Sound Money Theory
A research lab on monetary hardness. We will define what "sound money" means, compare the moneys people actually use across eight properties, check the historical data ourselves, and weigh the strongest argument against the whole idea before drawing any conclusion.
If you had to store the value of your work for thirty years, which money would you trust, and what property are you actually betting on?
Why this matters
Every time you save, you make a quiet bet: that the money you set aside will still command roughly the same goods and labor when you need it. A retirement account, a down payment fund, an inheritance left for a child: each one is a wager on a money holding its value across decades.
"Sound money" is the name one school of economics gives to money that wins that bet reliably. But the phrase carries assumptions worth testing. This deep dive treats it as a question, not a slogan: what makes money hold value across time, how do the candidates actually compare, and where does the hard-money argument run into its strongest objections?
What does "sound money" mean?
Start from function, before naming any school. Money does three jobs: it is a medium of exchange (you can trade with it), a unit of account (you can price things in it), and a store of value (it carries purchasing power into the future). "Soundness" is mostly about that third job: how well the money resists having its purchasing power diluted.
The Austrian School, following Carl Menger and Ludwig von Mises, argued that the best money is the one the market chooses through voluntary use, not the one a government decrees. Mises developed this in The Theory of Money and Credit (1912). The core idea: money whose supply cannot be expanded cheaply protects savers, because no one can quietly print claims against the goods they have stored.
A note on sourcing
The properties checklist below is a useful teaching synthesis. The classic short list of six is widely associated with Saifedean Ammous's The Bitcoin Standard (2018); we extend it to eight by adding censorship resistance, seizure resistance, and settlement finality, which matter for digital money. It is not a verbatim list from Menger or Mises, who wrote about money's origin and value in different terms. We use it because it is a clear lens, not because it is an ancient canon.
The monetary properties, and the tradeoff between them
Eight properties are useful for judging whether something works as money: scarcity (hard to produce), durability (survives time), portability (easy to move), divisibility (splits into small units), verifiability (easy to check as genuine), censorship resistance (no third party can block a payment), seizure resistance (no third party can confiscate it), and settlement finality (once settled, it cannot be reversed).
Here is the key point most explanations skip: no money is best on all eight at once. Gold is extraordinarily durable and final to settle but hard to move and hard to verify. A bank deposit is easy to divide and move but can be frozen or reversed by the bank. Bitcoin scores high on most axes, while a young and volatile market is the price it pays elsewhere. Which money looks best depends on which properties you weight most, and that depends on what you need money to do. The tool below lets you test that across seven things people actually use as money today.
Weigh the properties yourself
Uncheck the properties you do not care about for your use case (for example, a 30-year store of value versus everyday spending versus moving money past a hostile government). The fit score is the average of the scores on the properties you keep. Scores are illustrative, on a 1 to 5 scale, with reasoning in the caption.
| Money | Scarcity | Durability | Portability | Divisibility | Verifiability | Censorship res. | Seizure res. | Finality | Fit score |
|---|
What does "backed by" actually mean?
People often ask what a money is "backed by," as if every money must be backed by something. That question hides a confusion worth untangling, because there are two different kinds of money and only one of them is "backed" at all.
- Claims (IOUs). A bank deposit is a promise from the bank to pay you. A stablecoin is a promise from its issuer to redeem it for a dollar. A CBDC is a direct liability of the central bank. These are "backed" by the issuer's reserves and solvency, and they are only as good as that promise. If the issuer fails, or freezes you, the backing is what you are left arguing about.
- Bearer assets. Physical gold, cash in your hand, and bitcoin in self-custody are not promises from anyone. They are not "backed" by an external thing; they are valued directly for their own properties. Asking what gold is "backed by" is a category error: gold is the asset, not a claim on one.
A better question than "what backs it?"
Ask instead: "what gives it value, and who can take it away?" Fiat is valued because a government collects taxes in it and makes it legal tender. Gold and bitcoin are valued for their monetary properties and the network of people who accept them. Neither answer is "nothing"; they are different stories. The sharp difference is that a bearer asset has no issuer who can revoke it, which is exactly the tradeoff the matrix above measures as censorship resistance and seizure resistance.
Measuring hardness: stock-to-flow
"Hardness" means resistance to supply inflation. One way to put a number on it is the stock-to-flow (S2F) ratio: the existing supply divided by the amount produced each year.
Stock-to-flow
S2F = existing stock ÷ annual new production
A high ratio means new supply is small next to what already exists. An S2F of 60 means it would take about 60 years of current production to reproduce the entire existing stock.
After Bitcoin's 2024 halving, the block subsidy fell to 3.125 BTC, so annual issuance is small relative to the roughly 19.9 million already mined. Each future halving raises the ratio further, and issuance ends near the year 2140. On this single metric, Bitcoin is the hardest of the four.
Issuance falls on a fixed schedule
For Bitcoin, hardness is not really a single number; it is a schedule written into the protocol. The block subsidy (new bitcoin paid to miners per block) halves about every four years, stepping issuance down toward zero. Roughly 19.9 million of the eventual 21 million already exist, near 95 percent of the total.
Where the S2F story breaks down
"A high stock-to-flow ratio predicts the price. Bitcoin's rising S2F means a predictable rising price."
What the evidence showsS2F measures supply hardness. It says nothing about demand, and price depends on both. The popular "stock-to-flow price model" that circulated from 2019 onward predicted six-figure prices on a fixed schedule and visibly diverged from reality afterward. Treat S2F as a description of supply, not a forecast of value. Hardness is necessary for a store of value, but it is not sufficient.
The fiat era: predict, then check the data
Since the United States ended dollar-for-gold convertibility on August 15, 1971 (the end of the Bretton Woods arrangement), the dollar has been a pure fiat currency, backed by government and central-bank policy rather than a commodity. Hard-money advocates point to what happened to its purchasing power. Before you read the number, predict it.
How much of a 1971 dollar survives?
A dollar saved in 1971, under the mattress, buys how much in goods today? Drag to your guess, then reveal the figure from official inflation data.
About 12 to 13 cents. Using the US Consumer Price Index, one 1971 dollar has the purchasing power of roughly eight 2026 dollars, which means the 1971 dollar has lost on the order of 87 percent of its purchasing power (BLS CPI-U, as of mid-2026). You can check the exact figure with the BLS inflation calculator.
You will often see "the dollar lost 97 percent of its value." That figure is measured from 1913, when the Federal Reserve was created, not from 1971. Measured from 1971 the loss is closer to 87 percent. Both are real, but quoting the larger number against the 1971 date is a common error. Getting the base year right is part of reading the claim honestly.
The same era shows two other figures that hard-money advocates cite, and that are straightforward to verify:
- Money supply (M2): about $22.8 trillion as of April 2026 (Federal Reserve H.6 release), up from under $1 trillion in the early 1970s.
- Federal debt: about $39 trillion as of mid-2026 (US Treasury), up from roughly $0.4 trillion in 1971.
These numbers are not in dispute. What they mean is where the real argument starts, and that is the next section. Correlation between a policy regime and an outcome is not the same as proof that the regime caused it.
Side calculation: the portability tradeoff
Scarcity and durability are not free. Gold's weight is the price it pays on portability. Enter an amount to see it.
At a gold price of about $4,371 per troy ounce (spot, June 17, 2026). The same value in Bitcoin is carried by a 12 or 24 word seed phrase. That is the portability tradeoff in one comparison.
The strongest case against the hard-money view
A deep dive owes you the other side at full strength, not a strawman. Most professional economists do not favor a return to a gold-style standard, and their reasons are serious. Here are the strongest, each followed by how a hard-money advocate responds. Decide for yourself where each one lands.
1. A little inflation may be useful, and deflation can be dangerous
Mainstream macroeconomics generally targets low positive inflation (around 2 percent) for two reasons. Wages are "sticky" downward: in a downturn it is easier to let inflation trim real wages than to cut nominal pay, which workers resist. And a small inflation buffer keeps interest rates away from zero, leaving the central bank room to cut in a crisis. Under a fixed-supply money, a demand shock can tip into falling prices, and Irving Fisher's "debt-deflation" theory (1933) argues that falling prices raise the real burden of debt, causing defaults that push prices down further.
The hard-money response: mild, productivity-driven deflation (prices falling because goods get cheaper to make) is not the same as a debt-deflation spiral, and may be benign. Advocates also argue the "stickiness" and crisis-room benefits come at the cost of a permanent, hidden tax on savers, and that the flexibility is used to delay necessary adjustments rather than avoid them.
2. Economists overwhelmingly reject a gold standard
This is not a close call among experts. In a January 2012 survey of a geographically and ideologically diverse panel of leading academic economists (the Chicago Booth IGM / Clark Center Economic Experts Panel), not one agreed that replacing discretionary monetary policy with a gold standard would improve price stability and employment for the average American. Among those who answered, about 91 percent disagreed or strongly disagreed.
The hard-money response: expert consensus reflects the framework experts are trained in, and the question asked specifically about a government-run gold standard, which is not the same as a market-chosen money like Bitcoin that no government manages. Still, anyone reaching the opposite conclusion should be able to say why nearly every monetary economist disagrees.
3. The gold standard era was not a golden age of stability
The classical gold standard (roughly 1870s to 1914) had real virtues, but it also had frequent banking panics (1873, 1893, 1907) and sharp swings in prices, including a long deflation from 1873 to 1896. The era's stability is often overstated in hindsight.
The hard-money response: many of those panics trace to fractional-reserve banking and government intervention rather than to gold itself, and the long-run price level was far more stable than anything since 1971. Whether that defense holds is exactly the kind of thing this page wants you to investigate rather than take on faith.
4. Hard money has "no intrinsic value"
A common objection: gold and Bitcoin produce no cash flow and have little or no essential industrial use, so their value is purely what the next person will pay. On this view a monetary asset can fall a long way, even to zero, if belief in it evaporates, unlike a bond or a productive business with underlying earnings.
The response: since the marginal revolution of the 1870s, economists have largely set aside "intrinsic value" as a concept. Value is subjective and set at the margin by what people want, not baked into an object. Money is valued for its monetary properties and the network that accepts it, which is how gold held value for millennia with almost no industrial use. The honest caveat cuts both ways: a network effect that confers value can also reverse, so "no cash flow" is a genuine risk to weigh, not a settled non-issue.
The tradeoff, stated plainly
Strip away the advocacy on both sides and a genuine tradeoff remains. Hard money and flexible money optimize for different things. Neither is free.
What hard money offers
- Credible, predictable scarcity that no single party can dilute
- Protection of long-term savings from quiet debasement
- A unit of account that is not steered by policy
What hard money gives up
- No lender-of-last-resort flexibility in a liquidity crisis
- No policy lever to cushion shocks or fight unemployment
- For a young money like Bitcoin, high volatility during adoption
Your view on sound money is really a view on which side of this tradeoff you would rather live with, and for which purpose. Storing value across decades and managing a national economy through a recession are not the same problem, and may not want the same money.
Where Bitcoin fits, and what is still open
Against the eight properties and the S2F metric, Bitcoin makes a strong case as hard money: a fixed supply of 21 million, enforced not by a promise but by every node that validates the chain. Changing that cap would require near-unanimous agreement across node operators, miners, and the economy that uses it, and anyone who tried would simply create a separate fork the market is free to reject.
The 21 million cap, precisely
The cap is enforced by network consensus, not by any authority. To raise it you would need, in practice:
- Effectively every node operator to adopt the change
- An overwhelming majority of miners to follow
- The wider economy to accept the new coins as "bitcoin"
That coordination has so far proven impractical for supply increases, which is the source of Bitcoin's credibility on scarcity.
But honesty requires naming what is not settled:
- Volatility. Bitcoin has had multi-year drawdowns (2018, 2022). Over short horizons its purchasing power is anything but stable. The store-of-value case is a claim about the long run that is still being tested in real time.
- Scarcity is not the whole of "money." Some economists, including some Austrians citing Mises's own regression theorem, question whether fixed supply alone is enough to make something money rather than a scarce asset. That debate is live.
- The long-run security budget. As the block subsidy approaches zero, Bitcoin's security will depend on transaction fees. Whether fees alone will be sufficient is an open engineering and economic question.
None of these closes the case in either direction. They are the questions a serious learner should keep open while watching the evidence accumulate.
Three claims to handle with care
"Hard money means stable prices."
RealityHard money limits supply growth. It does not guarantee a stable price level, because demand and credit still move. The gold standard had notable price swings and panics. Predictable supply is not the same as predictable prices.
"Bitcoin is perfectly fungible."
RealityAt the protocol level every satoshi is equal, but every coin's history is public. Analytics firms and some exchanges treat coins differently based on where they have been, which is a real fungibility limit that physical gold does not share.
"The dollar's decline since 1971 proves Bitcoin will hold value."
RealityThe dollar's debasement is a fact about the dollar. Whether Bitcoin holds value over decades is a separate question that depends on future demand, and it has not yet been tested across a full lifetime of saving. One claim does not establish the other.
💭 Reflection questions
No answer key. These are meant to be argued in both directions.
1. In the matrix above, which properties did you keep, and which money won for you? Would a central banker managing a recession weight those properties the same way you did?
2. If hard money protects savers but removes the central bank's tools in a crisis, who gains and who loses from each choice? Are they the same people across a full economic cycle?
3. Nearly every monetary economist rejects a gold standard. Make the strongest version of their case in your own words before deciding whether you agree.
4. The stock-to-flow price model failed as a forecast. What does that tell you about using any single metric to predict an asset's value?
5. "Mild deflation from rising productivity is good; debt-deflation spirals are bad." Can a money have one without risking the other? How would you tell them apart in real time?
6. What single piece of evidence over the next ten years would most change your mind about whether Bitcoin is sound money? If you cannot name one, is your view falsifiable?
Sources and claim ledger
Data figures verified as of June 17, 2026. Empirical values drift; check the linked primary sources for current numbers.
- Purchasing power of the dollar since 1971 (about 87 percent loss; one 1971 dollar near eight 2026 dollars): US Bureau of Labor Statistics, CPI inflation calculator (CPI-U). The frequently quoted "96 to 97 percent" figure is measured from 1913.
- M2 money supply near $22.8 trillion (April 2026): Federal Reserve, H.6 Money Stock Measures.
- Federal debt near $39 trillion (mid-2026) and roughly $0.4 trillion in 1971: US Treasury, Debt to the Penny.
- Gold spot price about $4,371 per troy ounce (June 17, 2026), used for the portability calculation: public spot-price data (Kitco, World Gold Council).
- Stock-to-flow ratios (gold ~60, silver ~20, Bitcoin ~120): gold from World Gold Council above-ground stock and annual mine production; Bitcoin computed from circulating supply near 19.9 million and post-2024-halving issuance of about 164,000 BTC per year (Bitcoin protocol; mempool.space).
- Economist consensus against a gold standard (0 of 40 agreed; about 91 percent of respondents disagreed, January 2012): Chicago Booth IGM / Kent Clark Center, Gold Standard survey.
- Stablecoin and CBDC freeze capability (basis for the matrix censorship and seizure scores): stablecoin issuers Tether and Circle have publicly frozen tokens at law-enforcement and sanctions request (issuer transparency statements); CBDC designs are programmable by the issuing central bank by construction (central-bank design papers, e.g. BIS).
- Debt-deflation theory: Irving Fisher, "The Debt-Deflation Theory of Great Depressions," Econometrica (1933).
- End of dollar-gold convertibility, August 15, 1971; Bretton Woods (1944): standard monetary history.
- Austrian monetary theory: Carl Menger, "On the Origins of Money" (1892); Ludwig von Mises, The Theory of Money and Credit (1912); F. A. Hayek, "The Use of Knowledge in Society" (1945); Murray Rothbard, What Has Government Done to Our Money? (1963).
- The monetary-properties teaching synthesis (the classic short list of six, here extended to eight): Saifedean Ammous, The Bitcoin Standard (2018). A partisan source, cited as the origin of the framing, not as neutral authority.