Everything You Always Wanted To Know About Casino (But Were Afraid to Ask) by Muddy Waters
Dear Mr. Giscard D'Estaing:
Neither Casino nor its more vocal defending analysts have given us reason to adjust our estimate of Casino’s equity value of approximately €7 per share. On the contrary, since we announced our short thesis on December 17th, five of Casino’s six listed subsidiaries have dropped in value by 8% to 27%. With Rallye’s NAV less than zero at Casino’s current share price, we similarly have not seen a reason to change our estimate of its value.
Casino’s response stated that our short thesis contained “a number of false and misleading allegations”.2 Despite this bluster Casino’s response failed to address the core concerns we outlined. This response reminds us of the reaction from Noble Group (NOBL SP) to our April 2015 short thesis, calling it “inaccurate, unreliable, and misleading”.3 When we shorted Noble it was rated BBB- (like Casino). We criticized Noble for its leverage and aggressive accounting (similar to our criticism of Casino). Noble has since been downgraded by both ratings agencies and its stock has declined by 64%.
We find two aspects of Casino’s response to our short particularly interesting, which is part of why we’re writing this open list of questions. First, Casino took only approximately 3.5 hours from the release of our short thesis to threaten to sue us. In activist short selling, such a kneejerk reaction from a company is often a strong validation of the short thesis. (In our experience, only Sino-Forest might have been faster with a lawsuit threat.) Second, Casino cited S&P’s recent rating as support for a purportedly “solid” financial structure. S&P’s calculations contain material arithmetic errors that we did not previously discuss. We would think these errors would have been evident to Casino.
To encourage transparency and accuracy we’re publishing a list of 10 questions for you to answer on the January 14th call. We hope you will overcome your apparent aversion to answering pointed questions for this occasion. In reviewing transcripts, we note that you told questioners six times in calls held during H2 2015 that the items they wanted to know about were not the purpose, objective, or subject of the call. We count 13 times since 2013 that you told questioners the company does not disclose a particular number. There are also numerous instances of your using the phrase “by definition”, which we interpret as often signaling your discomfort discussing a particular topic. Should you wish to regain investors’ confidence or change our minds, you should answer these questions fully, openly, and without caveats.
Many of the questions get to the heart of whether there is truly a recovery in Casino’s France retail business. If our estimates of property sale contributions to EBITDA are correct, then in 2014 and H1 2015, the France business deteriorated substantially more than investors understand. What is the evidence of a meaningful and sustainable recovery in Casino’s France business beyond management’s bullish EBITDA guidance and recent miniscule market share gains?
The confidence you and the company have expressed in 2015 France guidance could be explained by Casino generating an abundance of EBITDA through property transactions. When a company controls the amount of profit it books, it’s a lot easier to forecast the results.
Casino touted a 20 bps market share gain in the month of November 2015 as a sign of recovery. We note that Casino’s average market share throughout 2015 was actually lower than its average market share throughout 2014 (11.51% vs. 11.55%). Average market share in H2 2015 was only seven bps greater than in H2 2014 (11.58% vs. 11.51%). Despite the retail genius behind Casino’s aggressive discounting strategy, which you were touting on calls as early as January 2013, Casino’s 11.6% market share as of November 29, 2015 is lower than the 11.8% share as of March 24, 2013.
The table below showing Casino’s market share versus its “Real” EBITDA (i.e., excluding our estimate of gains on sale in 2014 and 2015) strongly casts doubt on the notion that there’s any relationship between Casino’s present market share and EBITDA generated by selling products in stores.
To throw down the gauntlet, we think it highly likely that Casino’s discounting strategy has not generated much of an increase in profitability, and that the company is using property sale gains to cover this up.
Casino seems to be following the financial engineer’s handbook, which generally goes as follows:
- Control person(s) have significant equity stakes and / or equity incentives,
- They aggressively seek to grow earnings through debt financing,
- As debt grows, their focus grows to include gaming credit metrics – particularly operating cash flow.
Usually this ends quite badly for companies and investors. Noble Group followed the same pattern.
Your complete and intellectually honest answers to the following questions will go a long way toward allowing investors to determine whether our opinion, that Casino is a highly-levered, poorly performing business being hollowed-out and managed primarily to buy time for the massive debt its controlling shareholder has amassed, is reasonable.
Carson C. Block
Carson Block: Everything You Always Wanted To Know About Casino
1. How much EBITDA has been, and is expected to be, generated by property sales? This answer should individually cover 2014, H1 2015, H2 2015, and 2016.
The company’s failure to address our estimates of respective property sale gain contributions to EBITDA in 2014 and LTM of €140 million and €165 million gives us confidence our estimates are not high. However, the company’s continued sensitivity about discussing the numbers makes us concerned that the property sale gains included in EBITDA might be even greater than we estimated. There is no investor-friendly reason to refuse to provide this information. We continue to believe that despite Casino’s view the property sales are operating activities, when combined with Casino’s leaseback or repurchase, the sales are in fact financing transactions. We therefore can only interpret Casino’s steadfast refusal to date to provide this information as a tacit acknowledgment that investors should not in fact view them as operating activities, and should not attach a multiple to their EBITDA contributions. Bottom line: if there’s nothing to hide, then stop hiding it.
2. How is booking gains on sales of property so that they’re included in EBITDA consistent with Casino’s disclosed accounting policy or IFRS?
Casino’s response states that it “strictly applies” IFRS. If we’re correct that significant property sale gains are included in EBITDA, then we strongly disagree. Unless the company has a rationale we haven’t considered, this practice seems to directly contravene its policy on accounting for asset de-recognition. Casino’s disclosed asset de-recognition policy, which mirrors IFRS policy, reads as follows:
“An item of property, plant and equipment is derecognized on disposal or when no future economic benefits are expected from its use or disposal. The gain or loss arising from the derecognition of an asset is determined as the difference between the net sale proceeds, if any, and the carrying amount of the asset. It is recognized in profit or loss (“Other operating income and expense”) when the asset is derecognized”
If Casino adhered to the above policy and books property sale gains as Other Operating Income, these gains would not be included in EBITDA because they would be below the Operating Income line. As we discuss next, as of 2013, Casino appeared to account for these gains in accordance with the above policy, which further calls into question Casino’s accounting for these gains in 2014 and H1 2015.
3. Why did the company’s response imply that it has always included gains on property sales to Mercialys in EBITDA, when in 2013, the vast majority of the gains did not flow into EBITDA (because they were booked as Other Operating Income in accordance with the company’s disclosed policy for derecognition of assets)?
Casino’s response to our criticism of its apparent practice in 2014 and 2015 of inflating France EBITDA by including gains on sales of properties reads in part (emphasis added):
“As a fully-fledged operational activity, real estate development generates results, which have constantly been reflected in the EBITDA according to the Group’s accounting principles, and in compliance with IFRS standards.”
However, as shown below, in 2013 Casino accounted for these gains outside of EBITDA (by accounting for them as Other Operating Income). Therefore, Casino’s response appears to be an attempt to further mislead investors.
4. How much of the proceeds of property sales to Mercialys were, or are expected to be, included in Operating Cash Flow? This answer should individually cover 2014, H1 2015, H2 2015, and 2016.
It is likely that Casino’s OCF includes proceeds from the sales of properties to Mercialys. Casino’s investing cash flow shows proceeds from disposals of noncurrent assets in 2014 and H1 2015 that are far lower than the proceeds from property sales to Mercialys. In addition, Casino uses the indirect method to prepare its cash flow statement. In practice, the way that indirect statements are often prepared is by first preparing the Investing Cash Flow and Financing Cash Flow sections. The difference between the periodic change in cash and the sum of Investing and Financing cash flow then often becomes the OCF number. Any calculated change in OCF that can’t be mapped to a specific account is generally treated as “other” OCF in the reconciliation.
If Casino’s cash flow statement treatment of property sale proceeds is consistent from 2014 through the present (and including Casino’s 2016 guidance), then an explanation of what portion of the proceeds are included in OCF would suffice (e.g., proceeds net of property purchases, gross property sale proceeds, IFRS gains only).
5. On a LTM basis, and on the Company’s 2016 forecast, did / will Casino generate Free Cash Flow per the following formula (all items pertain only to wholly-owned subsidiaries and the parent, unless otherwise noted):
EBITDA (excluding gains / losses from asset sales)
- interest expense
- cash taxes
+ dividends (from non wholly-owned subsidiaries)
FCF formulae differ. We see a company that does not generate enough cash flow at the parent level to service debt and pay dividends. Casino disagrees. Casino’s guidance for its France / Parent 2016 FCF is bullish. We question whether this FCF guidance is inflated by anticipated property and other asset sales.
6. Describe in detail the means by which Casino can upstream cash from its foreign subsidiaries to its parent. This description should include the payment options (e.g., dividends, interest on loans from parent, repayment of principal on loans from parent, reduction of equity capital, etc.); the constraints on how, when, and how much Casino may make of each type of payment; the taxes applied to each type of payment; and any other limitations or leakage on these payments.
This information is crucial for investors to understand how much cash Casino really has to service debt at the parent, and to pay dividends. As we previously wrote, Casino’s consolidated financials are meaningless to understand the cash flow to the parent. We strongly suspect the reality of what can be upstreamed is far less rosy than the consolidated financials would imply. We invite the company to show otherwise.
7. Is Casino’s very temporary reduction of commercial paper outstanding at the end of 2015 an attempt to window dress the balance sheet? Are there other ways in which Casino window dressed its 2015 financials (e.g., taking longer than usual to pay suppliers)?
The table below showing the amount of commercial paper Casino has outstanding suggests that Casino has been window-dressing its accounts.
8. Is Casino aware of two fundamental errors in S&P’s leverage calculations: overstating the proportional EBITDA contribution from Via Varejo by €164 million, and overstating proportional cash by €2.6 billion (thereby understating net debt)?
We did not discuss these errors S&P made when we unveiled our short thesis because we hoped Casino in its response would point them out. Instead, Casino defended S&P without condition or caveat. Is Casino unaware of these errors; or, is Casino so intent on misleading investors that it failed to adequately caveat its defense of S&P?
First, the overstatement of Via Varejo’s EBITDA contribution occurred because S&P’s flawed methodology assumes CDB owns 100% of Via Varejo. S&P did not look through CBD's ownership of Via Varejo, which is only 43.3%. As Via Varejo generated €737 million of EBITDA in 2014, this error results in a €164 million overstatement of proportional EBITDA. Second, and even more substantial, S&P does not make any adjustment for the proportional cash (i.e., cash held at non wholly-owned listed subsidiaries). What this means is S&P only reduces Adjusted Net Debt by the non-controlling share of Adjusted Debt without regard to the cash that is not owned by Casino. In other words, S&P erroneously assumes Casino has access to 100% of the surplus cash, and adjusted only for the gross debt at the listed subsidiaries – not the net debt. Had S&P made an adjustment for proportional cash, Casino’s proportional adjusted debt would be €2.6 billion more than it used for its proportional leverage calculations. We expect Casino management would be aware of these material errors. If management is aware of these errors, we do not think it’s appropriate to direct investors to S&P’s rating as assurance of Casino’s financial profile.
9. If the store leases with fewer than 2.5 years remaining were not renewed, by how much would proportional EBITDA decrease?
We stated that S&P’s operating lease adjustment methodology greatly understated the proportional debt because Casino’s contracted future lease obligations do not realistically reflect the lease payments Casino should expect to make in order to substantially maintain its existing business profile, revenue, EBITDA, profits, etc. Our point is that while the S&P methodology might work well for understanding the effective debt of many retailers, due to Casino’s idiosyncrasies, it understates the debt. Answering this question will help investors compare Casino to other retailers.
S&P’s 2014 lease expense was €801 million, which divided into S&P’s operating lease adjustment of €1.9 billion equals a remaining average term of Casino’s leases of just under 2.5 years.
10. Given that the Fitch Group’s chairman has sat on Casino’s board since 2003, and that Mr. Naouri is on the board of the controlling shareholder of Fitch, how is it appropriate that Fitch rates Casino?
As we discuss supra, fundamental errors and questionably applicable methodology underpin S&P’s rating of Casino. The Fitch rating poses another question: How easily could a ratings analyst find the courage to downgrade a company when the chairman of her ratings agency is a long serving board member of the company she rates? We’re not sure there are a lot of fiercely independent personalities found among ratings analysts. Mr. Naouri’s service on the board of Fimalac Group, which exercises significant control over the Fitch Group compounds this problem.