Three Jobs of Money, and Bitcoin Fails All Three: Yermack's Verdict

Bitcoin is called a currency. It says so in the whitepaper: "a peer-to-peer electronic cash system." The word cryptocurrency has the word currency inside it. The industry calls it money.

David Yermack, an unglamorous and rather effective approach, decided to simply check that claim against the definition of money that economists have used for well over a century.

The definition is not controversial and it is not new. To be money, a thing must do three jobs:

1. Medium of exchange. People use it to buy things.
2. Store of value. It holds its purchasing power over reasonable stretches of time.
3. Unit of account. People quote prices in it and keep their books in it.

Yermack took Bitcoin through each test in turn.

The problem: a claim nobody had audited against the textbook

The rhetoric around Bitcoin in its early years was overwhelmingly monetary. It would replace the dollar. It would free people from central banks. It was the future of payments. Central bankers were asked about it. Economists were asked about it. Almost nobody had bothered to do the boring, methodical thing: hold it against the standard definition and see whether it fits.

It matters, because the label determines almost everything downstream. If Bitcoin is a currency, then it should be evaluated on stability, on transaction usage, on whether it can support an economy. If it is a speculative asset, it should be evaluated on returns and volatility, and it should be regulated, taxed and thought about accordingly.

The key idea via analogy: does the tool do the job?

Yermack's method is essentially inspection. Take the tool, look at what it is supposed to do, and check.

Test one: medium of exchange. Bitcoin barely functions.

For a currency, the number that matters is: how much stuff do people actually buy with it? Yermack looked at merchant transaction volume and found it was tiny. Averaged across the small set of merchants who accepted Bitcoin, actual consumer transactions worked out to well below one per day. Not one per day per merchant per hour. Below one per day, full stop.

The reason is that Bitcoin's other properties actively sabotage this use. Confirmation takes minutes, which is fine for a wire transfer and hopeless for a coffee. Fees are unpredictable and sometimes enormous. And the volatility means both parties to a transaction are taking a currency bet they did not want. Most merchants who accepted Bitcoin used a payment processor that instantly converted it to dollars, which tells you exactly how much confidence they had in holding it.

Test two: store of value. Bitcoin fails badly.

Yermack measured Bitcoin's volatility and compared it against real currencies. The gap is not close. Bitcoin's volatility is vastly higher than that of widely used currencies. Real currencies move by fractions of a percent on a normal day. Bitcoin routinely moved by many percent, and sometimes by tens of percent.

This is a genuinely disqualifying property, and not just as a matter of taste. If you are paid in a currency that can lose a third of its value in a week, you cannot plan. You cannot sign a contract. You cannot take out a mortgage. The whole point of a store of value is that it lets you move purchasing power across time without gambling, and an asset with equity-like volatility does not do that.

Yermack also examined Bitcoin's correlation with major currencies and with gold, and found essentially none. That is the point crypto advocates like to cite as evidence of diversification value, and it is a fair point. But for the currency question, it cuts the other way: a currency that has no relationship with any real currency, with gold, or with anything else in the monetary system is not behaving like money. It is behaving like a speculative asset with its own weather.

Test three: unit of account. Bitcoin is not used.

This is the test people forget, and it is arguably the most damning. Nobody quotes prices in Bitcoin. Merchants who "accept Bitcoin" price their goods in dollars and convert at the moment of sale. Nobody's salary is denominated in Bitcoin. No company keeps its books in Bitcoin. Even inside the crypto industry, profits and losses are counted in dollars.

And the everyday practicalities are awkward in a way that is easy to overlook: the pricing of ordinary goods requires many decimal places, and Bitcoin's own denomination conventions have never settled. A currency whose users cannot agree on where to put the decimal point is not doing the unit of account job.

Why it mattered

• It provided the sober benchmark. In a period of extraordinary hype, someone did the boring work of holding Bitcoin against the standard and reporting the result, in a form that central bankers and regulators could cite. That is a real service.
• It reframed Bitcoin correctly and early. Yermack's conclusion, that Bitcoin functions much more like a speculative investment than like a currency, has aged extremely well. That is essentially what it turned out to be. The people arguing Bitcoin would replace the dollar have gone quiet. The people describing it as digital gold or a risk asset have not.
• It shaped the regulatory and tax framing. Treating crypto as property or as a commodity rather than as currency has enormous consequences for taxation, accounting and regulation, and the intellectual case for that treatment rests on exactly the analysis Yermack laid out.
• It made volatility the central objection. Of all the criticisms one could level, Yermack identified the one that is both quantitative and structurally hard to escape. Bitcoin's fixed supply schedule, which its supporters celebrate, is precisely what guarantees that all demand shocks show up in the price. A currency with a rigid supply must be volatile. The feature and the flaw are the same thing.

The honest limitations

• The data are from Bitcoin's infancy. The paper examines a young, thin, largely retail market. Bitcoin's volatility, while still far above any real currency, has declined as the market has matured and institutionalised. The direction of the criticism holds; the magnitudes have moved.
• The three-part definition is a choice, not a law of nature. It is the standard textbook framing, but a reasonable person can argue that an asset can be economically important without satisfying all three legs, and that Bitcoin is better understood as a new category rather than as a failed currency.
• It does not address the store-of-value case on a long horizon. Yermack's volatility critique is about short-term risk. The "digital gold" argument is about decades, and the paper does not engage with whether an extremely volatile asset can nonetheless preserve value over very long periods. (Liu and Tsyvinski's later work casts doubt on the gold analogy for other reasons.)
• The medium-of-exchange failure may be a layer problem. Payment systems built on top of Bitcoin, and other blockchains designed for throughput, address some of the transaction speed and cost problems. Judging the base layer's transaction capacity is not the same as judging the whole ecosystem's.
• It is an appraisal, not a valuation. The paper tells you Bitcoin is not a currency. It does not tell you what Bitcoin is worth, and "not a currency" is entirely compatible with "extremely valuable."

The one-line takeaway

Yermack ran Bitcoin through the three standard tests of money, found that almost nobody spends it, its volatility makes it useless for storing value, and nobody prices anything in it, and concluded that whatever Bitcoin is, it is a speculative asset rather than a currency, a verdict that the decade since has largely confirmed.