The short version: opinions up, share price down
Palantir shares fell more than 7% on Friday, October 16, 2026, closing at $186.03, a drop of $14.47.1 It was the fifth losing session in a row, and the stock had hit a 52-week high of $212.94 earlier the same week.1 The fall came during a broader technology sell-off.1 That is the same week that analysts were publicly turning more positive on the company.1
For an ordinary investor, the lesson is that a stock can fall on the day its analysts get more optimistic. A price target or a rating is an opinion, not a measurement. A large part of this piece is about who said what, and how much weight each source deserves.
What the Palantir bulls actually said
Bernard Zambonin said he remains constructive on Palantir. He called customer expansion and net dollar retention central to the bull case.1 Net dollar retention is a simple idea: it measures whether existing customers spend more over time. He also acknowledged that the elevated forward valuation makes continued execution essential.1 In plain English, the shares are priced for things to keep going well.
Ryan Canady upgraded Palantir to Buy. He concluded that customer-built AI applications can remain complementary to Palantir's platforms rather than displace its core offering.1 This addresses a common worry that customers will build their own AI tools and no longer need Palantir. It is a judgement about the future, not a reported result.
The dossier names two further analysts in this wave, Fiorillo and Zaccardi, but gives no verbatim claim from either. We therefore do not attribute any specific argument to them.
A run of partnerships
Over a few days in mid-October, Palantir was linked to several new ecosystem announcements. These come from our event record, not from a named news article:
- On October 13, Ondas and Palantir announced Regna AI. It is a standalone private company meant to help governments, businesses and institutions develop and deploy AI models in sovereign environments, with Palantir expected to provide model-training infrastructure and engineering.2
- On October 14, Customers Bank announced a collaboration with Palantir to develop an AI-oriented banking operating model. Its forward-looking statements highlight implementation risks, reliability concerns and contractual flexibility.3
- Also on October 14, Surf Air Mobility signed Flying Zebra as its fifth commercial OperatorOS customer, with the software powered by Palantir. Surf Air receives a percentage of revenue from trips managed through the platform.4
The Customers Bank wording deserves attention. A partnership announcement is not revenue. The bank's own risk language on reliability and contractual flexibility suggests that terms may not be locked in. The dossier gives no contract values for any of these deals, so we cannot say what they are worth to Palantir.
The other side of the tape: defense
While Palantir collected upgrades, J.P. Morgan was describing a tougher picture for aerospace and defense. In an October 8 research report, lead aerospace and defense analyst Seth M. Seifman said most companies should deliver solid third-quarter results and express confidence in underlying demand. He added that good earnings may not be enough to reverse the sector's recent stock-market weakness.5 The bank cited slowing airline traffic growth, uncertainty over U.S. military spending and lingering supply-chain constraints.5
J.P. Morgan downgraded Leidos. Seifman's reasoning was that Leidos' relatively low valuation is not enough to justify an Overweight rating when other aerospace and defense stocks also offer substantial potential returns.5 That is a relative call, not a claim that Leidos is a bad business. Our entity data also records BigBear.ai as a competitor of Leidos, though the dossier does not say how that bears on this downgrade.
Seifman's note was not uniformly negative. He expected Howmet's 2026 earnings estimates to rise and its management to remain confident about growth in aerospace and industrial gas turbines.5 The weakness is in sentiment about the sector, not an across-the-board view that every company is in trouble.
Is this really a divergence?
Our narrative tracking flags it as one: Palantir's AI-platform story is improving in analyst commentary while defense sentiment sours. We present that as our own framing, not an established fact. Palantir sells to governments as well as businesses, and the dossier does not break down its government exposure. The 7% fall happened because of a broad tech sell-off, not a defense-specific event.1 So the two stories may be less separate than the headline suggests, and the data here cannot settle it.
A verified-data benchmark: how one bank's profits have moved
The checked, SEC-sourced numbers in our dossier are about Goldman Sachs, not Palantir. We include them because Goldman is also an institution whose analysts issue ratings. In the same period, one Goldman analyst, Andrea Newkirk, resumed coverage of Iovance Biotherapeutics with a buy rating and a $15 price target.6 For context on the firm behind such calls:
- Earnings per share were $22.87 in fiscal 2023, $40.54 in 2024 and $51.32 in 2025.7
- Second-quarter earnings per share were $10.91 in 2025 and $20.98 in 2026, so the figure came in at about 1.9 times the year-earlier level.7
- Capital expenditure was $2.316 billion in fiscal 2023, $2.091 billion in 2024 and $2.064 billion in 2025.7
- Cash was $164.259 billion at fiscal-year-end 2025 and $187.266 billion in the second quarter of 2026.7
These figures say nothing about whether Palantir is a good buy. A bank's strong earnings do not make its analysts right. A track record of correct calls would be the relevant evidence, and the dossier does not contain one.
How much to trust the sources
Our measured reliability scores cover the news feed that supplied the Palantir and Leidos pieces. In both cases, 58% of 927 checked claims from that feed held up against their sources.1,5 That is a middling record. It is why we lean on named analysts and the original wording of their statements rather than on the articles' framing. The Goldman figures above are in a different category, because they are canonical and checked against SEC filings.7 The Iovance item comes from a source where only 26% of 23 checked claims held up, a small sample with a weak result, so treat that rating as lightly supported.6
What to watch
- The November 3, 2026 congressional midterm elections. Our event record notes they could change congressional control and constrain military spending growth, and that some uncertainty may dissipate after that date.8
- Whether Palantir's bulls are proved right on customer expansion and net dollar retention. Zambonin tied the case directly to those measures, so the next reported figures are the real test.1
- Whether the October partnerships turn into disclosed revenue. So far the dossier shows announcements and risk language, not amounts.3
- Whether defense third-quarter results confirm Seifman's view that solid earnings may not be enough to lift the shares.5
