Lessons
Each lesson names a screen and a field, and never a number. The figures come from your own live screens, which is what makes a lesson written once still correct after the next set of results - and what makes it correct for a company on any venue this product reaches.
Where to start
Nothing in this product is withheld until a lesson is finished. This list changes what is offered first and never what is available.
Not started.
- Learn the keyboard10 min
Not started.
Not started.
Not started.
Not started.
- A valuation, end to end35 min
Not started.
A first look at a company you chose
10 min - 0 of 5 stepsGet oriented on one company using the screens in the order they answer questions, rather than the order they appear in the sidebar.
- SOVWThe headline price, the one-day change, and the as-of on its tooltip.
Everything after this is a comparison against that price, so knowing which instant it describes comes first.
- DESWhat the company sells and to whom, in its own words.
A valuation of a business you cannot describe is arithmetic.
- HPWhether today's price is near the top or the bottom of that range.
Context for the multiple you are about to read on RV.
- RVThe multiple, and the peer set it is being compared against.
The peer set is the assumption. If the peers are wrong the multiple is a correctly computed answer to the wrong question.
- NEWSWhether anything in the last five days explains the price you started at.
A figure with no story behind it is where a reader invents one.
Check yourself
Nothing here is marked and no answer is recorded. Each one is answered by a method rather than a value, because a value would be stale.
- What instant does the price on SOVW describe?
Hover the headline figure — the tooltip carries the source and the as-of. If they disagree with the stamp on HP, one of the two panels is serving a cached day.
- Which companies is the multiple on RV comparing against?
The peer list is on the panel. Ask whether you would have picked those six yourself; if not, that is the finding.
A valuation, end to end
35 min - 0 of 7 stepsAssemble a discount rate, apply it, and then find out how wrong it can be — on a company you chose, with every figure read from a live screen.
- BETAThe regression beta, the window it was fitted over, and the R².
Beta is the only input to the cost of equity that is measured rather than assumed, and the R² tells you how much of the movement it explains. A low R² does not make the beta wrong; it makes it uninformative.
- WACCThe risk-free rate, the equity risk premium, the cost of debt and the two weights.
This is the discount rate assembled in front of you. Note which premium is selected — this build offers two named ones and they do not agree.
- WACCHow far the WACC moves.
That distance is the honest uncertainty in every valuation that follows, and it comes from a choice nobody can settle.
- FARevenue, operating margin and free cash flow, and whether any of them was restated.
A cash flow forecast has to start from cash flow that happened.
- EEThe consensus for the next two periods and how many firms contribute.
Two contributors is not a consensus. The count is the confidence.
- DCFThe value per share, and the terminal-growth and discount-rate inputs it used.
Most of a DCF's value is in the terminal period, which means most of the answer comes from the two inputs you understand least.
- RVWhether the relative and discounted answers point the same way.
When two methods disagree the useful question is which assumption they disagree about, not which one to believe.
Check yourself
Nothing here is marked and no answer is recorded. Each one is answered by a method rather than a value, because a value would be stale.
- How much did the whole valuation move when you changed one premium?
Note the WACC before and after on the panel itself. Any DCF answer quoted without that range is quoted more precisely than it is known.
- How many firms are in the consensus you used?
EE prints the contributor count. Under five, treat the consensus as one opinion rather than a market view.
- Was any year on FA restated?
FDIF shows year-on-year filing changes; a restated period is a changed figure with no event behind it.
Two companies, walked through
25 min - 0 of 4 stepsCompare AAPL against one peer without letting the comparison do the arguing for you.
- RVThe peer's multiple beside AAPL's.
A peer you cannot justify is a peer chosen by the number it produced.
- FAWhether the two fiscal years cover the same twelve months.
Two companies' 'FY2025' routinely differ by six months, and comparing them is comparing different weather.
- OWNWho holds each, and whether the holders overlap.
Two companies with the same holders trade together for a reason that has nothing to do with either business.
- CUSTWhether either depends on a small number of customers.
Concentration is the risk most likely to be invisible in a multiple.
Check yourself
Nothing here is marked and no answer is recorded. Each one is answered by a method rather than a value, because a value would be stale.
- Do the two fiscal years line up?
Read the period end on each FA sheet. If they differ by more than a quarter the comparison is indicative and should be labelled as such.
- Could you have picked this peer before seeing its multiple?
If not, you have found a number rather than a comparison.
What changed in this filing
20 min - 0 of 4 stepsRead a company's own year-on-year edits as the disclosure they are.
- FDIFThe risk statements added since the prior annual filing.
A risk a company added this year is a risk its lawyers decided it now has to disclose. That is the closest thing to a company telling you what it is worried about.
- FDIFWhat stopped being a risk.
Removals are rarer than additions and are read less often, which makes them worth more.
- SECFThe surrounding paragraph for one added statement.
A diff shows you the sentence. Only the filing shows you what it qualifies.
- ECALWhat happened between the two filings.
A new risk usually has a date attached to something.
Check yourself
Nothing here is marked and no answer is recorded. Each one is answered by a method rather than a value, because a value would be stale.
- Was any added statement a rewording rather than a new risk?
Read it beside the removed one it most resembles. Filers reword; the diff cannot tell the difference and you can.
How to open and read a filing
15 min - 0 of 3 stepsFind a primary document and read the part of it that carries information.
- SECFWhich filings contain it.
Full-text search across filings is a different instrument from looking up one company: it finds who ELSE said the same thing.
- EDOCThe first three, which are ordered by the filer's own judgement.
Order is editorial and is the only ranking in the document.
- EDOCHow the company divides its own revenue.
The segments are the company's view of what business it is in, and they rarely match the sector the screener puts it in.
Check yourself
Nothing here is marked and no answer is recorded. Each one is answered by a method rather than a value, because a value would be stale.
- What period does this document cover, and when was it filed?
Both are on the cover page, and they are usually weeks apart. Anything that happened in between is in an 8-K, not here.
Learn the keyboard
10 min - 0 of 2 stepsDrive the terminal without the sidebar.
Ideas behind the screens
Longer than a glossary entry and shorter than a lesson. Every one names the specific wrong conclusion people draw from it, which is worth more than the definition.
- A filing
A document a company is legally required to give the SEC on a schedule. A 10-K is the annual one, a 10-Q is quarterly, an 8-K reports a single event within days of it happening.
Why it matters: It is the only company-authored text in this product that somebody signed under penalty of perjury. Everything else — estimates, ratings, news — is an opinion about it.
Commonly misread as: That a filing is a summary. It is a legal disclosure: it is written to be complete and unactionable, not to be read quickly, and the parts that matter are usually in the notes rather than the statements.
- A multiple
A price divided by something the company produced — earnings, sales, book value. It answers 'what is being paid per unit of this', which lets two companies of different sizes be compared on one axis.
Why it matters: Every valuation on a relative basis is a multiple, and every multiple is a ratio of two numbers that were each measured differently.
Commonly misread as: That a low multiple means cheap. A multiple is a price and a forecast in one number: it is low when the market expects the denominator to fall.
- A discount rate
The return a supplier of capital requires to leave money in this business rather than somewhere else, used to convert a future cash flow into what it is worth today.
Why it matters: It is the single input that moves a discounted valuation most, and it is the one input that is never observed — it is always assembled from other estimates.
Commonly misread as: That it is a property of the company. It is a property of the ALTERNATIVE: when the risk-free rate moves, every discount rate moves with it and nothing about the business changed.
- An estimate
A forecast published by an analyst at a firm, aggregated across firms into a consensus. It is a number about the future with a name attached.
Why it matters: A price reacts to the difference between a result and the estimate, not to the result. Without the estimate a reported figure has no sign.
Commonly misread as: That consensus is a prediction. It is the average of the forecasts that happened to be published, by the firms that happened to cover this name, at whatever moment each of them last revised.
- An as-of
The instant a figure describes. Not when it was fetched, not when the page was drawn — when the world looked like that.
Why it matters: Two figures on one screen with different as-of stamps cannot be subtracted, and this is the most common way a correct calculation produces a wrong answer.
Commonly misread as: That a fresh page means fresh figures. A cached daily close reloads instantly and is still yesterday's.
- Survivorship
Any list built from what exists today has already dropped everything that went bankrupt, was acquired or was delisted.
Why it matters: Every peer set, every screen result and every backtest in this product is assembled from live symbols, so all of them carry it.
Commonly misread as: That a historical average computed from a current list is a historical average. It is the average of the winners.
- A restatement
A company refiling a period it already reported, because the first version was wrong.
Why it matters: It means a figure you read last quarter can change without any event, and the number now on screen may not be the one a chart drawn earlier used.
Commonly misread as: That the accounts are a fixed record. Only the filing is fixed; the figures inside it can be amended for years.
- A minor unit
Some venues quote in a hundredth of the currency. London quotes GBp — pence — so a line at 1,528.80 is £15.29, not £1,528.80.
Why it matters: It is the difference between a correct figure and one out by a factor of a hundred, and nothing on the screen looks wrong when it happens.
Commonly misread as: That GBp is a typo for GBP. It is a different unit and this build treats the case as load-bearing on purpose.
What this tool cannot tell you
The per-screen manual carries the limits of each screen. These hold whichever screen is open, which is exactly the set a per-screen note can never state.
- Whether a price is right.
Nothing here forecasts. Every valuation screen assembles inputs you chose and shows the arithmetic; changing an input changes the answer, which is the point.
- What will happen after an event.
The events calendar knows the date. No feed in this product carries the outcome, and none of them would carry it in advance.
- Whether a filing is honest.
The diff shows what a filer changed. It cannot see what a filer chose not to write.
- Anything intraday, reliably.
The free sources this product reads are delayed by fifteen minutes or settle daily. The as-of on every figure says which.
- Anything about a private company.
Almost everything here starts from a filing or a quote, and a private company has neither.
- Whether a peer set is the right one.
Peers are assembled from a classification and a screener, both built from companies that still exist.
- What you should do.
This is a research tool. It has no view, and a tool that expressed one would be giving advice it is not licensed to give.