What Is Taufolio? Four Tests to Run on Any Stock Report

What is Taufolio? Tau plus folio. Tau carries four standards a report has to meet; folio is the bound, numbered set of pages where the work outlives the conversation. Four standards rather than four slogans: each is a question you can put to any stock report. Two of them you can settle yourself, from the finished text. The other two only the person who wrote it can answer, ours included.

What Is Taufolio, and What the Folio Does#

Financial product names usually claim nothing. Quantum, Apex, Synergy: the sound has to seem serious, and that is the entire job. Fair enough. A brand is not a commitment, and nobody sane audits a name against a filing. A name that lists four conditions is a different proposition.

Start with the easy half. A folio is a numbered leaf in a bound volume, so the second half of the name promises form: a report sits in the library under a company name and a date instead of scrolling out of sight with the question that produced it. A Full report breaks a company into 25 points and puts The gist of the report on top, so nothing forces you to read it all at once. I sort out which write-up suits which moment separately, in the piece on how a Monthly summary differs from a Full report.

The first half is harder, because it promises four things at once.

Can You Check It#

Here is the one you can run without asking anyone's permission. Open any finished piece of analysis, pick the strongest claim it makes about the company, and try to reach the document behind it.

That is the whole first test, and it comes back yes or no. A claim about margins with no link to a filing, an annual report or an earnings call leaves you holding someone's tone. Tone is sometimes right. It is never checkable, and a sentence with no trail looks identical whether it is true or invented.

Tau earns its place here. In its oldest use it was a mark someone put on a document to say they had seen it and stood behind it. A citation does the same for a paragraph. In Taufolio reports the findings that matter carry a link where they are made, and the Transcript opens from the Earnings brief so you can read the passage yourself. The link carries no timestamp from the recording: it points to the written record, not to a minute.

I trust this test more than the other three, because the answer sits in the text and nobody has to grant it to you. I take up what a citation changes separately, in the piece on what a source link does to analysis written by a model.

What Three Models Agreeing Actually Measures#

Tau has a second life in statistics, where Maurice Kendall used it in Biometrika in 1938 for a measure of agreement between two rankings (A New Measure of Rank Correlation). The second test puts that to the author rather than to the text: how many independent readings of the same material stand behind a conclusion? One model can be wrong with total conviction, because confidence costs it nothing. Three models and a judge answers that directly: three write the same analysis separately, and a fourth compares the versions and keeps what can be pinned to a source. A Full report pro costs 400 credits against 100 for the plain version, four times the price for a text written three times and checked once.

Agreement is not truth.

Three models will repeat the same mistake if they took it from the same document, so the second test measures how well a conclusion survives independent reading, not whether it is right. I walk through it step by step in the piece on how three independent analyses become one report.

How Long You Have After the Event#

The third test comes from engineering, where tau is a time constant: the number that says how fast a system answers a change. In market form the question is how long passes between an event and your understanding of it, and the clock starts earlier than most readers assume. A US-listed company has four business days from the event to file a current report on Form 8-K (General Instruction B.1 to Form 8-K), so at most a week separates the event from the moment you can read about it. The rest of the delay is yours.

Faster is not automatically better. The first hour after a release belongs to the price rather than to understanding, and the document reads the same at nine in the morning as it does three days later. The third test measures time to understanding, not time to reaction, and only in that form is it worth measuring.

On our side the test comes out like this: a Full report is produced in about five minutes and takes 40 to 60 minutes to read, so the time goes on reading rather than writing. That is why The gist of the report sits on top, and why Monitoring sends Breakthrough news only when an event can change the company fundamentally, typically zero to two times a month. A month with none is normal, and nothing in Monitoring costs credits.

A Full Turn, Not One Ratio#

A P/E is not an understanding of a company, and a price chart carries even less. So the fourth test asks about reach: did the report walk all the way around the business, or stop at one ratio? Tau is in this one too, since popular mathematics writes it as 2π, a full turn instead of half of one.

A Full report works through 25 points, from the business model and the moat, through management tone and analyst concerns, to risks and valuation, each with source material underneath. The same list works as a set of 25 questions worth asking any company, whoever writes the report. This is the second test you can settle unaided: the table of contents is visible before you read a word. The trap runs the other way here, because completeness is easy to mistake for accuracy. A report that covers 25 points has a paragraph for each, including the ones where the company gives you nothing interesting to say. That is why a full turn starts at The gist of the report rather than at point one.

The turn has a price and the credits show it: an Earnings brief is 80 credits for one quarter, a Full report 100 for the whole company. Twenty credits more covers the whole business instead of a single quarter, with proportionally more to read.

What the Name Leaves Out#

The fifth thing is missing on purpose. There is no recommendation in the name: a report describes the business, the risks and what management said, names no target price, and does not tell you what to do with your money.

The four tests are about how the text was made, never about what you do once you have read it, and they share one limit that is better said plainly: they measure process, not judgement. A report can carry a link under every sentence, survive three independent readings, appear in five minutes and cover the whole company, and still rest on a thesis the next quarter demolishes. I know of no test that catches that in advance, which is why a Full report produces an Investment thesis with a measurable condition under each assumption, so that after earnings you can see which assumption broke.

The four collapse into one rule: verifiability is a property of the text, not of its author, which is why it carries to the next report you open. There, the table of contents and one link will tell you most of what you need. A free account carries 100 credits a month, which is exactly one Full report, and the first steps in Taufolio and what stands behind each report are written up separately.