Palantir Technology (NASDAQ:PLTR) stock was a play on AI before it was even a “thing. Back in 2021, I wrote about it as the “God emperor of data analysis.” It seemed to have “cracked the code” for getting real insights from large databases.
Palantir built this capability while seeking military contracts, starting over a decade ago. It still has those contracts. But since the AI hype train left the station, it has tried a lot harder to get civilian business.
Technology Debt and PLTR Stock
The problem with this is that Palantir’s is a proprietary technology. Its systems are custom-built for customers like the Defense Department.
Military contracting is profitable, but it doesn’t deliver fast growth. Palantir’s proprietary and even political reputation can be a detriment when going after civilian contracts. Witness its problems with Britain’s NHS. When you live in a world of politics, contracts can disappear when the political winds change.
Then there’s the question of whether Palantir is AI at all.
Systems like ChatGPT are passive, waiting for users to pose questions to them. Palantir’s system is active, focused on directing military campaigns, hospital schedules and other systems. Much of the data Palantir uses also comes from machines, not people. The Foundry, as Palantir calls it, requires extensive set-up. Because the technology is proprietary, long-term value stays with Palantir, not the client.
This means PLTR stock may be more of a “machine Internet” company than a true AI play. Palantir expects revenue of $2.1 billion for all of 2023. That’s up just 10% from a year ago. There is net income, making it eligible for the S&P 500, but it’s marginal.
Should we be paying 17 times revenue for that?
Analysts Walk Away
There are indications analysts are souring on the Palantir story. Of 14 now following the stock at Tipranks, 6 are telling investors to sell it.
Stone Fox Capital issued a blunt warning recently. The hype from folks like Dan Ives of Wedbush is overdone, Mark Holder wrote. Palantir is getting minimum revenue from its “AI Platform,” which aims to give customers hands-on experience with its tools.
Coincodex, which calls itself an AI stock-picking platform, thinks Palantir could be worth $73 per share in a year. This is based on a bullish look at its new products, new contracts, and heightened demand. ChatGPT has given Palantir a likely price target of $20 to $40 this year.
Jefferies analyst Brent Thill, who uses actual intelligence, sees Palantir falling 20% this year. He sees its government business slowing down and little sign commercial contracts can make that up.
A bigger problem for the stock may be its high retail interest. Small investors own half the common. They could stampede as soon as Feb. 12, when Palantir is due to report its fourth quarter numbers.
The Bottom Line
Every technology, even the most important, goes through predictable cycles of boom, bust, and going beyond.
Buying the dip may work in the full market, but not always in individual stocks. Juniper Networks (NASDAQ:JNPR) still trades at less than one-quarter its year 2000 high. Citigroup (NYSE:C) is still 90% below its 2007 peak.
That’s not to say the Internet or residential real estate were bad ideas. Disappointment often follows technology hype, and not all companies survive. Even those that do can take years to get you back to even.
The point is, don’t buy the hype. Buy the numbers. Right now, my advice is to sell the numbers. They’re not good enough to get through the AI hype hangover.
As of this writing, Dana Blankenhorn did not hold (either directly or indirectly) any positions in the securities mentioned in this article. The opinions expressed in this article are those of the writer, subject to the InvestorPlace.com Publishing Guidelines