Why Is Palantir Down Today When Its Revenue Is Growing at 70%? The Answer Is Complicated

Over the past two months, the query “why is Palantir down today” has come up in search results so frequently that it has begun to feel more like a persistent grievance than a question. Despite quarterly results that most businesses would trade a lot to produce—70% revenue growth, 50% adjusted operating margins, record sales quarters, and a product suite that is, by any reasonable measure, deeply embedded in the organizations paying for it—investors who purchased at $130, $120, or even $100 have watched the stock decline. In any case, the stock is down. Additionally, the causes are constantly changing, which contributes to the situation’s genuine difficulty.

Anthropic was the most direct cause of the most recent leg down. A product based on multi-agent AI orchestration, which enables AI systems to plan and carry out complex workflows across multiple steps without human intervention, was released in early April by the private AI company, which is estimated to be worth $380 billion and is not publicly tradable. The market responded quickly and widely. On April 10, Palantir closed at its lowest level in more than a month after falling more than 7% in a single session. Around the same time, CrowdStrike, Adobe, Salesforce, and ServiceNow all reached 52-week lows.

Wedbush analysts dubbed it the “Anthropic fear trade” because they believed that if AI could automate enterprise workflows from the ground up, fewer traditional software subscriptions would be bought, fewer seats would be renewed, and businesses like Palantir would face structural challenges to the recurring revenue assumptions built into their valuations. That reasoning is not wholly irrational. Furthermore, it’s not totally proven. The market was the first to move, with questions to come later.

An already-burning fire was fueled by the Michael Burry element. In a post, Burry—who gained notoriety and wealth by betting against mortgage-backed securities prior to the 2008 financial crisis and has since established a reputation for being a bearish contrarian, sometimes right and sometimes a few years ahead of schedule—argued that Anthropic posed a significant threat to Palantir’s enterprise market share. Later on, the post was removed, which tends to increase rather than decrease its shelf life in contemporary markets. During the company’s most recent earnings call, Alex Karp, the CEO of Palantir, specifically named Burry and accused him of manipulating the market. For a CEO of a Fortune 500 company, that is not a typical move. Karp’s sense of obligation to act indicates that the impact was significant enough to call for a reaction.

Burry misinterprets what Palantir actually does, according to the counterargument put forth by Palantir bulls, some of them quite vehemently on Reddit threads that have taken on the feel of small community meetings. His criticism appeared to assume that having a large language model is what makes Palantir valuable. It doesn’t. The core of Palantir’s product is ontology and operational decision-making, which involves taking an organization’s disorganized, fragmented data and creating a framework on top of it that enables automated, auditable decisions to be made at scale. AI capability is added to that foundation by the AIP layer. That is obviously not replaced by an Anthropic multi-agent orchestration product. Depending on how businesses decide to implement it, it may enhance it, compete at the periphery, or cause no disruption at all. As usual, the market did not wait to find out.

The discrepancy between sentiment and operating performance is what sets Palantir apart from other growth stocks. When a company’s stock drops 37% from its peak, it usually indicates that something is wrong with the business, such as slowing growth, margin compression, or customer churn. None of those are present at Palantir. The expansion is quickening. The margins are growing. Contracts are long-term and sticky, and the clientele is expanding. The issue is valuation, which was priced at levels that required everything to work out for a long time and is now burdened with concerns about AI disruption that may or may not be warranted. The stock typically moves more quickly than the fundamentals support in either direction when you’re priced for perfection and uncertainty enters the room.

Even though ARK’s timing record has been inconsistent, Cathie Wood’s ARK Invest purchased shares during the decline, which is at least a sign of institutional conviction at lower prices. Before the worst of the selloff, Peter Thiel sold 2 million shares in March for about $140, but founder sales of that nature are typical and anticipated. Some interpret this as a sign that those closest to the company aren’t freaking out over the fundamentals because no unusual insider selling has been discovered during the April decline itself.

On April 13 and 14, the fear trade began to wind down. Palantir, Oracle, ServiceNow, and other companies all made significant recoveries in just two sessions after Wedbush’s Dan Ives declared the software selloff to be overdone. Whether or not actual enterprise spending data over the upcoming quarters supports or contradicts the Anthropic disruption thesis will determine whether that bounce continues. Which way that goes is still unknown. There are no indications that demand is declining, according to the company. The stock price presents an alternative narrative. Both of these things are real, and the argument presently resides in the space between them.