Value investing, made visible
This works out the second number. Every company is read from its own filings, scored on the quality of the business, and valued with a discounted cash flow — so you can see the gap between the two and argue with it.
Then it keeps score of you: what you hold, what you decided and why, and a rank that moves when the businesses change rather than when the market has a bad week.
And you are not doing it alone: publish a call with a reason behind it, and it is scored against the field rather than against zero. Being up while the market was up more is not being right, and the board says so.
Your book stays on this machine — no account, no sign-up, nothing about what you own leaves it. A call is the one thing you can publish, and it carries a ticker, a direction and your reasoning. Nothing else.
What you do here
Two of these are the reason value investing is slow work: find what a business is worth, then hold the position long enough for being right to pay. The other two are the reason it does not have to be lonely.
Every company is read from its own filings, scored on the quality of the business, and valued with a discounted cash flow. What you get back is the gap between that number and the price — and every threshold that produced it.
Positions, what you paid, and the reason you wrote down at the time. Your rank moves when the businesses you own get better or worse — not when the market has a bad week.
Publish a call with a thesis and it is scored against the field rather than against zero. Up 10% while the market did 20% is a call that lost, and the board says so.
A shared book, argued over before anything is bought. This one is designed and not built: it is the first feature that makes one person's positions visible to another, and that is not a seam to get wrong quietly.
Why this exists
Value investing is one idea worked out carefully: a share is a claim on the cash a business will produce, that claim is worth something calculable, and the market quotes a price for it every day that is often not the same number. Everything in this platform exists to make that comparison explicit enough to argue with — and then to slow you down long enough to act on it.
The gap between what you pay and what the business is worth. It is not optimism insurance — it is the admission that your estimate has an error bar. Benjamin Graham's original point, and the reason this platform asks for 30% below intrinsic value before it calls anything a buy: a discounted cash flow wide enough to need a margin is a discounted cash flow honest enough to admit it.
What the business is worth on its own cash flows, not what someone will pay for it tomorrow. Here that is a full discounted cash flow with a bottom-up cost of capital, revenue-weighted country risk, R&D capitalised, leases treated as debt, and terminal growth capped at the risk-free rate — because a company cannot outgrow the economy forever.
Two questions, kept on two axes. A great business at a silly price and a mediocre one at a giveaway price are different mistakes, and averaging them into a single score hides both answers. That separation is what makes this a matrix rather than a checklist.
Run the valuation backwards and solve for the growth the current price already assumes. It turns a share price into a sentence you can disagree with — "the market is asking for 25% a year" — which is far more useful than another opinion about what the shares are worth.
The model will score anything you point it at. Whether you understand how the business makes money is a judgement only you can make, so the platform asks you to record it — and lets you put a company in a too-hard pile as a decision, rather than meeting it again on every screen.
A number the filings do not carry is absent, never zero and never a fail. Coverage is reported alongside every score, because a data outage and a bad business should never look the same — and a screener that confuses them will hand you the wrong answer with total confidence.
Every threshold this platform uses is written down with the argument for it, and rendered in full on the model page — including the sixteen places it departs from the original hand-written rules, and why.
The model
A good business and a good price are different questions, and averaging them hides both answers. So they are kept on separate axes and the company lands somewhere in the field — which is the whole reason this is called a matrix. Top right is the only corner worth owning: a business that scores well, at a price below what the cash flows say it is worth.
Returns on capital, growth that is funded rather than bought, margins against the company's own industry. What the thing earns, before anyone mentions a share price.
Debt it cannot service, accounting that flatters, dilution, cash that never arrives. Six of these are fatal on their own; the rest subtract.
Whether the people running it own it, what they paid for their own stock, and where five years of cash actually went.
Price is scored on its own axis, against a discounted cash flow rather than against a multiple. Both numbers stay visible, and every threshold that moved from the original framework carries the argument for why.
The character layer
Each one is a rescale of a pillar the framework already computed. That is the whole rule, and it is enforced by tests: a stat can never make a claim the valuation did not already make.
The roster
A class is the shape of the six stats, not a label someone chose. A company with a fortress balance sheet and thin growth reads as one thing; the reverse reads as another, and the model says which without you having to hold both numbers in your head.
The ladder
Every scored company lands somewhere on this ladder, and so do you — the same cutoffs, applied to your book. The point is not the letter; it is that you can open any rank and read exactly which threshold put it there.
See it run
Everything above is a claim about how the model works. This is the model working — the real engine against real SEC filings, through the same endpoint the terminal uses. Nothing here is pre-computed, and there is no demo data anywhere in this platform on purpose: numbers that look real and are not are worse than no numbers at all.
Make it yours first
The framework and its numbers never change — only the room they are read in. Pick a palette, a shape and how much environment you want behind the data. It follows you into the terminal and you can change it there at any time.
The terminal opens on an empty dashboard until you run a screen. There is nothing to sign up for.
Click a company to open it · click a column header to sort · ⌘/Ctrl-click to add to the comparison
Pick a company from the screener, or type a ticker above.
Load a company to open the workbench.
Uses the companies in your current screen — run a wider screen first and this gets more meaningful.
Two to four companies on every criterion at once, on the same rows and the same scale — so the shapes are comparable rather than just adjacent.
Good for the question a single company page cannot answer: is this margin normal for the business or normal for the industry? Comma-separate the tickers above. Mixing listings is fine; everything is converted to one currency before it is compared.
One per line: TICKER, name, role, exposure —
exposure is one of pure, major, component,
optional. Saved on this machine.
On live data this fetches every name in the universe, so it takes minutes on the first run and seconds afterwards from cache. Start with 150.
Chip makers invest in the labs, the labs commit to buy from the clouds, the clouds fund the labs. This is that web, what it costs, and the revenue it has to earn.
Every dollar of debt the government issues has to be bought by somebody at some yield. This is the supply, and the somebody.
Each manager's last 13F, the stocks most of them own, and what changed in their latest quarter. A photograph of a few months ago, not a signal.
Loading…
What you actually own, kept with your account. The paper Solo Run is separate and does not touch this.
loading…
One per line:
TICKER, shares, cost basis (optional).
A discounted cash flow gives you one number, and the number is the least reliable thing it produces. Three of them are more honest than one.
Each one is a full valuation, not a haircut on the base: you set the growth, the margin and the cost of capital, and the spread between the three is the honest measure of how much you actually know about the company. Put a ticker in the box above to start from its own numbers.
Publish a call and it is marked against the benchmark from that day. Nothing about your portfolio, your size or your account is shared — only the ticker, the direction and the reason you give.
Choose a stock or an index and every beast it offers is ranked by how rarely it happens. Take one on; it settles on the closing price at the deadline.
Found a team or join one with its code. The team's record is the calls you publish under Community while you are in it, and the letters your Solo Run seasons earn — your grades, never what you hold.