Description

Thesis summary

Zensho Holdings, one of the largest restaurant operators in Japan, trades at ~60x earnings, 33x EBIT and 15x EBITDA despite near-zero organic growth (for years) and facing a domestic restaurant market that is both over-saturated and likely in structural decline. More importantly, near-term profitability should be decimated by a significant and unavoidable increase in labour costs, as the Japanese economy strains at full employment, forcing the government to rewrite labour laws with adverse consequences for those reliant on part-time labour. At the same time, domestic retail interest in the name – the sole ownership group responsible for the stock’s multi-year re-rating – may see less of an incentive to own the shares due to a potential change in the way shareholder discount coupons (‘yutai’) are granted to shareholders, and/or a lower relative attractiveness of Zensho's all-in yield versus JGB rates that may actually begin to rise. Many of Zensho’s direct competitors have reported ugly numbers and commented on these trends, yet Zensho’s stock price has remained elevated (for now). This should be temporary, and is explained by a lack of sell-side coverage and institutional ownership (either foreign or domestic). As a kicker, the balance sheet carries 4x of net leverage (even disregarding leases), increasing downside optionality (although this is a re-rating story not a solvency story). Adding it all up, I think street consensus numbers for 2019-20 are at least 25-30% too high, and I think the multiple can de-rate too. At 10x my estimate of 2019 EV/EBITDA – still a premium to where it has traded for most of its listed history – the stock should be >40% lower, and I believe that is a starting point for valuation discussion (it would still imply 33x P/E for a business where I think profits are shrinking over the next two years)

Zensho has a ~$3.5bn market cap, is listed in Tokyo (7550 JT), trades >$10mm/day, and has GC borrow, so the trade is highly actionable for most all funds. 

Background

Zensho is a diversified restaurant operator (not franchisor), running just under 5000 total restaurants across >30 brands and food categories. The largest exposure is to ‘beef bowl’ restaurants (gyudon, in Japanese), which are ~37% of group sales (but 57% of restaurants), via the Sukiya chain. Beef bowls are essentially a form of Japanese fast food, some beef on top of rice, the typical cost of which is ~500 JPY per bowl. Another 22% of group sales come from the ‘Restaurant’ segment – defined not by one particular brand but by the average ticket size being around 1000 JPY per person (hence casual dining is probably more appropriate). Another 23% of sales come from the ‘Fast Food’ segment – again, across many different brands. A small portion of total restaurant revenues (<5%) come from overseas restaurants; restaurant revenues constitute 86% of total revenues. The remainder comes from apparel/nursing, and supermarkets, and is not really material to the thesis.

It should be pretty clear then that Zensho is a very diversified operator – we are not making a discrete bet against any particular food category or brand here. Zensho runs beef bowl chains, noodle chains, Japanese style pubs (‘izakaya’), coffee shops, kaiten sushi chains, etc etc. In essence we are simply making a bet against the profitability of casual dining chains (QSR) in general and Zensho is one of the most liquid, and diversified, plays available (this is actually great for us as we don’t want to take brand-specific risk for this bet). Skylark (3197 JT) is another such diversified operator but is much less expensive and has much lower reliance on beef bowls. This is actually important, as will be discussed later.

As you can probably imagine, the restaurant business in Japan is a tough one: it is extremely crowded, it is extremely fragmented, and the overall market is not really growing. In 1995 consumers spent 28 trillion JPY on outside dining; in 2005 this number was 24.4tn, and in 2012 it was 23.2tn JPY. It may be a touch higher now (since the economy has recovered) but essentially demographics are dictating that less dollars are available over time for outside dining (Japan is ageing rapidly, and old people eat out less). According to AlixPartners, the amount being spent per-capita on dining out has basically remained unchanged in the 2003-2014 period, at 252k JPY/capita in 2014 vs 248k JPY/capita in 2003 – ie a CAGR of +0.1%. Meanwhile, the Japanese market is far more fragmented than other comparable markets – the top 10 listed companies claiming just 6% share (in the US this is 15%). This is because the preponderance of restaurants remain family-owned ‘mom and pops’ as chain concepts, in general, have proved far less appealing than in other places. Adding to the difficulty, the density of restaurants in Japan is also incredibly high (one restaurant for every 266 people, according to some 2010 data, versus one for every 547 people in the US). Everyone knows Japan is a great place to get a quality meal for a competitive price, regardless of cuisine – but clearly what is good for consumers is bad for the restaurant industry. I could go on to add some of the more recent trends making it tough for the food service industry – the rise of convenience stores as viable dining options, etc – but the overall picture should be pretty clear.

Within this context, it shouldn’t surprise that overall profitability is low and has been declining. Simply put, there are too many restaurants competing for price-discerning customers (this is especially true at the lower end of the spectrum that QSR occupies). Zensho, for example, despite being a top-5 listed player, saw operating margins peak at 6.3% in 2007 (when the business was significantly smaller), and has seen average OPM around 3% over the last 10yrs. Last reported fiscal (Mar’18) saw 3% operating margins as well. Note that the company – again despite all the advantages of scale – has seen gross margins contract from near 70% to mid-50s% over the past 10yrs, due mostly to price competition. While the company does not disclose the exact cost breakdown, I believe it looks something like this:

-          Food costs: 44%

-          Labour costs: 31% (of which at least 80% is part-time)

-          Rent: 8%

-          Depreciation: 4%

-          Others (mostly utilities/electricity): 10%

-          Profits: 3%

Long-term returns on invested capital have been pretty poor – low-mid single digit % over 10+ yrs, currently running 4-5% - and certainly reflective of the quality of the business.

OK, it’s not a great business, but does that make it a short?

At its heart I am making an extremely simple argument:

-          Labour costs have to go up, likely significantly;

-          Zensho (as all restaurant operators) are highly exposed to rising labour costs;

-          There is next to nothing Zensho et al can do about it

Labour costs have to go up

This point really is pretty straight forward. The Japanese labour market has never been this tight: the jobs to applicants ratio (currently 1.5) is the highest it has been in 4 decades; unemployment is at record lows of 2.5% and heading lower; and wages have started to go up (nominal cash earnings in May rose +2.1%, the fasted pace in 2yrs). While this wage increase may sound modest, it understates the impact somewhat, because wages for part-time workers were already rising faster (+2.3% on an hourly basis through March 2018), and large corporation wages rose +6.6% over the past year in aggregate. See the article here for more details: 

https://asia.nikkei.com/Economy/Labor-costs-hit-16-year-high-as-Japan-pays-workers-more

Again – this is basically structural and won’t change near-term (it’s a demographic problem). The situation is so serious that two things are happening: 1) in an unpopular move locally, the government has made it easier for foreign workers to immigrate to take some of these service jobs (though this is still way too small of a trend to affect the overall labour balance near term); and 2) the government is being forced to change labour laws to partially compensate.

This second point is particularly relevant to the restaurant operators. Japan has always had a two-tiered labour system: full-time employees and part-time employees. While FTEs generally received good benefits and treatment (lifetime employment, etc), part-time employees were the ‘forgotten man’ of Japanese labour, without access to the social benefits FTEs received. Japanese companies loved this two-tier system because they could maintain some operational flexibility with part-time hires (much easier to fire in tough times) whilst keeping up appearances to Japanese society by maintaining the employment guarantees to ‘real’ employees. However since the market is so tight, there is a huge amount of pressure on employers to offer more full-time contracts than part-time contracts. Recent labour law changes have partially enabled this, as well. This is a complex topic (am happy to answer specific follow-up questions) but essentially what is happening is the distinction between part-time and full-time workers is being blurred – to the benefit of current part-time workers – and this necessarily means the cost of part-time labour will go up. Your average busboy in a QSR will now become more eligible for a normal corporate job, so why would he stay being a busboy? This law change is really pretty new (legislated last month I believe), and so has not even contributed to the pressures being seen thus far; in the coming quarters it will simply be more kerosene on the bonfire of rising costs. In any case, the direction of travel is very clear: labour costs will be structurally higher for the foreseeable future.

Zensho is highly exposed to rising labour costs

This point is also really pretty obvious: direct labour cost is I believe 30-31% of revenues (and ~59% of total opex). Outside of food costs, it is the key determinant for profitability and most importantly I believe the vast majority of Zensho’s ~62k employees are simply hourly labour at or near the minimum wage, hence the flow through from higher costs will be near immediate.

There is next to nothing Zensho et al can do about it

Fair enough, you say, but why can’t the restaurants simply raise prices to offset rising costs? The quick answer is simply that this (casual dining) is a commodity product in an over-saturated, price-conscious market. Those who have tried to raise prices in the industry have seen an aggressive decline in same store sales as consumers simply migrate elsewhere. A couple of examples:

-          Torikizoku, a small operator of Japanese pubs, hiked prices 7-10% in August a year ago to offset costs. Since then customer traffic has fallen 6-11% and SSS have fallen 3-8%;

-          Toridoll, an operator of udon noodle shops, hiked prices ~5% in March/April this year. SSS fell 6% the next month, driven by a 9% decline in customer traffic;

-          Ringer Hut, an operator of pork cutlet shops, hiked prices 10%+ in October. After a couple of months of price-inspired positive SSS comps, customer volumes have fallen 5-6% in recent months and reported profitability was -20% yoy in the most recent Q any case as they still couldn’t fully offset rising costs.

 The other answer is that in Zensho’s case, they have been effectively raising prices for years already. While overall SSS have been flat to slighter higher the last 4 years, this hides the fact that average customer numbers have been going down since 2013, and that positive comps (for the Sukiya chain that is the largest contributor to the group) have simply been sustained by higher prices:

-          2013: SSS -7.8% (customer numbers -10.5%, average spend +3%)

-          2014: SSS -4.3% (customers -4.5%, spend +0.1%)

-          2015: SSS +2.6% (customers -3.8%, spend +6.6%)

-          2016: SSS +1.1% (customers -5.6%, spend +7.1%)

-          2017: SSS +1% (customers -1.7%, spend +2.7%)

-          2018: SSS +2.5% (customers -0.5%, spend +3%)

-          2019 thus far (June): SSS +0.8% (customers -2.6%, spend +3.6%)

In the face of rising costs and lower customer count for years, I am really skeptical they can keep raising prices ad nauseam as they have already pulled this lever multiple times over in past years (indeed they already raised prices on some items last year, with minimal effect).

What are the comps saying?

Rising labour and other costs have been affecting some of the smaller restaurant chains for at least a few quarters – this is why Zensho is such an outlier – but the pressures appear to be accelerating. For example, Yoshinoya, a $1bn market cap operator of beef bowl restaurants (ie a direct competitor to Zensho’s main Sukiya chain), reported shocking 1Q results (200bps of operating margin decline YoY despite a 4% increase in SSS). They blamed a combination of higher rice/meat costs, as well as ‘increased recruitment and training costs to secure the labour force, and an increase in personnel expenses due to the rise in the hourly wages for part time employees.’ Clearly all these issues translate directly to a huge portion of Zensho’s business – but since that report Zensho consensus numbers have not changed. Basically every other restaurant/café chain that has reported thus far has seen declining margins (even those with flat or positive SSS, like Yoshinoya), mostly due to increased labour and other costs. This is a theme that has befallen many apparel retailers, as well, so it does seem pretty endemic.

Thoughts on Valuation

I am not making any particularly draconian assumptions for this year/the out years, beyond my rising labour cost thesis. Looking at street ‘consensus’ (just two domestic sell-side houses), it looks like 1-2% SSS, and 3-4% new store growth, is being modelled the next two years (ie to Mar’19 and Mar’20). While these absolute growth rates look a bit high, I don’t have a huge problem with them (I am at more like a 3% topline growth rate with lower SSS but that’s splitting hairs). The issues I have are with the margin progression. Currently OPM is 3% and I think basically all cost inputs will be running strongly against the company this year and next. Beyond labour, the majority of costs are food related and imported (ie a weaker Yen hurts the company); meanwhile rents and utility costs are both going up too. Frankly I think labour alone could rise at a 5% CAGR the next two years and that alone would take ~5bn/yr out of operating earnings, or 50bps of margin – in other words I think OPM could realistically be closer to 2.5% this yr (vs 3% last year) and maybe 2-2.5% in 2020 (depending on how much they can pass through) – but somehow the street thinks margins will expand to 3.5% by 2020. I simply cannot comprehend how this is going to happen without massive price increases, which, as discussed, seem very unlikely.

 Even with some degree of cost pass-through (mostly in 2020), I am ~5bn and 7bn lower at the operating and EBITDA levels in 2019/20 than the street. Using FY19 to keep it simple, I think EBITDA will be more like 33bn (vs 38bn consensus) and hence 15x EV/EBITDA today, while on a P/E basis at this earnings level I think the stock is at ~60x P/E. I think a reasonable multiple would be the average of where it has trade the last 10yrs, or around 10x EV/EBITDA, implying 40% downside from today’s price to start the conversation. It goes without saying that this would still be, in absolute terms, a very expensive price for a negative growth asset in a structurally-difficult market, and still well above the historical low multiples with likely ongoing downside earnings momentum.

Why does the opportunity exist? Yutai and the shareholder discount coupon phenomenon

OK, you say, but this is all pretty obvious stuff, why does the stock trade at such an elevated valuation today? Why would investors start caring now? These are good questions and I think go to the heart of the opportunity. It is important to acknowledge two things, at the outset: a) this stock has traded at much cheaper valuations historically (indeed the last time OPM was <2%, the stock was around 9-10x EV/EBITDA) – so there is nothing ‘permanent’ about the stratosphere in which it now trades; and b) pricing these stocks is not just about traditional valuations but needs to take into account the shareholder discount coupon yield as well (known as ‘yutai’ in Japanese); and c) this is not a well-followed company and short-interest is very low today. Of these three, I think the yutai concept is the one requiring the most explanation.

Yutai is a fairly common concept in Japan: essentially the company – whether it be a retailer, a restaurant, an airline, basically any consumer facing business – gives a discount coupon to its shareholders, every year. Generally the way it works is, once (or twice) a year, the company will mail its shareholders a set of coupons dependent upon how many shares they own. Thus, in Zensho’s example, the current yutai is  2x a 1000 JPY coupon for every 100 shares owned. Since the stock price today is 2525, this means your average retail shareholder needs to invest 252k JPY ($2k USD) to receive 2000yen in annual value, ie the implied ‘yutai yield’ is (2x1000)/(100x2525) = 0.8%. To this we must add the actual dividend on the stock – currently 18 JPY/share. In other words, the total shareholder return – the 'all-in yield' comprising yutai as well as dividend – is 1.5%.

This is not that juicy, either in absolute, or relative terms (there are many all-in yields now north of 3% in the restaurant space, and a couple north of 4%). However note that for most of the recent past (basically all of 2015 and 2016) the average Zensho stock price was nearly half of where it is today, meaning the all-in yield was closer to 3%, at a time when JGBs were yielding nothing. Today, the absolute all-in yield at current prices has contracted to just 1.5%, and JGB yields have clearly bottomed out (there is talk now that the BoJ may even start to taper later this year). Essentially then this sector, and this name in particular, has been, I think, a massive beneficiary of yield-hungry money looking for domestic ‘safe’ names with any kind of yield pickup, even small. This is one reason why I think now is the time to look at trades like this – the absolute level of total return available to retail shareholders is very poor, it has already contracted about as much as it can, and now interest rates may start to move the wrong way. This is all happening at a time when cost pressures on the business are at maximum intensity and the reported numbers should start to roll over.

But there is a secondary reason why now is the time. The yutai was, until recently, distributed in the form of paper coupons that could simply be cashed in at ‘ticket shops’ – since the coupons did not have the shareholder’s name printed on them, the return you received was almost as good as cash (clearly you have to accept some small discount to allow the ticket shop to make a return). However from June this year, key market trend-setter Skylark changed the way they issued yutai to its shareholders, by issuing shareholder club cards, made of plastic (just like an airline rewards card). Going forward, Skylark shareholders will be unable to transfer their yutai coupons to others. This change was made after many years because Skylark realized their discount coupons were being monetized by non-shareholders and, in an effort to increase the usability and representativeness of their customer data pool, they decided to issue yutai on a registered basis in this form. While this has not yet been adopted by Zensho (to my knowledge), since Skylark and Zensho are market leaders in the casual dining space in Japan, it stands to reason this could happen at some point in the near-term as well. Clearly this would massively decrease the retail incentive to own these shares.

This is really important because retail shareholders own, by my estimation, at least half the float. Indeed, outside of the founding family (around 41% total stake), there is next to zero institutional ownership (either foreign or domestic) – which makes sense in the context of current valuations. But it is also explains what could happen if or when the yutai discount gets issued in a non-transferable form, or when domestic shareholders wake up and realize that earnings are going down and will not come back for a while and that JGB yields may actually start to go back up. In this context, I could easily foresee the stock trading back to levels more supported by fundamentals and institutional ownership – which to my mind means 50%+ downside and back to 2015 levels, or lower.

Risks

- stronger yen = better purchasing power for imported food costs

- retail investors simply not caring when reported numbers roll over because they like 1% all in yield

- somehow they manage to pass through rising labour costs or replace their 61k employees with robots

Catalyst

Quarterly reports

JGB rates

Publication of change in the yutai system

Passage of time