Bitcoin: Convexity Backbone

Bitcoin is the framework’s reserve asset and its main source of convexity, and much of the rest of the portfolio is arranged around it. The case for holding it rests on properties you can check: a fixed supply schedule, several independent valuation models, and an adoption curve that has so far delivered lopsided gains, well beyond the single-digit compounding of mature assets.

You can hold Bitcoin under your own keys or through regulated funds (spot ETFs), and pledge it as collateral for a loan. It is also violently volatile. The framework treats that volatility as the toll on the upside, and uses sizing to keep the toll affordable. Its inclusion is doctrine. Its accumulation pace follows valuation discipline, with its CIS informing conviction; both are written guidance you apply. It is scored like every other holding, but the score neither sizes nor sells it. Bitcoin sits outside the three postures and the concentration limits, and this part’s rules govern it. As of September 2026 the dashboard shows Bitcoin’s CIS (Part 6 explains the score) and can overlay the power-law trend and floor on its price chart; nothing in the software paces purchases.

Why this asset, and not gold, stocks or real estate? The framework’s objective is the most compounding a person can survive over decades of regime instability, and the backbone is judged against ten requirements drawn from it. The fair suspicion is that the list was written backward from the answer. Most items follow from the objective and can be checked today. Two, regime resilience and multi-decade survivability, are judgments about the future, and every grade in the table below is the framework’s judgment, not a measurement. Any asset that met all ten would qualify equally. The ten are:

Gold is the closest alternative at 7.5 out of 10. It earns partial credit on custody (physical metal carries dealer premiums and storage costs a fund share does not), on models and on regime resilience, and none on upside: it preserves purchasing power, with no order-of-magnitude gain. The S&P 500 follows at 6, and real estate and commodities trail, short mainly on convexity and custody. The ranking follows from the framework’s own objective, and a different objective could be served by a different asset. This framework needs all ten at once, and on these grades only Bitcoin has them. That is not faith. That is architecture.

If Bitcoin clears every requirement, why hold anything else? Plenty of holders argue exactly that: if you never sell, why not go 100 percent Bitcoin? In a long bull market the all-in holder wins in raw dollars; through a tenfold rise, a 100 percent holder ends with far more than a 15 percent one. The framework accepts that trade. It aims at the best outcome weighted by probability, for a real person who will meet job losses, health scares, family needs and a regime shift or two over three or four Bitcoin cycles.

Sequence-of-returns risk is the lasting damage a large drawdown does when it lands just as money has to come out, through a forced sale or a withdrawal. It grows when one asset makes up most of the portfolio. A concentrated holder has little else to sell, or to spend at a multi-year low, so volatility can only be endured. Income-funded dollar-cost averaging is the usual answer, but the paycheck that funds it can shrink or stop in the same downturn that sinks Bitcoin. Even a steady one arrives on an employment schedule, not an opportunity schedule.

An illustration on actual Bitcoin prices, with simplified sleeves: two investors start with $100,000 in January 2021 and hold their convictions through December 2025. A keeps buying Bitcoin from income; B follows the framework’s rules.

A finished about $106,000 ahead, as it should. B’s figures are a simplified illustration: its other sleeves are carried at a round $100,000 instead of being modeled holding by holding, and nothing here is a backtest. B’s 44 percent Bitcoin share is what one strong cycle happened to produce, above the 20–30 percent elevated range set out at the end of this part; read it as an outcome, never as a target.

The real difference is what each investor maximized. A maximized exposure; B maximized control. At the 2022 trough B had cash to deploy and income that did not depend on selling Bitcoin; A had neither. Maximum exposure wins in ideal conditions. Structured architecture wins across uncertain conditions. The second sentence is the framework’s design claim, and one cycle cannot prove it; this example shows what control costs in a strong cycle. The claimed advantage comes from structure instead of forecasting: cash inside the portfolio, exposure to more than one regime, and rules for when to buy.

If control comes from how a position is built, the way the reserve is assembled matters more than the price of any single purchase. The framework accumulates Bitcoin over years: purchases are systematic and paid from income, and the reserve is held through whole cycles.

The arithmetic rewards patience. A fixed contribution buys more units at a lower price (units received equal dollars divided by price), so the volatility that punishes a trader trying to time an exit works for the steady accumulator, who collects the most units exactly when conviction is hardest to hold.

Valuation discipline, set out in Why Bitcoin can be modeled below, sets the pace. When the independent models agree that price is more than 30 percent below their combined estimate, you raise contributions by about half or deploy dry powder. When price runs well above them, you slow discretionary buying and bank income as dry powder, while the systematic income-funded purchases continue. Neither case calls for selling the core reserve, which would bring back the three failures the design exists to avoid: sequencing risk from a mistimed exit, a tax bill on a realized gain, and the behavioral error of selling on price, whether into weakness or out of a rally. It is hard to mistime a sale that never happens.

This is the Accumulate phase of a longer lifecycle. Once the reserve is a large share of net worth, borrowing against it becomes an option (the Borrow phase, at the end of this part). Refusing to trade keeps that option open: a position never sold is still there to borrow against, or to pass to heirs with a stepped-up basis under current law. All of it assumes the asset itself survives, which is what the risk register examines next.

Whether the asset endures is partly a question of plumbing, and in 2024 and 2025 the plumbing changed. US regulators approved spot Bitcoin ETFs in January 2024; institutions that had dismissed the asset began serving client demand for it; and the policy debate moved from whether Bitcoin is legitimate to how it should be held. That cycle tested the rails this part depends on (regulated funds, institutional custody and lending against coins) at scale for the first time. One cycle cannot prove permanence, and better access does not shorten the list below. The reserve ranges set out below (a 15 to 20 percent target, 20 to 30 percent at higher conviction) are sized for both the upside and the ways it can go wrong.

Most of those are tail risks you guard against and rarely feel. The one every holder feels is volatility. Plan for Bitcoin drawdowns of 75 percent or more: every completed cycle through 2022 fell at least 77 percent, each less deeply than the one before (about 93 percent in 2011, 85 from 2013 to 2015, 84 from 2017 to 2018 and 77 from 2021 to 2022), and the 2025–2026 fall reached about 54 percent. Nothing guarantees that the falls keep getting shallower. At a reserve-sized position, even the worst of those is a survivable toll paid for the upside, and the exhibit below runs the arithmetic.

Sizing keeps the downside survivable; it does not tell you what the reserve is worth. How do you value an asset with no earnings and no cash flows? Bitcoin helps in one unusual way: its supply is fixed by code (21 million coins, with issuance halving roughly every four years), so the models only have to explain demand, which tracks adoption, liquidity and network effects. That makes its growth look more like a technology platform’s than a currency’s. Plotted on log-log axes against days since the genesis block (January 3, 2009), its price since 2010 has clustered around a power-law trendline, Price = A × (Days)^B, where A and B are fitted constants whose stability is debated. The framework reads it as context for where a cycle stands, never as a forecast; the pattern could break if adoption changes course. The corridor below sketches it, with bands above and below, on a representative price path.

No single model is enough on its own, so the framework reads five: four valuation models and one directional reading of liquidity, and trusts them most where they agree.

Fair value, in this part, means the estimate the independent models converge on: a judgment you form from them, never the power-law trend on its own. Convergence sets conviction and the pace of buying. As of September 2026 the dashboard draws two power-law lines on Bitcoin’s price chart, a trend and a floor (the Math page gives both formulas); it draws no euphoria band and computes no convergence. Despite its name, the floor line does not hold price up: closing prices have fallen below it several times since 2015, including the March 2020 crash and the 2022 bottom, and most recently for most of February to mid-September 2026. On September 29, 2026, Bitcoin at about $83,600 sat just above the floor (about $82,900) and about 45 percent below the trend (about $152,000).

A valuation needs a yardstick, and Bitcoin’s is the money already stored elsewhere: in gold, property, offshore accounts and weak currencies. Bitcoin’s market capitalization at the start of 2026 was ~$1.75 trillion (about 19.97 million coins at ~$87,500), roughly 6 percent of gold’s ~$30 trillion above-ground stock (World Gold Council). On September 29, 2026 it was about $1.68 trillion (about 20.1 million coins at about $83,600), still roughly 6 percent. The doctrine’s TAM estimate counts only the share of existing monetary assets Bitcoin could plausibly absorb:

The four pools add up to about $23 trillion of addressable monetary premium, the doctrine’s estimate and about 13 times Bitcoin’s start-of-2026 market cap. It leaves out new uses such as settlement-layer or DeFi applications. By default, when no research profile supplies its own figure, the dashboard scores Bitcoin against a smaller pool, about $11.5 trillion, so the default score does not depend on this estimate (as of September 2026; the Math page has the formulas).

Run one scenario on its own. If Bitcoin reaches 40 percent of gold’s market value over 20 years, that is a $12 trillion market cap: about $570,000 a coin across the 21 million cap, or about 6.5 times the start-of-2026 price of about $87,500. In a $100,000 portfolio, a 15 percent allocation ($15,000) would become about $98,000, adding about $83,000, or about 83 percent of the starting portfolio before anything the other 85 percent earns. It is one conditional path, and no forecast.

Buy-borrow-die means holding appreciated assets instead of selling them, borrowing against them for cash, and leaving them to heirs. It is the main reason the reserve lives in a taxable account. The framework keeps Bitcoin outside Roth by default for three reasons. First, the reserve is never sold, and under current US law (IRC §1014) an asset inherited at death generally takes a basis equal to its value at death, so a taxable holding passed to heirs leaves none of its lifetime gain to tax as income. Second, only a taxable holding can be pledged for a loan; using an IRA as security for a loan counts as a distribution (IRC §408(e)(4)). Third, Roth space is scarce and better spent on Torque that may rotate. None of this is special to Bitcoin: it works for any appreciated capital asset, depends on current law and on credit staying available, and carries interest cost and liquidation risk. Gifts made during life keep the giver’s basis (§1015), and pre-tax retirement accounts get no step-up: heirs owe income tax on what they withdraw. What the account choice is worth over decades is worked in Part 4 and in Part 5’s compounding example.

Custody. The framework’s default is the full reserve in self-custody, on hardware wallets, in a taxable account, bought through payroll DCA. Spot Bitcoin ETFs carry the same exposure through a regulated fund. As of September 2026 the dashboard treats spot ETFs like native Bitcoin, leaving both outside its concentration and position-size limits, but by default counts only native coins toward the Bitcoin share (the Math page has the detail). The Borrow phase is optional and comes late, once the reserve has grown into the mature share described below. Staying in the Accumulate phase for good is a complete plan, not an unfinished one. A loan supplies cash without a sale, but it adds interest cost and the risk of a forced sale: if the collateral falls far enough, or a payment is missed, the lender sells it, and that sale is a taxable event. Pledged coins also leave self-custody for the lender’s custody or a shared multisig for the life of the loan. The rules below exist to keep that risk survivable.

How large should the reserve be? It scales with horizon and risk capacity. A 15–20 percent target suits horizons of ten years or more and moderate risk capacity. A 20–30 percent elevated range suits higher conviction and longer horizons, and still needs valuation discipline. A 30–50 percent mature share may emerge after 15–20 years through accumulation and appreciation, never by rebalancing up to it, and that is when the Borrow phase becomes an option. These percentages describe outcomes under given conditions, not rebalanced targets. Run the drawdown arithmetic before you pick one: at a fixed weight, a repeat of the 77 percent drawdown of 2021 to 2022 costs about 11.5 percent of the portfolio at a 15 percent reserve and about 15.4 percent at 20 percent. As of September 2026, the dashboard proposes a 15 percent Bitcoin baseline at setup (10 percent under two of its six preset theses, Conflict Economy and Re-industrialization), which you confirm or replace. Its construction score, the Framework Integrity Score (FIS) of Part 6, carries a 10 percent Bitcoin reference within a 5 to 15 percent band, for information only: it costs no points and never triggers a sale. Bitcoin held under the dashboard’s own Bitcoin wrapper does add to the FIS wrapper drift, because that wrapper has no target (the Math page explains). Part 5 builds the rest of the portfolio around the reserve.