Questions, answered plainly

Three things, in this order. It shows each figure with its source: a filed figure with the formula, the inputs, the statement and the filing date, and a market-data figure labelled as such with the date it was supplied. It records why you own a company, in your words, and the numbers that must stay true. Then it re-tests those numbers nightly and emails you the first night one breaks.

When nothing breaks, you hear nothing.

People who pick their own stocks and hold them for years, researching three to ten companies a month. Not traders, and not anyone who wants the software to decide: there is no buy, hold or sell of ours, no verdict word, and no fair value or price target of ours on a company.

No. It gives you the filed figure, its formula, its source and its filing date, so the decision is yours and you can say why you made it. Then it tells you the day one of your reasons stops being true. The check runs every night at 23:30 UTC. A filing is picked up the night after it is published.

Nothing here is advice, a recommendation or a forecast.

More detail: Risk disclaimer.

A spreadsheet is as current as the last evening you spent on it. This reads new filings every night, keeps your reasons beside the numbers that carry them, and writes to you the day one breaks. The check runs every night at 23:30 UTC. A filing is picked up the night after it is published. Every filed figure also keeps its source filing, so you check the filing, not a cell you typed months ago.

No. Every score opens into the tests behind it, and every sourced figure into the statement it came from. You do need to be willing to read a balance sheet, and the academy starts from the beginning.

More detail: Academy.

Yes: Curious, which keeps your own work, a watchlist and saved screens with their export. A new account starts with 30 days of Professional, free and with no card, one trial per person, and when your free days end becomes Curious unless you choose a plan. Free with no account at all: the company catalogue and every company page, the investor tests, the rule test, the Benchmarks, the calculators and the academy. Calculators run in your browser; nothing typed is sent to us.

More detail: Screener, Investor tests, Test a rule, Benchmarks, Calculators, Academy.

Every US company we cover, one row each, with its market value and eight ratios: P/E, P/B, EV/EBIT, free cash flow yield, ROIC, dividend yield, debt to equity and the Piotroski F-score. Sort or filter any column. A missing figure reads n/a, with the reason. A ticker opens that company's page, with no account, and each one dates its prices and its figures.

More detail: Screener.

No. The academy teaches how the numbers are built and what they cannot tell you. A calculator is arithmetic on the inputs you type. Neither is a view on any security, and a result is only as good as its inputs.

More detail: Risk disclaimer.

A row for each value investor we cover, from Graham and Buffett to Greenblatt and Piotroski, with the tests they published. Where a test has a number, it is set against US-listed companies, with how many meet it, or fall in it, and the date of the data. They cover the tests, never what the investors own.

More detail: Investor tests.

No. None of them is affiliated with us or has endorsed the product. Each test is drawn from their published work, named beside it, and any number or measure of ours is marked OURS. Wording is theirs only where shown as a quotation. A result is arithmetic, never a claim about what they would buy.

More detail: Investor tests, Risk disclaimer.

Named sets of tests, 30 of them: 14 drawn from investors' published work and 16 of our own. For subscribers, Research runs them across the market, forensic scores beside each company; a company page shows each test's value, threshold, result and date; the disciplines panel on Today counts your holdings by result. Nothing is added into a score.

More detail: Methodology.

Meets: every test in the discipline passed, on figures we hold. Does not meet: at least one test failed on a figure we hold. Incomplete: nothing failed, but at least one test could not be measured. A missing figure never counts as a fail, and none of the three is a reason to buy or sell.

That a number or a measure is ours, not the investor's as printed: a threshold the author gave only in words, a stand-in for a figure we do not hold, or a test in one of our own disciplines. Our own disciplines carry no investor's name; where one follows an investor's idea, its details say whose.

Companies listed on NYSE, NASDAQ and NYSE American, plus 54 other markets with a Home Market Pack. Each market is listed at /markets. A company listed on a US venue that reports in another currency is included, with figures in the currency it reports in. A company that trades only on a market this site does not cover is not covered.

Both counts are at the top of this page, read live: how many companies we hold filed accounts for, out of how many you can look up. The methodology prints them too, by coverage tier.

More detail: Methodology.

No, and each company page says which group it is in. Verified: filed statements on disk, so figures trace to a filing and can be monitored. Metrics only: a value with no statement behind it, so it cannot be sourced or monitored. Listed only: no usable figures at all.

No. They are left out of the company catalogue, the named disciplines and the ranking pools, because a percentile rank only means something against comparable operating businesses. A fund has no Piotroski score to compute.

Because a price is true only on the day it was set. Prices come from a market-data vendor, stored with the date of their close, which can be days old, so they will not match your broker's live quote. Recomputing an indicator refreshes no price, so the two dates are shown apart.

The accounts each company files with the SEC, read from EDGAR and standardised into comparable statements. Every statement carries its fiscal year, period, filing date and reported currency, which is what lets a figure open into its source.

More detail: Methodology.

A failed or partial fetch never overwrites data we already trust. Where two sources disagree, the figure carries a mismatch badge. Where a value is missing, the product prints n/a, never a zero or an estimate.

Up to ten fiscal years on screen where the filings exist, and fewer where they do not. Nothing is padded or extended. Machine-readable filings began around 2009 to 2011, so that is the floor. Price history goes back five years.

A percentile rank against the whole scored universe, not a buy or sell rating and not a price target. A score of 80 ranked above about four fifths of that universe when computed. The rank is not within a sector, so compare a utility with a software company through the pillars and the figures, not the headline.

More detail: Methodology.

Five pillars, weighted: Value 35 per cent, Quality 30, Health 15, Growth 12 and Risk 8. A pillar is the plain average of the indicator percentiles inside it, and the score is the weighted average of the pillars that scored. A pillar with nothing we could rank is left out rather than counted as zero, and when the pillars that scored carry less than 80 per cent of the weight, no score is published. The methodology page prints the weights the engine uses.

More detail: Methodology.

A nine-point checklist of accounting improvement, published by Joseph Piotroski in 2000: profitability, cash flow, leverage, liquidity and efficiency, one point per test. The nine tests open on screen. It is not a valuation measure.

More detail: Methodology.

A 1968 distress model that blends five ratios into one figure. Above 2.99, distress is unlikely on this model; 1.81 to 2.99 is inconclusive; below 1.81 is the distress zone. It was fitted on US manufacturers, so read it as a prompt to check the balance sheet.

More detail: Methodology.

A 1999 earnings-manipulation screen built from eight indices. It flags the statistical signature of aggressive accounting, which tells you which notes to read first. It is not a finding of fraud, and it needs two consecutive years of statements.

More detail: Methodology.

n/a means the figure is missing. Zero is a number and missing is not, so a gap stays a gap everywhere, including in the formula view. n/m means not meaningful: return on equity on negative equity produces a number that means nothing, so it is flagged instead.

Yes. Open any figure to see the formula, the inputs, the statement they came from, the fiscal year, the filing date and the reported currency. A figure that cannot show that trail says so.

A deliberately simple two-stage discounted cash flow: one growth rate, a discount rate you enter, and a perpetuity for the terminal value. It shows how much of the result sits in that terminal value and saves versions of your assumptions. The value it gives is arithmetic on your inputs and yours alone, not a target price; we publish no fair value of our own.

More detail: Methodology.

Your reasons for owning a company, in your own words, plus the numbers that must stay true for them to hold. The numbers are what the service re-tests. The first takes about five minutes: the builder suggests up to three guardrails, filled in with the company's current values.

One condition and one threshold, phrased as the statement that keeps your thesis alive. "Return on invested capital stays at or above 20 per cent" is a guardrail. It is breached the night that statement stops being true.

Every night at 23:30 UTC, after the day's figures are recomputed. You hear within a day of the data reflecting a breach. It is not a price feed: a filing reaches us after the market has moved, and what it ends is holding a broken thesis for months without knowing.

No. A guardrail fires once, when it goes from holding to broken, and stays silent while it stays broken. If the value comes back inside your line, a recovery note goes to your inbox, with no email, and the guardrail re-arms.

Nothing fires, and the state is left as it was. A data gap is a fact about our pipeline, not evidence about your company, and an alert that cannot tell the two apart is worse than none.

Not on a single figure: monitoring is built on the thesis, where every line is tied to a reason. You can also watch a saved screen. It re-runs each night and emails you when a company enters or leaves it, and stays silent otherwise.

Besides breach alerts: an email when a company you hold, watch or wrote a thesis on files a delisting notice, sells shares outside a registered offering, changes its auditor or says past statements can no longer be relied on; the changes on any saved screen you watch; and an optional weekly brief on Sunday. The privacy policy says how to stop each one.

More detail: Privacy policy.

Not news or prices. It re-tests the figures you chose against filed statements. For a company you hold, watch or wrote a thesis on it also reads its 8-K filings by item: a delisting notice, an unregistered sale of shares, an auditor change or a non-reliance notice is emailed, and a director or officer change, a material agreement, a new debt obligation or a shareholder vote goes to your inbox.

Write that part in words, as a pillar with a review date, and your inbox reminds you when a review is due. Only the numbers are re-tested nightly, because only numbers can be re-tested honestly.

On the public pages, only if you agree: say no and neither Google Analytics nor Microsoft Clarity loads. Inside the product there are no third-party trackers. We record account milestones, such as a thesis created, and at most once a day that the disciplines panel was shown, in our own database.

More detail: Privacy policy.

On hardware we own and run in Spain, reachable only through an encrypted tunnel. The database and its nightly snapshots are on that machine, and restores are rehearsed on a throwaway copy.

More detail: Privacy policy.

Passwords are hashed with bcrypt at twelve rounds and never stored in any other form. Refresh tokens are stored only as SHA-256 hashes, are single-use, rotate on every refresh and are revoked on sign-out. Repeated failed sign-ins lock the account temporarily.

Stripe for payments, and card details never reach us. Resend and Brevo for email, Cloudflare for the network, Slack and Telegram for our own notices, and Telegram for the support chat. Google and Microsoft only if you agree, and only on the public pages. The privacy policy says what each receives.

More detail: Privacy policy.

Only if you tell us. Theses, guardrails, notes, positions and cost bases are optional, and they are yours. We use them to host your work, run the nightly checks, send your alerts and keep backups.

More detail: Terms of service.

Curious is free. Serious is $29 a month or $299 a year, and Professional $99 a month or $999 a year; each keeps everything in the plan below it. Tax is added at checkout where the law requires it. Stripe takes the payment, so card numbers never touch our servers, and your plan is checked against Stripe every ten minutes.

More detail: Pricing.

30 days of Professional, every step of the workflow, for a new account, one trial per person. No card at sign-up. When the 30 days end, nothing is charged unless you chose a plan. Without one, the account becomes Curious when your free days end, and you decide then, having used the thing, which steps are worth paying for; if you chose no plan, there is nothing to cancel.

It used to take a card at signup. That asked you to trust us with a payment method before the product had shown you anything on a company you care about, which is the wrong order.

No: the payment page has no code field. If you came through a partner’s link, their reduction is already in the price shown to you before you pay, and on the invoice. There is nothing to type.

No card at sign-up. When the 30 days end, nothing is charged unless you chose a plan. Without one, the account becomes Curious and the parts above it lock. Your work is kept exactly as you left it: theses, screens, saved valuations and watchlists are all still there, and what your plan does not include opens again on a plan that does.

You keep the free pages either way, signed in or not.

In Settings, under billing: one click, and it is done, with no email and no call. We ask why, and answering is optional. The renewal stops and you keep access to the end of the period you paid for, or of the trial.

Yes: the 30-day money-back guarantee on your first payment. Ask within 30 days of it and we refund it in full, with any payment you made in those days to change plan. No form, no questions, once per account, and never a renewal. That is more than the 14-day statutory withdrawal right.

More detail: Terms of service.

On Professional, in the first 60 days of a paid subscription, if a guardrail you set broke and we did not email you within a day of the data showing it, we refund your last two months. It covers our silence, not late, restated or missing data.

You claim it in Settings on the web. Where our own records show plainly that the guardrail broke and we then said nothing, the refund is issued automatically, in the same moment. Every other outcome goes to a person, who writes to you either way. Nothing is ever refused by machine.

More detail: Terms of service.

Your name: in Settings in either of our apps. Your email: write to us and we change it, on any platform.

Settings, on whichever of the three you are using: the web, the iOS app or the Android app. It is the same deletion, and the screen lists what goes and what we keep before you confirm. It happens at once, and any subscription ends with it, which also means you do not keep access to the rest of the period you paid for. If you would rather a person did it, write to the address in the privacy policy.

More detail: Privacy policy.

Results from Research and your portfolio data export as CSV from inside the app, and you can ask for a copy of everything we hold about you.

More detail: Privacy policy.

Use the reset link on the sign-in page. It emails a single-use link, and completing a reset signs out every existing session.

There is no team plan and none is planned. The product watches one person's reasons for owning something, and sharing a login would mean sharing a thesis board.

The app keeps a service record: what ran, when it ran and what it found, including a receipt for each thesis from each nightly pass.

Report it. Every research page has a report-a-data-issue link that goes to a person. Open the figure's source first if you can: knowing the statement and fiscal year turns a vague report into a fixable one.

Because that listing is in the metrics-only tier: an indicator value exists, but the statement behind it was never archived. Those values cannot be sourced, recomputed or monitored, and the product marks them.

It trades only on a market this site does not cover, or it is a fund, warrant or preferred share, which are excluded on purpose. Otherwise, send us the ticker and it goes to the front of the queue. New filings are read every night, but no date is promised for any company.

Four things to check. The value may have been missing that night, and then nothing fires by design. The guardrail may already have fired and be waiting to re-arm. Monitoring stops when a subscription ends. Or the email is in your spam folder.