Update 11 September 2014: My views on digital currencies, including Bitcoin, have evolved somewhat since this post. Interested readers might care to read two new Bank of England articles  on the topic. I was a co-author on both.
Original post is below …
Discussion of it is everywhere at the moment.
- Technically, it’s magnificent. It overcomes some technical difficulties that used to be thought insurmountable.
- As a medium of exchange, it’s an improvement over previous currencies (through the anonymity) for at least some transactions
- As a store of value (i.e. as a store of wealth), it offers nothing [see below]
- There are already many, many well-established assets that represent excellent stores of value, whatever your opinion on inflation and other artefacts of government policy
- Therefore people will, at best, store their wealth in other assets and change them into bitcoins purely for the purpose of conducting transactions
- As a result, the fundamental value of a bitcoin rests only in the superiority of its transactional system; for all other purposes, its value is zero
- For 99.999% of all transactions by all people everywhere, the transaction anonymity is in no way superior to handing over physical cash or doing a recorded electronic transfer
- Therefore, as a first approximation, bitcoin has a fundamental value of zero to almost everybody and of only slightly more than zero to some people
- Therefore, the recent massive run-up in the bitcoin-to-Any-Currency-You-Care-To-Name exchange rate is a pure bubble and last Friday’s 30% drop in the bitcoin-USD exchange rate  will not be the last
This thing is only ever going to be interesting or useful to drug dealers and crypto-fetishists. Of those, I believe that drug dealers will ultimately lose interest because of a lack of liquidity in getting their “money” out of bitcoins and into hard cash. That only leaves one group …
A note on money as a store-of-value: When an asset pays out nothing as a flow profit (e.g. cash, gold, bitcoin), then that asset’s value as a store-of-value  is ultimately based on a) the surety that it’ll still exist in the future and b) your ability to convert it in the future to stuff you want to consume. Requirement a) means that bread is a terrible store of value — it’ll all rot in a week. Requirement b) means that a good store of value must be expected to have strong liquidity in the future. In other words, there must be expected future demand for the stuff. If you think your government’s policies are going to create inflation, putting your wealth in, say, iron ore, will be an excellent store of value because the economy at large will (pretty much) always generate demand for the stuff.
That makes gold an interesting case. Since there isn’t really that much real economic demand for gold, using it as a store of value in period T must be based on a belief that people in period T+1 will believe that it will be a good store of value then. But since we already know that it has very little intrinsic value to the economy, that implies that the T+1 people will have to believe that people in period T+2 will consider it a store of value, too. The whole thing becomes an infinite recursion, with the value of gold as a store-of-value being based on a collective belief that it will continue to be a good store-of-value forever.
Bitcoin faces the same problem as gold. For it to be a decent store-of-value, it will require that everybody believe that it will continue to be a decent store-of-value, and that everybody believe that everybody else believes it, and so on. The world already has gold for that purpose (and gold has at least some real-economy demand to keep the expectation chain anchored). I’m not at all sure that we can sustain two such assets.
 All currencies are assets. They’re just don’t pay a return. Then again, neither does gold.
 Yes, yes. Saying that it’s “value as a store-of-value” is cumbersome. It’s a definitional confusion analogous to free (as in beer) versus free (as in speech).