Tesla Is the Most Undervalued Artificial Intelligence (AI) Stock Right Now, According to a Top Wall Street Analyst. I'm Not Sure I Agree.

Tesla (TSLA 0.62%) stock is down 12% in 2024, despite a raging bull market for the S&P 500 that has sent the index 23% higher this year. The company is grappling with sluggish electric vehicle (EV) sales amid growing competition and softening consumer demand across the industry.

But last Thursday, Tesla held its "We, Robot" event where it unveiled the long-awaited Cybercab robotaxi, which will be powered entirely by full-self driving (FSD) software. Autonomous driving could be a $1 trillion opportunity for the company over the long term, according to Dan Ives of Wedbush Securities.

For that reason, Ives recently told CNBC he thinks Tesla is the most undervalued artificial intelligence (AI) stock in the entire market today. Here's why I disagree.

Tesla's core business is going through a rough patch

Tesla delivered a record 1.8 million passenger EVs during 2023, which was a 38% increase from 2022. While it was a strong result, CEO Elon Musk has long said he wants to grow production 50% per year, on average, for the foreseeable future. That goal looks to be in jeopardy since deliveries aren't keeping pace.

In fact, Tesla only delivered 1.3 million EVs during the first three quarters of this year, which is down 2.3% year over year. In other words, deliveries are on track for an annual decline for the first time since the company launched its flagship Model S in 2011.

Demand appears to be softening across the EV industry right now as consumers opt for cheaper gas-powered vehicles amid tough economic conditions headlined by high interest rates. Plus, Tesla is facing a growing competitive threat, not only from legacy car manufacturers but from new, low-cost EV producers in countries like China.

Tesla intends to launch a low-cost model of its own next year, which could sell for as little as $25,000. That could help reignite the company's sales growth.

EV sales still account for 78% of Tesla's total revenue, so even though analysts like Dan Ives are focused on longer-term opportunities like FSD and the Cybercab, it's hard for investors to ignore the abrupt slowdown in the company's core business.

The Cybercab could be years away

Autonomous technologies like FSD are subsegments of AI. In fact, Tesla plans to spend $10 billion on data center infrastructure to train its FSD models this year alone. Those data centers will be filled with the very same graphics processing chips (GPUs) that are used to develop AI models like ChatGPT.

Dan Ives isn't the only one who thinks FSD is a trillion-dollar opportunity for Tesla. Cathie Wood's Ark Investment Management predicts the company will generate a whopping $1.2 trillion in annual revenue by 2029, with 63% attributable to FSD and the Cybercab.

Tesla is likely to monetize FSD in three primary ways:

1. Selling the software on a subscription basis to owners of Tesla's passenger vehicles.
2. Licensing the software to other car manufacturers for a fee.
3. Creating an autonomous ride-hailing network (like Uber), with Cybercabs hauling passengers 24 hours a day, seven days a week.

Tesla might also sell Cybercabs to the public for around $30,000 each, which means practically anybody could buy a fleet of them and start their own ride-hailing business.

Here's where things get murky. FSD is still in beta mode, which means Tesla customers can use it in their passenger EVs, but they must supervise it and be ready to take the wheel at all times. At the We, Robot event, Musk said the final, unsupervised version of FSD could be released next year, but he has been making that promise for the better part of a decade.

FSD will have to clear numerous regulatory hurdles before Cybercabs are allowed to roam freely on American streets, and since the vehicle is due to go into production by 2027 (according to Musk), Tesla has a very narrow window to get the software ready.

Several AI stocks look like a better value than Tesla right now

Tesla has generated $3.56 in trailing-12-month earnings per share, and based on its stock price of $217.80 as of this writing, it's trading at a price-to-earnings (P/E) ratio of 61.2. That's almost double the P/E ratio of the Nasdaq-100, which is currently 32.1.

It also means Tesla's P/E ratio is on par with Nvidia's. But there's one problem: analysts believe Nvidia is on track to grow its earnings 139% in its current fiscal year, whereas Tesla's earnings are forecast to shrink 27% in 2024. In other words, Tesla should be substantially cheaper than Nvidia, at least based on these near-term estimates, which implies potential downside for the EV stock.

Tesla is also significantly more expensive than other leading AI stocks, including Amazon, Microsoft, Apple, and Alphabet.

Data by YCharts.

In this context, it's hard to agree with Dan Ives' suggestion that Tesla is the most undervalued AI stock right now. The FSD software comes with regulatory risks, technological uncertainties, and even competitive threats from other companies like Waymo. As a result, nobody truly has enough information today to make an accurate forecast about its financial potential several years from now.

That's probably why Tesla stock sank 8% immediately following the Cybercab reveal. If the company can't find a way to excite investors soon, there is a risk its stock could further decline, bringing its valuation more in line with that of the Nasdaq-100 and other leading names in AI.