per se.  We don’t know if this will occur but are working with lawyers researching its prospects and believe there is merit to the activists’ claim that a REIT conversion is possible.  What we do know is management is interested in exploring all suggestions for increasing shareholder value and this includes the possibility that a REIT structure today makes sense.  We believe management’s public acknowledgement of this possibility will increase the value of the REIT call option, which is currently undervalued in CXW’s share price.  We think this revaluation should happen regardless of the ultimate outcome, which won’t be known for at least 6 months in our view.  The first indication may occur as soon as their May 3 earnings call.  As shown later in this report, we think CXW’s expected value today is at least $40, nearly 40% higher than where the stock is trading. 

The 13D filed by Corvex and Marcato on April 5 notes as follows:

Item 4. Purpose of Transaction

The Reporting Persons believe that the Shares are undervalued and are an attractive investment. The Reporting Persons have had conversations and meetings with each other and the management and members of the board of the Issuer to discuss the Issuer’s business, assets, capitalization, financial condition, operations, governance, management, strategy and future plans and will seek to have additional conversations with one or more of management, the board, other stockholders of the Issuer and other persons to discuss such matters. These discussions have reviewed, and may continue to review, options for enhancing shareholder value through various strategic alternatives, including, but not limited to, proposals for one or more of the actions described in subsections (a) through (j) of Item 4 of Schedule 13D. In particular, and without limiting the generality of the foregoing, the Reporting Persons have discussed, and will continue to discuss, among themselves and with one or more of management, the board, other stockholders of the Issuer and other persons proposals to convert the Issuer to a real estate investment trust (“REIT”) for U.S. federal income tax purposes and other strategic alternatives. The Reporting Persons have engaged experts in the legal, accounting and tax professions who believe that the Issuer could convert to a REIT without material disruption or changes to the Issuer’s current operations. The Reporting Persons believe that a REIT conversion would result in a significantly lower cost of equity capital, increased growth prospects and a material increase in value for all the Issuer’s shareholders based on current trading multiples of comparable publicly traded REITs. The Reporting Persons look forward to working expeditiously with the Issuer’s management and the board in order to effectuate this outcome.

We have spoken with management about these proposals.  During our call, management was very receptive to discussing the idea and clear that they are interested in all shareholder suggestions for value creation including the possibility that a REIT structure today makes sense.  They were unable to give answers to specific questions relating to the feasibility of an actual conversion, however, noting that an 8k disclosure would be required.  Our impression is management is interested in all shareholder ideas and is not on the defensive.  We think management is shareholder-oriented as demonstrated by:

1. Aggregate management and BOD stock ownership of 4.09% ($99m) excluding options with $30m owned by the chairman.
2. LTIP comp triple base salaries.
3. A history of savvy stock repurchases (repurchased 8%+ between Q4’08 and Q2’09 at an average price of $11.68; repurchased 10%+ in 2011 at an average price of $22.40).
4. The recent dividend initiation (3.3% yield at the outset).

Management addressed the question of a REIT in 2010 and has not updated their presentation since then yet circumstances have evolved in the past three years.  The hurdles to a REIT conversion provided by management in 2010 included control over growth and financing and a requirement to renegotiate contracts.  In our discussion with management, they highlighted that these hurdles existed as a result of one assumed precondition: the need to separate the REIT company from the operating company.  Management has not publicly addressed the possibility of converting to a REIT under the assumption that this is unnecessary—likely because they have not fully explored the idea.  Our research indicates that a Taxable REIT Subsidiary (TRS) may alleviate most of these hurdles.  The TRS rule was created as a result of the REIT Modernization Act in the late 90s.  Its application has increased in the past 2-3 years as the IRS has allowed the use of broad definitions in approving the way these and other structures pass the required tests.  Recent examples include Ventas, a healthcare REIT, which no longer needs to find tenants for their properties because they can own them through a TRS.  Another example is Coresite Realty Corp., a data center REIT, where the IRS ruled that a wholly owned TRS could provide the management services to the tenants.  Other examples of liberal application of the REIT tax structure in recent years can be found in infrastructure, gas treatment plants, power transmission structures and even investment companies.  Upon information and belief, we understand Corvex and Marcato have hired lawyers, accountants and tax experts to aid in their effort to educate and cooperate with CXW with an end goal of fully exploring this outcome via a 482 study, an IRS ruling and ultimately a REIT conversion upon passing all of the requisite tests. 

The attractions of a REIT include the following:

1)      Premium valuation is the elephant in the room.

2)      Value of the tax shield.

3)      Investment grade rating from the agencies.

An IG rating would give CXW a competitive advantage through a lower WACC as well as flexibility in relaxing restrictions on increasing the dividend payout ratio.  It would also make deals like the recent Lake Erie acquisition where they purchased a prison from the state even more valuable.  The company recently wrote a letter to 48 states proposing similar transactions as a way to relieve state budget shortfalls, representing a potential new era of growth.  These deals are unique because there are really only 2 buyers (CXW and GEO).  Each deal would be accretive similar to deals the towers and others do.  Their stocks go up when they come to market. 

A couple examples illustrate what we believe may occur here:

• IRM in 2010.  The stock went up 35% - 50% just by exploring the possibility of converting to a REIT—after Elliott Management Corp filed a 13D proposing the idea and Davis Selected Advisors among other large (passive) shareholders told management they liked the idea.  A final decision is expected from IRM by June.  Should they go through with a conversion, their stock may have another 30-60% upside in our view.
• A more recent example is EQIX.  The stock has grinded 50% higher YTD almost in a straight line as management has solicited shareholder and analyst input regarding capital allocation, tax management initiatives as well as the possibility of a REIT conversion.  Naturally, this has attracted more shareholders in favor of this outcome.  Should they follow through with a REIT conversion, there could be another 30-60% upside here as well.  From their Q1’12 conference call:

“As discussed on prior earnings calls, we're taking a serious look at our global tax strategies, which includes assessing the feasibility and the applicability of Equinix converting into a REIT structure. We have engaged advisors and commenced diligence to assess this tax structure.  While we consider this ongoing assessment to be very important to our global tax strategy, we want you to know that we'll continue to focus on profitably growing our business in both our core and non-core markets.  Now given the status of our assessment, at this time, we're not able to provide you with any additional update on certainty or timing. And on a related note, included in our SG&A guidance is an incremental $2 million of anticipated spend for the initial cost of the REIT analysis for 2012”

What’s interesting is that in both of these examples, the stocks increased materially just by management publicly considering a REIT conversion.  We think the likelihood is high CXW follows this course.

Background on a recent negative event responsible for the stock’s ~9% decline on April 24, 2012:

The state of CA’s Department of Corrections and Rehabilitation (CDCR) issued a 227 page report outlining their 5-year “realignment” plan, which has been ongoing for the past 12-18 months.  The full report can be found here: http://cdcrtoday.blogspot.com/2012/04/cdcr-releases-plan-to-cut-billions-in.html.  This report proposes many things—of which one is potentially very negative for CXW: the insourcing of CA’s 9,500 out of state beds, over 5 years, which CXW operates.  This represents approximately 20% of CXW’s current AFFO and 15% of its normal AFFO. 

We think CA is a unique situation because the state corrections system has been in receivership as a result of a prison population that had grown to 200% of design capacity in 2006.  Through “realignment,” Governor Brown is trying to shift the burden of a massively overcrowded state prison system onto local county prisons to comply with a Supreme Court mandate to reduce state system utilization to 137% by 2013.  Realignment is intended to avoid the need to build nine new state prisons at an incremental annual debt service and operating cost of $2.2B.  The reason moving inmates from state prisons to local jails reduces capacity utilization is because the 137% mandate only applies to the 33 state prisons, not the lower level county facilities.  The report shows impressive progress towards reducing the state prison population but what it does not discuss is the current capacity of the local prison system.  The problem is these local facilities are also overcrowded and are not built for long term inmates; they are short term facilities. 

The state’s proposal to eliminate their 9,500 out of state beds is contingent on their ability to successfully redeploy ~44k state inmates to a local system that only has ~6k available beds and cannot sustain this population.  There are other challenges as well including convincing the Supreme Court to relax the mandate that utilization be reduced to 137% and instead allow it to be 144% as well as raising taxes to continue funding realignment.  So what gives?  A combination of relief valves have been proposed including letting lower level offenders off the hook.  The problem with this is that local police and sheriffs don’t have the resources to deal with this.  The state is simply kicking that can down the road.  Sound familiar?  Ultimately, 2/3 of these offenders may end up back in jail within three years and realignment may be a disaster.  While the market has now assumed a high likelihood CXW will lose 20% of their earnings over the next 5 years—the hurdles for the state to achieve its goal make this far from certain if not unrealistic.  Ultimately, we believe the state will be unsuccessful with their plan because they are simply reshuffling inmates from state facilities to local facilities to sidestep the Supreme Court’s mandate that they reduce system capacity to 137%.  To be conservative, however, we assume the state is successful and we have haircut our assumptions accordingly.

Before discussing the REIT math, let’s review why we think there is a margin of safety here even if no REIT conversion occurs.  CXW trades at 11x run-rate AFFO of $2.65 (9% after tax FCF yield) after they complete their capx program this year.  This excludes the latent earnings power in their idle beds, which represents capx-less growth of 70c (+25%) over the next couple of years.  We view normal AFFO, therefore, as $3.35.  The stock trades at 8.5x normal AFFO, or a 12% FCF yield.  The market yields 6.5% on current earnings and 5% on normal earnings.  The 10-year bond yields 2%.  The average REIT yields 5%.  We think these gaps are too wide.  We view CXW as a mis-priced bond yielding 9% with an ability to grow 25% over time with no additional capital investment.  We think downside is low due to the magnitude of the earnings yield differentials noted above.  We also believe there is a floor on the stock given the chain of events likely to occur going forward whereby in conjunction with shareholders, the company will explore methods for enhancing the value of their stock.

Framing a range of earnings scenarios and valuations assuming no REIT conversion:

1. If no idle beds ever get filled and the worst case plays out in CA (~50c hit), AFFO becomes $2.15 vs. the current run-rate of $2.65.  15x this highly depressed number is reasonable to us.  15x $2.15 = $32.  This is what may be fair, not where the market may trade the stock (price and value diverge).
2. What is more likely, even if the worst case plays out in CA, is an offsetting effect through filling up their other idle beds over the next couple of years…holding AFFO flattish at $2.65.  A few points regarding growth potential: A) The private prison population is <10% of the total prison population and has increased penetration annually over the past decade.  B) Federal system-wide corrections capacity utilization is 139%; the BOP alone projects a need for 4k-6k beds/year for the next 3-5 years.  28 states are above 100% of capacity.  C) There are more than 200k beds in public facilities that are older than 75 years.  95k of these are over 100 years old.  Older beds are inefficient and more expensive to operate.  This compares to a private prison population of ~160k.  An example of growth is the state of GA, which recently closed an old facility and contracted with private operators.
3. There are many outcomes in the middle.  Half of CXW’s OOS beds could get cut (~25c hit) and idle beds fill up elsewhere, share buybacks, etc….AFFO grows slowly from $2.65 in 2013 onward—as it has in each of the past three recessionary years (9% EPS CAGR from ’07-’11).  We think this scenario is the most likely unless the State is able to convince the Supreme Court to relax the requirement that utilization must decline to 137% by 2013—which by the state’s own admission in its blueprint, will not occur.  15x this still depressed number may be fair and realized once it becomes evident demand for idle beds exists.  15x $2.65 = $40.
4. Normal AFFO remains $3.35 but it takes longer to achieve if OOS beds get cut.  We think 12.5x normal AFFO may be fair or conservative.  12.5x $3.35 = $42. 

We view CXW’s fair value without a REIT conversion as a range of $32 - $42. 

A REIT conversion represents a call option that costs little (free?) yet has a pay off of 40%-85%.

Here are relevant comps:

• The average REIT trades at 20x AFFO today with a range of 14x-32x.
• Healthcare REITs may be most similar although with increasing regulatory risk, they may now be inferior; they trade at 17x.
• We assume a reasonable range for CXW may be 15x-17x.  As an aside, CXW used to trade at big AFFO multiples (as high as 20x fwd AFFO) without being a REIT.  Why can’t it revert to historical multiples when growth resumes?  Or maybe it happens sooner with a new discussion of a REIT conversion?
• Regardless, 15x AFFO seems like a reasonable benchmark and 17x may be fair in a REIT scenario. 

Framing a range of values if CXW converts to a REIT:

1. 1.       Assuming worst case scenario in CA and no idle beds fill up ever: $40-$44 (15x-17x $2.15 + $7.50 tax shield*).
2. 2.       Assuming base case scenario in CA (half of OOS beds get cut; ~25c in AFFO) and some offsetting growth elsewhere: $49-$54 (15x-17x $2.65 + $9 tax shield*)
3. 3.       Value on normal earnings if converted to a REIT: $62-$69 (15x-17x $3.35 + $12 tax shield*)

*We make simplified assumptions for the value of the tax shield.  They pay about $100m in taxes.  We assume 75% of this may be eliminated.  100m S/O.  Roughly 75c of tax shield.  We capitalize this at 12.5x rather than 15x-17x to account for less certainty in our estimates and potential tax leakage.  12.5x 75c = $9.40 of tax shield value in base case with slightly more and less in bear and bull case respectively.

We view CXW’s fair value, if converted to a REIT, as a range of $40 - $54 on current earnings and $62-$69 on normal earnings. 

Putting probabilities on these scenarios is inherently uncertain but should become more clear as management responds to shareholder inquiries.  Management has insinuated that if it is possible to convert to a REIT, they will likely do so.  We infer from this that estimated probabilities of conversion will increase over time.  One can adjust assumptions accordingly—fair value with no conversion, fair value with a conversion and the probability of conversion.  If we assume a coin flip and pick the middle of the range of our estimate of fair value with a conversion ($47) and the low end of our fair value estimate ($32), the expected value of the stock today is $40, nearly 40% higher than where it’s currently trading.  This gives no credit for their latent earnings power.

What’s the variant view of the stock as it pertains to a REIT? 

1. Bill Ackman failed to convince management to convert in 2009/2010.  As an aside, CXW traded at 13x AFFO then vs. its current ~11x AFFO multiple.  We believe there’s a misperception here.  Ackman never filed a 13D—this was a passive investment.  Investors wrongly assume Bill was an activist here because of his activism elsewhere.  Ackman is one of the largest funders of the Innocence Project.  Headline risk mattered to him for personal reasons.  Pershing runs a concentrated $10B fund so liquidity was also prohibitive.  Finally, he was busy making a lot more money on General Growth Properties, which was a better use of his time. 

A)     The company itself failed as a REIT in the late 90’s and nothing has changed.  The company failed 15 years ago because 1) they split into an OpCo and a REITCo and because 2) they were overlevered at 6x net debt/ebitda.  Externally, at least two things have changed: 1) The TRS rule didn’t exist 15 years ago and its application has become more common in the past 2-3 years; A TRS should eliminate many of the previous hurdles and 2) The premium placed on yield is much higher today versus history as a result of ZIRP + a government bond bubble.

The downside is none of this could happen and we are left with an inexpensive stock with a current earnings yield of 9%, a dividend yield of 3% and prospects to grow earnings by 25% with no requirement for additional capital.  The REIT call option is hardly receiving much value in our view.

Fundamental risks:

A)     Contract delays and cancellations are the most common.  This industry has experienced a series of contract delays and a few rare cancellations over the past 18 months.  What the CDCR just announced can be interpreted as close to a worst case development for CXW.  Their largest state customer has proposed to eliminate ~10% of their revenue and ~20% of their earnings.  We take comfort that this known risk is viewed as highly probable today.  In our view, one of the largest shoes has effectively dropped and the stock proceeded to rise on the second and third day of trading.  Moreover 7% of the shares changed hand on that day and the incremental buyer is likely interested in a REIT. 

B)      Less than optimal capital allocation e.g. big capx.  This is more of a concern for GEO than CXW as in our opinion, CXW has a more focused business model and their compensation is more aligned with shareholder value.  Given the transition of the shareholder base, poor capital allocation is unlikely to be tolerated.

C)      Fundamental change to sentencing and law enforcement.  While support for reform has become more vocal, political will to do it is another matter.  Change here tends to be glacial.  Prisoners aren’t popular in backyards. 

More generic pushback: CXW has a high P/E (16x), a mediocre ROE (11.5%) and a high debt load (47% D/C; 2.6x net debt/ebitda).  Our response:

1. AFFO is how to interpret this company’s earnings; it is similar to EPS for most companies.  GAAP EPS is not that meaningful because maint capx is 45% of D&A (prisons last 50+ years and their real estate is stated at the lower of cost or FMV quarterly).  By definition, their real estate is worth more than stated on their books.
2. Their low double digit ROE is depressed because 28% of their equity is earning nothing today.
3. Their debt level (2.6x net debt/ebitda) may look high versus an average S&P company but it is not relative to competitor GEO, which is levered at 4.1x.  It also does not look high relative to REITs, which are levered 6x on average.  CXW’s leverage ratios are also inflated since they are under-earning what their assets should produce.  Lastly, their debt is mostly secured by their real estate.

Catalyst

A)     Management engagement with shareholder activists and their forthcoming public response.

B)     Transition of the shareholder base in favor of a REIT conversion or other value enhancing alternatives e.g. increased dividend payout, private equity buy out et al.