I was recently asked whether accounting provides a useful way to answer a question that has surrounded Bitcoin since its creation: is it really money?

There is an important distinction to make at the outset. The question is not whether Bitcoin can be accounted for using double entry. Of course it can. Almost any asset can be accounted for in that way. If a company buys Bitcoin, it can debit an asset account for the Bitcoin acquired and credit its bank account for the cash (dollars, sterling, or whatever) used to purchase it. There is nothing particularly unusual about that.

The more interesting question is whether Bitcoin itself embodies the reciprocal asset-liability relationship that characterises all of money, without exception. To put this another way, if someone owns Bitcoin and records it as an asset, whose liability is that Bitcoin? Who owes them something as a consequence of their ownership? Answering that question is where things become interesting.

Double-entry bookkeeping is based upon a remarkably simple idea. Transactions have two sides, and accounting recognises them both. If one person owes, another has a claim. If one person pays, another receives. If one person has income, another has a cost. The debits and credits are not arbitrary devices invented to make accounts balance. They reflect the relationships that economic transactions create. They record reality. 

Money provides perhaps the clearest example of this. Suppose a commercial bank lends someone £100,000. The bank records the loan as an asset because the borrower owes £100,000 to it. At the same time that the bank records the loan, it also credits the borrower's current account with £100,000, which they can now spend. That is what the loan was meant to achieve; its purpose was to give the borrower purchasing power they otherwise did not have. That deposit is a liability of the bank because the bank owes that money to its customer. The loan and the deposit arise simultaneously as part of the same transaction. They balance. And that is how all money is created.

When looked at from the borrower's perspective, the transaction and the relationship are equally obvious. The borrower has acquired a £100,000 bank deposit, which is their asset, whilst simultaneously acquiring a £100,000 obligation to the bank, which is their liability. There are accounting relationships on both sides because there are economic relationships on both sides.

The important point is that the £100,000 of newly created bank money is not simply an object that has appeared from nowhere. It represents a claim. The customer's deposit is the customer's asset precisely because it is the bank's liability. Those two things cannot be separated.

The same basic point applies to central bank money. Central bank reserves are assets of commercial banks and liabilities of the central bank. Again, the monetary asset is simultaneously somebody else's liability.

This is why I have long argued that money is debt. Modern money records relationships between people, banks, companies and the state. Money is not simply a valuable object that happens to circulate between people. It is an accounting relationship denominated in a currency, and double-entry helps us understand what that relationship is.

Bitcoin, though, is fundamentally different.

Bitcoin has a ledger, in the form of its blockchain, but the fact that something has a ledger does not mean that it has a double-entry accounting system. The Bitcoin blockchain records transactions and establishes who has the right to control particular amounts of Bitcoin. Its architecture is designed, amongst other things, to prevent the same Bitcoin being spent twice. What it does not do is establish the debtor-creditor relationship that exists when conventional money is created.

Suppose new Bitcoin is created as a mining reward. The miner acquires Bitcoin because the Bitcoin protocol says that they should. But who simultaneously acquires a liability to the miner?

The answer is that no one appears to do so.

No bank owes the miner the value of the Bitcoin. No government promises to accept responsibility for it. No issuer stands behind it. No organisation has recorded the newly created Bitcoin as its liability. The miner has acquired an asset, assuming Bitcoin is appropriately described as such, without another economic entity acquiring the corresponding liability.

That does not mean the miner's accounting records are wrong. If the miner is a business, it will record the Bitcoin it receives in its accounts, and there will be an appropriate corresponding entry, which is presumably as income. Similarly, if another company subsequently buys that Bitcoin, it will record both the acquisition of the Bitcoin and whatever it gives up in exchange.

But those are double-entry accounting entries about transactions involving Bitcoin. They do not demonstrate that Bitcoin itself embodies a reciprocal accounting relationship. A company can use double-entry accounting to account for a building, a painting, a bar of gold, or a field. None of those things is somebody else's liability simply because the owner records it as an asset.

That is why I think the crucial question is not, strictly speaking, “Where is the credit when Bitcoin is created?” An accountant can create an appropriate credit when recording the receipt of Bitcoin. The more fundamental question is: if Bitcoin represents a monetary asset, what is the corresponding monetary liability, and who owes it?

There appears to be no answer to that question.

That fact puts Bitcoin into a very different category from modern money. Gold is not somebody else's liability. Nor is a painting, a building or a field. These things can all be assets because ownership does not necessarily require another person to owe the owner anything. Bitcoin appears to share that characteristic. In accounting terms, it looks much more like an asset of this sort than like money created through a credit relationship.

This distinction matters because money is necessarily relational. A pound in my bank account represents a claim that I have on my bank. My asset is its liability. If I own Bitcoin, nobody owes me anything simply because I own it. What I possess is the ability, subject to the rules of the Bitcoin system, to transfer that Bitcoin to somebody else. Its value therefore depends upon finding somebody else willing to accept it at an agreed price.

Bitcoin's supporters might say that this is precisely the point. Bitcoin was designed to remove the need for banks, governments and other institutions standing behind money. They say that the absence of an issuer is not an accidental omission from the Bitcoin architecture. For many of its advocates, it is one of Bitcoin's principal attractions.

But that design choice has consequences. Remove the issuer and there is no issuer's liability. Remove the liability and there is no corresponding claim against an issuer. Remove that reciprocal relationship and what remains is something quite different from the credit money that actually powers a modern economy.

Bitcoin can certainly be traded. Its price can rise and fall. People can speculate in it, save it and sometimes exchange it for goods and services. But none of those things settles the question of whether it is money. Gold can be traded and speculated upon. So can paintings, houses and countless other assets. Exchangeability alone does not define money.

Accounting therefore provides a useful test. If Bitcoin is claimed to be money, ask what monetary relationship it records. Ask whose liability a Bitcoin represents. Ask who has the corresponding obligation to its owner.

There is no obvious answer to any of those questions.

That does not mean Bitcoin does not exist, nor does it mean that it cannot have a price. It means something rather more specific, which is that Bitcoin does not appear to have the reciprocal asset-liability structure that characterises modern credit money.

Bitcoin might be a digital asset. But if money is fundamentally a credit relationship, as the accounting evidence strongly suggests it is, then Bitcoin fails that test. And that is why I do not think Bitcoin is money. So, what value does it have, and why? 

PDF of article