Own what no issuer controls

One preference, and everything that follows from it

Every model published here inherits a single preference: own things whose supply nobody can decide to increase. That is the whole worldview. It is not a forecast, it is not a prediction about any date, and it is not a claim that the alternative is worthless.

Ranked by how hard the supply is to change, the preference reads: bitcoin, then gold, then productive businesses, then everything else. Bitcoin’s issuance schedule is fixed and independently verifiable by anyone running the software. Gold’s above-ground stock has grown slowly for centuries and no committee sets the rate. A share in a productive business is a claim on what that business produces rather than on anyone’s promise to pay.

A government bond is none of those things. It is a promise to pay a number of units of a currency, and the issuer of that promise is also the issuer of the currency. That is the structural objection, and it is the only one being made.

This is a preference, not a proof

It is worth being exact here, because the blunt version of this argument is everywhere and it is wrong. Lending to a government does not always cost you purchasing power. Inflation-linked government bonds exist, and they are contractually tied to measured consumer prices rather than to a fixed number of currency units.

As of August 2026, a ten-year US inflation-linked government bond held to maturity contracts to pay roughly 2.4% a year above measured consumer prices. Somebody who wanted a defined amount of purchasing power on a defined date could have it, from the very issuer this worldview declines to lend to.

What the preference costs

A worldview that is only ever described by its advantages is marketing. Here is the other side, stated before you need it.

Concentration

Refusing an entire asset class narrows what is left. Fewer things in the plan means each one matters more, and it means the plan can be wrong in fewer but larger ways.

Deeper falls

Bitcoin fell roughly 77% from its 2021 peak to its 2022 trough. Any plan holding a meaningful weight of it inherits a share of that, and the published models print the arithmetic on their own pages.

No fiat promise

Nothing here promises a defined amount of euro or dollar on a defined date. A near-term goal in a currency wants an instrument that promises that currency, and saying so is more useful than selling volatility with a deadline attached.

Long flat stretches

Gold sat still for the better part of two decades after 1980. Hard assets are not obliged to reward patience on any particular schedule.

What this is not

• Not a prediction that any currency fails, on any date. The preference does not need a collapse to make sense; it only needs the supply of one thing to be decided by people and the supply of another not to be.
• Not a claim that measured inflation is secretly enormous. The honest numbers are in the next article, and they are smaller and more interesting than the version usually shouted.
• Not a reason to hold nothing that produces cash. Every published model holds productive businesses, because a plan holding only scarce things is a position rather than a portfolio.
• Not advice, and not an assessment of your circumstances. This is published editorial, identical for everyone who reads it.

The next article gives the two numbers that are usually merged into one, and shows why merging them is the single most common dishonesty in this corner of the internet. Read debasement, measured honestly before you decide whether you agree with any of this.