Palantir Faces Skepticism Amidst AI Overvaluation Concerns

134

For months now, investors have been warned about the overhyped nature of enterprise AI software. Palantir Technologies has traded up to $20 on multiple occasions in the last year, but the company has lacked the actual growth metrics to warrant the stock rally. My investment thesis remains Bearish on the stock, trading at $16 and far above a realistic value for their growth rates.

Overhyped AI

While enterprise AI software companies from Palantir to C3.ai (AI) and analysts like Dan Ives of Wedbush routinely discuss the surging demand for AI and the strong use cases, corporations aren’t spending aggressively on enterprise AI software. Unlike OpenAI, Microsoft (MSFT) and other generative AI providers building the technology by spending billions on chips and cloud software, corporations are still only taking fliers on generative AI and LLMs while most enterprises seem uninterested in aggressively adding to the current IT budget.

The generative AI providers can spend billions to build the business of the future, while enterprises need a direct use case, and those takes time. Palantir is building the AIP platform to provide the tools to handle LLM functionality and reliably harness the technology via enterprise software to block out some of the hallucinations not viable for an enterprise especially a defense organization.

The odd part is that AI startup Anthropic is forecast to reach up to $1 billion in revenues in 2024. The company was only on the verge of $100 million in annualized revenues in 2023 showcasing how the companies building the technology are thriving.

Again, the problem here is that the developers of the generative AI tools and LLMs are raking in massive revenues while the platforms and AI tools being built to harness the powers of LLMs safely for enterprises aren’t driving the same surge in revenues. Or at least, the revenues haven’t flown into the analyst or corporate projections yet.

Palantir has built the AI Platform, or AIP, to create interest along with bootcamps to allow customers to get hands-on-keyboard interactions. Still, the company isn’t seeing the actual revenue flow to the revenue projections.

Alex Karp, CEO & Co-Founder, has the following quote on the corporate website:

AIP will allow customers to leverage the power of our existing machine learning technologies alongside…large language models, directly in our existing platforms.

As an example, the defence sector needs AIP to control the powers of LLM and AI in an ethical and legal environment with documented procedures. The platform provides security and guide rails for sensitive and classified data.

According to TechCrunch, the CIO of Principal Financial Group provided the prime scenario of why enterprise AI software deployment isn’t happening very fast. The financial services company have curated 25 use cases, but the company only plans to initially push 3 into production.

Be Patient

Palantir remains a strong enterprise data software company. The question is always about paying the right price for the stock, especially considering the company ran into growth challenges in the last year in the existing government and commercial businesses.

Jefferies just downgraded Palantir to underperform and slashed the price target to only $13, down from $18. The analyst doesn’t appear negative about the developing AI business, as much as the slowdown in the legacy commercial and especially government business.

Palantir reported Q3’23 revenue growth of only 17% and guided to Q4 growth in the 18% range. All the while, analyst like Jefferies aren’t coming up with surging revenue growth similar to the numbers reported for Anthropic.

Going back to the Q3’23 earnings call, management discussed a surge in demand as follows:

The potential market for AIP and the trajectory of possible AIP growth for our business is massive. We almost tripled the number of AIP users last quarter and nearly 300 distinct organizations have used AIP since our launch just five months ago. We will continue investing meaningfully in boot camps as our go-to-market strategy for AIP.

Palantir only lists 453 total customers at the end of Q3. The commercial clients are only at 330 suggesting AIP has already matched the traditional machine learning business, as far as customer counts.

The consensus analyst estimates forecasts Palantir will struggle to generate 20% revenue growth over the next couple of years. Anthropic is busy going from virtually no revenue to $1 billion within a year while Palantir is only forecast to add $0.44 billion in revenues during 2024 on top of a $2.2 billion revenue base.

Even these AIP customers leaving the boot camps excited about the product take time to convert into actual revenue, especially sizable amounts. The concerning part is that analysts like Jefferies aren’t seeing signs the business is turning these boot camps into material revenues.

The stock currently trades at over 13x the current 2024 revenue targets. As highlighted in previous research, Palantir needs revenue growth over 30% to come anywhere close to justifying the current multiple based on the $16 price target.

At the $13 price target of Jefferies, the stock gets more interesting. With 2.3 billion shares outstanding, Palantir would still be worth $30 billion with a 2024 revenue target of $2.66 billion.

The stock doesn’t get appealing to where investors can generate solid returns until falling towards $10 with a $23 billion market cap, while still trading at nearly 9x sales. Even this price could be aggressive for Palantir unless one truly believes sales growth will ramp up towards the 30% range.

Takeaway

The key investor takeaway is that Palantir is still far too expensive due to AI hype. Enterprise customers appear very interested in AI software, but the companies don’t appear very excited about adding to the current IT budget.

Investors need to wait for the stock to trade at a more reasonable valuation with an interest starting at $13, but ultimately Palantir needs to fall to $10 before one should get aggressive on the AI story.