Description

ARCO was written up a couple of times in the past – please reread past posts.  The write-up will focus on the current virus situation and a differing take on valuation.  Here is link to the IR site – with the various presentations/financial filings.

https://www.arcosdorados.com/ir/

Investors should remember that their scorecard is not computed using Olympic-diving methods: Degree-of-difficulty doesn’t count.

I don't look to jump over 7-foot bars: I look around for 1-foot bars that I can step over.

Warren Buffett

In looking for investment ideas in one of the more challenging investment climates in years/decades, it does pay to remember the Oracle’s advice.  Arcos Dorados violates both rules in spades.  Flipping burgers and opening stores in demographically attractive markets (550mm+ people, 42%<25 years old) should be easy, correct?  Not when the jurisdictions include Venezuela, Argentina, Brazil and 17 other markets with varying degrees of exchange rate risk.  The company discusses attractive returns (new stores (20%+), reimaging existing stores (mid-teens), McCafes (~30%) and Dessert Centers (100%+)), but these returns are nearly impossible to verify from the outside.  And despite strong local currency improvements in nearly every market, dollar translated gains have been mercilessly chewed up by depreciating LATAM currencies.  And no matter how many stores the company opens, how high operating margins go, the stock is going to move up or down based upon the movement of the Brazilian Real…and this has moved in one direction over the past couple of years.  ARCO listed in the United States in 2011 at a time when US investors were clamoring for emerging market exposure (It really happened).  But, following Lava Jato and the jailing of an ex-president/numerous members of Congress and brutal economic downturn, it is fair to say that sentiment has moved180 degrees from the ~18x EV/EBITDA multiple days of 2011. 

So, this was the tale of woe prior to the coronavirus, and the stock reflected these concerns and then some.  Now, a new existential threat has emerged. While the situation is fluid, large parts of ARCO are shut down, with the exception being ~50-60 percent of Brazilian stores and the vast majority of Mexican restaurants as well as various smaller markets.  While the virus is beyond problematic in the United States, it could be potentially lethal in less prepared LATAM markets whose economies were already struggling prior to the crisis.

To avoid confusion, this write-up is advocating a long position versus a short.  Why? This is still a pretty good business, and the stock is insanely cheap – there is a credible case that the real estate alone is worth the entire enterprise value of the company.  Additionally, while ARCO’s main business is operating restaurants, there is a vastly underappreciated predictable annuity stream from sub-franchised locations. At the beginning of 2020, one could argue the real estate value and the sub-franchise stream (at conservative capitalization rates) were worth most of the enterprise value – now one could be getting both for free.  As noted, ARCO has had strong execution despite the horrible economic environment. If there is anything close to “normal” economic environment, ARCO is likely worth multiples of the current stock price.  

But, can ARCO survive the virus?  The answer is yes for a couple of reasons.  First, MCD has publicly adopted a “no franchisee left behind” mantra and ARCO is clearly among the most important, accounting for ~5% of total MCD royalty payments.  ARCO is also clearly one of the more important growth drivers for the company. The last 3-year agreement planned a more aggressive rollout adding ~100 stores annually for the next 3 years, albeit with substantial flexibility on timing/total amount.  Based on conversations with MCD team, MCD has a high degree of confidence in the ARCO management team and has zero desire to take over operations in the volatile LATAM region. It is highly likely that all royalty payments will be deferred as long as needed and MCD will likely agree to whatever tweaks are needed in the store build plan to accommodate ARCO.  

ARCO entered this crisis with substantial business momentum and a clear recent history of operating execution – February was the best month in the company’s history.  All results below are in US dollars and include the meaningful currency depreciation as well as the economic impact of the worst economic downturn in Brazil since the Great Depression.  Admittedly, the current downturn will likely be worse.  

In its largest market (Brazil), ARCO has outperformed Burger King in recent quarters after a prolonged period as a share donor.  ARCO has a larger base than BK, operates in multiple geographies, has more freestanding stores (capable of delivery or drive through – dine in options have generally been closed throughout the country since March 23) and is therefore arguably better positioned for the crisis than its largest competitor despite BK’s net cash position.  ARCO has made substantial progress on delivery, offering the service in 1,500 stores across 15 markets and noting that sales are incremental to existing stores. Its application has been downloaded 34 million times and allows far more efficient direct marketing to its targeted customers. ARCO’s substantial investment in self-severe kiosks has been a driver of improved comps in recent quarters and this investment payoff should continue when diners can return.  

ARCO’s leverage is 1.6x (versus a targeted range of 2-2.5x) as of 12/31/19 – the calculated ratio includes the derivative value (more on this in a second).  The company’s outstanding notes mature in 2023 and 2027, with 50 percent of the principal fully hedged at implied rates of 3.12-3.20 (versus ~5.27 rate today).  ARCO relies on local suppliers for 90 percent of its food and paper (F&P) expense. Around 20 percent of ARCO’s of total F&P costs are exposed to FX fluctuations (~16% in Brazil), and ARCO uses rolling hedges to cover ~50% of estimated exposure over the next 6 to 12 months ARCO (First half Brazilian expenses are hedged around 4 reals/$ versus a current spot of ~5.25 reals/$).  

The cross-currency hedges on interest/principal were put in place following ARCO’s debt scare back in 2014/2015 as the restaurant faced rapidly declining sales as Brazil entered the Lava Jato recession and overall debt levels were inflated following a more aggressive expansion.    The company got religion on cost cuts/productivity improvements following the 2014/2015 experience and enters the current crisis with a far better balance sheet/operational state. To gross simplify, ARCO had done little to optimize costs/productivity in the periods prior to the 2014/2015 downturn and instead relied on enormous comparable store sale gains driven by price increases – the downturn changed this.  It is also clear that the noticeable improvement in operating metrics generally occurred after Woods Staton stepped down as CEO in October 2015 and was replaced by Sergio Alonso. Substantial progress was made during Alonso’s tenure, but the improvements were masked by the currency depreciation. Marcelo Rabach replaced Alonso as CEO midday through 2019. Alonso was generally considered the best operator of the entire ARCO team, having previously led the Brazil division and then subsequently leading the rather remarkable turnaround in the NOLAD division.  The great hope has been that if the macro environment could stabilize, the dropdown to ARCO’s operating income would be substantial given all the cost cuts/productivity improvements.   While one can question some of the capital allocation decisions (for starters, kill the dividend – no rational investor owns ARCO for a yield and Woods can use other funds to buy his Miami Big Macs), the senior team is very competent from an operating perspective.  The executive team has almost exclusively worked at McDonalds their entire career and have produced strong results even in the most difficult of operating environments.  

For better or worse, the virus will not be ARCO’s first rodeo dealing with recessionary conditions and covenant violations.  During 2014-2016, ARCO was able to secure waivers on its debt with its banks and under its master franchise agreement with MCD – ARCO remains highly confident both will work with the company again when covenants are inevitably broken given the restaurant shutdown.  ARCO has $50mm of undrawn revolver capacity (JPM/BAC), and believes it potentially has access to substantial (possibly up to ~$100 million) uncommitted capacity.  

And then there is the real estate.  ARCO purchased multiple locations in Mexico, Brazil and Argentina when it first came to South America in the late 1970’s/early 1980’s.  ARCO monetized ~$120 million of real estate back in 2015/2016 during the severe economic downturn and actually stopped the sale process as it achieved targeted debt ratios sooner than anticipated.  At the time of its 2011 IPO, the real estate was appraised at ~$900 million. While the company did not publicly disclose the updated values in 2015, it indicated that the updates suggested meaningful appreciation since the IPO.  The company says that an assumed valuation of nearly $1 billion is very reasonable and all of the real estate is currently unencumbered. It is our understanding that the Board actually considered a proposal to monetize more of its real estate within the past 1-2 years but ultimately decided against a sale/leaseback proposal.  With a lagging share price, it would have seemed that this might have been reconsidered prior to the virus outbreak. Regardless, even assuming a discount of 20-50% and an LTV of ~40% and it seems that ARCO can get substantial financing from its real-estate holdings if needed.    

So how bad is the current cash burn rate?  To simplify, it appears that most of the country’s SLAD market and Caribbean markets are in quarantine – so no sales of any kind, not even delivery.  Roughly 60 percent of Brazil’s stores are open for delivery/drive through/takeout. Terrifyingly all but 20 of Mexico’s stores are open, and the country was actually comping up year-over-year in March – perhaps you’ve seen those videos of President AMLO hugging people?  The above is not fully accurate and is changing regularly - there are other countries open on a limited basis, but, let’s simplify. ARCO notes that roughly 20-30% of its costs are fixed and the rest variable so its cost structure offers a certain degree of flexibility during the crisis.  The company seems to view labor and SG&A as the main components of the fixed bucket but this “fixed” is a relative term and there is likely room to cut and/or possibly receive some government assistance in cutting/deferring. Occupancy expense would appear to be a “semi-fixed” cost but it appears that ARCO has made substantial progress in conversations with landlords and likely will not have to pay for mall based restaurants that are closed.   For the below analysis, we make multiple assumptions/simplifications:  

-Assume sales for the first 2 months of 2020 rose 7% but then suffer from 30% FX depreciation on translation.

-Brazil and Mexico stay open at current sales levels for 10 months.  Brazil sales volume averages 35% of 2019 productivity (typically 70 percent of sales are from dine-in customers – delivery should be able to offset some volume).  Mexico accounts for 60% of NOLAD sales and comps at 90% of 2019 levels. We then translate both Brazil/Mexico sales at existing exchange rates. 

-Assume all other operations are closed but fixed costs are in local currency which declines 30% y-y.

-Cut 10 percent of “fixed costs”

-Defers all royalty payments for one year

-Defers all capex for one year except $50mm maintenance capex 

-Does not pay taxes and working capital remains constant.  

-Interest expense down on the reduced swap payments

Obviously, every caveat applies about “assumption upon assumption.”  It is far from a pretty picture, but with $122 million of cash on hand, $50 million of revolver capacity with a path towards nearly $100 million more in commitments, $1 billion of unencumbered real estate and a rich uncle willing to pull all stops out to ensure survival, ARCO can likely last many months in absolutely brutal virus conditions. Obviously, the hope is that economies start opening up in the next couple of months and sales momentum builds in the back half of the year.  With ~1,377 free standing and in-store restaurants out of ~2,300 base, the company will be in a position to ramp up drive through/delivery in multiple markets as life (hopefully) returns to normal. Large portions of the above analysis are wrong as there is assumption upon assumption; the assumptions have been run by the company but they can only offer high level comments that essentially amount to “we are in a position of strength operationally and financially and have the unequivocal support of MCD – we will get through this.” 

ARCO’s stock should be able to quickly double from current levels if there is any sign of improvement in any of its markets.  In anything that would be considered a more normal environment the upside is far higher. Historically ARCO traded at roughly 7-8x EBITDA – if the company can get back to 90 percent of 2019 levels, this would imply a price of $8-9.  It’s worth noting that Burger King Brasil’s growth story had captivated investors as shares traded at 12x EV/EBITDA as recently as January of this year despite less favorable investor sentiment towards Brazil.  

Alternatively, if ARCO gets through the current crisis without monetizing its real estate, the upside is far higher.  As noted, the real-estate value is substantial. That said, we believe investors substantially underestimate the annuity stream coming from its sub-franchised restaurants, with the largest grouping in Brazil.  ARCO is typically paid 12% of sales by the sub-franchisee and then pays out 5% in rent, locking in margins of 7%. Subs-franchisees are also responsible for royalty fees, but ARCO is the guarantor should a sub-franchisee not make a payment.  Based on our understanding, there have been few defaults in ARCO’s history – this would also be a nice statistic for the company to publish. Generally, the subs are high-net worth Brazilians and there is a rigorous screening process to become a sub-franchise.  Any royalty relief granted by MCD will extend to sub-franchisees. While the margin is lower than owned locations, the stream of payments is consistent and likely more likely more highly valued by investors than the core operating income. Another way to look at ARCO’s valuation is to assume value for the real-estate, capitalize the “landlord” stream and then look at the implied EV/EBITDA of the core business (operating restaurants) after assuming some addback for rental expense on monetized real-estate.  At current prices, the exercise would look like the following.  

At $8 per share it would look like this:

On a GDP/store basis, ARCO has noted that it could more than double the current footprint over time.  At current bombed out prices, a precise estimate of upside is unnecessary as there are lots of ways to get to a substantially higher stock price.  

So what can go wrong?  An awful lot. The virus is terrifying – the US was unprepared for this.  LATAM is woefully underprepared in terms of healthcare coverage/availability, fiscal resources to combat economic weakness and testing levels.  If there is any good news here, it is that outside of Brazil and Mexico, many other LATAM countries have instituted hard quarantines/lockdowns. It is also possible the warmer weather provides some relief prior to treatment advances and (hopefully) a vaccine, although the Ecuador videos call this into question.  

It would take multiple pages to just summarize the macro challenges in each country.  There was progress on Brazilian pension reform during 2019, and after coming out of the worst downturns since the Great Depression, Brazil was expected to be one of the stronger economies within the LATAM region during 2020.  The virus has clearly changed this. While the real has already declined 30 percent year-to-date, emerging markets continue to suffer capital outflows and there easily could be additional downside. The dark joke about Brazil being the country of the future (and it always will be) has some strong historical wisdom/experience to support the view.  

Argentina accounted for ~16% of sales in 2018 and given the depreciation likely accounted for 10% or less of sales in 2019 (the 2019 20F has yet to be released – it is painful how slow it is for this document to be released).  Argentina has traditionally been a strong market for ARCO, but the country is in the middle of severe downturn and likely debt restructuring. Venezuela has been a basket case for years. That said, the Venezuela operations are self-funding, and ARCO has not had to inject any money into the country.  If there is good news, it is that ARCO has a clear history of pricing power – increasing checks at above the rate of inflation during good times and at or slightly below during downturns. During periods of economic stress, ARCO will discount to ensure the stability of traffic. The ubiquity of its phone application should allow more efficiency during this downturn as ARCO can better target its discounts towards likely shoppers. 

Without question, the F/X volatility could jeopardize returns – it certainly has hurt operations in the past and the macro headwinds could make this the wrong stock, in the wrong location at the wrong time.  But, the above valuation suggests this is substantially priced into shares. McDonalds is one of the most valuable brands in the world, and we think the exclusive franchise to this brand for 550+ million people with substantial growth potential is still an amazingly  valuable asset.  

Catalyst

Catalysts:

• Survival 

• Delist from the US and move to Brazil – there is little reason to stay listed in the US; the world has changed from 2011 and the Burger King experience suggests multiple expansion is more likely in a locally listed name.

• Stop listening to US based investors regarding the dividend…why hedge if you create naked dollar exposure?  Nobody is buying ARCO for a 0.something yield.  

• Make store returns more transparent – show returns on store openings by year opened and provide a bridge of local currency returns vs. dollar reported 

• Put out 20F before summer, release Q4 earnings prior to spring and make sure IR website actually works