Palantir Technologies (PLTR 1.44%) shares advanced 340% in 2024 as the data analytics company reported strong financial results amid growing demand for its artificial intelligence (AI) platform. That made it the best-performing member of the S&P 500 (^GSPC -0.22%) by a wide margin.

However, Palantir has tumbled 30% since peaking near $125 per share on February 18. Insider selling and possible Pentagon budget cuts have factored into the drawdown, as have concerns about how tariffs will impact the U.S. economy.

History says Palantir still has a long way to fall. Here's what investors should know.

History says Palantir could decline more than 70% from its current price

Wall Street has been sounding the alarm on Palantir's valuation for months. Last year, Gil Luria at D.A. Davidson said the stock traded at an "unprecedented premium" to other software companies. And Malik Ahmed Khan at Morningstar wrote, "We view the long-term growth implied in Palantir's current valuation as unrealistic."

Similarly, Gregg Moskowitz at Mizuho commented, "Valuation cannot and should not be irrelevant, and we find it increasingly difficult to justify Palantir's high multiple." In February, Brent Thill at Jefferies told CNBC, "We've never seen a multiple like this."

My own research led to the same conclusion. Palantir attained a peak price-to-sales multiple (P/S) of 107 on Feb. 18. I reviewed the valuations of more than 50 software stocks over the last 20 years, and only six achieved a P/S above 100 during that period. All of them eventually declined at least 65%, as detailed below:

• Bill Holdings traded at 103x sales on Sept. 8, 2021. The stock eventually declined 87% and remains 86% below its high today.
• Cloudflare traded at 114x sales on Nov. 18, 2021. The stock eventually declined 83% and remains 46% below its high today.
• SentinelOne traded at 106x sales on Sept. 16, 2021. The stock eventually declined 83% and remains 75% below its high today.
• Snowflake traded at 184x sales on Dec. 8, 2020. The stock eventually declined 73% and remains 61% below its high today.
• SoundHound AI traded at 111x sales on Dec. 26, 2024. The stock eventually declined 64% and remains 59% below its high today.
• Zoom Communications traded at 124x sales on Oct. 19, 2020. The stock eventually declined 90% and remains 87% below its high today.

To summarize, only six software stocks excluding Palantir (to my knowledge) have achieved valuations above 100x sales in the last two decades. Every single one eventually declined sharply, with an average peak-to-trough decline of 80%. None of them have yet hit a new high.

Past performance is never a guarantee of future results, but I can apply that information to Palantir to make an educated guess about what may happen with the stock. Specifically, Palantir peaked at 107x sales on Feb. 18, 2025. The share price was $125 at the time. It may eventually decline 80% to $25 per share if its performance aligns with the historical average.

Palantir stock has already declined 30% to $88 per share. If the stock does eventually fall to $25 per share, that leaves an implied downside of 71% in the future.

Image source: Getty Images.

Palantir is an excellent business but an absurdly expensive stock

CEO Alex Karp says Palantir's data analytics platforms more effectively operationalize artificial intelligence than any other software products on the market. Put differently, he thinks Palantir has a superior solution, as measured by its ability to help clients move AI projects from prototype to production, and several industry observers agree.

The International Data Corporation (IDC) recently ranked Palantir as the market leader in decision-intelligence software. And Forrester Research has recognized the company as a leader in machine learning platforms, awarding its Artificial Intelligence Platform (AIP) product higher scores than similar solutions from Alphabet, Amazon, and Microsoft.

That leaves Palantir well-positioned to capitalize on a large and growing market. IDC estimates AI platform sales will increase at 40% annually to reach $153 billion by 2028. But not even the best business is worth buying at any price. The market may get carried away sometimes, but valuations matter eventually.

Wall Street expects Palantir's adjusted earnings to increase 37% in 2025. That makes the current valuation of 215 times adjusted earnings look absurdly rich. Admittedly, Palantir beat the consensus earnings estimate by an average of 14% in the last six quarters, but the stock would still be extraordinarily expensive, even if that pattern continues.

Here's the bottom line: Palantir is an excellent business but also a very expensive stock despite trading 30% below its record high. I think investors tempted to buy the dip should wait for a better entry point. However, if you prefer to buy the stock today, keep your position very small.