The Knowledge Bank
Read the work behind the work.
Essays, papers, and study notes from the house's own research — written to be understood, and read aloud if you prefer to listen.
What the Knowledge Bank is
The Knowledge Bank is the house's content estate: a growing body of articles and essays on financial analytics, analytics in general, decision science, and verification. Everything here grows from our own research — our papers, our engines, our study notes — and exists for education. It is a place to build understanding, not a place to be told what to do. Nothing in these pages recommends, advises, or predicts. The work is meant to sharpen how you think, so that your own conclusions are yours, arrived at by your own effort.
How an article earns its place here
An article enters the Knowledge Bank only if it is written from primary work: our own research papers, our engines, our study notes. Every source is named on the page. Every limit is stated plainly in the piece — what the data does not cover, what the method cannot claim, where the evidence stops. Nothing is published that we did not produce ourselves, and nothing is filled in to look complete. If something is missing, the page says so. That is the standard, and it applies to every shelf.
The shelves
Financial Analytics — How markets are measured: the instruments, datasets, and methods behind financial analysis, examined from first principles and traced to their sources.
Analytics in General — The wider craft: how data becomes signal across domains, and what the recurring patterns of analysis teach us wherever they are applied.
Decision Science — How conclusions are actually reached: the structure of judgment, the traps in reasoning, and the discipline of deciding under uncertainty.
Verification & Proof — How we know what we know: checking claims against first sources, testing methods against their limits, and the practice of proving rather than asserting.
Every page here can be read aloud — listen as well as read, whichever suits the way you work.
Start here
New to the Bank? The eight articles form one path — each step assumes only the ones before it.
- Accountable Only When It Is Verifiable — first, what proof means in this house.
- First Source of Truth — where a number is allowed to come from.
- What a Quant Engine Actually Computes — the machine, in plain terms.
- One Ratio, Three Stories: The DuPont Decomposition — the first instrument.
- The Buyback Flattery Test — the second instrument.
- Prediction = Data × Assumptions — the frame every forecast stands in.
- Agents That Show Their Work — systems that hand you receipts instead of asking for trust.
- Reading a DCF Without Hidden Defaults — the full valuation, every input on the table.
The first-article slate
Eight articles open the Bank. Each grows from a named file in the house's corpus — our own papers, engines, and study notes — and each states its limits inside the piece.
-
Accountable Only When It Is Verifiable
Verification & Proof
why trust in agent-produced work has to be checkable by a stranger — and the sealed record, BLOCKED verdict included, that shows what that means in practice.
-
First Source of Truth: Where a Number Comes From Is Part of the Number
Financial Analytics
how the house decides where financial data may come from — first sources, four verdicts, and why a missing number stays missing.
-
What a Quant Engine Actually Computes
Financial Analytics
a plain tour of the machine — trends, DuPont, forensic scores (Altman, Beneish, Sloan), valuation with every assumption on screen — and why it outputs numbers, never verdicts.
-
One Ratio, Three Stories: The DuPont Decomposition
Financial Analytics
how a single return-on-equity figure splits into margin, turnover, and leverage — and what each leg can quietly hide.
-
The Buyback Flattery Test
Financial Analytics
how to compare a company's total growth with its per-share growth, and read what management actually paid for its own shares.
-
Prediction = Data × Assumptions
Decision Science
the house's working frame for every forecast — change the assumptions and the prediction confesses what it was made of.
-
Agents That Show Their Work
Verification & Proof
sealed receipts and named checks — how an automated system's work becomes something a second party can audit instead of merely believe.
-
Reading a DCF Without Hidden Defaults
Decision Science
a valuation with every input on the table — and the reverse question worth asking of any price: what growth is already assumed inside it?