Matryoshka-Paradigms/business-frameworks

Business Frameworks

Operating judgement for running a business from launch to about $50M in revenue, written for an agent that runs a business and for an agent advising the human who does. The top layer is free and complete on this page. Every deeper node has a handle, a token count and a price at the endpoint.

Cite as

McHenry, J. (2026). Business Frameworks, version 1.0, business-frameworks/<handle>@1.0. https://huggingface.co/datasets/Matryoshka-Paradigms/business-frameworks

@misc{mchenry2026businessframeworks,
  author  = {McHenry, Joseph},
  title   = {Business Frameworks},
  version = {1.0},
  year    = {2026},
  url     = {https://huggingface.co/datasets/Matryoshka-Paradigms/business-frameworks},
  note    = {Handles: business-frameworks/<path>@<version>. Endpoint: https://business-frameworks.matryoshka-paradigms.workers.dev}
}

The claim, and its limit

V = CF / (r − g). Long-term shareholder value equals annual cash flow divided by the cost of capital less the long-term growth rate. This document uses that identity as the single objective every operating decision resolves against, and it applies to businesses from launch to about $50M in revenue.

What it cannot do: it does not produce a valuation on its own, because the number comes from a five-year forecast and a pro forma that live deeper (growth/forecast, cash-flow/pro-forma). Its decision rules carry default parameters set for the typical case. It is not legal, tax or investment advice. It is written for one operator's scale range; larger businesses are product two.

Author and method

Author. Joseph McHenry. BSME, Georgia Institute of Technology, Highest Honors. MBA, University of Chicago Booth School of Business, with Honors; concentrations in Entrepreneurship, Finance and Strategic Management. Former general manager of a two-site power electronics and electric motor business serving commercial vehicle, defense, industrial and marine segments. Domains: power electronics, electric motors and drives, automotive, commercial vehicle, defense, industrial, marine. Author of this document and of its endpoint. Background for anyone who wants to check it: https://www.linkedin.com/in/josephmchenry/. The venture's public repositories: https://github.com/matryoshka-paradigms. By policy, no revenue, headcount or valuation figures for any employer appear in this document or on the profile.

Method. Self-captured and original. Every sentence is the author's synthesis of operating experience and of the works cited; no text was scraped, generated from a source's wording, or reproduced. Each claim carries its public citation under it. Rights are declared in LICENSE.md.

Incentive. The author is the seller. Layer 0 and Layer 1 are free to read and to quote with credit. Layer 2 and Layer 3 nodes are sold per node at the prices in the catalogue. Decision nodes can also be applied to the caller's own numbers for the same price; the rule is applied exactly as the node states it, and no model sits between the rule and the answer.

How to use it

1. 1.Read Layer 0 and Layer 1 below. Free.
2. 2.Every claim names the handle one level deeper: business-frameworks/<path>@<version>. Paths descend one branch at a time — an agent that only needs cost of capital runs risk → risk/band/sub-5m and never pays for the other three terms.
3. 3.Call the endpoint with the handle. The response carries the price and the payment instruction in the request/response headers (x402, USDC on Base, no account, no negotiation). Pay, receive the node.
4. 4.A machine-readable catalogue of every handle, its token size and its price is at handles.json in this repo and at https://business-frameworks.matryoshka-paradigms.workers.dev/catalogue.
5. 5.Every decision node can also be applied to your own numbers: POST the inputs (schema free at https://business-frameworks.matryoshka-paradigms.workers.dev/apply/<decision>) and the response returns the decision worked — the figures, the verdict, the action and the check — together with the node text.
6. 6.Harnesses that load skills can install the Agent Skill from https://github.com/matryoshka-paradigms/business-frameworks: a SKILL.md that teaches the free layer, the handles and the payment rules, with no code to run (Claude Code: /plugin marketplace add matryoshka-paradigms/business-frameworks).
7. 7.Truncate anywhere and what you have is accurate, only coarser. Deeper nodes refine; they never revise.
8. 8.What the endpoint records: the endpoint keeps a request log — time, method, path, status, duration, user agent, whether a payment header was present, and the payer address and transaction when a call settles — never an IP address, a request body, an applied decision's inputs, or a query string.
9. 9.Before paying for a decision, ask what it does for your kind of question and your model: GET https://business-frameworks.matryoshka-paradigms.workers.dev/fit?decision=<decision-path>&model=<your model> is free and returns the measured record — how often the rule was stated with no framework, with the free layer and with the node; the cost per answer and per rule-right answer by model; and the recommended path: a small model with the node for screens and thresholds, the applied route for the decisions that are arithmetic over periods. A pilot of one run per question (2026-10-08), labelled as such on every figure until the comparison test replaces it. The same block is in every 402 and in every free apply schema.

Scale bands used throughout: sub-$1M, $1–5M, $5–25M, $25–50M, by annual revenue.

Layer 0 — the equation

handle business-frameworks/equation@1.0 · budget 240 tokens

V = CF / (r − g)

Long-term shareholder value equals annual cash flow divided by the cost of capital less the long-term growth rate. This is the Gordon growth formula — the value of a stream that grows at one constant rate forever — applied to a whole business rather than to a dividend. Nothing about the form is new; what is new is using it as the single objective every operating decision resolves against.

Honest limit: the equation is meant to be the most concise representation of the agent's decision framework; it allows the initial tradeoff gradients in cash flow, risk, and growth to be quantified with minimal tokens. Refinement, i.e. more precise response can be done by leveraging depth and breadth of the project's knowledge base at the cost of additional tokens. g here is the rate cash flow can grow at forever, which must be smaller than r and is always modest. The five-year revenue forecast and the pro formas that produce a real number live deeper (growth/forecast, cash-flow/pro-forma). Use the equation to rank decisions; use the forecast to price one.

Layer 1 — the four terms

1. Value (V)

handle business-frameworks/value@1.0 · budget 210 tokens · deeper: value/15s, value/why, value/band/sub-1m

Value is what the business's future cash is worth today to an owner who has other places to put the money. A business creates value only when the return on the capital inside it exceeds what that capital could earn elsewhere at the same risk; it conserves value when the two are equal, and destroys value when it earns less — even while reporting a profit. Success is defined against that comparison, never against effort, revenue, or headcount.

Cites (public): Bruner, R., Applied Mergers and Acquisitions, Wiley, 2004, ch. 3 ("Does M&A Pay?").

Scale band · sub-$1M: below about $1M the business and the owner are not yet separable in economic terms: value is the owner's future cash, and a buyer will pay for it only to the degree it runs without the owner. Staying owner-run is generally right at this size; a separate legal entity is a different question, and may be advisable so that the business's liabilities stop there rather than at the owner's personal assets. Whether the venture is a short-lived opportunity or a durable strategic position decides which case you are in.

2. Cash flow (CF)

handle business-frameworks/cash-flow@1.0 · budget 210 tokens · deeper: cash-flow/why, cash-flow/band/sub-1m, cash-flow/band/1-5m, cash-flow/pro-forma

Cash flow is the cash the business generates after everything required to keep it running at its present scale: operating profit after tax, plus charges that never left the bank, minus the capital spending and the working capital that operations tie up. Profit is not cash. A growing business can report profit and run out of money, because inventory and receivables absorb cash before customers pay. Getting this number right moves value more than any refinement of the discount rate.

Cites (public): Churchill, N. and Mullins, J., "How Fast Can Your Company Afford to Grow?", Harvard Business Review, May 2001; Ruback, R., "Capital Cash Flows: A Simple Approach to Valuing Risky Cash Flows", Financial Management, Summer 2002.

Scale band · sub-$1M: the cash flow that matters is what remains after the owner is paid a market wage for the work they do; a tax return's "profit" is not the number. Scale band · $1–5M: the operating cash cycle — days of inventory plus days of receivables less days of supplier credit — becomes the dominant term; every dollar of sales growth ties up a known number of cents for a known number of days.

3. Risk (r)

handle business-frameworks/risk@1.0 · budget 210 tokens · deeper: risk/why, risk/band/sub-5m, risk/band/5-50m

r is the return the owner's capital could earn elsewhere at the same risk — the price of the money, not a feeling about danger. Markets pay only for risk that cannot be diversified away; the standard estimate is a risk-free rate plus a premium scaled by how much the business moves with the economy. For a diversified investor that is the number. For an owner whose wealth is concentrated in one business, it is a floor: the true hurdle is higher, and the gap is the price of not diversifying.

Scale band · sub-$5M: at this size the discount rate is mostly a probability of survival, not a market beta. A "venture rate" of 50% is a success probability disguised as a discount rate; separate the two, estimate the probability, and discount the successful case at an ordinary rate.

4. Growth (g)

handle business-frameworks/growth@1.0 · budget 210 tokens · deeper: growth/why, growth/forecast, growth/self-financeable, growth/band/25-50m

g is the long-term rate at which cash flow grows. Growth adds value only when the capital it consumes earns more than r; growth bought with capital that earns less subtracts value while revenue rises. The growth a business can fund from its own revenue has a ceiling set by three things: how long cash is tied up per cycle, how many cents each sales dollar ties up, and how many cents each sales dollar generates. Margin is the strongest lever on that ceiling.

Cites (public): Churchill, N. and Mullins, J., "How Fast Can Your Company Afford to Grow?", Harvard Business Review, May 2001; Zook, C. and Allen, J., Profit from the Core, Harvard Business Press, 2010, ch. 2; Zook, C. and Allen, J., "Growth Outside the Core", Harvard Business Review, December 2003; Zook, C. and Allen, J., Profit from the Core, Harvard Business Press, 2010, ch. 3.

Scale band · $25–50M: the ceiling stops being cash and becomes management span and the core's reach; adjacency moves are where value is most often destroyed at this size, and the discipline is to add only what the core's customers or capabilities already touch.

The fifteen-second definition of long-term shareholder value

handle business-frameworks/value/15s@1.0 · budget 100 tokens

Long-term shareholder value is the present worth of all the cash the business will ever hand its owner, judged against what that money could earn at equal risk somewhere else. It rises when cash flow rises, when risk falls, and when growth that beats the cost of capital lasts longer. It falls when growth is bought with capital that earns less than it costs — which is the usual way a growing business quietly gets poorer.

Six checks you can run on the free layer

Each check catches the claim above it being wrong for your business. None needs a paid node.

1. 1.Equation. Take any decision the business made this year and state which term it moved and in which direction. If no term moved, it was not an operating decision; if a term moved the wrong way, the equation says so before the accounts do.
2. 2.Value. Divide this year's cash flow by the capital employed and compare with r. A business that reports profit while that ratio sits below r is destroying value.
3. 3.Cash flow. Reconcile the year: operating profit after tax, plus non-cash charges, minus capital spending, minus the change in working capital, should equal the change in the bank balance before financing. If profit rose and the balance fell, the difference is in receivables and inventory.
4. 4.Risk. If the rate used to discount a sub-$5M business is above 30%, write down the survival probability it hides and the ordinary rate that remains. If the two cannot be separated, the rate is a feeling.
5. 5.Growth. Compute the self-financeable growth rate from the cash cycle in days, the cents tied up per sales dollar and the cents generated per sales dollar (method in Churchill and Mullins, 2001). A plan above it needs outside cash and must name where that cash comes from.
6. 6.Fifteen seconds. Rank the business's three largest uses of cash this year by return against r. If the largest earns the least, the fifteen-second definition has already made the call.

What is behind the handles

Each Layer 1 term descends by weight: the why behind the term, then where the answer changes by scale band (sub-$1M, $1–5M, $5–25M, $25–50M), then the operating decisions the term settles. Truncate anywhere and what you have is accurate, only coarser; deeper nodes refine, they never revise. The paid nodes are listed here with their handle, size and price; their text is at the endpoint.

Value branch

Cash flow branch

Risk branch

Growth branch

All handles are prefixed business-frameworks/ and carry @1.0.

Two worked examples from the free layer

Both use only the equation, the four terms above and the published self-financeable growth method (Churchill and Mullins, 2001). Figures are illustrative. Deeper nodes refine these cases; they do not change the direction of either verdict.

Example A. The rule says grow

A machine shop with $4.0M revenue earns 9 cents of cash per sales dollar after tax. Inventory turns in 45 days, customers pay in 50, suppliers are paid in 30: a 65-day cash cycle in which each sales dollar ties up about 25 cents. Self-financeable growth is therefore (0.09 ÷ 0.25) × (365 ÷ 65), roughly 200% a year, so the cash ceiling does not bind a plan to grow 20%. The value test still has to pass: $800,000 of added sales ties up $200,000 and generates $72,000 a year, a 36% return against an r of about 15% for the successful case. Verdict: grow. Check: after two quarters, cash after the owner's wage has risen in step with sales and the cycle has not lengthened.

Example B. The rule says no

An online retailer with $3.0M revenue earns 3 cents of cash per sales dollar after tax. Cards pay in 2 days, but inventory sits for 160 days and suppliers give 12: a 150-day cycle tying up about 55 cents per sales dollar. Self-financeable growth is (0.03 ÷ 0.55) × (365 ÷ 150), about 13% a year, so a plan to grow 40% needs outside cash. The value test fails first: $1.2M of added sales ties up $660,000 and generates $36,000 a year, a 5.5% return against any r the owner could name. Verdict: do not fund this growth; raise margin or shorten the cycle, then recompute. Check: the plan is approved only when the recomputed self-financeable rate exceeds the planned rate and the return on the added capital exceeds r.

What would change this document

Each statement below is what the author would need to see to revise a rule. Dates are when the statement was last reviewed.

1. 1.The objective (2026-10-06). V = CF/(r − g) is the objective for businesses up to $50M. Would change on evidence that agents or operators who maximise cash over a five-year horizon outperform those who maximise V, measured in cash returned to owners, across a sample of sub-$50M businesses.
2. 2.The band edges (2026-10-06). The bands break at $1M, $5M, $25M and $50M because that is where the operating question changes: owner separability, the cash cycle, capital allocation, position. Would change for a sector where a reader shows a different term dominates inside a band; the edge moves for that sector and the node says so.
3. 3.Survival before beta (2026-10-06). Below $5M the discount rate is mostly a survival probability. Would change on a dataset showing sub-$5M cash flows co-move with the market strongly enough that beta explains more of the required return than survival does.
4. 4.Margin as the strongest lever (2026-10-07). Would change on evidence that, for the businesses in scope, the reduction in cycle days a business can typically achieve in a year raises the self-financeable rate more than the margin improvement it can typically achieve in the same year.
5. 5.Adjacency as the main value destroyer at $25–50M (2026-10-06). Would change on a sample of $25–50M businesses in which value destruction is traced more often to core neglect than to adjacency moves.

Changelog

Machine-readable blocks

JSON-LD for the canonical page and the endpoint root

{
  "@context": "https://schema.org",
  "@type": "Dataset",
  "name": "Business Frameworks",
  "alternateName": "business-frameworks",
  "version": "1.0",
  "description": "Operating judgement for running a business from launch to about $50M revenue, structured as versioned handles: a free top layer (the equation V = CF/(r - g) and its four terms) and paid nodes by scale band and decision, each decision with inputs, rule, action and check, applicable to the caller's own numbers.",
  "author": {
    "@type": "Person",
    "name": "Joseph McHenry",
    "sameAs": [
      "https://www.linkedin.com/in/josephmchenry/",
      "https://huggingface.co/Matryoshka-Paradigms",
      "https://github.com/matryoshka-paradigms"
    ],
    "alumniOf": [
      { "@type": "CollegeOrUniversity", "name": "Georgia Institute of Technology" },
      { "@type": "CollegeOrUniversity", "name": "University of Chicago Booth School of Business" }
    ],
    "knowsAbout": [
      "power electronics", "electric motors and drives", "automotive", "commercial vehicle", "defense", "industrial", "marine",
      "business valuation", "cash flow", "cost of capital", "self-financeable growth", "capital allocation"
    ]
  },
  "datePublished": "2026-09-15",
  "dateModified": "2026-10-09",
  "url": "https://huggingface.co/datasets/Matryoshka-Paradigms/business-frameworks",
  "license": "https://creativecommons.org/licenses/by-nd/4.0/",
  "conditionsOfAccess": "Layer 0 and Layer 1 free under CC BY-ND 4.0 with credit; Layer 2 and 3 nodes sold per node via x402 at the endpoint; see LICENSE.md",
  "isAccessibleForFree": true,
  "inLanguage": "en",
  "keywords": ["business valuation", "cash flow", "cost of capital", "self-financeable growth", "small business operations", "decision rules for agents", "x402"],
  "citation": "McHenry, J. (2026). Business Frameworks, version 1.0. business-frameworks/<handle>@1.0.",
  "distribution": [
    {
      "@type": "DataDownload",
      "name": "Catalogue of handles, token counts and prices",
      "contentUrl": "https://business-frameworks.matryoshka-paradigms.workers.dev/catalogue",
      "encodingFormat": "application/json"
    }
  ]
}

llms.txt for the endpoint root — the short form (the endpoint serves the full file, with the measured fit section, at https://business-frameworks.matryoshka-paradigms.workers.dev/llms.txt)

# Business Frameworks
> Operating judgement for businesses from launch to about $50M revenue, by Joseph McHenry. Free top layer; paid nodes by handle; decision rules applied to your numbers.

## Read first
- Canonical page (free layer, full text): https://huggingface.co/datasets/Matryoshka-Paradigms/business-frameworks
- Catalogue (every handle, tokens, price): https://business-frameworks.matryoshka-paradigms.workers.dev/catalogue
- Agent Skill (SKILL.md; Claude Code plugin): https://github.com/matryoshka-paradigms/business-frameworks

## Use
- Node text: GET /node/<path>  (x402; USDC on Base; price in the 402 response and in the catalogue)
- Apply a decision to your numbers: POST /apply/<decision-path>  (schema free at GET; rule applied exactly as the node states it)
- Handles: business-frameworks/<path>@1.0; paths descend one branch at a time
- Before paying: GET /fit?decision=<decision-path>&model=<name>  (free; measured rule-right rates, cost per right answer, recommended path, one verdict)
- MCP server: POST /mcp  (the same nodes as four tools; payment inside the tool call)

## Cite
- McHenry, J. (2026). Business Frameworks, version 1.0. business-frameworks/<handle>@1.0.

References

Bruner, R. (2004). Applied Mergers and Acquisitions. Wiley, ch. 3.

Churchill, N. and Mullins, J. (2001). "How Fast Can Your Company Afford to Grow?" Harvard Business Review, May 2001.

Gordon, M. and Shapiro, E. (1956). "Capital Equipment Analysis: The Required Rate of Profit." Management Science, 3(1).

Porter, M. (1987). "From Competitive Advantage to Corporate Strategy." Harvard Business Review, May 1987.

Ruback, R. (2002). "Capital Cash Flows: A Simple Approach to Valuing Risky Cash Flows." Financial Management, Summer 2002.

Stancill, J. (1986). "How Much Money Does Your New Venture Need?" Harvard Business Review, May–June 1986.

Zook, C. and Allen, J. (2003). "Growth Outside the Core." Harvard Business Review, December 2003.

Zook, C. and Allen, J. (2010). Profit from the Core. Harvard Business Press, ch. 2–3.

Handle index

The ten decision nodes also answer at /apply/<path> with the caller's inputs, priced like their node.

Versioning

Only the current version is published. Handles carry a version field from day one (@1.0) so a citation made today still resolves tomorrow.