« Central Bank Communication and the Term Structure | Main | Health Spending and System Characteristics in Canada and Spain »


Feed You can follow this conversation by subscribing to the comment feed for this post.

For example 1, if Bitcoin is universal medium of exchange, why would someone sell Bitcoin for Canadian dollars? How does the the CAD stay relevant, even as a store of value, when its liquidity is no longer sought after?

louis: In Case 1 Nobody sell Bitcoin for Canadian dollars, and nobody uses Canadian dollars as a store of value, because Canadian dollars don't exist. In Case 2, "Canadian Dollar" is just a word, which people use to talk about prices because everyone else uses that word to talk about prices, and everyone uses that word to mean what the Bank of Canada says it means, because everyone else does. "Litre" is just a word too. Litres don't actually exist. We sell gas in $ per litre.


This doesn't directly address your post and bitcoin specifically, but more for a central bank issued crypto-currency. Taken more or less verbatim from an argument I was having on a WhatsApp group, with some light editing:

1. Miners create "money" through validating transactions. For Bitcoin, there's two essential features of interest - the price of validation, and the speed of validation. Both are regulated by the community and the mining software. Both are fixed at any given point of time, though the former is reduced at discrete intervals.

2. That implies that there is no supply "curve" for Bitcoin. It's a vertical line (or horizontal, depending on how you arrange the axes). Higher demand for transactions won't induce more suppliers to enter the market, since the price is fixed. Price per block is fixed at any point in time, but decreases as more coins are mined. Increasing [miner] revenue depends on increasing the number of transactions validated.

3. If that's the case, demand and supply of coins are regulated by two things - speed of transaction (difficulty), and the value of the coin itself.

4. From the characteristics, there will be an upper limit to supply at any given point in time. As long as demand intersects with the supply curve, no issues. Coins will be mined at the rate demanded, but the price will vary according to demand. If however it exceeds the available supply of miners and their hardware, transaction demand will be rationed via even higher prices of coins, or delays in validation

5. How would a central bank manage such a system? I think in two ways.

6. Keep a coin reserve to regulate market liquidity (coin price), which is what I think some in the community are advocating. But this is effectively an adoption of Bretton Woods, with all that implies - a Keynesian world where monetary policy doesn't exist, and demand management is through fiscal policy and real economy adjustments. There's always the risk of running out of "reserves", causing periodic crises a'la Greece or the Asian Financial Crisis.

7. Another way to regulate such a system is to dynamically vary the price per block and/or the speed of transactions to influence the supply of miners. In other words, a digital analogue of the current monetary system, but with speed and transaction profits as supplements/replacements for policy interest rates. This removes the primary attraction of having a decentralised clearing system, but still reduces some of the counterparty risk. Instead of open market operations to adjust liquidity, central banks adjust difficulty/profit rate of transactions daily, or even by the minute.

8. One way to get both the advantages of cryptocurrencies, while avoiding deflation/volatility/rationing - raise the limit on mining enough to cater for both growth, as well as variance in transaction volume. Something like 7%-9% a year. In other words, build in a little inflation.

hishamh: I am not very knowledgeable about the technical details of Bitcoin; I leave that to the computer people who are more keen on it. So don't trust my answer to your question. But this *might* help.

When we draw a supply curve we put price on the vertical axis (by convention, even though it's sorta the wrong way round, for historical reasons). But when we are talking about a durable good, like houses or Bitcoin, what precisely do we put on the horizontal axis? We could put the stock of houses, or the flow of houses built per year (or per month or week or whatever). It is perfectly possible that the flow supply curve could be horizontal (perfectly elastic) but the stock supply curve slope up (or vice versa). Think about building houses in a U-shaped valley (only a 3D valley) for example, where the steeper the land the costlier it is to build. You start out in the bottom of the valley, where the land is flat, but as the stock of houses increases the cost of building new houses gets bigger as the land you are building on gets steeper.

I *think* the stock and flow supply curve(s) for Bitcoin are a bit like that. The flow supply curve is horizontal, but it shifts up as a function of the stock.

(If electricity is the main cost, I have never understood why Bitcoin isn't mined in the same places aluminum is produced -- near big hydro electric dams.)

It would be good to have a money that has a horizontal stock and flow supply curves, because it means the price would not change if demand shifted. But that only works properly if the flow supply curve extends into the negative quadrant, so the flow can be negative and the stock can fall over time if demand falls. We "eat" the money. Something like your point 7 could maybe do this (though I'm not sure how that could make the flow negative). My point 2 (Dictionary Money) would I think be a better and easier way to do this.

"In Case 1 Nobody sells Bitcoin for Canadian dollars, and nobody uses Canadian dollars as a store of value, because Canadian dollars don't exist"
Then how would the Bank of Canada buy bitcoin to reduce amount in circulation, if it has nothing of value to exchange for bitcoin?

"In Case 2, "Canadian Dollar" is just a word, which people use to talk about prices because everyone else uses that word to talk about prices, and everyone uses that word to mean what the Bank of Canada says it means"
This is pretty powerful, especially given the role of contracts in modern economic life. But I wonder about the limits on the Bank's flexibility. For example, the Venezuelan gov't can declare that a Bolivar is worth $1, but that doesn't mean anyone really respects that. What circumstances would cause the Bank of Canada to lose traction in managing macroeconomic targets?

Hi Nick, good post. Example 1 sounds a lot like how David Glasner describes the gold standard, with large nations (like France) capable of throwing the whole world into a recession if they try and buy large amounts of gold.

Its easy to see how case 1 works. The CB varies the price of bitcoin relative to other goods by varying the quantity of bitcoin.

Case 2 seems to have the CB varying the price of bitcoin relative to other goods by decree with no change in supply and demand conditions for either. If the ratio of bitcoin to other goods that the public wants to hold has not changed - won't the price level (expressed in CB dollars) just change to reestablish the original bitcoin/other goods ratio ?

Is there a flaw in Case 1?

The idea presented (in Case 1) is that the CB will trade assets for bitcoin, thus reducing the supply of bitcoin. My question is where the CB might get assets? Without any tax base and without the ability to print money, I can't help but think that the CB would no longer have any source of asset increase.


louis: I don't know. But if the Bank of Canada did Dictionary Money at purely random, every morning doubling or halving the meaning of the word "dollar" in terms of Bitcoin, people might stop using the word "dollar" and switch to Bitcoin language instead, because it's less hassle. It's sorta similar to what we sometimes observe in hyperinflations, only in reverse, where menu prices get quoted in the more stable currency, regardless of which currency is used as medium of exchange.

JP: My example 1 is based very much on my understanding of what David Glasner has been saying about the Bank of France. And I was searching on his blog for a simple post that I could link to to refer readers to. But my search skills aren't very good, and I couldn't find the particular post I was looking for. And I was scared of saying "David said this..." without a particular example so readers could see for themselves whether I was understanding him right. But let me know if you find a good one, so I can update.

MF: it all depends on whether $ prices are sticky or perfectly flexible. If perfectly flexible then yes. If sticky (menu costs etc.) then we get real effects.

Roger. Yes that's a problem, because central banks might need to use all their current assets to buy back their currency to prevent it becoming worthless when demand for it falls to zero when Bitcoin takes over. The government (taxpayer) is the ultimate answer.

Thanks. So assuming that nothing else changes and that in the long term prices are flexible then case 1 would lead to a permanent change in prices expressed in bitcoin, while in case 2 prices (expressed in bitcoin) would eventually go back to where they started as a result of prices (expressed in dollars) adjusting ?

MF: If prices are sticky in the short run but flexible in the long run, then eventually, in the long run, all prices expressed in dollars (roughly) double, and all prices expressed in Bitcoin stay (roughly) the same.

Maybe this one ?


maybe this one ?


Nick, it's hard to find a small and succinct post, but here are two:


JKH and JP: thanks. I have updated the post. Sorry you got caught by the spam filter.

The US CB yesterday announced that it would let $10 Billion in bonds held roll off when they expired. In other words, the CB will give the bonds back to the borrowers in exchange for cash.

From the macro perspective, where might this cash come from? I think it will come from reserves held by the CB in the name of member banks. Now try to extend this to the micro economy and we notice that bank reserves represent deposits owned by bank customers.

This background makes me think that the net result of this CB policy will result in less cash (in the form of less bank deposits) in the hands of the public. This is a form of monetary tightening and should result in higher interest rates.

Those of us who think that "banks create money when they lend" expect to watch federal government borrowing remain unchanged. This will leave the rate of money creation nearly unchanged at the same time that the CB is removing cash. Not a problem if we also expect the velocity of money to increase. In other words, a smaller amount of cash will service the same amount of debt creation.

Cash should begin to move around rather than just remaining parked at the CB earning IOR. This condition should put upward pressure on interest rates but light pressure due to the huge amount of reserves in place.

The US CB seems to be moving to stabilize the value of it's currency.

Next, I guess we could ask "stabilize the value of it's currency" against what reference? Hmmmm.

Central bank offers a service. The user checks in their own bitcoins and checks out the central bank branded, custodial bitcoin. Everyone spends the branded bitcoin which clears just like dollars, no block chain needed, and users can check in their custodial bitcoin for their own when they want. Central bank can manipulate flow by charging a fee on check in and check out.

And the obvious starter idea. The government could tax what ever activity it does, in the coin used for that activity. The central bank deals with four or five major digital currencies. Why not? Government taxes observable activities; if it taxed in the currency used the number of visible activities to tax nearly doubles. More efficiency because multiple currencies niche and expand price discovery revealing more taxable activities.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad