Description

AUTO is a Norway-based automation company which we believe has broken hearts of many investors in the past 1.5 year period as persistent macro headwinds did not allow AUTO deliver steady top-line growth. We believe current stock price levels offer a very reasonable risk-return profile in a high-margin business which should benefit from the structurally non-automated world of warehouses.

While founded in mid-1990s, the real money-making business for AUTO began around 10 years ago as it took AUTO enormous amount of time to prove customers that their concept of “cubic storage” can solve a lot of storage/retrieving pain points. When all set and done, the private equity army has been constantly behind the door of AUTO’s office, offering money time and again – initially by EQT, then by THL and eventually by SoftBank.

Business overview

AUTO designs and sells their own, proprietary ASRS (automated storage and retrieval system). Simply speaking, ASRS are computer-controlled systems that enable automatic placing and retrieval of goods in the warehouse. AUTO’s solution, as mentioned above, is called “cubic storage”, while there exist a couple of other alternatives like Shuttle systems, Autonomous mobile robots (AMR) and Mini load. Considering that storage and picking operations are the largest OPEX parts in any warehouse, so far AUTO has been focused on solutions for this part of the warehousing process. Other parts of warehousing (loading/unloading of truck, depalletization, sorting, packaging) are out of scope so far. Reportedly, ASRS solutions typically enable around 50% cost reduction, with the majority coming from labour cost decrease and increased yield (less risk for injury, 0 exposure to “bad days” of an employee, and eventually “more for less in a given amount of time”).

In AUTO’s cubic storage, products are stored in containers/bins (those can carry up to 30kg of goods; bins can be 22cm, 33cm or 42.5cm tall). These containers are stacked on each other and eventually form a cube of up to 6m in height (and no matter how long or what form it takes). When certain goods are needed, robot(s) “walk” on top of the cube, finds the correct bin with the needed goods and delivers it to an operator in the warehouse. Thereafter, the good is sent to a customer. It is the densest solution on the market, meaning you don’t need to leave any aisles as robots move only on top of the grid. Here is a simple visualization:

This type of system cannot be fit for every type of product (think of the bin size and weight constraints), but it turns out that this type of solution resonates with a lot of well-known names, especially when you are an e-commerce player, for whom delivery time benchmarks are only becoming shorter. Here is a list of end-markets served by AUTO in 2023:

AutoStore makes business in a relatively asset-light manner – their core focus stays on product development and updates, design, assembly and leads generation. When it comes to project implementation and integration, so far AUTO has been doing it through distribution partners. There are 23 distribution partners, ow the largest ones are Swisslog (part of KUKA group), Demantic (part of KION group), Element Logic, Bastian Solutions (part of Toyota), Kardex and others. All required materials and parts of the robot are supplied to AutoStore’s assembly plants in Thailand and Poland. 2023 proved being „somewhat different“ in terms of sales channels as reportedly AUTO has generated ~30% of total sales from its’ own business development-led activities while the remaining 70% was the business from distribution partners. Previously, the ratio was around 10/90, but so far it early to say, in our opinion, whether AUTO is trying to dis-intermediate completely or it is a natural result of „slow macro“. We hope to hear more on this topic at analyst day held in September, where fresh vision on addressable market and product development will be revealed. Considering historically high margins in the business, we so far see rather little incentive for AUTO to dis-intermediate completely – eventually, the current distribution model allows AUTO not to pay substantial money for excessively large salesforce and to focus on what matters – product quality and development. At the same time, it is useful to keep in mind that AutoStore’s solution cannot be considered a one-stop-shop – i.e. when a large end-customer comes to distribution partner with the need for a complex solution (besides just ASRS system), they can get it, and enjoy the situation of „everything from one place“.

Eventually, highly standardized and modular (i.e. fitting any kind of warehouse, of any shape) product AUTO sells, and the distribution model they utilize, allows them to generate ~25% returns on capital. At the same time, the company claims that the average payback period for customer stays in between 1-3 years.

Current times

Currently, AUTO is living through the hangover in form of ongoing digestion of the post-covid e-commerce rapid growth (and certain over-investment in warehousing space). While previously AUTO showed explicitly their exposure to e-commerce (which was at ~70% of total revenues), now we can only guess where this number stays, but probably some suggestions can be drawn from the chart with end-markets shown above. Obviously, Apparel and 3PL, Other retail, are very closely related to e-commerce business. While currently it hurts in terms of revenue growth slowdown (as rather little greenfields are being added and customers’ decision-making slowed down), we have little doubt that long-term trends in terms of continuous money flow from offline to online retail will continue, and thus making AUTO proposition very relevant, as all of us buying stuff from online stores in general expect very short delivery times (and this trend is only intensifying).

Overall, AUTO was telling investors during IPO (back in 2021) that cubic storage part of the ASRS market will grow at 33% per annum until 2026. This projection can easily be incorrect considering how slowly 2023 was going on (revenues +10% yoy for AUTO) and for 2024 consensus is expecting at best flat year (first 6-month revenue growth was at -10% yoy, but projects execution is told to be year backloaded). Nevertheless, AUTO managed to prove that in times of high inflationary pressure and supply chain volatility, they were able to execute certain pricing actions last year (implemented aluminum surcharge and executed general price increases), thus recovering gross margin to the historical ~70% level and adj EBITDA to ~50% level (and those levels are considered long-term normal, according to the management).

Competition

Attabotics and Ocado are considered direct competitors, while technology-wise AMR solutions are probably the closest ones. Not much is known about Attabotics as the company stays private.

Ocado’s solution is mostly an exposure to e-grocery segment (at least it has been so far), and they implement a subscription-based model (which AutoStore has implemented as well in 2023 with the launch of “pay per pick” concept). As we understood, Ocado’s solution can be considered as a one-stop-shop for food retail (as they also have solutions beyond ASRS), but also it takes 1-2 years for the system to be installed. Eventually, Ocado as an investment is a tough call, in our opinion, considering that it consists of a lot of other businesses (like e-grocery JV with Marks & Spencer and others). On top of that, Ocado is still far from profitability point on a group level, not even mentioning the AUTO-like margins, while brokers are slowly anticipating Ocado to ask shareholders for another round of equity injection (management is not confirming that they will go this way).

When it comes to Ocado, then AUTO and Ocado share quite some past. For example, it all reportedly began back in 2012, when Ocado was a customer of AUTO. But already in 2020 companies were facing each other in court, with AutoStore’s filing of patent infringement cases across different geographies. Eventually, while hopes were high that AUTO may win something out of legal battles, the settlement agreement has been reached with AUTO agreeing to pay out ~250m USD to Ocado in instalments over 2 years. We were absolutely confused seeing that, but it may be a hint that when companies faced each other in court, it eventually turned out that AutoStore’s patents were worded too vaguely to win those legal battles. While money-wise 250m USD for AUTO is around 2-3 years or cash flow given away, the settlement statement is clear that all parties can continue to utilize technologies as before.

Shareholding structure

Softbank continues to be #1 shareholder with ~38% stake, followed by 28% of THL (who was selling at the beginning of the year a bit more than 5% shares outstanding). THL has made enough money on this investment already (when it sold part of its' stake to SoftBank back in 2021), so in theory creating an overhang risk. Ideally for these shareholders, we think, would be the scenario if someone comes to AUTO with an M&A proposal at an attractive price, especially taking into account that SoftBank is roughly 50% under water with their stake at the moment, while for THL it would take too long to sell the remaining 28% on the market.

Valuation

While thinking about the future, we see opportunities for AUTO to grow. A lot of warehouse spaces are still not automated and eventually companies will go back to expanding their warehousing space. Also, take it into account that e-comm penetration globally has not yet likely peaked. Where we differ from the company’s IPO view is that we do not expect the company to grow at ~30% per year, like they tried to pitch it previously. We land currently at ~10% revenue CAGR and also pencil in slow margin dillution going forward, factoring in presence of competitors (eventually adj EBITDA margin, according to our thinking, could land to 39% level by 2028 vs 48% seen in 2023 and ~52-54% seen in 2018-2021). Still, with plenty of opportunity to grow and a high-margin business nature, we can easily think on 15-20x EV/EBITDA for AUTO as being fair (though the higher-end rather being achievable in case they show at least a couple of ~15-20% years of growth without dilluting margins too much). Applying this range of multiples, we see the fair level for the stock pice at 20-25 NOK/share by end of 2028, implying 12-18% IRRs.

Catalyst

Improving market outlook

New product launches