Description

Symbotic is currently the best performing SPAC of this entire cycle trading at $30, a 3x from its SPAC completion in June 2022, despite ~85% redemptions when the deal closed. For comparison, the median 2022 SPAC is down ~80% – meaning Symbotic trades at $30 while the median of its SPAC cohort trades at $2. The stock is up >100% since March 2nd alone when The Bear Cave published a brief short report, though there has been no other fundamental news. This is obviously not the short thesis, but it is what catches the eye and makes you ask whether Softbank found thee diamond in the rough, or whether there is an opportunity here. 

This is a real business, but it is a misclassified business with a relatively thin float and June lock-up expiration looming. It is a $17B company trading for 16-17x FY23 revenue (FYE 9/30) and ~80x FY23 gross profit. They remain Adjusted EBITDA negative with 20% gross margins.  

The warehouse automation company is more of a construction business than a technology business – with 20% gross margins, a cost-plus revenue model, and 80-90% of revenue coming from Walmart. They work with steel and machines, not really with software, and they do long-term construction-oriented projects, not deploy zero-marginal cost software. The promise of long-term higher-margin software contracts after the systems installation is appealing, but not evident in the data given hundreds of millions in past Systems revenue but only $3.7M in software revenue in FY22. 

Symbotic was created within C&S Wholesale Grocers and is led by CEO Rick Cohen, the 70-year old decabillionaire owner of C&S. Symbotic hired Michael Loparco as CEO in April 2022, though he resigned in November 2022 - and Rick Cohen reassumed the CEO role. What’s notable about Loparco is that he came from Jabil, a $12B EV manufacturing services company that trades at 5x EBITDA, presumably because his experience in services was highly relevant to Symbotic.  

Additionally, there is extremely high revenue visibility – their Walmart agreement (cancelable by Walmart) runs through 2030 – which makes Symbotic a nice business at the right price but also one that is relatively easier to value than most SPACs. At a ~$17B market cap, the risk / reward here is extremely attractive with several near-term catalysts looming. 

The borrow is not great (20%, locate varies), but there are options available and I think this is an idea worth preparing for when opportunity arises in the next 1-3 months. There is a retail element, and a thin-float element, so while there are several catalysts over the next 6 weeks, this could be volatile. I generously estimate fair value at ~$14 (50% downside) and realistically think the business is worth $4 - $8 per share (70%+ downside).   

I am going to cover several topics here:

• Product and Market

• Relationship with Walmart

• Valuation

• Shareholders and Timeline

• Risks

Product and Market

Warehouse automation is a broad term that has different meanings to different customers. This 10 minute video about Amazon’s warehouses highlights several key points: 1) “warehouse automation” encompasses a number of different products in different settings, and 2) each warehouse clearly has different needs, emphasizing that there is no off-the-shelf warehouse automation solution. 

We can break down the market into a 2x2: point solution vs. systems, and cases vs. items. Cases vs. Items is sometimes also classified as distribution centers vs. e-commerce fulfillment, or pallets versus individual SKUs. 

Point solutions are hardware products that serve a specific purpose, such as conveyor belts, robotic arms, or Roomba-esque robots. The Berkshire Grey (recent ~$400M take-private by Softbank post-SPAC) product page highlights some of these products – Pick and Pack, Pick and Sort, Mobile Robots. The Amazon video highlights some of their Kiva products (~$500M acquisition in 2012). 

A system is just an integrated collection of point solutions, meaning the point solution vs. system distinction is almost more a procurement and operations question. Do you buy the hardware and installation services from a few vendors, or do you buy from a single vendor that offers integration? 

Symbotic says their differentiation is that they provide a system with some third-party hardware, some internally developed hardware, some externally developed hardware with Symbotic IP, and the software and operations sheen to make it all work together. From my checks, Symbotic is really the only player selling a system for cases, or a system for larger regional distribution centers. This Symbotic video shows the difference – they’re working with pallets in large wholesale distribution centers. Symbotic was initially built within C&S Wholesale Grocers and so was tailored to wholesale distribution needs. 

You can also see in the image below from their materials that this is a steel-heavy business and requires detailed construction. Some of the hardware is proprietary (two main products are the autonomous robots, and the palletizing arms), though not especially unique. The “software” is not software in the classic sense – it’s just required for any of the system to work and makes up a tiny portion of revenue. 

Their filings list several other competitors who are a combination of hardware and service providers. A description of some of those competitors is below from the Goldman initiation report. 

These are mostly un-sexy, low-multiple businesses. Dematic was acquired in 2016 for $3.3B by KION Group, a company that today has a total market cap under $5B. Intelligrated was also acquired in 2016 by Honeywell, for $1.5B. 

Relationship with Walmart

Symbotic has worked with Walmart since 2015. In 2021, they expanded their partnership to deploy 88 “modules” (think of this as a measure of scale) in 25 Walmart regional distribution centers. In 2022, they expanded the agreement to 188 modules in all of Walmart’s 42 regional distribution centers. It’s unclear how many modules there are in those 42 centers (e.g. could 188 go to 200? 300?), but 42 is 100% of Walmart’s US RDCs. Through past investment and warrant issuances, Walmart owns approximately 60M shares of Symbotic - a little under $2B. Note that the S-1 shows ~60M, and CapIQ shows 15M. I think the difference is the various share classes and that 60M is the correct number. 

In Symbotic’s FY22 (FYE Sep 30), Walmart comprised 94.4% of revenue, or ~$518M of the $549M.  

What’s also interesting is we know Customers B and C are likely two of Albertson’s, Target, or Giant Tiger – their only other customers (other than C&S, where revenue is de minimis). We know that Customer B did $41M in FY20, $71M in FY21, and less than $50M in FY22. We know Customer C did $39M in FY20, and less than $23M in FY21. This revenue is obviously not recurring and the bull case of long-term software contracts does not appear to be real – ~$350M in revenue in FY20+FY21 translated into $3.7M in FY22 software revenue. 

Back to Walmart – Walmart is responsible for the vast majority of the existing backlog that should be recognized through the 2020s. Importantly, Walmart is able to cancel the agreement, meaning this backlog is not a guarantee. Walmart featured Symbotic in their most recent investor day presentations – along with several other automation companies – which probably contributed to the recent run-up in the stock. 

There are several ways you could interpret the Walmart agreement. If you’re bullish, this is the ultimate validation. This deal, combined with the IPO, should bring more attention to Symbotic and ultimately more business. And if Symbotic is successful in its new product development, Walmart might expand the agreement even more. 

If you’re bearish, like me, you see the risk of extreme customer concentration, particularly given Walmart has the right to cancel, and you recognize that Symbotic is essentially a Walmart contractor, not a high-multiple software business. Additionally, being a Walmart vendor is a notoriously challenging position and Walmart is certainly aware of the power they hold over Symbotic. They will get what they need at whatever cost, and they are unlikely to offer tons of margin to Symbotic. 

Valuation

The high visibility from the Walmart contract actually makes the business less challenging to model than most growth companies. They have disclosed several key inputs about the model:

• They are focused on existing customers and “not looking to really grow backlog”

  • They mentioned maybe adding 1 or 2 customers per year

  • This is because deploying Walmart’s systems is still an enormous challenge and resource drain

• They have three business segments: Systems (>95% of revenue, mostly hardware and installations), Software and Support (1%), and Operations Services (4%)

  • Operations and Services is not intended to generate gross profit

  • Software is higher margin over time (they’ve said 60%), though has been between $2.5M and $4M over the last 3 fiscal years and is only a MSD % of backlog

  • Systems is currently in the 20s% gross margin (fluctuates) and they’ve indicated that should push towards 30%

• Their growth strategy assumes they eventually get into new verticals (e.g. auto parts, food) and new geographies 

  • This can be accounted for in backlog growth (1-2 customers per year)

  • Currently 5 customers

• New products (e.g. Breakpack, a solution to address the “Items” section of the 2x2 above) are hard to model, though obviously offer upside

  • One former was not enthusiastic about new products or software as they repeatedly emphasized the DNA of this company is in warehouse installation and operations

    • Software and new robotics technologies are a different skill set than construction and C&S Wholesale Grocers experience (related, but not quite the same)

So let’s build up the valuation. Let’s first waterfall the Walmart backlog (starting 9/30/23) over the next ~7 years. I am being generous to them in most assumptions and provide scenarios at the end that are more bearish. 

Then we look at new customers assuming they book almost $2B of backlog in FY23. They closed a deal with United Natural Foods in September and we only know it was less than $900M. Let’s give them credit for 2 of these per year, and be generous on the sizing of them, and then grow that amount by 15% per year (sensitivities at end). 

Let’s roll it all together and include GAAP OpEx, which they say will only grow a little bit (though I think MSD is generous again by me). 

Note that I am materially above consensus for FY23, FY24, and FY25. 

Assume 10% cost of equity and 20% tax rate. We get a generous $14.5 per share, and several scenarios where the equity is much less (exit PE multiple on the left): 

So to summarize, a $14.50 price today requires:

• Successful on-time completion of the Walmart contract with gross margins expanding from ~20% today to 42% by 2030

• The non-Walmart business grows from $30M in FY22 to ~$3B in FY30, also with 42% gross margins

• GAAP Opex CAGRs through 2030 at 6% while Revenue CAGRs at 22%

• And/or material contribution from new products and services

• An exit multiple of 15x for a then >$4B, MSD growth business 

There are many ways the stock is worth materially less, including:

• Walmart cancellations

• Non-Walmart business grows slower than the 100x in 8 years that I’m modeling

• Gross margins stay down near 20%, or perhaps 30% – not the 42% that I’m showing

• Opex grows much faster than 5% per year. Even bringing Opex growth in-line with Revenue growth chops about $10 off the present value per share

• Faster than forecast share dilution

Shareholders and Timeline

There are approximately 550M shares currently outstanding (excluding RSUs). The CEO, Rick Cohen, owns approximately 77% and is locked-up until around June 7th. Walmart owns about 60M shares, or ~11%. And Softbank (at IPO) owned about 30M shares, or ~5%. Walmart and Softbank unlocked at the 6 month mark, though it’s not clear yet what (if any) they’ve sold. 

On the Q1 earnings call in January, CEO Rick Cohen was asked if he would try to expand the float so that institutional shareholders could buy in. He indicated that they would likely raise equity and had no intention of keeping the company predominantly owned by himself, Walmart, and Softbank. This was when the stock was around $14, or >50% below today’s price. They do not need the cash, but he pointed out they would have plenty of use for it.  

Symbotic reports earnings on Monday, May 1. Given their high visibility, I would not expect earnings to disappoint or be a driver of downside in the near future. They could however announce an equity offering in conjunction with earnings. They also are holding an analyst day on May 18th, which will provide more clarity into long-term targets. And then Rick Cohen unlocks on June 7th. 

This strikes me as a company with very little fundamental upside given the ~$17B market cap and 80x 2023 gross profit multiple, and huge fundamental downside. The opportunity exists due to a thin float, though that very well could change over the next 6 weeks. I would expect volatility until we get a larger share float, and likely a slow bleed after that event down to a more reasonable valuation. 

Risks

1. Walmart expands contract - either internationally or with new not-yet-released Symbotic products

   1. This is possible, though it’s not obvious how much fundamental value this drives 

   2. The current agreement runs through their entire US RDC footprint and through 2030; there’s not much to add and the present value of any 2030+ 25% margin commitment is very low

2. They land large customers at a rapid rate

   1. They’ve indicated they are not trying to do this

   2. The Walmart contract prohibits them from working with one unnamed retailer, presumably Amazon or maybe Target (a current or perhaps former customer)

      1. Present value of multi-year $500M - $1B agreements is very low. $100M - $200M of gross profit over 2-4 years

3. Software revenues do ramp and contribute a material 60% margin revenue stream

   1. I do expect software revenues to ramp, though they only kick in once systems are delivered and the software contracts then run (supposedly) for 15 years

      1. Software is a MSD portion of the backlog (let’s say $1B) spread out mostly from ~2026-early 2040s, meaning the Walmart software portion is maybe $70M in revenue per year and ~$40M in gross profit per year (again, far into the future)

         1. The PV of this stream is most generously $1B, unless we start to see much faster ramp this year (Q1 ending Dec ‘22 was $1.2M in software rev)

4. Breakpack (e.g. “Items”) or other technologies significantly broaden the TAM

   1. Possible, but this is probably more than accounted for in the price

      1. What would you pay for a venture-backed, pre-product, pre-revenue robotic arm company working on sorting individual SKUs? 

Catalyst

Earnings on May 1

Analyst day on May 18th

77% of the float unlocks on June 7th