ARBE stock: buy it now before the definition of the next 12 months
Robotic TREE (NASDAQ: ARBE), a 2021 SPAC working on automotive RADAR systems, is facing what could be the crucial 12 months of its lifespan. If next year goes as planned, it could be a billion dollar business. Anything less, and it could be just another busted SPAC. Several major automotive players will be making purchasing decisions over the next 12 months that will tie them to products for 3-5 years and ARBE is on the shortlist of companies totaling $7.5 billion in procurement contracts. sale.
I first wrote about Arbe in January and described them as a high return, low risk bet. This article builds on the first, and it may be helpful to read the first article to fully understand what Arbe does.
Many startups have recently entered the automatic speed camera market, often described as millimeter speed cameras; start-ups from China, Korea, Israel and the United States have entered what appears to be a lucrative market. However, Arbe was the first and is leading the technical challenge as its products will enter mass production in early 2023.
ARBE has developed a high-resolution imaging radar system on a chip that it is trying to sell to automakers as a competitor to Lidar and an enhancement to existing automotive radar. (My first article on ARBE covers the product in detail) ARBE hopes its chips will be a crucial part in the drive towards autonomous driving. Arbe’s advantage is its intellectual property. Its flagship product has 48 transmitter x 48 receiver channels with a resolution of 1 mm at 1 degree. It can detect objects of all sizes (and their speed) up to 300m away, giving a full 360 degree point cloud that ARBE’s software can interpret. Its maximum range is 800m for objects. Being RADAR rather than LIDAR (it uses radio waves, not light waves), it works in all weather conditions; unlike LIDAR, it is unaffected by fog, snow, rain and sunlight. RADAR can also see through many objects giving information such as:-
“a dog is running behind this fence towards the road,”
“Visibility is only 10m due to snow, but there is a wrecked car 25m ahead of us,”
“a cat is sitting behind this trash can.”
RADAR is not as accurate as LIDAR; however, the Arbe radar is accurate to 1mm, which is sufficient for all automotive applications. The increase in information from the Arbe solution includes speed at the pixel level.
The ARBE product looks like a significant improvement over existing automotive RADAR systems from companies like Infineon (OTCQX: IFNNY), NXP (NXPI) and Calterah. It also appears to be better than other offerings from startups pursuing the same market, of which Uhnder appears to be closest to the market.
Uhnder, a Texas-based startup, is developing a 4D digital radar product. They collaborate with Magna (MGA) and have already signed Fisker (FSR) for their product. Full specifications are not still available on the site, and it is difficult to be sure of the product. Uhnder uses terms like “virtual channels”, of which there are over 100 as the product appears to be in development, it’s hard to be sure what exactly it will offer.
ARBE is unique in its approach, as it develops its high resolution 4D RADAR chip in collaboration with GLOBALFOUNDRIES (GFS) based on the 22nm GFs architecture. ARBE’s chipset can decode the 4D point cloud (position + speed for each point in the cloud) and transmit this information to the car’s main information system. Arbe also develops the algorithms needed to interpret the information received.
The automotive radar market was around $20 billion in 2020 and growing at a CAGR of 10%. Current car radar helps with collision avoidance, cruise control, park assist and some blind spot detection.
The LIDAR market is expected to reach $3.4 billion in 2026, growing at a CAGR of over 20%.
Arbe’s 4D radar system will target both of these markets with its products capable of fulfilling both roles simultaneously.
A Crunch 12 months
Now is the critical moment for ARBE and its competitors; Innoviz (INVZ) stole a march on the rest with its recent significant design wins (as predicted in this article, Lidar Review leading to Innoviz’s buy recommendation. Kobi Marenko (CEO of ARBE) said during the earnings call after the second quarter results that between September and March, they expect ten decisions from different automotive companies with which Arbe has completed its tests; each decision represents revenue between $30 and $100 million a year. As soon as he said that, the CFO jumped in, saying those numbers were identified customers and they thought revenue would grow. Arbe always guided toward $300 million in revenue. per year, requiring three big wins out of the ten they bid on.
Five analysts follow Arbe and, using their forecasts (all of which assume Arbe will meet its $300 million revenue target), SimplyWall.st has produced the following fair value calculation from a cash flow model discounted cash.
This fair value is a little higher than my model predicts with the given turnover. Arbe guided +60%.
Late in the conference call, analysts asked management about the ten expected contract decisions and the CFO’s intervention, saying things could be a lot bigger. CEO, Kobi, said if you add all the RFPs to the RFPs, they’re bidding on $7.5 billion in revenue between 2023 and 2029. His seriousness was hard to judge when he said that he expected to win 95% of those bids. I suspect he was honest.
Adjusting the financial model for 95% of the $7.5 billion revenue results in a stock price well over $300, it’s hard to estimate the costs of such a huge ramp-up without more information .
Where are they now?
GlobalFoundries is an essential part of the ARBE sales pitch. A customer received the first large shipment directly from GF in the last quarter. A second customer will receive a delivery during the current quarter. GF tests the chips in the factory, but this verification is not yet automated and must be carried out before the start of mass production. There is also a problem with the automatic approval of grades; however, this is about to be resolved by GF. ARBE plans to move to full mass production in early 2023. All potential ARBE customers have spoken directly to GF to ensure the product is of the correct standard and that GF supports ARBE’s sales plan . During the earnings call, Kobi said potential Tier 1 customers have production guaranteed by GF and that GF has established a designated area dedicated to production and sampling of the ARBE chipset within their factory. GF’s commitment is significant; Global automakers are used to dealing with GF by trusting them, without GF ARBE being just a small Israeli startup.
HIRain, the first Tier 1 ADAS provider in China, announced that he is undertake projects with major Chinese OEMs with imaging radar based on the ARBEs chipset. HIRain expects to reach full mass production in 2023. HIRain has been shortlisted for several high profile tenders with major Chinese manufacturers.
Tree has a confirmed customer, AutoX, an Asian delivery vehicle maker with big backers, AutoX is making a delivery robot using Arbe Radar which is heading for mass production in 2023. When launching Arbe, they made a big deal out of this 400,000 unit deal however, in the latest earnings call it was described as a small customer. It is an acknowledgment of how Arbe has progressed.
Arbe product range
In the last quarter, Arbe added a second product to its flagship offer, the Lynx, launched in July; It is a 360 degree long range radar system with 24 transmits x 12 receivers. The product complements the flagship Phoenix 48 Transmission x 48 Receiver radar, and the two devices can work in sync to provide full coverage around the vehicle. It comes with the same perception abilities using the same AI and has 4th degree of speed as standard. Arbe expects Lynx to be used in the back and corners of a vehicle, with any individual vehicle being able to have a Phoenix and 5 Lynx plus one or two backup radars on board.
Lynx can be used for forward radar on driver assistance rather than fully autonomous vehicles.
Arbe products are very inexpensive compared to competitive LIDAR, we don’t have an exact price, but a Phoenix RADAR should cost a tenth of the cost of an equivalent LIDAR and the Lynx much less than a Phoenix. LIDAR is falling in price, but is unlikely to compete with RADAR in the foreseeable future.
The competitive situation looks good for Arbe, great products, low costs and high customer engagement.
Do they have the money to generate income?
As with many of these recent SPACs, finances are strong. The balance sheet summary shows $75 million in cash and no debt. Cash burn in the last quarter was close to $10 million (vs. $6 million in the first quarter); they have a cash trail of about seven quarters.
The company’s forecast calls for EBITDA of -34 to -38 million dollars this year. The guidance is in line with my Arbe model and suggests they will have ample liquidity as long as they deliver orders this year.
This is Arbe’s moment, they have a product that looks competitive and superior to the competition. However, I am not a car manufacturer and can only rely on published information; we need to see confirmed orders to be sure.
Many potential customers will make their final decisions over the next 12 months. Wallstreet analysts expect Arbe to win many of these orders and give the following revenue forecast.
Arbe suggested he could win orders for 95% of $7.5 billion
I first bought ARBE at $8.51. I added in April at $6.64 and again last month at $5.80. My average cost is $6.71.
I’m going to hold on for the next 12 months and hope to see orders and revenue coming in.
I recommend buying Arbe now.