Issuer: Sears Holdings
Rank: 2nd lien
Price: ≈ 94
Yield: ≈ 12.9%
Ratings (S&P/Moody's/Fitch): B-/CAA2/CCC+U
Sears Holdings (NASDAQ:SHLD) is a well-known story in the investment community. Once a mighty US retailer and profiled as a business case in many well-known management books, SHLD has been struggling since 2010 to run its operations efficiently given the rise of e-commerce, the oversupply of retail properties developed over the 1990-2000s and the actions of a management team that saw the changing environment coming some time ago and has not invested enough to keep operations running in a profitable way.
First of all, I should warn that this piece of research is not an investment recommendation on SHLD's common equity. The merits and demerits of SHLD's common equity have already been explained elsewhere in Seeking Alpha at length, and I do not want to contribute to that (quite often) heated discussion. Rather, in this piece of research I will deal with SHLD '18 secured bonds. I will argue that these SHLD '18 secured bonds can have a place in a proper diversified distressed debt portfolio, and the investment case can be predicated on the short maturity of the notes, the availability of enough collateral to cover the obligations, the involvement of Mr. Lampert and Mr. Berkowitz as shareholders of the company and, especially in Mr. Berkowitz's case, as unsecured creditors too, and, finally, the lack of better alternatives with attractive risk/reward attributes in the (very) expensive US high yield space.
Most of the research that deals with SHLD only focuses on the poor recent performance of the retail operation. The initiatives taken by the management over the last years (close unprofitable stores, invest in the 'Shop Your Way' platform, reduce general costs and net working capital needs) have not translated into better cash-flow generation, and the retail operations have been burning between $1.5bn and $2bn. per year over the last couple of years, which has been covered through asset sales. The table below shows SHLD's main operating and financial metrics: sales have been trending down, from $8.1bn. in 4Q'14 to $6bn. in 4Q'16, and gross margins and selling expenses have not improved. In reality, sale-leaseback transactions (e.g. properties spun-off into Seritage) have weighed down on margins, with the subsequent result that SHLD has roughly the same levels of debt as two years ago but with a smaller asset base. However, it is important to note that SHLD still has enough assets to fund additional cash outlays, given its 1,250 stores with a (estimated) footprint of 145Msqft. at the end of 2Q'17, plus the iconic brands Kenmore, DieHard and Sears Home Services plus a sizable amount of receivables and inventories ($3.7bn.).
Now I am going to deal with the debt structure. This is not an easy task, given the speed i) at which SHLD has been burning cash and ii) at which the different debt tranches have been refinanced/replaced in the last couple of years to meet these cash requirements, and the different levels of subordination across the capital structure. It is well-known that vendors have recently been becoming stringent in financing SHLD's operations, but Mr. Lampert has had enough dry powder to cover the gap so far.
The following table shows how SHLD's debt is structured. The numbers have been taken from SHLD's latest 10Q (2nd quarter of 2017), and have been complemented by numbers from Bloomberg and from Fairholme's latest semi-annual report. Unfortunately, SHLD's latest 10Q does not display the same level of detail as displayed in the 10K, so the information presented here may have some minor mismatches, especially in the case of ESL and Fairholme's stakes:
The first lien debt is comprised of: i) the domestic credit agreement that has in turn three tranches (one of them with the 'accordion' feature depending on SHLD's working capital levels) and ii) several loan facilities given by ESL over the last few quarters that have first liens on determined properties (69 properties in the case of the '17 loan facility and 21 properties in the '16 loan facility) and very high coupon rates, higher than the ones attached to the domestic credit agreement, despite having a pledge on specific properties. Although as per the latest 10Q ESL has distributed some of these stakes to other lenders, ESL is still the biggest lender with more than $700M in its books. Finally, it should be noted that the pension liability reported by Sears in its balance-sheet ($1.7bn.) has not been included in this table, but it should be considered senior to the '18 secured notes in the capital structure, given that it has a springing lien on i) the REMIC properties (125 properties, most of them Sears full-line stores) plus an additional $100M of real estate assets that were recently pledged after the sale of the Craftsman brand and on ii) the trademarks of Kenmore and DieHard, plus a fraction of the sales (2.5%-3.5%) for the next 15 years derived from Craftsman products sold by Stanley Black & Decker. I believe the pension liability is overcollateralized given the value of the assets, and SHLD will be able to cover such liability with only a fraction of those assets, as we will see below.
The next part of the capital structure is comprised of the second lien loan and the senior secured notes. $1bn. of senior secured notes was initially issued in October '10, and in August'15 some of them were tendered, so the current amount outstanding is $300M. As per the latest 10Q, ESL had $11M and Fairholme $46M, and because they are presumably going to hold the securities until maturity, the actual amount of notes outstanding is therefore less than $250M. The notes mature in October'18. The notes rank pari passu with the $300M second lien loan, which was recently amended to create and additional maximum borrowing capacity of $500M, using some of this additional capacity at the same time with $330M of proceeds from ESL and Fairholme, mainly. The notes are secured by a security interest in certain assets consisting primarily of domestic inventory and credit card receivables.
Finally, we have the debt junior to the '18 secured notes, which comprises a small amount of capitalized lease obligations (although such leases could be considered in some bankruptcy cases senior to the notes), the $625M '19 unsecured notes and a collection of unsecured notes maturing in several different dates (2017, 2027, 2028 and 2032) that were issued before the Sears/Kmart merger (at the end of the 1990s, in the good old days when Sears was still a prime company and could afford to issue new debt well far into the future) and were also partially tendered several years ago, at the beginning of the 2000s. It is worth noting that both ESL and Fairholme are the major holders of the '19 unsecured notes, having effectively together more than 80% of the amount outstanding (SHLD approved a pro rata rights offering allowing its stockholders to purchase the debt with warrants attached). I claim that this is one of the important features of the investment thesis, given that the major shareholders are also major creditors of SHLD's unsecured debt, which is subordinated to the '18 secured notes.
Real estate (and Craftsman) transactions and other REIT's valuations
The bull case for SHLD equity is always based on the vast amount of real estate that SHLD has in its books. Although a large chunk of the real estate has been sold to cover the cash burned by the retail operation, it still has many properties that will be gradually liquidated to cover the debt and the losses going forward.
Given that SHLD's footprint is so vast and no real estate appraisals have been made public, I have gone through the most relevant transactions executed by SHLD over the last two years, in order to get the average amount in $/sqft. obtained by Sears for each of its properties. The list of transactions is shown below:
- The JV transactions: all of them were executed in April 2015, and SHLD sold 12, 10 and 9 Sears full-line stores to General Growth Properties, Simon Property Group and Macerich, respectively. Taken the dollar amounts for the 100% value of the property and the sqft reported by Seritage Growth Properties for each of these joint-ventures, the $/sqft is between 133$/sqft and 191$/sqft.
- Seritage Growth Properties: the spin-off of SHLD was born with 235 properties (a combination of regular Kmart properties and Sears full-line stores), and given the $2,7bn. fetched by the transaction and the 36,8Msqft commanded by the properties, the average value was 73,5$/sqft. One of the most common assertions in the blogosphere and in the investment community is that the best properties were sold in the Seritage transaction (and even so they could only fetch 73,5$/sqft!), so it would not be realistic to assume going forward the same multiple for the remaining properties. Although there can be some merit to that claim, Fairholme reported in its 2015 Annual Letter that the quality of the properties pre-Seritage and post-Seritage held by SHLD was approximately the same (see graph below). Although I understand that some of the readers may not give any credibility to Fairholme's analysis given its timing in its SHLD investment (as Mr. Berkowitz has publicly admitted several times), I have little doubt that their analysis is quite right, and it has been proved accurate in the recent transactions done by Sears since SRTG spin-off, where the multiples fetched have been considerably higher, as we will see below. Given the large amounts of sqft. that Seritage had to redevelop given its reliance on SHLD as a mayor tenant (versus the investment needed when you only acquire a few properties), it is no wonder that the transactions were made at a discount to other (smaller) transactions.
- 3 full-line stores: this transaction was closed in the 1H'17 according to the latest 10Q. The amount raised for the 3 full-line stores was $104M, and because neither the size of the stores nor the location have been reported, I have assumed an average full-line store size of 139ksqft (as reported in the latest 10K), which yields 417ksqft in total. Therefore, the $/sqft fetched in the transaction was 250. The sale was completed through a sale-leaseback transaction with a one year lease.
- 2 Kmart stores: this transaction was closed in the 1H'17 according to the latest 10Q. The amount raised for the 2 Kmart stores was $48M, and because neither the size of the stores nor the location have been reported, I have assumed an average Kmart store size of 95ksqft (as reported in the latest 10K), which yields 190ksqft in total. Bear in mind that Kmart stores are in general smaller than Sears full-line stores, and that SHLD only had one Kmart super center left (the super centers used to be bigger than Kmart regular stores, around 130-170ksqft.), so I have assumed that the stores sold were regular ones. Therefore, the $/sqft fetched in the transaction was 252,6.
- Distribution centers: Sears sold 4 distribution centers in total during 1H'17. For one of them they fetched $23M and for the other three they fetched $89M (through a sale-leaseback transaction), so the average selling price per distribution center has been between $20M and $30M. SHLD had 31 distribution centers at the end of the previous year, and some of these properties were owned and some of them were leased, so it is prudent to apply a discount to the remaining 27 properties to take into account this fact.
- Craftsman transaction: Although not part of the real estate assets, it is worth mentioning that Craftsman brand was sold in March this year for $900M net present value (for details of the transaction see the table above). Because Craftsman sales are not reported, it is difficult to get an idea about the multiple fetched in the transaction. For my valuation, I have assumed that the remaining two brands, Kenmore and DieHard, are sold at heavy discounts to the Craftsman amount ($900M). Although both brands have been suffering Sears' reputation and have been losing market share over the last few years, they are still valuable: Kenmore brand is reportedly the third brand by market share and it could be sold to a foreign company wanting to set foot in the US market (SHLD has also reached an agreement with Amazon recently to sell Kenmore products in Amazon) and DieHard brand is very well-regarded among the US public and has entered into the tire business recently. I think my valuation is very conservative, but as a creditor, 'it is better to be roughly right than precisely wrong'.
As an additional sanity check, I took a look at the EV/sqft. of the main REITs in the mall space. I recognize that many of these REITs have several differences with SHLD's properties, among them: i) portfolios of properties mostly concentrated in top-notch locations, ii) fully-developed properties, where no capex upfront is needed, iii) reliable tenants and iv) no cash-burning operations as in SHLD case, but I also use this analysis to get a reasonable floor valuation for SHLD acreage.
Summing up, in our valuation I will use 100$/sqft as our base case scenario, 75$/sqft. as our bear case and 125$/sqft. as our bull case, and I will leave my readers to decide whether the bull case (in case they are interested in the equity story) is too conservative or not. For the distribution centers, the amounts in every scenario (bear, base, bull) will be $7,5M, $15M and $20M per property, respectively.
The following table shows the assumptions as well as the value of the assets under three different scenarios. Many of the assumptions have already been explored, so I will focus on those that have not been discussed yet:
- Cash and restricted cash: as shown in the balance-sheet.
- Accounts receivable and inventories: I have applied recovery rates to the balance-sheet amounts. In the bear case, I have assumed 70% for receivables and 50% for inventories. Given the diverse nature of SHLD's inventories, it is difficult to calculate an average recovery rate, but if assume recovery rates of 50% for home appliances and hardware and higher recovery rates for pharmacy products are assumed, it is not far-fetched to assume an average recovery rate of 50%.
- Owned/leased properties: there is no breakdown between owned and leased properties in the latest 10Q, so I assumed the same owned/leased split as the one at the end of the previous years, and I applied that split to the sqft held by SHLD at the end of the second quarter, which per my calculations was around 145Msqft. Therefore, I reckon that there are around 38,7Msqft owned and 106,8Msqft leased. The valuation assumptions for the owned properties have already been discussed above, and the dollar values range from $2.9bn in the bear case to $4.8bn in the bull case. For the leased portfolio, given the difficulty of getting a number, I have assumed heavy discounts to the $/sqft assumed for the owned properties: for instance, in the bear case I have assumed a discount of 80% on 75$/sqft., which would yield a mere 15$/sqft. The only valuable information we have about the leases is the following slide, which tells us that most of the leases (especially Kmart's) have below market rental rates and expiration terms with renewal options for more than 25 years:
- Hoffman Estates and Troy Building: for these two properties I took the acreage from the latest 10K and I applied the same $/sqft amounts that I applied to the owned properties.
- Other assets: mostly prepaid expenses and other unidentifiable assets. I have applied a recovery rate of 50% in the bear case. It is worth mentioning that I have not assigned any value to Sears Home Services nor to Innovel. I have not assigned any value either to the reinsurance business: it is my understanding that most of risks insured are in-house, with very little third-party business (I was trying to find if SearsRe still insures the risks of some of the properties sold, such as those owned by Seritage, but I did not find any explanation in Seritage's initial public offering prospectus).
- Cash burned from ops. 2H'17-'19: I have calculated the cash that will be burned by SHLD operations until the end of '19 under certain assumptions (120 stores closed every year, gross margins of 22% and selling and G&A costs as a share of sales of 28%, average cost of debt of 11% and LFL growth in sales of -2%). Although the secured notes mature in October '18, this number can be useful to investors interested in studying the '19 unsecured notes or the common equity of the company. But for the senior secured notes (the topic of this article) I have calculated that SHLD will burn $1.6bn by the end of '18, which must be subtracted from the current asset value. Alternatively, if SHLD files for Ch.11 before the end of '19, a small share of this cash burned could be thought of as the legal and financial costs that the estate will have to bear during the restructuring/liquidation process (although given the involvement of Mr. Lampert and Mr. Berkowitz, one would assume a short restructuring period).
All in all, in the bear case scenario SHLD still has assets worth $5.8bn., enough to cover the pension liability and the first and second lien debt. The base and bull case scenarios would provide additional upside and coverage to the notes, and in both of these scenarios the unsecured debt would also have full recovery rates.
Comparison to Mr. Berkowitz's net asset value
Given the lack of coverage of the name in the Wall Street analyst community (you know - no investment banking fees, no sell-side coverage), it is difficult to benchmark the previous valuation against others. However, Mr. Berkowitz published in the 2015 Fairholme Annual Report a very useful snapshot of what they thought SHLD was worth at the time. I have already discussed above people's misgivings about the accuracy of Mr. Berkowitz's valuations, and although in my view Fairholme's valuations are a bit high with the available information I have been able to access, the asset base provides a solid coverage, but the timing of his investment has partially eroded such asset coverage. I think that for that reason Mr. Berkowitz is now in the board of the company to speed up the liquidation process, otherwise the NAV will be gradually burned every year and little will be left (if any) to SHLD's shareholders.
Fairholme reported total assets of $28bn at market value as of 3Q'15, of which:
- The easiest part to assess the change in value between then and now is that related to Kenmore, Craftsman, DieHard and Home Services, because Craftsman has been sold for $900M, so assuming that the rest of the brands have kept their value (a challenging assumption, given SHLD's deteriorating conditions) the value for the brands would still be around $4.2bn, substantially higher than my $900M in the best case scenario. I do not have any sound basis to judge whether Fairholme's valuation is high or not (I personally think it is), but I should point out that the equity investment case relies largely on the final realizable value of these brands.
- Real Estate as of 4Q'15 (no detailed information for 3Q'15) comprised 1,672 stores, around 189Msqft, 28 distribution centers, Hoffman Estates as well as other minor property assets. Given that now there are around 145Msqft in stores, and assuming the quality mix has not changed, Fairholme's valuation is above my current bull case estimate of $11bn.
- Additionally, $7bn. were reported in the balance sheet as 'inventory, cash, and accounts receivable' as of 3Q'15 (mostly inventories), so Fairholme's valuation ($4.6bn.) was giving an implicit 63% overall recovery to these amounts, which is reasonable.
- Finally, I was not able to find out what 'other assets' meant: other assets balance-sheet carrying amount was $1bn. as of 3Q'15, and comprised mostly prepaid expenses. Again, this item should be considered (unfortunately) paramount for the equity story but not so much for the secured notes, given the rest of the collateral.
It is very difficult to find in today's environment investment opportunities that give appropriate rewards to bear (credit) risks. Most of the US high yield space is priced for perfection, with very little downside protection (protection should be understood here either as high coupons or as discounts versus nominal amounts) in case things go wrong. For instance, I reported some months ago that the fracking industry is basically a cash-burner industry no matter what, but most of the debt of these companies is rated in the upper echelons of the high yield spectrum, yielding in many cases less than 6% for notes maturing in 2022 - when, if properly priced, they should trade at distressed levels. I think this kind of protection can be found in SHLD '18 secured notes, which in this case are already being priced taking into account the poor performance of the core business. Such notes could be part of a diversified portfolio of distressed securities. The thesis is based on: i) the availability of collateral and full recovery in case of Ch.11 event, ii) the activist shareholders with skin in the game and also a substantial share in the unsecured debt tranche, iii) the short duration of the notes, which provide protection against SHLD's current poor operations limiting the amount of value that can be burned in one year (the short duration could actually be thought as a 'catalyst' of the investment case) and iv) the current price of the notes, yielding around 13%.
Upside risks: i) better performance in the retail operation, so a lower cash funding gap, ii) asset sales at premium valuations (especially the brands).
Downside risks: i) Ch.11 event that can delay the full recovery of the amounts and lower the yield of the investment, ii) a general economic downturn that can preclude asset sales at reasonable prices, yielding thus lower recovery rates.
Disclosure: I am/we are long SHLD, '18 NOTES.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.