To The Moon

$NVIDIA(NVDA)$

The "ChatGPT moment" for robotics is just around the corner—at least according to NVIDIA's CEO, who recently stated that the robotics industry could become the largest technology sector the world has ever seen. So, what happens when NVIDIA invests in a robotics company?

Today, we're diving into a robotics technology company in which NVIDIA holds shares—one that has surged 800% since last summer when news broke of NVIDIA’s investment. If you want to find out which publicly traded stocks NVIDIA holds, you can check their 13-F filings.

NVIDIA has invested in Serve Robotics stock

As confirmed by their SEC filings. The next filing will be released in February, so for now, we’re working with the most recent one from November. By comparing this to their first filing in February, we can see a few key changes: NVIDIA added Applied Digital (which we recently covered) and removed TuSimple after the company moved to China and shifted its focus. And, most notably, they added Serve Robotics, which we’re discussing today.

But here’s the question—why is NVIDIA’s investment in Serve Robotics only now appearing in filings when they’ve actually been involved with the company for a while? A 2022 article confirms that NVIDIA invested $10 million in Serve Robotics back then. Based on that, we can extrapolate share prices: $10 million for 2.676 million shares comes out to about $3.74 per share. Later, they added 62,500 shares—relatively insignificant. Most recently, they acquired just over a million shares at $2.42 each. Compare that to today’s trading price of $17 per share, and the math checks out.

Bottom line: NVIDIA increased its stake in Serve Robotics by 38% when a $2.5 million convertible note came due—essentially a rounding error for a company of NVIDIA’s size. Yet, how many analysts actually looked at the context behind this investment? Many articles covering it are just noise.

Could Serve Robotics Become the Next NVIDIA?

Motley Fool loves running that headline with any company name. Then there’s Seeking Alpha with "A Dirt-Cheap AI Stock With Huge Upside." But what does that even mean?

Let’s start with that “dirt-cheap” claim. When a company isn’t profitable, traditional price-to-earnings metrics don’t apply. Instead, analysts often use price-to-sales. We take it a step further with a Simple Valuation Ratio (SVR)—market cap divided by annualized revenue.

Annualizing revenue can be tricky when earnings fluctuate quarter to quarter, so we use the last 12 months. First off, we don’t even consider a company unless it has at least $10 million in annual revenue—that’s the minimum threshold to show some real traction. Around that level, we can start evaluating gross margins (not operating profits) to assess potential profitability.

For Serve Robotics, their trailing 12-month revenue is about $1.68 million. With a market cap of $917 million, that gives us an SVR of 546—even higher than quantum computing stocks, which are already at absurd valuations. For comparison, our catalog average SVR is 6, and anything above 18 is considered overpriced. Serve would need $50 million in revenue just to bring its SVR down to 18.

Right now, the stock is being hyped by people ignoring these fundamentals. But let’s examine the real investment thesis—its partnership with Uber, which is big news since Serve Robotics was originally spun out from Uber.

Uber And SERV Robotic

Serve Robotics spun out from Uber in 2021 as an independent company. This year, they plan to manufacture 2,000 robots for Uber Eats as part of a major expansion deal. However, they went public via a reverse merger, a process that skips the traditional IPO vetting, which is never a great sign.

They’ve also announced a partnership with Shake Shack, but let’s be real—partnership announcements alone don’t mean much. Until Serve Robotics actually executes and delivers results, these deals are just press releases. The companies signing these agreements have little to lose and all the leverage.

Looking at Uber, we recently analyzed whether autonomy is a threat or an opportunity for them. Uber clearly wants to leverage autonomous technology to improve its low-margin business, which is a key part of the Uber bull thesis. Whether they can successfully integrate autonomy remains to be seen, but they’ve partnered with multiple autonomous tech firms—including Serve Robotics.

So far, Serve has deployed an average of 48 robots actively completing deliveries—a small but noteworthy step in their journey.

SERV Robotic Progress So Far

According to the company’s earnings data, Serve Robotics had an average of 48 robots actively making deliveries per day during the summer quarter of last year. These robots were operational for about eight hours per day, a number expected to increase with their third-generation models.

However, one major issue stands out—negative gross profit. We’ll analyze that shortly.

Serve Robotics has three main revenue streams:

1. Software services – Licensing software to Magna, a major manufacturer also producing their robots.

2. Delivery services – Revenue from deliveries.

3. Branding fees – Potential advertising revenue from robot branding.

Now, let’s do some quick calculations. Dividing Serve Robotics’ Q3 2024 delivery services revenue by the total operating hours gives us $241 per hour—the same value as the previous quarter, which is oddly consistent. This suggests that their business model doesn’t charge per delivery but likely operates on a fixed-fee basis with one primary client (presumably Uber).

Serve plans to scale their robot fleet 33x. If the same revenue structure holds, this implies about $3.7 million in quarterly delivery revenue or roughly $15 million annually. However, the company projects that once all 2,000 robots under the Uber Eats deal reach full utilization, they’ll generate $60–80 million in annual revenue—four times our estimate.

The key question remains: How profitable is this model? It appears Uber is paying a fixed cost, but whether that leads to sustainable margins is yet to be seen.

Profitability?

So, essentially, it’s being subsidized for now. Serve will eventually need to transition to a variable cost model, and they’ve stated they can reduce that cost to around $1 per delivery. Looking at their other segments, the Software Services component seems somewhat variable, while branding revenue is questionable—how much value can really come from that? Time will tell.

Uber is betting on autonomy to improve its profit margins, but the question remains: how much of that revenue will they be willing to share with Serve, especially when Uber holds all the power at the negotiation table?

This whole scenario isn’t very compelling, particularly considering the level of dilution happening. In their most recent quarter, Serve announced expansion into new geographic areas, raising an additional $86 million—which they claim extends their operational runway through the end of next year. Of course, this comes at the expense of existing shareholders being diluted. The numbers show that shares outstanding have nearly doubled in just eight months, which is not a great sign.

Serve Acquire Vebu

Serve Robotics recently announced the acquisition of Vebu, a company behind the Autocado, an avocado processing robot. This raises a bit of a red flag. The latest news I found indicates that Chipotle has debuted the Autocado in a single store. Serve Robotics' reason for acquiring it is to expand beyond delivery into back-of-house restaurant automation.

However, this seems like a bad move. Why shift focus away from expanding your core delivery services, the very reason investors are backing you? Venturing into unrelated areas often leads to distractions, and history shows that this can be a warning sign. These types of pivots should only happen once a company has reached economies of scale.

An interesting note is that Buck Jordan, the founder and CEO of Vebu, is joining Serve Robotics. He was also an early investor in Serve. This makes me wonder if there are other reasons behind the acquisition beyond just seeking operational synergies.

The Last Mile Thesis

The last-mile delivery thesis still faces many hurdles. There are significant questions about whether this model will be both economically and socially viable. For example, think about deliveries to apartment buildings: when using food delivery services, many people would likely pay a bit extra for someone to take the food up in the elevator, knock on their door, and hand it over. Robots aren’t capable of doing that. And what about when these robots aren’t actively delivering—where do they stay?

Another issue is the potential for vandalism or disruption in areas with less oversight. Can robots safely navigate through neighborhoods where they may be at risk? And will restaurants be willing to adapt to this new delivery method?

For Uber Eats, there are three delivery options: leave at door, meet at door, and meet outside. The robot delivery option clearly falls under the "meet outside" category. Uber claims customers in regions served by robots can choose a robot delivery option, but after the initial novelty, will customers consistently opt for robots over human delivery, especially if there’s a price difference?

When you look at the regulatory challenges, there’s a lot of confusion. Supply Chain Dive discusses how states are struggling to decide how to regulate sidewalk delivery robots, which is where many of the policy challenges arise. While the technology may be available, it might not solve the social problems involved.

We’ve been tracking delivery robots for a while. I personally wrote about them back in 2018, observing Starship robots delivering food on the UC Berkeley campus, but the students weren’t overly impressed. Today, the leader in the space is probably Starship. They claim to be 99% autonomous, with millions of deliveries completed, far ahead of any competitor. Then you have Coco Robotics, which has completed over 300,000 deliveries since 2020, and Serve Robotics, which has delivered in the tens of thousands.

Starship has some impressive stats:

• Global service areas

• Focused on campuses and industrial sites

• 8 million miles driven, 7 million deliveries

• 150,000 daily road crossings

But even with those numbers, the technology still faces challenges. CNBC highlighted that while the technology is advancing, widespread adoption has been limited due to setbacks in implementing these fleets.

Looking at Amazon and FedEx, two major players, Wired reported in 2020 that both companies were pushing to get delivery robots onto sidewalks, but by 2023, Supply Chain Dive pointed out that they still hadn’t figured it out. The question remains: Will mass adoption ever happen? And that’s a key point to keep in mind as we consider the future of delivery robots.

Conclusion

Key questions for Serve Robotics investors, is the idea of last-mile robotic delivery economically viable? I’m not convinced yet, and I’m content to wait and see how this unfolds. Another question is, what’s the best form factor for these robots—should it be wheeled or a humanoid that moves like a human?

Then there are the margins. The last-mile delivery value chain already operates on razor-thin margins, so there likely won’t be 70–80% gross margins to be made here. What’s the cost-benefit of an autonomous car versus a smaller autonomous robot once self-driving vehicles are available?

The idea of someone handing you your food sounds great, but there are still too many unknowns to determine if this model is even feasible or economically viable. Perhaps it’s best to leave it to Uber to figure out the most economically viable option. As the industry giant, Uber has the capital, capabilities, and reputation to dominate.

SERV doen’t meet my investment check list, the The last-mile delivery model is still in the experimental phase, and there are significant questions around whether it will be economically sustainable. With already tight margins in the delivery industry, there’s little room for high-profit potential.

Disclaimer: I want to make it clear that I am not a financial advisor, and nothing I say is intended to be a recommendation to buy or sell any financial instrument. Additionally, it's important to remember that there are no guarantees or certainties in trading or investing, and you should never invest money that you can't afford to lose.

@Daily_Discussion @TigerPM @TigerObserver @Tiger_comments @TigerClub