An executable philosophy
The Adaptive Convexity Framework
A tax-aware method for compounding through unstable markets, built to survive being wrong.
The framework
From philosophy to governance
Six Parts in four movements.
Philosophy
Architecture
Bitcoin: Convexity Backbone
Why Bitcoin is the portfolio’s main source of convexity, and how it is valued, accumulated, held, and borrowed against.
04Tax Architecture & ROC Strategy
Which account each holding belongs in: Roth for Torque, taxable for Bitcoin and return-of-capital income, and pre-tax only when it pays.
Before you read
Three pictures the framework thinks with
Two show the framework’s habits; the third, the shape of results it is designed for. All three are sketches, not market data.
The framework’s first habit is knowing where you stand. Growth and inflation are one common way to sort the market’s weather, and leadership rotates across it. Part 2 reads growth, inflation, and policy together to form the macro thesis that positions are judged against. The map below follows one path through four quadrants of that weather. The path is illustrative, and so is the point marked “Now”: it shows the habit of re-reading where you stand, not a measurement of today.
L1 · Capital Has Weather: Same assets, different regime, different behavior
The same portfolio behaves differently in different macro regimes.
A conceptual four-quadrant map of growth versus inflation, with quadrants labeled stagflation, reflation, deflation and goldilocks. An illustrative path traces capital from a disinflation boom through a deflation scare, reflation and an inflation shock, and ends at an illustrative “Now” marker near the center.
Illustrative regime path; every waypoint, including “Now”, is placed by hand. Leadership in each quadrant is the framework's reading of common market experience, not a measured result. Quadrant names follow common market usage, not any one published model. Conceptual diagram. Illustrative framework exhibit, not historical data. Leadership labels are the framework's judgment.
Sources: Rule stated in Part 2 · Identification, evaluation, and governance; Rule stated in Part 1 · The traditional playbook was built for a regime that has ended; ACF dashboard · Regime reading from the direction of rates, inflation and unemployment (software behavior as of September 2026).
The second habit is patience about entries, and it governs the positions you pick. Part 1’s order of operations gives a new position its full weight only once momentum confirms its CIS score, and Part 5 builds it in stages as the evidence earns them, then lets it run while conviction holds. The Bitcoin reserve runs on a different clock: Part 3 buys it on a schedule and leans in when it is cheap, and Ballast income can fund those purchases in a dislocation. The window below follows one position you pick, from compression through confirmation to a run that lasts until its thesis, its momentum, or the concentration limits say otherwise.
L2 · The Window Opens: A new position waits out the range, adds on confirmation, then lets the move run
A new position earns its full weight only after the market confirms the thesis.
A conceptual price path for one new position. It stays range-bound inside a soft compression band, clears the range and is confirmed, then accelerates upward into an asymmetric release.
Conceptual path for one new position, with no price levels implied; the Bitcoin backbone and Ballast run on different clocks. Conceptual diagram. Illustrative framework exhibit, not historical data.
Sources: Rule stated in Part 5 · Momentum: conviction requires confirmation.
Those two habits work one decision at a time. Over many years, the six Parts’ rules are meant to change the shape of your results as a whole, though none has been backtested against a benchmark. The picture below holds the average fixed to show what the framework changes around it: fewer deep losses, more small shortfalls, and a longer right tail. The wash on the left is what position limits, stops, and tripwires take away; the wash on the right is what convex positions left free to run add.
S2 · Bend the Tail: Hold the average fixed and see what the framework changes around it
With the average held fixed, ACF is designed to thin the deep-loss tail and lengthen the gain tail, paying for it with more small shortfalls.
Two outcome distributions with the same area and the same average. The symmetric reference is a normal bell centered on the base case. The ACF curve is thinner in the deep-loss tail, piles up just below the base case, is thinner through moderate gains, and runs longer into the far right tail.
Conceptual outcome distributions with the same area and average; illustrative, not measured returns. Conceptual diagram. Illustrative framework exhibit, not historical data.
Sources: Rule stated in Part 1 · The traditional playbook was built for a regime that has ended; Rule stated in Part 5 · Three behaviors, one portfolio; Rule stated in Part 2 · The intellectual foundations, thesis-agnostic.
Part 1 starts where the framework does: why the traditional playbook is failing, and the nine steps that replace it.
Resources
The Framework in Pictures
Every chart in the framework, grouped by Part.
The Framework in Math
The arithmetic underneath: weights, clamps, penalties, and how a score becomes a size.
Glossary
The framework’s vocabulary, defined where it is taught.
The Software
The framework running as software, with an interactive demo you can try.
FAQ
Convexity, clarified.
What is the Adaptive Convexity Framework?
A method for running a portfolio through unstable markets, with two scores to keep you honest. Rather than betting on a point forecast, it is built to survive being wrong, while keeping room for the outsized gains a regime shift can deliver.
What does convexity mean here?
A payoff that curves upward: a favorable move earns more than an equal adverse move costs. The framework sizes and places positions to keep that shape, and Part 6 scores it.
Who is it for?
Anyone who wants to own the reasoning behind their portfolio. If you doubt that a stock fund and a bond fund are enough, start with Part 1. If Bitcoin is already your biggest holding, Part 3 covers how the framework values it, holds it, and borrows against it. If you run a concentrated book of stocks, Part 5 covers the sizing bands and the 15 percent default cap on a single position, and Part 6 covers the score that picks the band. If you live on what the portfolio pays, Part 4 covers return-of-capital income and which account each holding belongs in; the framework does not yet publish a withdrawal policy. Advisors can check the dashboard’s formulas on the Math page. The tax chapters use US accounts: the logic travels, the rules may not.
What will I come away with?
A working method: how to size a position before you buy it, which account it belongs in, what score it has to keep, and what you will do when the thesis breaks.
How much of this does the software do?
Some of it. As of September 2026, the dashboard computes the two scores, CIS and FIS, and flags some rules; many others are written rules you carry out. The Math page marks which is which, and the Parts say so where it matters. The dashboard records every trade and never blocks one; it places no orders.
Is it free to read?
Yes. All six Parts are free, and you’re welcome to share them.
Is this investment advice?
No. The framework is educational and analytical. It explains a way of thinking about portfolios; it does not recommend specific securities, allocations, or actions for your situation. Where the Parts name an instrument, such as STRC as the income Ballast in Parts 4 and 5, or use default bands, they show the method at work as of the date given; none of them is a recommendation for your portfolio. For decisions about your own money, talk to a licensed professional who knows your circumstances.