GROCERY OUTLET HLDNG CORP (GO) GO S
January 30, 2023 - 11:05pm EST by
pathbska
2023 2024
Price: 30.01 EPS 1.15 1.32
Shares Out. (in M): 100 P/E 26 23
Market Cap (in $M): 3,005 P/FCF 0 0
Net Debt (in $M): 294 EBIT 0 0
TEV (in $M): 3,300 TEV/EBIT 0 0
Borrow Cost: General Collateral

Sign up for free guest access to view investment idea with a 45 days delay.

Description

We and our affiliates are short Grocery Outlet. We may buy or sell shares without notification. This is not a recommendation to buy or sell shares.

Grocery Outlet has been written up twice as a long idea on VIC but we find it to be a compelling short at current levels and want to offer a different perspective on the key thesis points.  

For more background on the business, we would point you to the prior GO write-ups linked below. Both outline some of the key factors of how GO’s model differs from the typical grocery store concept.

 https://valueinvestorsclub.com/idea/Grocery_Outlet_Holding_Corp/1598515561

https://valueinvestorsclub.com/idea/GROCERY_OUTLET_HLDNG_CORP/7301124626

 We will focus the write-up on where our thinking and work differs from the bull narrative.

 Bear thesis

 1.    GO’s target of +10% unit CAGRs (and 4,800 total unit potential – i.e. 10x today’s number) is unrealistic given there are already multiple warning signs around IO new unit economics. To the extent that GO scales back its unit growth ambitions and/or cuts its annual growth targets, the TAM story that bulls love risks becoming broken and will likely result in a significant rerating of the stock (trades for 15x EBITDA; 30x EPS on 2022 consensus numbers today)

Both prior VIC write-ups reference “store growth potential” and/or “long runway for [unit] growth” as key tenets of the GO long thesis and why the stock deserves to trade at premium multiple similar to other high unit growth / large potential TAM retail names. By contrast, our work suggests there is already evidence that GO’s store TAM will prove to be materially overstated and the growth targets will eventually be cut.

Our diligence suggests GO’s recent IO cohorts have struggled to make an acceptable ROI on their stores. There is evidence that GO’s systemwide AUVs are becoming increasingly bifurcated between “legacy core” markets and “new” markets, where legacy market AUVs remain healthy while newer store AUVs are increasingly challenged. We can see signs of this in GO’s new-store-productivity (NSP) metric trends and we have also spoken with several IOs who have described the wide performance differential between top quintile and bottom quintile stores, with a disproportionate amount of the newer stores falling into the latter category. We have seen this dynamic play out several times in the past with high unit growth stories and it is almost always a red flag around for how scalable the concept will prove to be. We would note that GO (while private) already has a history of exiting certain markets that may not be well known (there were stores in CO, AZ, and even Hawaii at one point). We find it to be exceptionally notable here given GO is still relatively early in its self-described growth phase with <500 stores primarily in just 3 states.

A second way to look at this would be the unit economics for a potential new IO. The long-term scalability of the GO concept and the ultimate store TAM will be dependent on the concept’s ability to attract new IOs with compelling unit economics. Based on our work, this already seems questionable. Multiple former IOs have outlined 4-wall math for their stores that supports the idea that newer stores have struggled to earn a sufficient ROI to keep their IO’s engaged (admittedly, the example given below will have adverse selection bias but the mere fact that multiple IOs have decided to leave the system already despite GO suggesting they are still in the very early stages of unit expansion suggest there are problems here).

Nonetheless, here is a rough outline of the unit economic math we have heard from former IOs that squares reasonably well with some of the 3rd party research we have read. IO with stores that do AUVs in the low/mid $7mm range (which we believe is disproportionately true for the newer cohorts) are not making enough to justify the 90-100 work weeks they will put into the stores during the early years (and potentially even longer), especially when you consider that most IOs are husband/wife teams (i.e. this pre-tax profit is often household income).

3rd party channel checks in S. California (a newer growth market for GO but one that is very competitive) found that many IOs in that market were marginally profitable and were being supported by GO’s “Temporary Commission Adjustment Program” (TCAP) which allows IOs at underperforming stores to receive a temporary higher commission split in exchange for a future receivable due to GO. [We won’t go into a deep dive on GO’s TCAP program in this write-up but we’d encourage you to review GO’s disclosures in the 10K/10Q around how TCAP works, what have been the trends in loan balances, etc. Suffice to say, we believe the growth in TCAP loan balances directly relates to the unit economic challenges for new IOs in the GO system). One IO interviewed by a 3rd party research firm suggested that 30-40 of the GO stores in S. California (out of 91 total) were likely being supported by TCAP. We cross-checked this assumption with a former GO executive and our own IO checks and found it to be reasonable. We have separately heard that potentially as many as 20% of total stores carry a TCAP loan. Those interviews also suggested that some IOs have been willing to stick with marginal stores in exchange for an implicit/explicit promise from GO that they will be allowed rights of first refusal to buy the stores of more successful legacy IOs when they retire and/or to move to a “better” location as GO continues to expand. It goes without saying that this strategy will become increasingly problematic as the absolute number of stores grows – there won’t be enough “good” stores to go around at some point.


As further evidence of this unit TAM skepticism, consider that GO established its initial East Coast market presence back in 2012 via its acquisition of Amelia’s, a regional discount chain with ~15 stores in eastern Pennsylvania at the time. In the subsequent 10 years, GO has only managed to add +11 net new stores to its East Coast footprint (5.6% CAGR). If GO had expanded at just its overall target level (i.e. +10% new units per year), there should have been ~40 stores by now. You’d normally expect a new market expansion to allow for above-average growth given the white space opportunity available but in this case, GO only managed to expand at half its targeted rate over a 10 year period (i.e. not a small sample size)

 

2.      GO’s unique business model has important structural limitations

While GO is sometimes described as having an attractive franchisor-type business model deserving of a premium multiple, we would note that comparison is flawed. Unlike the asset-lite franchisor models most investors will think of, GO is responsible for funding the build out of its store footprint, not the IOs. This means store growth is highly capital intensive and FCF conversion is quite low.

A second limitation vs. a more traditional franchisor model is that GO’s IOs are typically only allowed to own/operate a single location. This is supposed to keep the IO laser-focused on running the best store they can, but it presents a herculean challenge to scale given the need to effectively recruit new IOs for each new store GO wants to open on a 1:1 basis. While GO references the large pipeline of interested IOs, we have heard the recruiting and training effort for qualified IOs is already challenging. As the store base grows, and GO is forced to find new markets to enter, we think the task of maintaining +10% CAGRs will be much harder for GO than true franchisor peers who allow franchisee groups to own/operate clusters of stores in a given area.

A third limitation is GO’s sourcing model itself.  As referenced in prior write-ups, GO differentiates itself from traditional grocers in that roughly ~50% of their inventory is acquired at deep discounts via close-out buys from branded CPGs. The close-outs are mostly well-known national brands but often relate to products near expiry date, packaging changes, and/or low velocity SKUs that were under-performing at mainstream grocery for whatever reason. As a result of passing along some of the associated savings with these deeply discounted buys, GO claims it can offer its shoppers savings of 40%+ vs. traditional grocery stores on a similar basket of items (although a GO store will carry a more limited assortment with far fewer SKUs that a typical mainstream grocer). This all sounds great until you learn that GO already acquires ~30% of all close-out CPG inventory (as one former executive told us). CPGs are certainly not looking for ways to structurally grow close-out inventory, so as GO continues to scale its store base, we think they will quite literally run out of product to source in this advantaged manner.

 

3.      Bulls that are anchored to GO’s historical SSS trends may be (negatively) surprised in 2023. Unlike most of its history, we believe GO’s strong 2022 comps were mostly driven by food inflation – this will not be sustainable going forward and will make for much tougher comparisons in 2023+

While GO has not consistently broken out the components of SSS historically (although the SEC sent them a letter in August 2022 requesting they do so; something the company promised it will start to do in 2023 … https://www.sec.gov/Archives/edgar/data/1771515/000000000022008737/filename1.pdf ), we can estimate a reasonable proxy for basket growth based on BLS data for food-at-home CPI. As seen in the charts below, historical average food-at-home inflation of +LSD spiked to +MSD/+HSD in late 2021 and was running closer to low/mid-teens during the summer months. GO’s reported SSS trends improved in tandem with this.

With food inflation moderating, and even flipping to deflation in certain areas, we think this SSS tailwind will disappear and GO’s comp will be more tied to underlying traffic in 2023. Bulls believe GO should benefit from trade-down dynamics in the same way that it did in 2008/2009, but we would note that most of the foot traffic estimation data we’ve seen suggests that 2022/2023 is not playing out the same way and traffic benefits at discount stores like GO or the dollar stores is tracking below the way it did during the GFC.


Why Does this Opportunity Exist?

1. Unit growth/TAM story that bulls love (UBS headline: “GO white-space opportunity is unique in food retail”)

2.   Anticipated beneficiary of trade-down (perception as “safe-haven” name)

3.  Mistaken perception that GO is effectively a franchisor model and should be comped against other franchisors

We would also point out a history of insider selling with zero insider buying activity since the IPO.

 


Risk/reward

The most recent VIC write-up from earlier this month described GO as trading at a “sensible valuation” not far below the current stock price – we would beg to differ and call it an expensive stock trading at significant premium to most grocery & retail peers (26x 2023 consensus EPS / 13x consensus EBITDA) due its perception of having one of the better unit growth stories/TAMs in retail. We think Street estimates are too high and the multiple can re-rate significantly if we’re right about the unit TAM cut and/or SSS deceleration.

Assuming we’re correct on the structural problems with GO’s scalability, we think the stock could easily re-rate to <15x EPS which on our numbers would get us to a $14 type stock, or 50%+ downside. If we’re wrong and GO maintains a premium multiple, we see 30x EPS or a $36 stock (~20-25% upside risk) over the next 12 months.

 

 

I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise hold a material investment in the issuer's securities.

Catalyst

Slowing comps, continued evidence of NSP challenges, continued unit growth below 10% target

    show   sort by    
      Back to top