Lesson 8 of 8
What this does not establish
Every signature here has a dull business explanation producing the same number. Here they are, and why that is not a reason to stop looking.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- Each of the six readings in this course has at least one ordinary business explanation that produces an identical number.
- That is not a defect of the readings. It is why the output is a figure to watch rather than a conclusion.
- A reader who cannot state the dull explanation for their own finding has not finished reading.
This lesson is the one that makes the other seven honest, and it is short because the material is a list. For every signature in this course, the same numbers arrive for reasons that have nothing to do with credit.
Six readings, six ordinary explanations
The cycle reading
- Capital spending rises far from the consumer.
- The asset base grows while output does not.
- Inventory builds while sales fall.
- Work in progress lengthens.
- Interest is capitalised into long assets.
- Cash and profit diverge.
The ordinary explanation
Same numbers
- A customer was won and capacity was contracted for.
- A plant was bought and its output is consolidated elsewhere.
- A product transition held old stock while new stock was built.
- A permitting delay stalled projects already started.
- The company happens to hold assets that qualify under the standard.
- The depreciation charge is large because the asset base is.
Both columns produce identical figures in the accounts. Nothing in the statements separates them.
Why the two columns cannot be separated from the statements
Because a filing reports amounts and classes, not reasons. Two companies can report the same capital expenditure, the same inventory build and the same divergence between cash and profit, with entirely different stories behind them, and no line in either filing distinguishes the stories.
Some of the difference is disclosed in the management discussion or in a segment note, which is why the research process course puts reading the narrative before computing anything. Some of it is not disclosed anywhere and is settled years later by whether the assets earned.
The confounding that catches people
The sharpest version is not that a dull explanation exists. It is that the dull explanations are more common. A company builds capacity because it won a contract far more often than because a credit condition changed, and a screen that flags capacity builds will therefore flag mostly the former.
The statistics course calls this a base rate problem, and the arithmetic is unforgiving: a reading that is right about a rare cause will still be wrong most of the times it fires, however good the reading is.
What survives
The figures survive, and so does the question. What does not survive is the conclusion.
That is the correct output, and it is what the whole product is built on: a reading of this kind produces a figure to watch and a level at which something has changed, which is a condition. It does not produce a verdict, and a condition written down is worth more than a verdict, because next year it is still checkable.
Check yourself
Why can a filing not separate a cycle explanation from an ordinary one?
- Because the disclosures are inadequate
- Because a filing reports amounts and classes rather than reasons, and both explanations produce identical amounts
- Because the two explanations are the same thing
Show the answer
Answer: B.
Some of the difference reaches the management discussion or a segment note. Much of it is settled years later by whether the assets earned.What is the base rate problem here?
- That filings are published too infrequently
- That the ordinary explanations are more common, so a reading that fires on the signature will mostly be catching them
- That the cohort excludes small companies
Show the answer
Answer: B.
Being right about a rare cause is not enough. If the common cause produces the same observation, most firings will be the common cause.What is the correct output of a reading in this course?
- A judgement about where the company is in a cycle
- A figure to watch and a level at which something has changed, which is a condition
- A ranking of companies by exposure
Show the answer
Answer: B.
A condition is still checkable next year. A verdict is not, and it was never supported by the statements in the first place.Sources
- Micron Technology, Inc., annual reports on Form 10-K for the fiscal years ended 1 September 2022, accession 0000723125-22-000048, and 31 August 2023, accession 0000723125-23-000054.
- The empirical record lesson of the case against course, for what the published tests of this theory actually found, reported with each study's sample, period and method.
What this leaves out
- This is a heterodox tradition. The mainstream account of the same phenomenon differs, and nothing here is a forecast or a view on any security.
- The list of ordinary explanations is the list this course met. It is not a proof that no further explanation exists for any of the six readings.
- The two readings offered for the filed pair are both consistent with the figures. Presenting two is not a claim that there are only two.
- Nothing here is a view on Micron or on any security.