Extendicare represents one of the best catalyst oriented value investments I have come across in quite a while.  The catalyst for value to be realized at Extendicare is a result of yesterday’s surprise announcement that the board has hired Citigroup to explore strategic alternatives which may include a spin-off or a sale of the Company’s US business by the end of the year.  Bizarrely, this announcement was made during market hours and the existing shareholder base and sell side analysts are missing the substantial value that will be created through this separation.  The Company operates in the skilled nursing industry in Canada and the US, and owns the vast majority of its underlying real estate.  From a valuation perspective, there is a large divergence between valuation of Canadian skilled nursing facilities and US skilled nursing facilities.  In addition there is a wide spread between the value of skilled nursing operating companies vs. skilled nursing real estate assets.   Extendicare happens to be a combination of all three of these assets; they are (1) a Canadian skilled nursing company, (2) a US skilled nursing operator and (3) an owner of significant real estate assets in the US and Canada (93% ownership in the US).  Because Extendicare owns all of these types of assets together, there is a massive divergence between market value and intrinsic value which will be unlocked by a separation of assets through a spin-off or sale transaction. Based on a split up value of Extendicare, I estimate total value to be in the range of $12.00 - 21.50 per share which represents upside of 87%-232%.  Furthermore, I have a very hard time imagining scenarios that will result in capital impairment and given that a separation of assets has already been announced, the likelihood of achieving one of the upside scenarios is quite high.

Recent Events

In order to understand the reason this opportunity exists, it is important to note the recent history at the Company.  Extendicare spent most of its recent history as an income trust focused on paying dividends, and as such was valued as and primarily owned by income oriented shareholders.  On July 1, 2012, the Company converted from an income trust to a corporation, but elected to maintain its dividend at the rate of $.07 per month.  As such, this was essentially a non-event for the Company’s shareholders and as you can see from looking at the stock price chart over the past year, the stock has traded in a narrow range of $7.50 to $8.50 per share.  In late April, the Company shocked its dividend-hungry shareholder base by electing to reduce the monthly dividend distribution from $.07 per share to $.04 per share.  As you might expect, this event caused a significant dislocation in the value of Extendicare stock as the downdraft in shares nearly mirrored the decline in the dividend rate.  While I viewed the $8 price as a significant discount to intrinsic value, the current share price is a really remarkable opportunity, particularly in light of the announced strategy to unlock value through an unwinding of the corporate structure.

In short, this opportunity exists because of (1) the combination of three disparate types of investment assets under one company and (2) a shareholder base focused on nothing but dividends as opposed to intrinsic business value.

Business Description

Canadian LTC

In Canada, EXE operates a total of 88 LTC facilities (49 of them owned) representing 11,467 beds (6,745 owned).  These facilities are located predominantly in Ontario and to a lesser extent in Alberta, Manitoba and Saskatchewan.  The Company’s Canadian operations generated LTM Revenue and EBITDA of $733.4mm and $72.0mm, respectively and have had very steady occupancy in the 97% range.

Within the Canadian market there are three large players, which include Extendicare, Revera (private co. of similar size to EXE) and publicly traded Leisureworld (TSX:LW).  Each of these major players commands a significant share of the market, with Extendicare and Leisureworld in particular playing a major role in the 77,582 bed total in the province of Ontario.

Canadian skilled nursing assets (more commonly referred to as long term care or LTC assets in Canada) have a much different set of investment circumstances and valuation than US skilled nursing assets.  As opposed to US skilled nursing assets, Canadian LTC assets have the characteristics of (1) high occupancy rates (essentially 100%, with a long waiting list), (2) an extremely stable funding picture, (3) high barriers to entry and (4) strong popular support for these facilities.  The combination of these factors results in a supply / demand balance that is very favorable to the LTC industry in Canada and valuation multiples afforded to Canadian LTC assets that are similar to REITs. 

Canadian LTC’s are funded primarily by the Ministry of Health and Long Term Care (MOHLTC), and the funding picture has been extraordinarily stable over time.  One reason the funding picture is so stable is that the cost to the MOHLTC of keeping patients in a Canadian acute care hospital is approximately 4x the cost of providing services through an LTC facility.  As such, the MOHLTC has a strong interest in maintaining and encouraging the health of Canadian LTC facilities. Furthermore, the wait list to get into Canadian LTC facilities in Ontario is about 30% of the size of the total number of beds, which means that occupancy rates near 100% will be the norm well into the future.  Further, based on demographics and new supply coming online, the gap between supply and demand is only expected to widen.  One of the issues with building new supply, according to CIBC, is that the incentives offered by the government on new builds have not been adequate to support rising supply.  As a result, there has been just modest growth in units over time.

The following report, though dated, gives a lot of great data supporting the long term thesis of the Canadian LTC industry:

http://www.oltca.com/Library/march11_cboc_report.pdf

US Skilled Nursing

In the US, the Company operates 158 facilities (virtually all SNFs) representing 15,361 beds, and the Company owns the underlying real estate related to 93.3% of those beds.  In addition, the Company owns 21 SNFs in Kentucky (the “Kentucky SNFs”) that the Company does not operate, but leases out to a third party under long term lease for a net of $15mm per year.  The Company’s SNFs are geographically diverse, but centered in the Northern states of Michigan, Wisconsin, Pennsylvania, Ohio, Washington and Indiana.  

The US SNF industry is quite large and contains a number of significant players, many of which are public and will be discussed individually later on.  Many of the players in the space have reorganized their assets into OpCo/PropCo structures in order to separate real estate from operations and hence maximize the value of assets awarded by the market. 

From a regulatory perspective, the industry has gone through significant cycles over the past 15 years.  You may recall the last down cycle in the late 1990s where significant reimbursement cuts forced some players into insolvency.  We are currently near the tail end of significant reductions in Medicare reimbursement for SNFs, and while there is healthy debate over the likely course of the industry over the coming years, it is fairly clear that the financial markets are saying that the industry is an important part of the health care continuum and is as relevant as ever, despite the government not compensating the industry adequately for the value that they provide.  For those who are new to the space, the industry is basically structured along a continuum based on severity of illness (acuity level). 

One could summarize the continuum along the following lines:

• Acute care hospitals:  sit at the top of the pyramid to treat the most severely ill (and at the highest cost)
• Long Term Acute Care hospitals (commonly referred to as LTACs), which is where very ill patients typically are sent after the first 5 days at an acute care hospital and typically remain for 30 days or so.
• Transitional Care Centers:  this is the next step in the chain and it is a bit of a murky line between these facilities and SNFs.  In fact, more commonly SNFs are now starting to provide some of these same services in order to bridge the gap and take in patients straight from the LTACs.
• Skilled Nursing (SNFs):  This is where our subject company, Extendicare plays in the market and is characterized by average stays in the range of a few months and Medicare reimbursement rates in the range of $475.  This is part of the “red meat” of our health care system; it serves a crucial function of being able to provide for a very large number of ill to moderately ill and elderly patients at a relatively low price as compared to hospitals.
• Assisted Living Centers (ALCs):  These facilities provide a longer term solution (about 2 years or so on average) and you can think of these facilities as being hotel-like with the addition of basic medical and personal services.

The Company is a significant player in the SNF industry and is not particularly notable for anything other than (1) being a Canadian company with significant SNF assets and (2) its significant real estate ownership.  Otherwise, Extendicare is regarded as a competent player, well run, with good facilities (not above or below average).  While there is debate over the future course of reimbursement, I will not delve into that issue deeply in the course of this analysis.  I happen to believe in the crucial role of SNFs in the continuum and that we are at a low point in valuation and reimbursement levels, but that is more opinion than fact.  I will keep this writeup focused squarely on the objective analysis of how the market currently values SNF assets and how the market is likely to value Extendicare’s US business once it is either sold or otherwise separated from the Canadian business.

Capitalization

If you are reading this writeup you may have already taken a quick look at some of the financial statistics relating to EXE.  You are probably asking about the seemingly enormous debt load of the Company. I too had a dubious view of their capital structure when I first took a look at it, so allow me to dispel your fears.  Now that I understand it, I actually view it as a fairly conservative capital structure and I am an investor that is sensitive to leveraged balance sheets.  I refer you to pg. 92 of the Company’s 2012 annual report (Note 11, Long-Term Debt). 

The first point relates to the largest chunk of the EHSI capital structure, the HUD Mortgages.  The HUD mortgages were arranged over the past 18 months, have a term of 33 years and an effective interest rate of 4.33%.  Importantly, the HUD mortgages are underwritten conservatively with a loan to value ratio of 65%.  The HUD mortgages are arranged on a property by property basis, with each individual loan in an entity that is non-recourse to anyone including the US subsidiary (e.g. if an individual property were to default, it would not impact any other property let alone the US subsidiary or the Company as a whole). 

The second point I will make about the capital structure is that EHSI (the Company’s US subsidiary), is a bankruptcy remote entity from the perspective of the parent company (Extendicare Inc.).   You will see language about a guarantee of the relatively small EHSI Credit Facility on pg. 94, but in a murky detail I will note that the guarantee is made by Extendicare Holdings, Inc., which is actually the US holding company for the US subsidiaries.  This is confirmed in the Org. Chart on pg. 7 of the Company’s 2012 Annual Information Form and I have confirmed such with the Company as well.

The third point relates to the level of indebtedness, and as such I refer you to pg. 93 of the annual report, which states “at 12/31/2012 EHSI had approximately 55 unencumbered centers valued at an estimated $250mm”.  Now, as I will show later in the report, I think that estimate is quite low, but as you can see the US business is capitalized conservatively with long term mortgages at LTVs of 65% and a significant number of additional unencumbered assets.

In the Canadian subsidiary, the largest portion of debt is CMHC debt (the Canadian Mortgage and Housing Corporation), which has terms similar to that of the HUD debt – extremely long maturities at a weighted average rate in the low 4s. 

At the Canadian parent, there are 2 series of convertible debentures which are convertible into stock (way out of the money), but within the bounds of my valuation expectations.  These are publicly traded and the most junior debt securities in the capital structure, and they trade above par yielding less than 6%, adding additional support to the notion that the capital structure is not a concern.

http://web.tmxmoney.com/quote.php?qm_symbol=EXE.DB

For reference, here is a summary of the capitalization structure of the Company:

EHSI Debt
HUD Mortgages				521.6
PrivateBank Loans				33.8
Other Debt				14.7
Total				$570.1
Canadian Subsidiary Debt
CMHC mortgages				189.2
Non-CHMC Mortgages				15.5
Finance Lease Obligations				110.3
Construction Loans				39.7
Total				$354.7
Canadian Parent Debt
2019 Debentures			$11.25	120.9
2014 Debentures			$19.90	117.3
Total				$238.2

 

Valuation

As stated at the beginning of the writeup, I think the current share price represents a remarkable valuation disconnect.  What follows is an objective discussion of the various comparables that Extendicare Canada and Extendicare U.S. will be compared to once the businesses are separated by the end of 2013.  The Board has noted that the separation may take the form of a sale or a spin-off – as such I think it is very relevant to think about scenarios where a potential buyer looks to monetize the real estate value that is inherent in the US business, whether it be a REIT or a private equity investor. 

Over the course of this discussion I will refer to the Excel schedules contained at the following link:

http://www.mediafire.com/view/?c5kc3jkavoo3tbo  (once you access this link there is a “download file” option in the bottom right corner of the page)

US Real Estate

I will start with the largest anomaly inherent in the Extendicare valuation dynamic: the market value of skilled nursing beds as implied by publicly traded US Reits that own SNFs. 

I refer you to the Excel schedules, tabs labeled Property Detail and Comps.

Consider for a moment Sabra Health Care REIT (SBRA).  Sabra is the result of the Opco/Propco reorganization of Sun Healthcare, whereby the operating business was acquired by Genesis Healthcare and the real estate assets spun into the Sabra REIT.  Consider for a moment that 87.5% of SBRA’s rental income comes from skilled nursing, and they own 10,826 beds and their enterprise value is $1.6 billion.  SBRA owns a little more than half the number of beds that EXE owns at its US subsidiary alone, has no US operations, no Canadian operations/real estate, yet it is valued at almost the exact same amount as Extendicare as a whole.  This does not make any sense.  Assuming that 87.5% of Sabra’s enterprise value comes from the 87.5% of its rental income that comes from SNFs, and applying that value to SBRA’s 10,826 beds yields a per bed valuation of $130,000.  Applying that $130,000 figure to EXE’s 16,095 US owned beds yields a valuation on EXE’s US real estate of $2.1b.  This figure in and of itself is far in excess of the entire valuation of Extendicare today, and is the easiest way of showing the enormous spread between valuation of similar assets in the market when they are separated out for investors to see in daylight.  Frequently, you might have valuation disparities between SNF assets from one state to another that accounts for the disparity in per bed valuation – comparing the states where EXE has SNFs to where SBRA has SNFs, shows that 4 of Extendicare’s 8 most important states (KY, OH, MI and MN) are also among the top 8 states for SBRA.  These are highly comparable assets valued extraordinarily differently by the market.

The next best comparable is LTC Properties, which generates a little more than half of its rental income from SNFs.  Since it is not a pure SNF play like Sabra it is not as clean of a comp, but still quite useful as another data point.  The same methodology applied to LTCs valuation (assuming half of the enterprise value comes from the half of the rental income to LTC generated by SNFs), applying that value to LTCs 10,072 beds, yields a valuation of about $96,000 per bed.   Using this lower number per bed and applying it to EXE’s 16,095 US beds also yields a valuation in excess of EXE’s entire current enterprise value.  How comparable are these assets to EXE’s?  Guess who is listed as one of LTCs top tenants?  You guessed it, Extendicare.  This is further damning evidence that the market is valuing cash flow coming out of Extendicare very highly when it is dolled up in the form of a dividend paying REIT, and valuing it quite differently in the case of a Canadian corporation with a confusing mix of assets.  Put differently, this shows how the market values lease payments coming from Extendicare, which is important to the extent that a real estate oriented buyer is interested in carving out the US real estate assets.

US Skilled Nursing

As discussed above, there is significant debate about the long term outlook for the US SNF industry.  Despite the uncertain outlook of the industry the comps are plentiful, allowing us to see exactly how the market is valuing various SNF related enterprises. 

The US Comps can be placed into 2 buckets: (a) those that have significant real estate assets as well as operations (owned and operated), and (b) those that have reorganized into an Opco/Propco structure.  Clearly, those that have separated into the Opco/Propco camp will achieve a much higher aggregate value for their assets as opposed to the owner/operator players.  Which is the more relevant for Extendicare?  In the near term, the only announced transaction is to separate Canada from the US and this may take the form of a spin or a sale of the US business.  To the extent a sale transaction occurs, I would bet the highest bidder is one who has an agenda for the real estate value.  Therefore, given the extraordinarily high percentage of owned real estate at EXE (higher than any owner/operator) and especially given the meaningful real estate income stream generated by the 21 Kentucky SNFs leased out to a third party for $15mm per year, I think it is very appropriate to start thinking one move ahead on the chess board about how valuation can ultimately be realized, especially in the event of a sale to a real estate minded investor.

OK, so first off the owner/operators.  For this set the best comps are Skilled Healthcare and The Ensign Group.  These players each own 77% of their bed count, while Extendicare owns 93% of their beds.  I would argue that the higher ownership level of EXE should afford them a higher than average multiple, but for purposes of this analysis I will use the average.  I look at value based on both an EBITDA multiple and based on a per bed valuation and take the average of those methods.  As a third check on valuation I look at asset value as determined by the Company’s lender via loan to value ratios and the Company’s disclosure of value for unencumbered properties.  I do not use this method as an input in the overall valuation of the Company, but I think it is another interesting sanity check on value – you will see this method in my spreadsheets under the title “Back of the Envelope Asset Value”.

A summary of valuation by method is shown below and full detail behind these calculations is presented in the Excel sheets.

Valuation Summary
EBITDA Methodology		$852.5
Per Bed Methodology		$1,238.2
Average of Methods		$1,045.4

 

For the Opco/Propco reorganization scenario, I valued the owned beds based on an average of the REIT comps shown above and then assumed out of conservatism and for ease of discussion purposes I assumed there is no Opco value.  This Propco value is as follows:

Propco Value
Owned Beds			16,095
Value / Bed Adjusted for CAD			$114,594
Total Real Estate Co. Value			$1,844.4

 

Before continuing, I would like to note that my valuation of the US business appears to be the area where there is the biggest divergence between my valuation and that of the 1 sell sider who has attempted a break up valuation.  Where we differ is that I give EXE credit for its substantial ownership of real estate and try to take that into account in my valuation, whereas the sell sider is happy to apply a 7x multiple to EBITDA and call it a day.  There is no doubt that EXE’s US business is underperforming right now and is earning a margin well below its US peers despite its stronger real estate position.  I think it is overly punitive to ignore the real estate assets and merely assume an inline multiple on a depressed earnings number.   The sell side approach of applying a 7x EBITDA yields a valuation in the mid $700mm range.  This simply does not make any sense in light of the $550mm balance of HUD mortgages – this debt was arranged recently (some of it as recently as a month ago), and is secured by just about half of the Company’s owned properties at a 65% LTV.  It appears that the lenders think that just half of the Company’s US real estate alone is worth in excess of the sell sider’s mid $700mm range valuation. 

 

Canada

On to Canada.  This one is the easiest to look at as it is the most stable stream of income and there is an extremely close comp in Leisureworld.  I do not attempt a valuation scenario based on a separation of the Canadian real estate assets vs. operating company, as that structure is not likely to occur.  The simplest way to look at this is the EBITDA multiple afforded to Leisureworld, an extremely close comp.  Chartwell Retirement Residences is a decent comp as well and trades at a similar level, but Leisureworld is by far the best.

I apply a 14.6x LTM EBITDA multiple to EXE’s Canadian business (this represents a 20% discount to Leisureworld’s LTM EBITDA multiple to be conservative).  The resulting valuation is $1.1b for the Canadian LTC assets. 

I want to make the important point that the Canadian business, as valued above, covers 80% of the current stock price.  To clarify that - the valuation outlined above of $1.1b, less all of the Canadian debt (both at the Parent and the Canadian sub) would yield a value of $5.15 per share.  This is highly relevant because, though they have not formally announced this, to me it is quite clear that the Canadian debt will stay with the Canadian business and all of the US debt will be spun off or otherwise attached to the US business.  Hence, you are pretty well covered by the Canadian subsidiary alone, and then you will get the upside of the US business. Furthermore if you look at the pattern of events and listen to the dividend reduction call from a week ago, the Company hints that the dividend reduction was designed to reflect the income from the Canadian business and not rely on the uncertainty around the US business.  It may very well be the case that the Canadian entity sticks with the $.04/month distribution after the spin off/sale of the US business, lending further support to the argument that you are covered by the Canadian business alone.

 

Valuation Summary

My basic valuation scenarios are summarized in the chart below – a scenario based on a spin off of US operations into a new publicly traded entity, and a scenario based on getting value out of the real estate either through an Opc/Propco split or through a sale of the US subsidiary to a real estate oriented investor:

SUM OF THE PARTS VALUATION
		Owner/	Opco /
		Operator	Propco
US Business - Owner/Operator Scenario		$1,045.4
US Business - Opco/Propco Scenario			1,844.4
Canadian LTC Business		1,053.5	1,053.5
Total Enterprise Value		2,098.9	2,897.9
Less: Net Debt		1,060.8	1,060.8
Equity Value		$1,038.0	$1,837.0
Per Share Value		$12.07	$21.36
Appreciation from Current		87.7%	232.2%

 

Final Thoughts

This is a unique set of circumstances and a rare opportunity to participate in a spin off situation that virtually nobody seems to be on top of. The bizarre sequence of events (dividend reduction, spin announcement during market hours, conversion from income trust a year ago) have all come together to create the opportunity that exists in the market today. There was actually one caller on the dividend reduction call from a week ago who hinted at this scenario and who noted that there is $800mm of value on the table here ($9.30 a share or so).  I brushed it off when I first heard it, assumed that would never happen – but here we are today.  I look forward to seeing how this one plays out from here.

Revaluation of the assets upon separation of US / Canadian businesses before YE 2013.