You use it every day and could probably not explain what it actually is. Most people can’t. It is the most important financial lesson nobody is taught, and everything else — inflation, saving, why your raise did not feel like a raise — sits on top of it.
The problem money was invented to solve
Money is a tool that solves one specific problem: the double coincidence of wants.
Before money, a fisherman who needed shoes had to find a shoemaker who wanted fish, at the same moment, in the same place. That constraint kept trade slow, local, and limited to communities small enough that everyone knew what everyone else needed. Money removed the friction by acting as a universal middleman — something everybody would accept, so nobody had to match wants directly.
Which raises the question the rest of this article is about: what makes something acceptable to everybody?
The three tests
Economists broadly agree that something is money if it does three jobs.
- Store of value — it holds purchasing power across time, so work done today can be spent years from now.
- Medium of exchange — people accept it in trade without asking what it is really worth.
- Unit of account — prices can be quoted in it, so things can be compared.
These are not arbitrary. They are what thousands of years of trial and error converged on, and the useful thing about them is that they are a test rather than a definition — you can apply them to anything and get an answer.
Apply them to gold and it passes all three, for millennia. Apply them to a gold-backed dollar and it passes all three. Apply them to today’s dollar and it passes two: universally accepted, universally used for pricing, and losing purchasing power every year as a matter of stated policy. That is not a controversial claim. Central banks target it.
Hold that result. It is the whole argument in one line: the money you hold fails one of the three tests, deliberately, and nobody mentioned it.
How money got here
The earliest economies ran on direct trade, with all the friction described above. What emerged next was not designed by anyone.
Commodity money. Cattle, shells, salt and eventually metals became widely accepted because they shared the same traits — durable, divisible, portable, scarce enough to hold value. Salt was used so widely that the word salary descends from the Latin salarium, payment in salt. No government invented any of this. Millions of people independently solved the same problem and arrived at similar answers.
Gold and silver. Gold won the long competition: it does not corrode, it takes a coin’s shape easily, new supply arrives slowly, and it is recognized across cultures that share nothing else. Silver served as the smaller denomination for daily purchases.
The genius of a gold system was not the metal. It was the discipline. A government could spend what it taxed or what it borrowed, and not much more. When Roman emperors tried to escape that by mixing cheaper metals into their coins, they got inflation, economic decline and eventually the end of the empire. The lesson repeated for centuries and was rarely learned.
Paper and the promise. Paper money began as convenience, not deception. Carrying gold was heavy and dangerous, so banks issued receipts for gold held in a vault. The paper itself was worthless; its value was the claim it represented, and anyone could walk in and redeem it.
That worked exactly as long as the promise held. The temptation to issue more receipts than gold proved irresistible to every government that ever had the option. France broke it with the assignat in the 1790s. Weimar Germany broke it in the 1920s. The United States broke it in 1971.
What happened in 1971
On August 15, 1971, President Nixon ended the dollar’s convertibility into gold. Overnight the dollar became fiat — from the Latin for “by decree,” money that has value because the state says so. Since that day nothing physical has backed it. Its value rests on trust in the United States government and on the Federal Reserve’s management of the supply.
The consequence is measurable rather than theoretical. The dollar has lost roughly 87 percent of its purchasing power since then — not because goods got better, but because each dollar became a smaller share of a much larger pile. Federal Reserve data put M2 at under $1 trillion in 1971 and above $21 trillion by 2024.
Worth sitting with how unusual this is. For roughly five thousand years of recorded monetary history, money was a thing — salt, silver, copper, gold. For the last half-century it has been a number in a ledger, managed by institutions with no hard limit on how many they can add. That is the experiment we are all inside, and it is younger than most of the people running it.
Sound money and fiat money
Austrian economists draw the line at supply control. Sound money has a supply no single authority can expand at will. Fiat money can be created in whatever quantity the issuer chooses, and issuers consistently choose to expand it — to fund spending, to manage downturns, to avoid the politics of taxing.
Each new unit dilutes every existing one. That is the mechanism behind the hidden tax called inflation, and it falls hardest on the people holding cash rather than assets: savers, retirees, and anyone whose wages move slower than prices. Who receives the new money first, and who receives it last, is its own distributional question.
Putting Bitcoin through the same test
“Bitcoin isn’t real money” is a claim you will have heard from a friend, a news anchor or an advisor. The better move is not to argue about it but to run the same three tests that gold and the dollar just went through.
Store of value. Bitcoin has a fixed cap of 21 million coins. No person, company or government can raise it. New coins arrive through mining on a schedule that halves roughly every four years until issuance ends around 2140, which makes the supply the most rigidly defined of any monetary asset. Gold’s supply grows about 1.5 percent a year from mining; Bitcoin’s issuance rate is below that and falls further at each halving.
The standard objection is volatility, and it is a fair one. A thing that moves twenty percent in a month is doing a poor job of storing value over that month. The counter-argument is that a monetary asset still being priced in by the market is volatile while that repricing happens — gold swung violently for three decades after 1971, from $35 to $850 to $250, before settling into the role people now assume it always had. Whether Bitcoin is on the same path is exactly the open question, and anyone who tells you the answer with confidence is selling something.
Medium of exchange. Bitcoin settles to anyone, anywhere, without a bank, a processor or permission, with final settlement rather than a reversible promise. The Lightning Network makes small payments practical at near-zero cost.
Is it widely accepted? Not remotely as widely as the dollar, and the honest picture includes retreats as well as advances. El Salvador adopted Bitcoin as legal tender in 2021 and then reversed that status in early 2025 under an IMF arrangement, making merchant acceptance voluntary. Treating that as a footnote would be dishonest; it is the clearest real-world test to date of mandating a medium of exchange, and mandating it did not work.
Unit of account. This is the weakest of the three by a distance. Almost nobody prices goods in bitcoin or satoshis. Some Bitcoin-native businesses do, and the community thinks in sats, but that is a rounding error against global commerce. Historically the unit-of-account function is the last one a new money acquires, long after the other two — so its absence tells you where Bitcoin is in the sequence rather than whether the sequence will finish.
What the test settles, and what it doesn’t
On the three tests as economists have applied them for five thousand years: Bitcoin clearly passes one, partially passes one, and largely fails one. Today’s dollar clearly passes two and fails one. Gold under a gold standard passed three.
That is a more interesting result than either side of the argument usually wants, because it does not hand anyone a win. It says the dollar and Bitcoin fail different tests — and which failure matters more depends entirely on what you are using the money for. If you are buying lunch, unit of account and acceptance are everything and the dollar is obviously better. If you are trying to move purchasing power thirty years into the future, the test the dollar fails is the only one that counts.
What the framework does not do is tell you what to own. It is a way of asking a better question than “is this real money” — namely, which job am I asking this money to do, and does it pass the test for that job? That question you can actually answer about your own situation. Nobody can answer it for you, and this is education rather than advice.
Keep reading
- How money loses value — the debasement mechanism in detail
- The dollar since 1971 — the measured record
- The Cantillon effect — who gets the new money first
- When governments control money — what the historical record shows
- Halving House — the rest of the sound money series
Monetary history and supply figures as cited; Federal Reserve M2 data as of 2024 and El Salvador’s legal-tender reversal confirmed against reporting of the 2025 IMF arrangement, read 2026-09-20. This is educational writing about how money works, not investment advice.