The model in one paragraph
For each business line we take today's revenue (last 12 months, or the latest quarter or run-rate if you choose), grow it for 10 years at your two growth rates (years 1-5 and 6-10), and move its operating margin in a straight line from today's level to your year-10 target. "Bets" (robotaxis, Optimus, orbital compute and so on) ramp in from zero to your year-10 revenue, multiplied by the chance you give them. We tax operating profit (default 21%), put your price/earnings multiple on year-10 profit to get the company's value in 10 years, discount that back to today at your required return (risk-free rate plus a risk premium), shrink it for new shares issued (dilution) and add net cash. That is your value today.
Facts, assumptions and opinions
Everything on Stonkometer carries one of three labels. Fact: numbers and statements from SEC filings or named reports, linked on each page. Assumption: a default we chose so the model can run (rates, margins, scenario values); you can change every one in Advanced mode. Our view: our own judgement, kept to a minimum and always marked.
Defaults
| Setting | Default | Type | Where it comes from |
|---|---|---|---|
| Currency | Nominal US$ | Fact | All four companies report in US dollars, so there is no FX conversion. |
| Inflation | Included in other inputs | Assumption | Growth rates and the risk-free rate are nominal. |
| Risk-free rate | 5% | Assumption | the 10-year US Treasury yield was 4.99% on 26 Oct 2026, rounded. |
| Risk premium, SpaceX | 5% | Assumption | Our choice. |
| Risk premium, Tesla | 5% | Assumption | Our choice. |
| Risk premium, Anthropic | 6% | Assumption | Our choice, higher for a young private company. |
| Risk premium, OpenAI | 6% | Assumption | Our choice, higher for a young private company. |
| Tax on profits | 21% | Assumption | US federal corporate rate; effective rates differ. |
| Dilution, SpaceX | 1.5% a year | Assumption | Our estimate of stock pay and fundraising. |
| Dilution, Tesla | 1% a year | Assumption | Our estimate of stock pay and fundraising. |
| Dilution, Anthropic | 2.5% a year | Assumption | Our estimate of stock pay and fundraising. |
| Dilution, OpenAI | 3% a year | Assumption | Our estimate of stock pay and fundraising. |
| Horizon | 10 years | Assumption | Fixed. |
| Magic revenue margin | 30% | Assumption | Our choice. |
| Moonshot exit multiple | 20x | Our view | A mature-business multiple, applied once a moonshot starts paying. |
| Today's margins by business line | see each company | Assumption | Our estimates from the latest segment results. |
| Bear / middle / bull values, bets | see each company | Assumption | Our starting points, not forecasts. |
The needle
Your value divided by today's market value (or last private round): under 0.6 is Bubble, 0.6-0.9 Pricey, 0.9-1.1 Fair, 1.1-1.5 Cheap, above 1.5 Bargain.
Market's view
We search for the extra growth, added to every business line in our middle case, that makes the model equal today's price. It answers "what is the market assuming?" without anyone having to guess.
Magic revenue (aliens will buy it)
Some valuations can't be explained by any business line you can name. The magic revenue slider adds unexplained revenue in year 10 at a 30% margin, and the aliens line tells you how much of it today's price needs on top of your numbers. For Anthropic and OpenAI it is called "the AI will figure it out". If you use it, it shows up in your exports and share links, so everyone can see the leap of faith.
Moonshots
Moonshots are markets that may never exist, valued like options: chance x the company's yearly profit if it happens x a 20x multiple, discounted from the year it starts paying. Switched-on moonshots add their value today to your model, on top of the 10-year projection. All moonshot numbers are our view.
Implied yearly return
If you paid today's price and your year-10 value came true, this is the yearly return you would earn (ignoring dividends). Compare it with the return you require.
Scenarios
Assumption Bear, middle and bull are Stonkometer's own starting points, not forecasts and not anyone's consensus. CEO and company cases are our reading of public targets (for example Tesla's 2025 CEO Performance Award); each says where it comes from.
Limits
- One exit multiple hides a lot: real valuations use cash flows, reinvestment and risk that differ by business.
- Private-company figures (OpenAI especially) are self-reported and unaudited.
- Run-rate revenue annualises a short window and can overstate the year.
- Prices can be delayed. Nothing here is a recommendation.
Glossary
- P/E (price / earnings)
- How many dollars investors pay for $1 of yearly profit. 20x is typical for a mature company; fast growers get more.
- Operating margin
- Profit from the business before interest and tax, as a share of revenue.
- Required return (discount rate)
- The yearly return you want for the risk. A higher number makes future profits worth less today.
- Dilution
- New shares from stock pay or fundraising, which shrink existing owners' slice.
- TAM
- Total addressable market: everything everyone could sell in a market. It is a ceiling, not a forecast.
- Run-rate
- A recent period's revenue scaled to a year. Anthropic's uses the last 28 days x 13 plus subscriptions x 12.
Not investment advice. Stonkometer is an educational calculator, not investment advice. Valuations depend entirely on the assumptions you choose. We are not a broker or adviser, and nothing here is a recommendation to buy or sell any security. Do your own research and consider a licensed adviser. Full disclaimer