Imagine there is competition between two currencies. The two currencies are identical in every respect, except one is backed by assets and the other isn't. Which one would win the competition to become the preferred medium of exchange?
I think the one that is backed by assets would win. For two reasons:
1. The issuer of the backed coin could use those assets to help stabilise the value of the coins in response to fluctuations in demand. If demand falls, the issuer could use the assets to buy back some of the coins in circulation. If demand rises, the issuer could sell more coins in exchange for more assets. The issuer can make the supply curve of the coins more elastic, so that fluctuations in demand cause smaller fluctuations in price. Other things equal, people prefer a safer asset to a less safe asset.
2. The issuer of the backed coin could use the returns on those assets to pay interest on the coins. Or to buy back coins so the owners of the coins would see capital gains. Other things equal, people prefer an asset that pays a higher rate of return to one that pays a lower rate of return.
Other things equal, we would therefore expect to see competition between currencies lead to issuers of those currencies having assets backing those currencies, and paying out all the returns from those assets (minus administrative costs) to those who own those currencies. Free entry and competition drives down the profits from issuing currency to zero. Only currencies with 100% backing would survive.
But are other things equal?
There could be network externalities and a first mover advantage. Like Microsoft, VHS vs BetaMax, and the English language. If UnbackedCoin got on the market first, and if everyone wants to use the same coins as everyone else, and translation costs are high enough, it could survive against competition from BackedCoin. Customers cannot coordinate their simultaneous switch from UnbackedCoin to BackedCoin.
One more post in response to Brad Delong's email query and arguments about competition for Bitcoin.
I recommend Tyler Cowen's post.
[Update: I think JP Koning got there first.]
(Mike Sproul deserves all the blame for making me think this way.)
Worthwhile Canadian Initiative 
Wouldn’t the one NOT backed by assets win? Less likely to be used as a store of value and therefore more likely to be exchanged for other goods….. Is this not the lesson of bimetallism? The bad drives out the good.
Robert: Gresham’s Law works if the two currencies have fixed exchange rates. Bimetallism was a fixed exchange rate between gold coins and silver coins. These two currencies would have flexible exchange rates.
Backed vs unbacked: this is exactly why I keep saying that Ripple (or something like it) is going to be the end of Bitcoin:
http://jpkoning.blogspot.ca/2013/03/selling-out-of-bitcoin-ledger.html
I would think the unbacked currency would win the circulation battle. The reasons are three:
1. If an owner has both backed and unbacked currency, he would trade the unbacked first. This reduces his risk of loss of value.
2. More of the unbacked currency will be produced. This happens because the issuer does not need to limit production to available real assets; production can follow political sale of desired, good, projects.
3. A long term holder-with-choices would place assets into the backed currency. That action would withdraw the backed currency, making it more a reserve currency, not a trading currency.
Only when a currency becomes unacceptably weak, does it receive reduced usage. Actual rejection must occur, not simply preferred unloading.
JP: I have updated the post, to link to yours.
When some Bitcoin type operation manages to stabilise the value of its coins relative to the national currency, the operation becomes indistinguishable from a commercial bank.
That is, coins issued by bank X and with “bank X” imprinted on the coins are no different to notes issued by bank X, which in turn are no different to loans extended by bank X: they are all liabilities of the bank. And in all three cases the bank will nearly always want collateral in exchange for lending out any of the above three. So to that extent, all three are backed by real assets.
The existence of some other asset (e.g. lots of gold in the bank’s vault) doesn’t add much: i.e. the bank will still want collateral from those wanting coins / notes / loans. Indeed, as Robert above suggests, that gold backing would be a COST, which might make bank X uncompetitive. And if the bank has interest yielding assets (e.g. property instead of gold) that’s a separate business surely? It has nothing to do with, and isn’t a necessary part of the coin / note / loan issuing business.
You could argue that US currency is backed by their military. Ultimately, all other resources are vulnerable to theft, invasion, redefinition under law, obsolescence, and all the other hazards of the world, but a whompin’ big military can offset many of those hazards, and can additionally assist in redefining other countries laws, tweaking their behaviour with mini-invasions, supporting their adversaries, and basically acting as the biggest thumb on the biggest scales in the world.
People wonder why the US wastes trillions on a military an order of magnitude bigger than anyone else’s. That’s why.
Noni
Nick:
Your first-mover advantage argument is plausible, but then we look around at modern-day dollars, pounds, euros, yen, etc, and find that they are listed as liabilities of the central banks that issued them, that those central banks hold assets against them, that no dollar is ever issued without getting a dollar’s worth of assets in exchange, and so on. It sure looks like backed currencies win.
A couple of weeks ago I even heard some Federal Reserve officials publicly discouraging people from using bitcoin. They said it is safer to use a BACKED CURRENCY LIKE THE DOLLAR (!!)
MajroMax: “In that regard, Bitcoin is technically cool but I don’t think it’s a long-term medium of exchange.”
I disagree. If there were no threat from competition, or from governments banning it, I think it could survive fine, despite its fluctuating value (and those fluctuations would diminish a bit over time as people learn more about the future demand to use Bitcoin). Because some people want to use a cryptocurrency. The danger is from competition from better backed competitors.
Ralph: “When some Bitcoin type operation manages to stabilise the value of its coins relative to the national currency, the operation becomes indistinguishable from a commercial bank.”
Bingo!
A commercial bank is like a central bank that fixes its exchange rate against another central bank.
But you lost me on the rest. A commercial bank that pays less interest and/or has higher fees and/or worse service than competing banks will lose the competition for deposits. Unless they use the returns on their assets to do those things, they will lose out in competition, because their customers will switch.
Noni: I don’t get that. Zimbabwe had a (relatively) strong military, but that didn’t help the Zim dollar.
Mike: the backing of the Bank of Canada helps to make the supply curve elastic, to prevent the Lonnie fluctuating in value. But holders of Loonie currency don’t see the interest on those assets.
JPK,
Your post that Rowe linked to states
“But that’s not the sort of ledger space I’m talking about. What happens when we bring in bitcoin-quality ledger space that has an anchor? I’m talking stable-value crypto-currency, not the sort that dangles and has a null value. These new alternatives will copy the best aspects of bitcoin, specifically its fast, efficient and safe record-keeping abilities. But rather than just providing blank tokens, they’ll twin the ledger with some intrinsically valuable item.”
Why does a currency need to be backed by an object of value and why does that linking prevent fluctuation in demand for the currency?
Its like when people claim we should use gold over bitcoin because gold has “intrinsic” value due to its uses in certain production processes. This doesnt give gold intrinsic value, the value of gold is determined by the sum total of demand for the commodity regardless where that demand comes from.
Gold is currently about $1200/ounce, if there was no demand for Gold as a currency/speculative commodity what portion of the price would be due to the demand for “real” uses of gold? If 10% of gold is purchased for industrial use then what stops your fluctuation argument from applying to gold when it only a small portion is used for “real” projects.
You dismiss the ledger and claim it points to nothing, but when it is traded against real goods that ledger becomes a full account of the trade ratios between goods without any interference from non-trade demand or the wishes of those who provide a monopolized currency. That’s the value of bitcoin, it removes the barriers that are created by those who are attempting to control the intertemporal trade in goods to fund their economic stimulation schemes.
I’m having trouble thinking through all of this.
The classic example is cigarettes in prison. Would you say that cigarettes are “backed” by the number of men that actually smoke them? And that would provide a floor on their value?
Now suppose cigarettes have a competitor in baseball cards. There’s a few guys in prison that collect the cards for sentimental reasons – but fewer than the number of guys that smoke. The floor for baseball cards is lower. But there are advantages to cards as medium of exchange (just assume that’s the case). For example, maybe they’re easier to hide when there’s a shakedown by the guards.
In any case, does the higher floor on cigarettes necessitate that it wins out as a medium of exchange? Is this example analogous to your note above?
Nick:
“holders of Loonie currency don’t see the interest on those assets.”
Maybe they do. It’s just that the interest is burned up by costs. Assume costlessly-issued bonds pay 3% per year. Then along comes another bond, called the Loonie, that has printing/handling costs of 5% per year. The loonie earns a 3% yield just like other bonds, but the 3% yield is burned up by the 5% costs, leaving the loonie yielding -2% per year. The loonie has a high enough convenience yield (aka liquidity yield) that people are willing to hold it despite its low net yield.
Of course, a checkable loonie would have lower printing/handling costs, and would have a higher yield. Central banks are just now discovering this as they start paying interest on reserves.
Ed: in the case of cigarettes, there is a demand curve for cigarettes from smokers, and a demand curve for those who want to use them as currency. Smokers have a flow demand curve, and currency holders a stock demand curve. If the demand for cigs as currency drops, the price drops, smokers smoke more, so the stock supply falls over time, preventing the price falling as much.
So the cigarette case is very much like a backed currency with an elastic supply curve, but a bit more complicated because of stocks and flows.
Gold is like cigarets too, except the flow supply from new mining is very small, so fluctuations in monetary demand cause big fluctuations in price for a long time.
It’s all in elasticity of demand and supply, and stocks vs flows. A currency that is 100% backed can have a perfectly elastic supply that adjusts the stock very quickly.
@Mike Sproul:
Mike: “Maybe they do. It’s just that the interest is burned up by costs.”
If that were the case, then, OK, it is equivalent to the owners of Loonies getting the interest, in kind.
But, as a matter of fact, it isn’t. Costs are relatively small. Most of the interest gets paid to the government.
Majromax: If a dollar is pegged to 1 oz or 1 CPI basket, then that provides an anchor, and backing a dollar with a dollar is no longer meaningless.
Nick: Costs are pretty hard to judge. I think printing alone burns up 1% per year. Then there’s vault storage at maybe another 1% per year, then the cost of all those armored cars, the cost of tellers counting out bills, the cost of chasing counterfeiters, etc. Nineteenth-century note-issuing bankers used to say that paper notes weren’t profitable, but were just a good form of advertising.
“Imagine there is competition between two currencies. The two currencies are identical in every respect, except one is backed by assets and the other isn’t. Which one would win the competition to become the preferred medium of exchange?
“I think the one that is backed by assets would win.”
Is there enough information to tell? The mandats of the French Revolution were backed by land, and quickly flopped. OTOH, the currency of the Pennsylvania colony was also backed by land, and was a success.
You did have Greenbacks vs. gold backed US Dollars during the Civil War. The Greenbacks dropped in value by about 1/3, but eventually recovered parity. {shrug}
As for Bitcoin, two analogies spring to mind: wampum and US Colonials. Wampum was a success, Colonials flopped.
Correction: Colonials were not US currency. The US did not exist yet.
Majormax,
“That means that the only thing giving Bitcoin value over Bitcoin2 is the network effect — there’s no extrinsic reason I should choose Bitcoin over any particular alternative. But when the currency’s value is governed solely by sentiment, that’s not stable.
On the other hand, backing by a sufficiently credible entity does give a currency (potentially small, but nonzero) intrinsic value”
Where do governments get their credibility from? The network effects that have developed over hundreds of years from the vast financial infrastructure that has been built up around the state. If bitcoin gains this level of credibility then there is no reason to believe bitcoin can’t be a sufficiently credible monetary institution, its just a matter of time for the bitcoin infrastructure to be put into place.
Intrinsic value (thanks for clearing up the definitions) might provide a lower floor but if that floor is really low (as in the case of gold) then the same criticisms apply to currency with intrinsic value. Fiat currency has intrinsic value only in so far as the monetary authority correctly manages it. The whole point of moving to bitcoin is that savers dont want to ride the QE inflation train and now potentially have an easily accessible currency to hold their savings.
I know the NGDPers like Nick dont want that to happen because they believe that certain points in history require the sacrifice of savers for the greater good of the economy but the fact is that in the face of massive US QE programs savers are going to perform that experiment regardless of what the monetarist economists want.
If the NGDPers are wrong then a massive shift to bitcoin can easily shift to bitcoin the credibility that the central monetary authorities have lost.
Edit: “no reason to believe bitcoin CAN’T be a sufficiently credible monetary institution”
(is there an edit function I dont know about?)
[I edited it for you NR.]
Ian,
“Where do governments get their credibility from?”
The system of laws that they create and govern by. A market system without a legal system will not get you very far and lawyers don’t work for free.
Min:
“The mandats of the French Revolution were backed by land, and quickly flopped. OTOH, the currency of the Pennsylvania colony was also backed by land, and was a success.”
A dollar backed by 1 oz worth of land will be worth 1 oz, and a dollar backed by 0.2 oz will be worth 0.2 oz. The assignats, and later the mandats, were initially well-backed by land, but as the land was sold off, and paper money was printed and spent like crazy, there was less land per unit of currency, so the currency lost value. Pennsylvania, on the other hand, held enough assets to adequately cover the money it issued, so the money held its value.
“Colonials flopped” You probably mean Continentals, which flopped because their quantity outran the assets (i.e., taxes receivable) backing them.
Majromax:
The CPI basket is defined by physical amounts of key commodities, so pegging to it is tantamount to pegging to 1 oz.
I’m going to by my usual navel-gazing self and dig into the two key words here:
“backed” by “assets”
second one first:
“Asset:” So what is an asset? Traditionally, its been rare metals, because that’s what technologically made sense for pre-industrial societies. But couldn’t assets be lots of thing? The United States owns 635 million acres of land; could you have a land-backed currency where every dollar is worth 100 square feet of land? What if the currency was backed by shares in a holding company that owned corporations that issue mortgages or make cars (to pick two totally random examples)? Or what if all patents and copyrights, instead of reverting to the public, reverted to the government, which licensed them for free but retained the right to alienate them and then used them to back the currency? What if a bunch of folks with guns and tanks and maybe even thermonuclear weapons, as well as large, unpleasant, remote buildings in which many people were forced to stay for a very long time, told everyone that unless they used the currency those folks liked that those folks would use those guns or put them into those large, unpleasant, remote buildings? Is “avoidance of violence or loss of freedom” an “asset?”
“Backed:” this is all about credibility, no? If you said that every currency unit could be theoretically redeemed for a gram-and-a-half of gold, but you issued lots and lots of currency and didn’t really have very much gold, is your currency “backed?” What if you really did have all the gold, but nobody believed that if push came to shove you’d actually redeem all that paper for delicious, delicious gold? Is your currency “backed?” What if you had all the gold and people believed you’d redeem all the paper for all the gold but somebody else in the country next door was trying the same scheme but nobody believe HE’D redeem the paper for gold so suddenly all the gold starts getting sucked out? What if that could happen at any moment?
I guess where I’m going is, if the government has a whole lot of guns, and they can credibly threaten to use them on people if they don’t use the government’s currency, then that currency is just as “backed” by an “asset” as any other.
I had a response to Ian posted, but I think it got spam-trapped, possibly due to a link to an externally-hosted bitcoin value chart. I’d much appreciate it if that spam-trapping could be looked into.
[Summary in lieu: bitcoin has no natural monopoly because the infrastructure can be redirected to other cryptocurrencies, it doesn’t work -now- as a means of savings because it’s suffered several large drops in value in the past year, and bitcoin can only be monetarily credible if and only if monetary policy can be reduced to a simple equation on quantity of money.]
@Mike:
Majromax,
“[Summary in lieu: bitcoin has no natural monopoly because the infrastructure can be redirected to other cryptocurrencies, it doesn’t work -now- as a means of savings because it’s suffered several large drops in value in the past year, and bitcoin can only be monetarily credible if and only if monetary policy can be reduced to a simple equation on quantity of money.]”
Several large drops…that occurred after several large increases. No one is saying the bitcoin of today is the optimal cryptocurrency. Its highly volatile, there are many other competitors, etc. But where was the internet 15 years ago? Completely indistinguishable from what we have today. Once it becomes less volatile it could easily become a vehicle for savings.
Frank,
“The system of laws that they create and govern by. A market system without a legal system will not get you very far and lawyers don’t work for free.”
Where do those laws gain credibility? From the credibility of the government that enforces them, which comes from the credibility that citizens have in their governing institutions. Have you guys never heard of civil war? Citizens can lose faith in their governments, governments are backed by nothing other than faith in the governing body. If faith based institutions can gain credibility after developing for hundreds of years there is no reason why bitcoin cant either, especially considering its non-physical form which allows for people from all over the world to support it essentially cost free.
Majromax and Roger Sparks: I found your comments in spam.
Ian Lippert: “Where do governments get their credibility from? The network effects that have developed over hundreds of years from the vast financial infrastructure that has been built up around the state.”
Gotta say I loved that bit.
Majromax: ” Its designer and community confuses certainty of quantity with certainty of value, so the reasoning on why Bitcoin in particular has value is circular. (It has value because it’s limited, so its value can never go down (much!))”
With any intrinsically worthless medium of exchange there are always two equilibria:
1. if its price is positive, it may be demanded as a medium of exchange, and at some price demand = supply (as long as supply is limited).
2. if its price is zero, it can’t be used as a medium of exchange, so demand is zero, so it’s a free good.
Squarely rooted:
I find it best to think of backing this way: The tax man will demand 5 oz of silver from you every year. If you don’t pay you’ll go to jail. But the government just built a road, and it paid the contractors with 1 oz certificates that the tax man will accept in lieu of actual ounces. If the present value of all the government’s tax collections is 1000 oz, then the government can issue up to 1000 certificates without causing inflation. But if it issued 2000 of those certificates, against only 1000 oz of taxes receivable, then 1 certificate=0.5 oz.
Majromax:
If the central bank pegs a 2% inflation target, (pegged against a basket with specified amounts of various goods), then the money is tied to goods.
I haven’t had a chance to read all the comments, so I apologize if this is a repeat …
It seems to me that the major problem with bitcoin is the built in deflation. In my amateur opinion, that’s what makes it unworkable. Especially if people start to consider it a store of value as well as a medium of exchange. The supply constraint and hoarding will cause a recession/depression and Nick will be screaming “I told you so!”
Now if some clever person figures out a way to automatically make sure that the supply of some future electronic crytocash system is perfectly balanced to achieve e.g. 4% NGDP growth, then we might be on to something. Until then, meh. It’s vaguely interesting from a technical point of view but that’s about it.
Nick:
” The two currencies are identical in every respect, except one is backed by assets and the other isn’t.”
Is the second currency a pure bubble asset, or is it fairly valued and backed by the present value of future payment services provided? My guess is that bitcoin today is a mix of the two. I guess that sometime ago bitcoin was fairly valued and it was completely backed by the discounted value of future payment services.
“2. The issuer of the backed coin could use the returns on those assets to pay interest on the coins. Or to buy back coins so the owners of the coins would see capital gains. Other things equal, people prefer an asset that pays a higher rate of return to one that pays a lower rate of return.”
Other things equal is the tricky part here. Other things equal should not include the price of the coin. Is expensive backedcoin better than a cheap unbackedcoin? We have no idea. Historically, when people wanted exposure to bubble assets, they bought them pure without any stabilizing add-ons.
If you have an unbackedcoin, you can create a substitute backedcoin yourself at home by putting an unbackedcoin and a treasury bill into an envelope – you can pay yourself interest. Is there any benefit of doing this combination on the level of the coin? We have no idea.
Nick:
“The issuer of the backed coin could use those assets to help stabilise the value of the coins in response to fluctuations in demand. If demand falls, the issuer could use the assets to buy back some of the coins in circulation. If demand rises, the issuer could sell more coins in exchange for more assets.”
There are two types of stabilization, and it is unclear which you have in mind.
1. The first type is concerned with the arbitrage between the markets for the coin and the underlying assets. The benefits of this type of stabilization are limited. Closed end funds work reasonably well. Bitcoin is like a closed end fund.
2. The second type uses two tranches of liabilities, one is safer and more liquid, the second is riskier and less liquid. Central banks and commercial banks are good examples. But this type of stabilization has got nothing to do with backing. We could imagine TrancheCoin with two classes of coins, one tranche safer and more liquid, the other tranche riskier and less liquid. TrancheCoin would enjoy the benefits of this type of stabilization.
Isn’t liquidity, moneyness premium/”temporary store of value”/stability the prime concerns? It’s probably true that BackedCoin will win on these features (for the reasons you mention), especially, the competition/replication/counterfeiting issue. Even for cash counterfeiting is an issue (for example in some small stores they won’t accept cash denominations more than $20; they prefer credit cards or debit).
Suppose that we had money backed by land, called Thalers (pronounced TAHLERS). Since “Thal” means “valley” in German, that’s not a bad name. 🙂
Perhaps one problem with it as a medium of exchange is that if thalers are used to buy land from the issuer of currency, then those thalers should be retired. That’s pretty crappy monetary policy, isn’t it?
Preferred medium of speculation, followed by preferred medium of collectors? I hear Confederate money is quite collectible.
Thanks Nick.
“Why does a currency need to be backed by an object of value and why does that linking prevent fluctuation in demand for the currency?”
Ian, I don’t have much time, but the comment below covers my thoughts on the gold vs bitcoin comparison:
http://jpkoning.blogspot.ca/2013/12/tales-from-litecoin-universe.html?showComment=1388165024432#c4580164208435040691
Comments above have it right I think. What’s an “asset” in this context? A government’s main “asset” is the power to tax (or redistribute property, or acquire property). While all these have limits, they are essentially political rather than related to any notion of what the government “owns”. But do we need to theorise when there is ample history to look at? For the first 2000 years of money there was no coin, just a unit of account. And any token, public or private, was backed by the ability of the issuer to exchange it for some tangible good. What was the proof of that ability? Continued exchange. Likewise, metallic coins are not goods in themselves, just tokens. They were used on a par with other tokens, and their value derived from their use in exchange, not from the ability to use them to pay taxes (which until early modern times in Europe were mostly collected in kind). It might help to think of two companies issuing commercial paper. One has a small asset base, but a high cash flow, and regularly redeems its paper and re-issues it without difficulty. The other has a low or irregular cash flow, but a large asset base. It sometimes misses a payment or delays redemption. Which one would command a premium? Which would be most acceptable as collateral?
Nick,
Thanks for saying “bingo”, but I’m not 100% sure I’m right: I’m just thinking aloud.
You claim that a bank must “use the returns on their assets to do those things..” else it loses out to the competition. As I understand you, you’re saying that where a bank has a big pile of income yielding assets, and uses that income to subsidise its customers, the bank will beat the competition. My answer is: “sure, but that’s cross subsidisation”. A garage or restaurant owner could do the same: i.e. subsidise their business out of income from assets that are not relevant to or necessary for running the garage or restaurant.
Majromax,
JPK
Vaidas: “There are two types of stabilization, and it is unclear which you have in mind.”
Neither of those two types.
I am talking about simple Open Market Operations. The issuer of BackedCoin is just like a regular central bank, that has some monetary policy target, like a price level or inflation target. If it sees the value of BackedCoin dropping below the target, it does an open market sale of assets, buying its own coins, to reduce the stock in circulation. If it sees the value of BackedCoin rising above target, it does an open market purchase of assets, selling its own coins, to increase the stock in circulation.
Nick, that is a second type – central banks have two classes of liabilities – monetary base and equity which usually is owned by Treasury. The profits or losses of OMOs go to the owners of equity. It is possible to replicate this with UnbackedCoins by changing the distribution of the bubble asset plus intangibles between two classes of coins.
Nick, consider UnbackedCoin with two classes of coins – EquityCoins (EC) and LiquidCoins (LC). ECs pays dividends in LCs. LCs pay interest on reserves in LCs. By changing IOR and dividend rates, UnbackedCoin can replicate OMOs.