Four questions I am forcing myself to answer before I buy Palantir. Steal them for your own tempting stock.

I am bullish on Palantir.

I know this price is not asking me to be right. It is asking me to be early.

Those are two different bets, and I keep pricing them as the same one. When I catch myself wanting something for a reason that has nothing to do with the price I am paying, I stop and write it down. This is that.

I run two systems. The Compounding Portfolio holds 8 to 9 businesses I would own for decades, bought at a price that respects the business. The Income Portfolio sells premium, never buys it, 45 days out, small and often. Palantir is not obviously either. The business might be a decades business. The price is not a decades price. That gap is exactly where good investors lose money on good companies.

This is what momentum addiction feels like from the inside. Not stupidity. A genuinely extraordinary growth story, still carrying the fresh-paint smell of a chart nobody has finished writing yet, arriving at exactly the price where my own discipline is supposed to say no.

So before I place the order, I am doing what I tell every founder in the Sales Gym to do before a big deal: find the person who will tell me no, and listen to them.

The stock, plainly

Picture the chart. Palantir hit an all-time high near $207.52 last November, at 329 times GAAP earnings and a market cap of $493.8B. It sits at $133.36 now, market cap $317.74B, down 36% from that high. P/E: 145.67. Price to sales: roughly 60x trailing revenue, 115% above its own ten-year median of 26x. Say $317.74B out loud. It does not get quieter the second time.

The growth underneath that multiple is not fake. Q1 2026 revenue hit $1.633B, up 85% year over year. US commercial revenue grew 133%. US government revenue grew 84%. Full-year guidance was raised to $7.65 to $7.66B, better than 71% growth, with a Rule of 40 score of 145, a number almost no software company at this scale has ever posted. This is a real platform, winning real deals, growing faster than almost anything else in public markets.

It is also priced like the growth is guaranteed to continue at this rate for years, and priced by a market that has been wrong about that exact assumption before, on this exact stock, twice in the last eighteen months.

Four things a skeptic would make me answer

The multiple is doing more work than the business. The valuation is a bet layered on top of the business, not a description of it. At 60x sales, Palantir needs years of continued 60%-plus growth just to grow into today's price, let alone earn a return from here. If the multiple simply reverts halfway back toward its own ten-year median, the stock is meaningfully lower with the business performing exactly as well as it is performing right now. Growth and return are not the same thing. I keep confusing them here specifically.

One budget cycle is nearly half the revenue. US government revenue was $687M of Q1's $1.633B, and Palantir's government business is concentrated in a small number of agencies inside Defense and the intelligence community. A continuing resolution fight, a change in procurement priorities, or a single large contract slipping a quarter would not be a rounding error here. It would show up in the number the whole valuation is built on.

The person who built it is selling on a schedule. CEO Alex Karp adopted a Rule 10b5-1 plan to sell up to 9,975,000 shares through September 12, 2026, on top of $4B already sold across 2024 and 2025, including 493,025 shares in February and 397,744 shares in May. Pre-scheduled sales are not automatically a red flag; RSU vesting requires them. Pre-scheduled sales at this pace, on this stock, at this multiple, are still the founder taking chips off the table faster than almost any other insider in the market is buying.

The moat is a bundle, and bundles get unbundled. Palantir's real edge is that Foundry, Gotham, and AIP work together as one accredited platform, not as separate features. Microsoft Fabric is already bundling AI orchestration directly into existing Azure enterprise agreements, the same unbundling playbook that hit Snowflake and Databricks once hyperscalers decided the data layer was worth subsidizing. Anduril is doing the same thing to the defense side, with software-native unit economics the legacy primes never had. The moat argument is real today. It has never once been permanent for a platform company at this size.

A sharper filter: what would Gavin Baker do here

Here is the test I run on every trade I am tempted by: find the investor most fluent in exactly this question, and see where they actually put sized money, not where they put opinions.

Gavin Baker runs Atreides Management and has spent real time in public comparing exactly this kind of growth story to what he considers the next tier up. I could not find an on-record Gavin Baker quote calling Palantir a buy or a pass at today's price, and I am not going to invent one.

What he has actually said is more useful than a hot take anyway. On Invest Like the Best, Baker pointed out that Anthropic added roughly $11B in annualized recurring revenue in a single month in early 2025, a number that exceeds what Palantir, Snowflake, and Databricks each took roughly a decade to build. He called it a discontinuity, not an acceleration, and used Palantir specifically as his benchmark for what "the old way of building an extraordinary SaaS business" looks like.

That is not a verdict on Palantir. It is a pattern worth sitting with: the sharpest public voice on this exact category uses Palantir as the standard for great, decade-long, patient compounding, not as the standard for the next discontinuity. If I am buying it hoping for a discontinuity, I may be pointed at the wrong layer of the story.

Five scenarios, three to six months out

| Scenario | Price target | Move from $133.36 | What has to happen | |---|---|---|---| | Super Bull | $230 | +73% | Blowout Q2 print (early August), a marquee new commercial mega-contract, Rule of 40 expands further, short covering accelerates the move | | Bull | $190 | +42% | US commercial growth holds above 100%, no government budget scare, stock re-rates toward the ~$190 analyst consensus | | Base | $145 | +9% | Growth continues near current pace, multiple stays rich but stable, stock chops in a wide range | | Bear | $95 | -29% | A government contract slips or a commercial growth deceleration surprises, multiple compresses toward 35-40x sales | | Super Bear | $65 | -51% | A broad AI-software repricing event hits every "story stock" at once, multiple resets toward the low-20s on sales, 52-week low support fails |

What I am actually going to do

Every investing group chat has someone who bought this under $100 and will not stop mentioning it. That is not a reason to buy it at $133.

The platform is real, the growth is real, and watching it compound without me would sting. Pretending otherwise would be its own kind of lie.

What changes is the size and the entry. This does not go in the Compounding Portfolio as a full position, because it fails the price-respects-the-business test used for every name in it at today's multiple. If it gets taken at all, it is a small, named speculation, sized so the Super Bear scenario does not touch anything that matters, entered after the Q2 print in early August gives one more real data point on whether commercial growth is holding, not before.

The bitter truth first: it is better to own a tenth of a position entered with clear eyes than a full position bought on a feeling that cannot be named. Discipline needs the same grip on a growth story: firm enough to hold a real position, never tight enough to choke a business that might still compound for years.

Conviction that a stock can keep compounding is not the same conviction that today's price is a good place to start. Being honest about which one you actually have is the whole exercise.

If you are wrestling with the same itch on your own tempting stock, run it through the same four questions before you place the order: how much of the price is the multiple versus the business, how concentrated is the revenue in one decision-maker's hands, are the people with the best information buying or selling on a schedule, and does the moat survive the hyperscaler unbundling playbook that has already hit two comparable companies. The stock will still be there in the morning. Your capital only gets one shot at compounding correctly.

$50K did not become $4M+ by winning big on stories like this one. It became that by asking these four questions every time a good story showed up wearing a rich price tag.

I built the Compounding Portfolio and Income Portfolio systems inside the Sprint Club to stop exactly this: paying an extraordinary price for an ordinary entry point. Come run your own tempting stock through the same four questions: strategysprints.com

Happy hunting. Simon & The Sprinters 🐬⚡️🐆