Next On The Low CAPE List: Russia

Russia has a little more variety in its ETF options than the first three countries (Greece, Argentina, Ireland) so I figured I’d take a deeper look at them. Excluding wacky 3x Super-Ultra-MegaBear nonsense, there are 4 major ETFs:

  • RSX – Market Vectors Russia ETF
    Probably the best of the three large cap Russian ETFs for reasons discussed below.
  • ERUS – MSCI Russia Capped Index Fund & RBL – SPDR S&P Russia ETF
    Why are these together? In terms of composition they’re fairly similar – they both weight the largest companies very heavily. The Big Three (Gazprom, Sberbank, and Lukoil) are 44.9% of ERUS’s assets and 40.88% pf RBL’s assets. That’s also why I like them a lot less than RSX. ERUS/RBL are closer to an actual cap-weighted index but their immense concentration in the biggest names (both for the Big Three and the % of AUM devoted to their top ten holdings) seems to introduce too much idiosyncratic risk. It seems like for a CAPE-based allocation strategy you want your investment spread out a bit, lest poor performance by one or two mega-caps drag it all down. Besides, with that level of concentration it’s getting close to the point where you might as well opt for a DIY approach with the relevant ADRs and save yourself the .6% management fees.
  • RSXJ – Market Vectors Russia Small-Cap ETF
    This is the only dedicated small-cap ETF for Russia. The other three do give investors exposure to some of the companies on this list, but it’s much more limited. On RSX, which has the lowest concentration in its top ten holdings (60.18%) only 10% of AUM are devoted to small-caps. Again, that’s probably closer to a true cap-weighting but with such a top-heavy market “true” makes me uncomfortable. This seems like a worthwhile inclusion to balance that out.
  • Tracking sheet: I’ve set up a tracking sheet to watch all the low CAPE names, plus everything else I write about from here on. I’ve had this post on the shelf for a while (oops) so the entry for this post was actually from a couple weeks ago.

    Lies, Damn Lies, and Argentine Statistics

    Right after I hit “publish” it occurs to me: is Argentina’s CAPE being calculated with the correct (meaning real) CPI figures? And if not how much does that swing things? Since I didn’t run the numbers I have no idea; if it was done with official numbers then the rest of this is relevant and if it was done with adjusted inflation stats then not so much. Good mental exercise in any case I suppose.

    The Argentine government has apparently been suppressing the real inflation figures for a while so everything back to mid-2007 is about half what the real value is (possibly even more distorted, but I’ll go with that for the moment). I wondered what effect this would have on a country’s CAPE since inflation-adjusting the earnings is sort of an important part of the calculation, so I built a spreadsheet to try and test it out.

    Now, I have absolutely no idea what the actual earnings were for the components of the Merval Index so I estimated a few earning sets based on what I did know for certain: the CAPE based on official figures, the current Merval value, and the official CPI figures. Based on that I worked backwards to get earnings sets that would result in the known CAPE value (5.17). I made three sets: one where earnings were assumed to grow slightly faster than inflation (I + 2%), one where they grew slightly slower (I – 2%), and one where they were completely stagnant in nominal terms.

    The results (“current” in terms of the index value is current when the CAPE was calculated):
    Argentina base CAPE
    Argentina declining earnings CAPE
    Argentina stagnant earnings CAPE

    Using the “true”(-ish) inflation makes the resulting CAPE even lower. I was puzzled for a moment – how would fake stats = better value? – but then I got it. The entire reason you inflation-adjust earnings in CAPE calculations is because otherwise previous years’ earnings would be too small in today’s dollars. So with that in mind, using the much higher CPI values to adjust the late 2000s period meant that their earnings were comparatively more valuable in real terms and that average real earnings for the 10Y period were higher in the “real” inflation scenario. In that case, Argentina is even more cheap that it initially appeared – though it would be even more imperative to hedge out the peso exposure. Interesting!

    Fun with CAPE (Don’t cry for me Argentina)

    I Like Charts Too Much

    I was tinkering around with CAPE stuff myself when I saw a bunch of CAPE-themed posts popping up (especially on Greenbackd), so this felt appropriate… The data all comes from Shiller’ssite and all returns below are real returns. I have everything on a single spreadsheet that I’ve uploaded here so you can make fun of me like R&R later. I tried to make it easy to tinker with, and even added some sparkline charts to visualize the y-o-y changes because they’re just cool.

    One thing that struck me was how much of a difference dividends make in the final result. The 5Y and 10Y CAGR scatterplots are a world apart with vs without dividends. Observe:
    CAPE 5yr no dividends
    CAPE 10yr no dividends
    CAPE 5yr
    CAPE 10yr

    Leaving out dividends basically shifts the entire chart down a few percentage points. With that in mind, I wish people writing about CAPE expected returns were clearer about whether/how they included dividends in the results because excluding them introduces a big downward bias to the results [for reference I calculated yearly total return as (Real Index Price Y1/Real Index Price Y0 - 1 + Dividend Yield Y0)]. The market’s dividend yield also seems to relate pretty closely to its CAPE:
    CAPE dividend yield

    CAPE doesn’t seem to be terribly useful as a market timing tool (Wes Gray seems to have demonstrated that in case anyone was curious), but when I saw the results charted above my first instinct was that CAPE is really a tool for asset allocation and expectation management. The folks at GestaltU seem to agree. I haven’t seen others touching as much on the issue of returns spreads and hurdle rates (everyone seems to like them some sweet regression models), so I guess I’ll dive in with another graphical avalanche. Histogram attack!

    CAPE 1st quintile 5yr
    CAPE 1st quintile 10yr
    CAPE 2nd quintile 5yr
    CAPE 2nd quintile 10yr
    CAPE 3rd quintile 5yr
    CAPE 3rd quintile 10yr
    CAPE 4th quintile 5yr
    CAPE 4th quintile 10yr
    CAPE 5th quintile 5yr
    CAPE 5th quintile 10yr

    Whew… A few points:

  • 5Y CAGR results are a lot more variable than 10Y (duh). I thought about checking to see if there’s any pattern or mean-reversion effect that would improve performance – like for example testing the current CAPE at the X-year mark to determine whether to stay/go – but I feel like it would be easy to end up with data minig issues. Eyeballing it from a 1st quintile starting point, it seems like the 10Y results tend to be poorer than the 5Y results (and with a lower absolute return, <10% CAGR) when the CAPE at the 5Y mark is in the third quintile or so. It's an intuitive rule (move on to greener pastures when valuations are getting unappealing), but it needs a real test. Eyeballs are so unreliable.
  • There’s a lot of variability to the returns at every level and even among periods with (roughly) the same CAPE. That sounds like a “duh,” but most models I see related to CAPE/long-term valuation levels don’t include any measure of their margin of error when they provide their 10Y (or whatever period) return estimates. Empiritrage totally did that right after I wrote this. It does however remain true of the more sophisticated blended model from GestaltU that I linked above. It claims to explain about 80% of returns (pretty slick, over-fitting?) but the authors don’t spend a lot of time talking about the error ranges on the results. Even a small difference in CAGR grows to a significant one after 10+ years.
  • Of significant interest to me was performance of the different 5Y/10Y periods against different hurdle rates.
    CAPE 5y failure rateCAPE 10y failure rate
    A breakdown into quintiles is arbitrary (why not sextiles? or vigintiles?) so I can’t read too much into it, but it’s not rocket science either: low valuation good, high valuation ouch. Since we’re firmly in fifth quintile territory, the thing I’d like to figure out is: how much better than average do you need to be to capture a worthwhile return (at least 6% real) in this environment? Past a certain point it’s only ego keeping you in the game. Everyone can’t be above average in any market – even more of a problem in an ultra-low return environment. This has clear implications for anyone who absolutely requires a specific minimum rate of return, so it’s funny that I haven’t seen anything suggesting that pension managers are tinkering with valuation-based asset allocation – mostly just folks trying to make up the difference with “alternative” asset classes. Which seems like reaaaaaally rough going. Valuation-based allocation’s not going to be equally valuable for all investors – CALPERS probably has more more money to invest than the three lowest countries below have market cap! – but it seems like it should definitely be a factor. Since you can’t guarantee above average performance (I mean, we all feel like we can, but…) it ought to be helpful to be in a situation where average is still good.

  • Real-world applications
    To me that final bullet point gets at the real utility of CAPE, which is guiding you on where to allocate your money and what kind of odds you’re looking at. This table below comes via Mebane Faber’s site (if you happen to be one of my 13 readers Meb and you mind me posting this let me know…also, yay) and shows the current CAPE:


    The green countries are all places where the odds are very much in your favor. I’ve only got quintile info for the US, but I’d hazard a guess that all of those qualify as deep 1st quintile for their respective CAPE histories.

    Greece I’ve already touched on before. There’s an ETF for that, even though it’s packed with financials. On the bright side the Cyprus banks on it (original index components here, new composition here) like Bank of Cyprus and Cyprus Popular Bank aka Laiki Bank have already blown up. So, uh, yay? They’re not gonna drag it down any further, for whatever that’s worth. Last time I checked credit markets in Greece looked grim and I’m going to go out on a limb and guess that hasn’t changed much, so my previous conclusion stands: the best bets are cash-flow positive companies with strong balance sheets (hard to go wrong with them in general, really…). Probably still plenty of good opportunities out there; Don Loeb is on the hunt, dontcha know?

    Then there’s Argentina. Argentina is a complete mess right now. For details: Sober Look. TL;DR version, the Argentine government is meddling in the economy too much in ever more clumsy ways and it’s going as crappily as you might expect. True inflation (official stats are fudged) is unknown but sky-high. The only passive vehicle for investing in this market is ARGT, the FTSE Argentina 20 ETF. It’s not really a “true” Argentine ETF because it has only 60% exposure to Argentina. The exposure is even more limited in terms of actual Argentine businesses since only about 30% of its holdings (by market value) are actually based in Argentina, so I feel like it wouldn’t really benefit from the low CAPE effect. There are ADRs for several promising companies, but you’d need to be able to hedge the currency to make it feasible otherwise devaluation of the peso against the USD would just eat your returns. Unfortunately the only peso futures I found with a quick search are on the ROFEX exchange…in Argentina. D’oh. With that in mind, this opportunity is better suited to big investors who can hedge with OTC derivatives. It’s going on the tracking sheet anyways for curiosity’s sake.

    Next lowest is Ireland, which is a lot easier to invest in and is almost as undervalued. There’s an iShares ETF (composition and data here) and the Euro is a lot easier to hedge (or ignore if you feel it’s secure) than a wild card like the peso. As far as individual companies go, Wexboy has a pretty exhaustive rundown of pretty much the entire Irish market (most recent post here) so there’s no point in getting into that here when he’s already done it better.

    I could go on – I’ve never looked at Russia before, seems neat – but I think I’ll save that for another post. In the meantime here’s a link to a tracking sheet where I’ve recorded all the ideas mentioned above along with a benchmark for comparison.

    Cigar-butt safari results: Don’t hunt in China

    Previous post

    Data: Google doc

    The second set of companies I recorded after the first post was far enough along (and I, uh, actually remembered that this experiment was still running) so I figured it was time to follow up on the original post. And then I left the post on the shelf for a month because that’s how I roll apparently.

    One result was clear: Chinese net-nets are ass. For the first company set, the spread between the averages for Chinese and non-Chinese companies was was a stupendous 51% over a 14 month period.
    The spread was “only” about 21% on the second set, but hey it’s only been 10 months. You need to be a long-term investor to capture that kind of underperformance. Back before I started this I’d sort of wondered if maybe the result would be an underdog victory for the Chinese net-nets – a triumph of simple quant rules over “ick” factor. Nope, crushing and puzzling defeat.

    I say puzzling because this means a ton of suckers were holding on long after the rest of the world wised up to the whole reverse merger fraud routine (for a laugh: a gem from my past…ah, the folly of youth). You’d think pretty much everyone would have bailed at that point but I guess hope springs eternal. I wonder what it would cost to short these, or if you even could. That’s a killer long-short…

    The outstanding returns on super-teeny net-nets didn’t hold up (hey, that Wesley Gray guy was right, who would have thought?) but the big returns on net-nets over $100M persisted even on the second company set…albeit with the minuscule sample size of 1 and a Chinese company at that. Not terribly meaningful. Still, I’m curious to follow that criteria in the future since it at least makes a modicum of sense. And hey, these firms might have actualy liquidity so you could implement that filter with more than $12. I tried running a test on it via but the backtester wasn’t finding any stocks that fit the strategy after 2003. I’ve seen’em with my own two eyes, so either I haven’t figured out the system yet or they might be outside the universe of stocks listed in the database. I’ll have to tinker some more but either way it seems like a fun tool.

    I was disappointed by how poorly the F-score performed as a predictor of the results. In both groups the F-score didn’t have any predictive power without factoring out Chinese firms, which were basically the investment equivalent of concrete overshoes and often had great F-score courtesy of their great fake financials. After filtering out the junk, the spread between average US net-nets and the high F-score batch was only about 5% in each group. I’d expected to see something bigger considering how shaky net-net firms are by definition – testing for financial soundness seems like it should add a little more.

    I’ll probably keep rolling this experiment forward once I gather the motivation to enter in a third set. It’s been a while since I pulled the initial sets, so it’ll be interesting to see how the changing market climate affects the results.

    Quick HMG update: no more treasure on the balance sheet

    From my cursory examination this morning I noticed two things about HMG that got me temporarily excited: (1) the property sale was for $24.4M for a gain on sale of $20M, meaning they were carrying Miami property (Coconut Grove) at waaaaay below current market value and (2) the company was founded in 1972, so maybe there was more property being carried at very low values.

    Sadly, that was not the case (note the hotel value though, depreciated down to $532k, wow):
    HMG properties

    The properties overall have been depreciated down significantly from their initial cost, but the Grove Isle properties (now sold) were the ones with the biggest value gap. A quick look at the Grove Isle website makes it clear that book value was a laughable understatement of its true value.

    I don’t get the same feeling from the Monty’s Raw Bar site. Add in the property’s marginal profitability and the mess that is HMG’s related party transactions (that section should not be so long for such a small company) and it’s a pass.

    Having seen this play out, I do wonder if there’s value in digging around in small REITs…something to explore in the future.

    Weatherford and the relative value of insider transactions

    I’ve been tracking WFT since I first mentioned it and so far it’s up about 14% versus 4% for the S&P 500. Blindly following insiders seems to have done well in the short term, +10% over the benchmark. What interested me was the performance of a small bank stock that I’d added to my watchlist for a similar reason – that one (TSBK) is is up even more against the S&P (+14% since Dec 31st). That’s an absurdly small sample size, so it indicates absolutely nothing, but it led me to an interesting question: is the informational value of insider transactions higher for small cap stocks? It makes intuitive sense because one of the theoretical advantages to investing in small cap stocks is that they have a much smaller following within the analyst community and are therefore priced more inefficiently. If that’s true it seems like it follows that insider transactions would be less widely tracked and more valuable.

    I don’t know a lot about the research into the insider effect, but I don’t recall seeing anything about size effects. If it’s an unexplored topic I’m not sure how to track and test this in an efficient manner, but it would be interesting to look into.

    Sears and economic machines

    When I was researching Dillard’s I got interested in trying to value Sears (SHLD). At first I thought the task was pretty daunting – Sears has a pile of real estate that probably represents much of its real value and I don’t know anything about valuing real estate. But then I started to think about one of Ray Dalio’s papers and the idea of economic machines. This is pretty much the opposite of profound, but it reminded me that ultimately value is determined by transactions (which is kinda unrelated to the actual content of his paper, but hey…) and the real estate portfolio is just the sum of of these hypothetical transactions. Like I said, “duh,” but it helped focus me on the subject in a way that I haven’t seen from others. Most people focus on the valuation from the seller’s (SHLD) end and try to slap an approximate multiple on the real estate, but I thought it would be interesting to approach it from the buyer’s end.

    Take for example this sheet I found from Manual of Ideas. I should add that I’m not downplaying the quality of this work. This is the best quick breakdown I’ve seen on Sears (a lot better than the Fairholme one, which I though was kinda lame) and my favorite part is how it gives a lot more clarity on the range of sales prices/outcomes. Thing is, it doesn’t raise the question of who would/could drop $2.4B+ on SHLD’s assets. To that end I’ve got some preliminary work to share, although it’s still definitely a work in progress and focuses on the Sears/mall-based property.

    The biggest real estate transaction SHLD has made in recent years was the sale of 11 locations to GGP in late 2012 for $270M, so I figured that the first step should be to assess the big mall REIT players. To that end:
    SHLD mall owners A few quick observations here. Although these companies don’t encompass the entirety of the US mall real estate market, with a combined EV of $91 billion and holdings of over 624 million square feet (I believe that’s direct holdings, not including unconsolidated joint ventures) they’ve probably got a majority share and the best access to the capital markets. There are about 1,397 US malls (give or take a few ghost towns) and 66% of them are less than 1M sq ft, so I figure this is a pretty good sample.

    Looking at that EV total puts a $14B SHLD ($10B for Sears) real estate valuation in perspective. At that price who could afford to be the buyer? In the Fairholme presentation Berkowitz points out that the square footage owned/leased by SHLD exceeds the total holdings of the gargantuan Simon Property Group. That sounds like a lot of “value” in theory but in practice who would have the resources to unlock it? What effect would that have on the market and on the buyer’s financial condition? Ditto for the square footage amounts. Even accounting for other holdings that presumably take the total shopping mall square footage above 1 billion sq ft, releasing the footage occupied by Sears (about 116M sq ft) puts in the neighborhood of 10% of total footage on the market. If you look at the square footage amounts of recent transactions – even accounting for the large SPG purchases without reference to footage size – the market for mall real estate just doesn’t look that deep, at least as far as the big REITs are concerned. You’ll also notice that with the exception of GGP most of the transactions were for entire properties rather than individual anchors. Also, the footnote to the GGP purchase makes it sound like the purchases were required as part of a remodeling project – meaning GGP was a forced buyer. My gut says that’s why the Fairholme (geez, I feel like I’m picking on them…) catalyst page was so wimpy. It’s hard to visualize a practical one for the kind of valuation that puts SHLD on the high end of the range.

    Mall REITs aren’t the only buyers for real estate though. Many other mall-based department stores own at least some of their property and several, like Dillard’s, own large amounts of their own real estate. Here’s a breakdown of department store counts since FY 2002 (year ended Jan. 31, 2003):SHLD store counts
    For reference, the SHLD figure is for mall-based Sears Domestic store and SKS is for core Saks locations rather than Off 5TH outlets. A general trend is that most mall-based department stores have either held steady or shrank store counts since 2007 whereas the off-mall retailers like TJX and KSS added 750 locations during that period. A few have made recent acquisitions or managed healthy growth over the past decade, so I’ll look at those in greater detail.

    First up is BONT. Its store count ballooned during the 2000s after two big acquisitions that each doubled (!) the size of the company. Currently BONT has a D/E ratio of 25.28, $1 billion in debt, and negative earnings. They do love a good acquisition but I think it’s safe to say they’re not in a position to make any major purchases at the moment. Even if they wanted to, the debt markets are going to keep them on acquisition detox for a while.

    Macy’s is in a much better place financially but its management looks a lot more conservative. Since the May Department Store acquisition in 2005 (side note, I’ve found it fascinating watching the ongoing consolidation of the department store industry as I did this research) they’ve generally shrunk the store count slightly each year and recently they’ve been allocating free cash flow to paying down debt or buying back stock.

    JWN is interesting because it’s profitably increasing its store count during a period where others have been cutting back. A quick note about that, however: the numbers above are for full-line stores rather than Nordstrom Rack. Rack growth has been the biggest area of JWN’s expansion in recent years (doubling store count over the last decade from 55 to 108), but the Rack stores are much smaller than full-line stores and not really relevant to the question of whether they’d be in the market for a large number of additional full-line stores. The very disciplined growth in full-line store count (3 per year average in this period, with 8 as a “big” year) makes me think that’s not their style.

    Last but not least is Belk, probably the best candidate. The chairman/CEO aims to increase sales from $4B to $6B over the next 5 years, so if that sales growth corresponds to store count growth they could be in the market for as many as 150 new locations. To evaluate how well that might work, you’d need to look at mall composition within their shared territory. Since it’s local and somewhat familiar to me, I listed out the mall composition for the anchors in every mall listed for Florida on Wikipedia:
    SHLD mall composition
    This table is one of the reasons I had trouble seeing anyone other than Belk as a potential buyer/lessee for Sears’ mall real estate. Some combination of DDS, JCP, and M (frequently all three) is already operating alongside Sears in the majority of its current locations but there are a lot of potential new cities/markets available to Belk with no overlap.

    The next thing I want to do is examine the Kmart real estate situation so I can get a complete picture and try to estimate the overall value, but from what I’ve seen so far I think that the final value for SHLD’s real estate would have to be discounted even lower than the low end of the MOI estimate due to the difficulty of finding buyers. The pool of potential tenants for anchor spaces is still pretty limited as you can see above because the vast majority of anchor spaces are still occupied by department stores. There are a few others like Bass Pro Shop and Dick’s Sporting Goods, so I’ll want to check out their expansion trends when I’m researching Kmart/strip mall real estate.

    Dillard’s Part 2: What can I learn from Dillard’s?

    Closer examination reinforced my initial impression about DDS – this wasn’t actually a full-fledged turnaround. That’s a little disappointing, because I was hoping to use this to learn more about how to identify successful retail turnarounds. That’s a tricky business (ask Ron Johnson *rimshot*) and I’ve thought for a while that it would be interesting to study a group of successful turnarounds to see what they have in common. “Turnaround” is too strong a term to be applicable here because DDS didn’t have a fundamental operating issue since it was steadily profitable for the decade pre-recession (and then some) with no history of decline. There were a few useful lessons though.

    I think it did give me a useful template for evaluating a retailer that stumbled. There were two interrelated problems at Dillard’s: slow inventory adjustment going into the 2008 Christmas season leading to a big drop in gross margins and misplaced managerial attention. Even though 2008 had already been a bad year going into the holiday season DDS failed to adjust inventory levels accordingly. That comparative margin performance chart wasn’t pretty:

    dds gm price

    Thing is, that was a temporary problem. There wasn’t a long-term trend of margin compression (if anything they held strong just prior to the recession). Once management signified that they were really focused on the problem – take your pick about whether the annual report/calls or the immediate Q1 bounceback qualified as sufficient evidence – outsiders had no reason to expect the problem to continue.

    The acquisition of Mercantile Stores seems to me like a case of empire building, which was symptomatic of how management took their eye off the ball. Given the $2B purchase price, what was the ROI there? Seems like it can’t have been too high since they’ve been reducing the store count ever since (from a high of 328 down to 304 at the end of FY 2011). Again, I think the key here was looking for signals that management was now focused on areas (like expenses) that they had the most control over rather than more slippery areas like growth.

    Another thing I got from studying Dillard’s was a distrust of “real estate plays” in the retail business. SHLD, I’m looking at you. At first I was interested looking at how much of their own real estate DDS owned, thinking that the asset values of that could have provided good downside protection. Then I tried to work out who would buy that real estate and I came up empty. Dillard’s is an anchor tenant and it would be selling to other anchor tenants. Thing is, who would that be in the event that Dillard’s was dying and had to sell? I imagine every local mall and think of who Dillard’s competitors are: JCP, SHLD, and M come to mind. But they’re not expanding store counts and haven’t been for years. Neither are others like BONT or Belk. For Dillard’s to monetize the property they’d need a big buyer but it doesn’t seem like there’s one in the market right now. Most of the growth seems to be in smaller strip mall-based stores rather than larger anchor tenant type stores.

    DDS department store sharesDDS department store vs strip

    It’s funny, as I was writing this I noticed this post from Oddball Stocks in my feed: Turnarounds: cut expenses, or increase volume? DDS turns out to be a great example of the first path. Aggressive cost reductions and tight controls rapidly turned the financial situation around and restored profitability. Something to take note of for future turnarounds.

    Dillard’s, Part 1: What could an investor have known at the time?

    An investor in 2009 needed to know three things: was DDS financially stable, was management responding sensibly to the oncoming recession, and was the company undervalued with a margin of safety?

    Midway through 2008 ratings agencies began to downgrade Dillard’s [sticking to Moody's here because they provide historical ratings and press releases]. Moody’s dropped the DDS corporate family rating from Ba3 to B1 in April, then down to B2 in January’09 with outlook negative.

    For some perspective, here are the financials. The balance sheet as of Q3 2008 and at the end of FY 2008:
    DDS Q3 2008 balance sheetDDS Q4 2008 balance sheet

    Cash flows as of Q3 and full year:
    DDS Q3 2008 cash flowDDS Q4 2008 cash flow

    Here’s a press release from 10/31/08, shortly after Moody’s put DDS on review for the second downgrade. This sums up the situation and addresses a few of the major issues facing the company in the 2008/2009 period.

    -We maintain a $1.2 billion revolving credit facility with JP Morgan Chase Bank as the lead agent. The credit agreement expires December 12, 2012. There are no financial covenants under this facility provided availability exceeds $100 million. Even at peak working capital requirements in late November, availability should well exceed $500 million.

    -After we pay the $100 million 6.625% notes maturing on November 15, 2008, total maturities of long-term debt over the next two years are less than $26 million.

    -We have announced the closure of 20 under-performing stores in 2008 to date. This action will result in over $50 million less working capital requirement from inventory alone. We are continually evaluating the store base and anticipate more closures in 2009.

    -New store openings in 2009 have been reduced to two compared to ten in 2008. Capital expenditures for 2009 have been trimmed to approximately $120 million from approximately $204 million in fiscal 2008 following $396 million in fiscal 2007. Rating agency metrics ignore the greatly enhanced cash flows from this significant reduction in capital expenditures.

    -We expect over $50 million of savings in advertising, selling, administrative and general expenses in fiscal 2008 as a result of measures implemented earlier in the year. We continually review our expenses for additional reduction.


    The most relevant point for the immediate situation was the debt maturing at the end of the year. As mentioned in the press release $100M was maturing November 15th. After that, DDS was out of the woods in the short term:
    DDS debt maturities
    The next two years amounted to <$26M as the press release pointed out. 2011 and 2012 only had about $56M each. That's only about 20% of the OCF of even a poor year for DDS, so that would have been manageable if DDS was unable to roll it over for some reason.

    Historically the company was a cash generator with significantly positive cash flow back to 1997, which was as far as I checked. That's a mundane positive, but in a deep recession it's hard to get excited about a company with a history of relying on infusions of cash to funds its operations. Following the big expansion period at the end of the 90s when they acquired Mercantile Stores Company and jacked up the store count from 250 to 328 in the space of two years (and that's after selling 26 stores post-deal to raise cash - the net gain would have been 104 otherwise!), DDS has generally funded its capex internally (exceptions in '07 and '05). Excess cash went to pay down the acquisition debt ($2B) or buy back shares.

    Regarding capex, Dillard's doesn't need it in the literal sense, but requires it in the staying-relevant-and-not-looking-like-Sears-inside sense. If you look at the capex in the FY 1999-2007 period (the post-MSC acquisition period) it averaged $295.8M. Even if the company dialed down its growth plans – which it had by this time – it probably needs at least $200M annually to keep current. It could get by with less, but only as a short-term measure.

    Operating leases are something to keep in mind since they tend to make a retailer’s balance sheet look more pristine that it is in real terms. Luckily DDS owned 77% of its stores outright, with some level of partial ownership of another 9.5%.

    Finally, I wanted to know what the industry as a whole looked like comp-wise. Behold:
    DDS comps
    DDS was the cheapest of them all in P/B terms. It was also the second cheapest by P/S, behind only BONT and its miniscule holy-shit-we-have-a-giant-pile-of-debt-and-here’s-a-recession market cap (BONT leases 239 of its 280 stores, so effective leverage is even higher than it looks). If you go by EV/S instead, DDS is once again the cheapest. Compared to its peers DDS’ D/E of .536 put it in the bottom half of the group. The four companies with negative net income for FY 2008 were generally speaking the cheapest as a result. The exception there is SHLD, which I would speculate is near the bottom because, y’know…it’s Sears. Initially I didn’t see why DDS was crushed so much relative to most of its money-losing peers. BONT made sense – it was highly leveraged, so I can see investors tossing it overboard when there’s a banking crisis. I broke out the gross margin trends and the cause became a lot clearer:
    DDS gross margin
    The companies that had investors deserting in droves were the ones that suffered huge declines in gross margin. DDS and SKS were in the gross margin hall of shame, with declines of 4.14% and 7.17% respectively. Add in BONT and their collective chart looks like this: Chart

    data by YCharts

    Whereas the other five looked like this (someone forgot to tell Kohl’s about the bear market…): Chart

    data by YCharts

    So, I think it’s fair to say that gross margin is a lot of what the valuation hinged on at the time.
    dds gm price

    Was management making the right moves? I would say tentatively yes (2008 annual report). Even though gross margin is a big issue, I think it’s important that management was prioritizing cost savings because that’s the factor they would have the most control over. Not rocket science, but a really important basic point for a company looking to rapidly improve its operations – if they’re overlooking that it’s a big “look elsewhere” sign.

    The first test is how management performed at meeting the goal outlined in the press release above. $50M+ in savings isn’t a lot relative to the size of total expenditure, so it makes a good benchmark to evaluate performance. If DDS failed to reach even that meager goal, red flag. If it exceeded, it’s a sign management has their eye on the ball. The results are under the year ended Jan. 31, 2009 below:
    The actual reduction ended up being $131M. For perspective, this is by far the lowest – by over $90M – SG&A has been since 2003. I also checked how volatile SG&A has been in the past since a big move in a volatile category is meaningless if it’s been historically jumpy. The standard deviation pre-2009 was only about $43M, making a 3+ SD move. I’d count that as significant. So management wasn’t asleep at the wheel and they were targeting an additional $200M in reductions for FY2009. What could an investor project for the future? A few scenarios below, alongside the actual results:
    DDS scenarios
    Case A is what it would have looked like if sales declined 10%, gross margin reverted to the recent historical average (33.63%), and management hit their goal of $200M in expense reductions [the D&A and interest expense estimates came from the FY 2008 annual report]. Case B is the ugly case, what it would have looked like if sales plunged 20%, margins remained at the ugly FY 2008 levels, and expense reductions totaled only $100M. Management had exceeded their expense target for ’08 so I think Case B is overly pessimistic, but where investors stand would depend on how confident they feel about margin mean reversion in the context of the recession and how much they trusted management.

    The first concrete indicator of how the situation was progressing came in mid-2009 with the Q1 2009 10-Q. As it happens, DDS killed it. Gross margin was up to almost 35% and the annualized SG&A reduction was over $200M. At that point it was definitely clear that management was handling the crisis well.

    For valuation, I like tables like this that visually emphasize the distribution of the possibilities:
    DDS valuation 1And in percentage terms, relative to EV at $7/share:
    DDS valuation 2
    If you believed that gross margin was likely to revert to pre-recession levels, the valuation at $7/share was compelling. If you were a “show me” sort of fellow and waited to see the Q1 2009 10-Q, the valuation table looked like this ($10/share):
    DDS valuation 3
    A lot more red, but in this case an investor would have the comfort of knowing for a fact that management had gotten margins way up and expenses down in line with their targets. For reference, the current EV of DDS is 4.593B – way off the far optimistic end of this table. More on that later though…

    To answer the article’s initial question, investors could have responded affirmatively to all three questions. How soon they could have done it just depended whether they were more comfortable trusting in margin reversion following an eventual return to business as usual in the overall economy or whether they needed to see hard evidence first.