What Are the Four Functions of Money — and How Does Bitcoin Score?

Ask ten people what money is, and most will say roughly the same thing: it's what you use to buy stuff. That's one job. Money has four.

In 1875, the British economist William Stanley Jevons wrote them down in Money and the Mechanism of Exchange, and the list has held up for 150 years. Money is a medium of exchange (you trade it for other things), a unit of account (you price other things in it), a store of value (you can hold it without it evaporating), and a standard of deferred payment (you can sign a contract today that pays back in it tomorrow). Anything that does all four is money. Anything that does some but not others is partial money — useful at one thing, useless at another, and the arguments about whether it counts as “real” money usually come down to which function the arguer cares about.

This is also the framework most Bitcoin arguments implicitly use without naming it. “Bitcoin is too volatile to be money” is a complaint about function three (store of value over short windows). “Nobody prices things in Bitcoin” is a complaint about function two (unit of account). “Nobody spends Bitcoin” is a complaint about function one (medium of exchange). They're separate critiques. They sound like one critique only because the framework is missing.

Scoring Bitcoin honestly on each function turns out to be more useful than picking a side. The answers aren't all “yes” and they aren't all “no.” Money has never been a binary.

The four functions, in plain English

Medium of exchange is the function people think of first. It's what makes money a step up from barter. Without money, a shoemaker who wants bread has to find a baker who needs shoes — and who wants shoes today, in the size the shoemaker happens to have made, for the quantity of bread the shoemaker is willing to accept. With money, the shoemaker sells shoes to whoever wants them and uses the proceeds to buy bread from whoever sells it. The economist's term for what money removes is the double coincidence of wants. The intuitive term is friction.

Almost anything can serve as a medium of exchange under the right conditions. In World War II POW camps, cigarettes did the job: small, durable, divisible, universally desired. Nobody issued them as currency. Nobody backed them with gold. They passed the test by being acceptable — and that is the entire test. A medium of exchange is whatever the parties on both sides of a trade agree to hand back and forth.

Unit of account is the function people notice least, but it might be the most important. It's the denomination people think in. In the United States, prices are quoted in dollars. In Japan, yen. In the Eurozone, euros. The unit of account is the mental ruler.

A money can be a strong medium of exchange and a weak unit of account. Black-market dollars circulated in the Soviet Union for decades, but Soviet citizens priced things in rubles, even when they paid in dollars. The dollar was a tool; the ruble was the language. The unit of account is the function that's stickiest, because it lives in people's heads, not in any institution.

Store of value is the function under attack any time inflation shows up in the news. It's the money's ability to preserve purchasing power between when you earn it and when you spend it. A money that loses its purchasing power between Monday and Friday isn't storing much. The Confederate dollar is the textbook failure case: a functioning medium of exchange and unit of account inside the Confederacy in 1862, essentially worthless by 1865 — prices in the South rose more than 9,000% over the war, and a Confederate dollar at war's end bought less than two cents in gold. Holding it didn't preserve anything.

Standard of deferred payment is the function that gets the least press, because it lives inside contracts. A 30-year mortgage, a 10-year corporate bond, a 5-year lease — each of those is a promise to deliver a specific quantity of money at a future date. The function asks: is the money stable enough that “$1,000 in 2035” still means something predictable in real terms?

Argentina is the classic case where this function fails. Argentine real estate is routinely priced and traded in U.S. dollars inside Argentina; rental contracts now legally allow payment in foreign currency following Milei's December 2023 deregulation, and many leases are written directly in dollars. Before the 2023 repeal, the workaround was inflation-indexed peso contracts. Different mechanism, same root problem: the peso isn't stable enough for two parties to agree on what a multi-year deal is worth in real terms. The contract layer breaks before the spending layer does.

Bitcoin's strongest function: store of value

Of the four functions, this is where Bitcoin's case is most concrete.

Over its lifetime since the genesis block in 2009, every rolling 4-year holding period for Bitcoin has produced a positive return. Every cohort. The weakest of those positive returns still substantially outpaced the dollar's loss to inflation over the same window.

That “every cohort” claim includes some uncomfortable cases.

Consider the buyer at the December 2017 peak, around $20K. The 2018 bear market cut their position more than 80%. They held anyway. Four years from entry, the price was near $47K. Or the buyer at the November 2021 peak, around $69K. The 2022 crash wiped out about two-thirds of their position. They held anyway. Four years from entry, the price was around $83K. Every other “worst-timing” cohort falls between those two.

Even those worst-timed buyers ended up with more real purchasing power than someone who held dollars over the same window. Over four years, the dollar has been losing 15-20% of its purchasing power. Bitcoin doesn't have to soar to win that comparison. It only has to not shrink.

Over long enough horizons, the dollar shrinks; bitcoin's purchasing power, measured in dollars, has expanded. The direction has held across every 4-year window so far.

The protocol-level case is what makes the historical case credible going forward. Bitcoin's supply is capped at 21 million coins, written into the consensus rules at genesis, enforced by every node on the network. The issuance schedule is fixed and halves every four years until it terminates near the year 2140. There is no central bank, no policy committee, no discretionary authority. The thing being “stored” is denominated in a unit that mathematically cannot be diluted.

The honest counterargument is volatility. Bitcoin's price against the dollar swings by 5% in a day routinely and 50% in a year occasionally. Over a week, it is not a stable store of value. Over a decade, the story is different — but anyone who needs predictable purchasing power next week should not be relying on Bitcoin for it. Volatility is a function of monetization stage, not a permanent feature, and the relevant timeframe matters. If you're saving for retirement in 30 years, Bitcoin's 7-day volatility is noise. If you're paying rent on the 1st, it's the whole story.

Bitcoin's mixed function: medium of exchange

The base layer of the Bitcoin network can process roughly 7 transactions per second, with blocks settling every 10 minutes. That isn't anywhere close to what the world's everyday retail and online payments would need. Visa, by comparison, processed roughly 293 billion payment transactions in 2024 — about 9,000 per second on average, with peak-load capacity in the tens of thousands. The base-layer limitation isn't an oversight; it's a deliberate trade-off. Every node on the network has to validate every transaction, and the only way to keep that feasible for ordinary computers is to keep the throughput modest.

The trade-off is intentional because Bitcoin pushes everyday payments to a second layer. The Lightning Network is a network of payment channels built on top of Bitcoin that settles transactions in fractions of a second, at sub-cent fees, with the base layer used only to open and close channels. Real adoption is happening: Strike uses Lightning for instant cross-border remittances at a fraction of incumbent fees; Cash App enables peer-to-peer Bitcoin transfers for tens of millions of U.S. users; Steak 'n Shake added Lightning checkout at all U.S. locations on May 16, 2025; El Salvador made Bitcoin legal tender in 2021 (a status the government substantially walked back in early 2025 under an IMF agreement, but the experiment generated several years of real-world retail data).

What hasn't happened, yet, is that ordinary people in the United States or Europe routinely buy coffee with Bitcoin. The infrastructure exists. The habit doesn't. Medium-of-exchange is a function that develops slowly because it requires both sides of a transaction to be ready at the same time — every additional merchant accepting Bitcoin is one more place a Bitcoin holder can spend without converting; every additional Bitcoin holder is one more customer the merchant might attract. The chicken-and-egg problem is real but not permanent.

Bitcoin's weak function: unit of account

Watch any Bitcoin holder describe their position and listen for the unit. “I have about $20,000 of Bitcoin.” “It's up 40% this year.” “That coffee would cost about $5 in BTC.” Even the most committed Bitcoiner converts mentally to fiat before speaking. The unit of account is the function that lives in people's heads, and most heads still think in dollars.

There is an emerging exception, and it's part of why this site is denominated in satoshis. As bitcoin's per-unit price grows, the base unit becomes unwieldy for everyday prices. A $5 coffee at recent BTC prices is some tiny fraction of a bitcoin — a number with so many leading zeros that human numeracy gives up. The same coffee priced in satoshis (each one a hundred-millionth of a bitcoin) is a few thousand sats — a number a human can hold. A companion article, Sats vs Bitcoin, walks through why the unit shift from “bitcoin” to “sats” is already happening for everyday prices, and why it doesn't depend on any price prediction.

The unit-of-account function is the last to develop in any monetary transition. People think in whatever money they already use every day. New monies get converted into the old unit, not the other way around.

Demanding that Bitcoin function as a unit of account today asks for something that didn't happen even for the dollar on the dollar's own timeline. Decades after American independence, U.S. merchants were still pricing things in Spanish pieces of eight. The unit takes its time.

Bitcoin's weakest function: standard of deferred payment

Almost no consumer contract in the world is denominated in bitcoin. Mortgages are fiat. Salaries are fiat. Leases, bonds, insurance policies, child support orders — all fiat. A small number of Bitcoin-collateralized loan products exist (Unchained Capital, Strike's Bitcoin-backed lending, on-chain DeFi platforms), but they're niche.

The tension is volatility's mirror image of the store-of-value strength. A 10-year loan denominated in BTC means dramatically different things to borrower and lender depending on which way BTC moves. If BTC's purchasing power doubles, the borrower owes twice as much in real terms — that is, the bitcoin they have to pay back will buy twice as much stuff as the bitcoin they borrowed. If it halves, the lender gets back half. Two parties trying to write a multi-year contract in bitcoin today have to agree on something they fundamentally don't know.

This function will be the last to flip. Bonds and long-term contracts follow the unit-of-account function with a multi-decade lag, because they require both parties to agree that the unit is predictable enough for the time horizon involved. Until volatility decreases substantially — which it has been doing as the market deepens, but remains higher than fiat — the standard-of-deferred-payment function will remain Bitcoin's weakest.

Why scoring “out of 4” misses the point

The temptation is to add up the scores. Strong on one, weak on two, mixed on one — does that mean Bitcoin is one-quarter money, or one-half money, or none of those?

It means it's a new money in the second stage of a four-stage adoption pattern, and that's exactly the shape new monies have when they're young.

The pattern is older than Bitcoin. Long before any cryptocurrency existed, the historical record of how new monies emerge — from shells and beads to silver and gold — followed a recognizable arc: collectibles became stores of value, became media of exchange, became units of account, in that order, over generations or centuries. The computer scientist Nick Szabo laid out this arc in his 2002 essay Shelling Out, drawing on archaeology and anthropology rather than crypto. Vijay Boyapati applied the same sequence to Bitcoin explicitly years later in The Bullish Case for Bitcoin. What's notable for our purposes isn't who named the stages — it's that the order has held across every monetary good that ever made the transition.

The four functions emerge sequentially, not simultaneously. They have to. A money that isn't yet credible as a store of value can't plausibly serve as the unit a 30-year bond is written in. A money nobody trusts to hold for a year can't be the denomination people think in.

Gold's transition through these stages took millennia. The U.S. dollar took most of the 19th century to displace Spanish pieces of eight as the unit-of-account inside the United States. Bitcoin's first block was mined in January 2009 — barely more than a decade and a half ago. By the historical pattern, the store-of-value stage is exactly the stage it should be in, with the medium-of-exchange stage just beginning to emerge.

The framework doesn't say Bitcoin will complete the transition. It says the shape of its current scoring is the shape new monies have during this part of their adoption curve. Whether it makes it the rest of the way is an open question. Mixed scoring isn't a contradiction. It's a snapshot.

What the clock measures: the first crossing

This site has dozens of basket items on it — coffee, eggs, rent, a movie ticket, a dental cleaning, the median new car. Every item has a dollar price today and a calculation underneath: how many satoshis does it take to buy this thing today, and when will that number cross the dollar price threshold? When it does, the item is at parity.

The clock isn't measuring whether Bitcoin is “money.” It's measuring one specific function — store of value — item by item. The implicit question is: which items pass the threshold first, and how fast does the rest of the basket follow? Every item card is the same question, applied to one good. The headline countdown is the median.

If Bitcoin's store-of-value function keeps strengthening, the medium-of-exchange function will follow. If the medium-of-exchange function follows, the unit-of-account function comes after it. And if the unit-of-account function arrives, the standard-of-deferred-payment function eventually catches up — though by then, the people writing 30-year contracts in sats won't be talking about whether Bitcoin is money. They'll just be using it.

The order has held for every new money that ever made it the whole way. The clock measures the first crossing.