Bitcoin Box

Ads

Alternative Distributed Currency Creation and Distribution Models and Implementation

author: Vitalik Buterin
published: 2011-08-15 13:50:35 UTC

Bitcoin's economic model is one of deflationism: while the monetary supply is currently rapidly growing, the growth rate will slow down in 2013 and reduce by 50% every 4 years until it plateaus at slightly less than 21 million in 2140. This is Bitcoin's main difference from conventional currencies like the US dollar, whose money supply increases at some central bank's whim, and it has aroused much controversy, with inflation vs deflation being the main argument discussed on the bitcoin.org economics forum; an article by Theodore Minick on Bitcoin Box itself discusses the issue here. There are many arguments on the issue from both sides, one fearing hyperinflation and the other fearing a deflationary spiral. Regardless of which side is right, there is sufficient disagreement for alternative controlled currency growth models to be proposed that are based on inflation rather than deflation.

There are two ways to implement this type of alternative "inflate-a-coin": a fixed growth rate and a growth rate dependent on some external factor. I will assume fixed growth to mean exponential, since linear fixed growth will eventually hit a 0% growth rate just as Bitcoin's current growth rate will, so it's equivalent to inflationism in the short term and a plateau in the long term. The problem with fixed growth is setting the growth rate. There are two scenarios:

1. The fixed growth rate will be smaller than the growth rate of the economy - such a currency model will essentially be a less severe model of deflationism.
2. The fixed growth rate will be larger than the growth rate of the economy - inflationism.

The problem is that it is impossible to determine in advance what a good growth rate will be. Perhaps with technological progress the size of the economy will even shrink as people prefer to take advantage of technology through increased leisure rather than increased production and consumption. Perhaps a credit expansion will occur at the same time as economic stagnation, causing severe inflation and potentially a hyperinflationary spiral. Or the growth rate will end up being too low and we'll be stuck with deflation anyway. Also, the growth rate of the economy can be very difficult to measure: is a change from $10,000 cars to $100,000 teleportation pods inflation or deflation? This question is in fact key - the fact that different goods are being produced is why there is controversy about what the inflation rate is, with estimates between 2% and 11%. Ultimately, the effect of a fixed growth rate is that it decreases the risk of a speculative bubble in the currency (ie. a deflationary spiral) but increases the risk of hyperinflation; which is worse is a topic of heated economic discussion and is outside the scope of this article.

To solve the issue of calculating a good inflation rate, another alternative is a regulated growth model. Central regulation is out of the question since that particular market is already saturated with hundreds of fiat currencies; however, decentralized or automatic regulation has also been proposed. Some possibilities are:

• Currency creation rate based on network hash power. This is impractical since an inflation rate as fast as Moore's law will quickly lead to hyperinflation.
• Currency creation rate based on transaction rate or volume. The problem here is that a larger economy will not have a larger transaction volume - if the economy is twice as large, prices will be half as high so values will be twice as low. Using the number of transactions seems to bypass this, but a change in the way people use bitcoin can cause far greater changes than a change in the economy's size: anonymizing bitcoin tumblers make a very large number of transactions, use of bitcoin banks reduces the number of transactions, use of microtransactions increases the number of transactions, and so on, so a change in the way people use bitcoin will cause the system to think that there is massive inflation or deflation occurring when no such thing is the case.
• Currency creation rate based on some outside economic statistic - impractical because the source of the statistic is itself centralized, and might disappear or change its methodology and the currency would be left with no idea of how many new units to create.

All of these systems, however, fall apart because they attempt to measure inflation and any attempt to measure inflation is intrinsically subjective. The characteristic that the measurement depends on might cease to become relevant, or unexpected non-inflation-based factors might alter it. This brings us to another form of decentralized regulation: a democratically controlled currency creation rate. However, this is only viable if the controllers have the best interests of the economy at heart and are not guided by selfish interests; unfortunately, everyone is:

• Miners vote - miners will vote to maximize profits to themselves, so hyperinflation will occur. There is the option for new coins to be transferred to currency holders and not miners, but this is really just a form of anti-demurrage, which has no effect (more on this later).
• Currency holders vote - currency holders naturally want to maximize deflation, so this will create a plateau.
• Addresses or transactions vote - easily exploitable, because of Bitcoin's inherent pseudonymity.
• Physical identity based democracy - undesirable for obvious reasons, one of them being that a central authority is required to oversee the elections and have the power to control the rate of coin creation.

There is another system that is proposed to encourage spending: demurrage). Essentially, there is a fee for having bitcoins that is automatically deducted once every period (ie. cash holdings decay exponentially), the proceeds from which are destroyed. The issue is that demurrage does nothing: if the total money supply were to drop by a factor of 2, and everyone's cash holdings were to drop by a factor of 2, then that's really just changing the unit - cutting everyone's money supply from 2X to X is equivalent to creating a new unit, Y, defined as one Y being worth 2X, and switching the unit currency is measured in from X to Y. Everyone has the same value they did before, it's just denominated differently. Thus, a system of demurrage with the destroyed money being replaced by miners is equivalent to fixed rate inflate-a-coin.

There is another set of proposals that merit consideration: that for a bitcoin backed by gold or some other commodity. Doug Casey is a proponent of such a Bitcoin and the arguments for it range from the psychological:

L: Do they have value in themselves?
Doug: There's the rub; I don't see that they do. Bitcoins are just an electronic abstraction. They can't be used for anything else, nor are they made of something that can be used for anything else. They are like one of those knots in a string that disappear if you pull hard enough on the ends of the string. They are not backed by anything at all.

To the practical:

But I predict a decline in their acceptance and value over time - if not an outright collapse - unless they are linked to something with near-universal value.

Doug Casey also throws the argument that Bitcoin is just a 'greater fool' game, which I discussed here - my own conclusion is that it is in fact an 'equal fool' game, which is in fact perfectly reasonable and sustainable.

The issue that I am covering today, however, is implementation of such models. Gold-backed bitcoin sounds good in theory, but falls apart in its implementation - such a system inherently requires centralization. Doug Casey advocates GoldMoney as a backed alternative to Bitcoin that already exists, but it is a far cry from being an actual currency: it doesn't even try to advertise itself as one! The Liberty dollar was an attempt at a gold-backed currency, with anonymous accounts and transactions, but its centralization doomed it as the US authorities took it down. Gold must be stored somewhere and there must be some way to prove that gold backing exists, and this proof requirement inherently leads to traceability.

There is, however, another option for a backed bitcoin: one backed by not a commodity, but a service. Namecoin is one such example, with there always being a guaranteed demand for namecoins since the function of URL registration is built into the block chain. This form of backed bitcoin has potential, but it is ultimately backed by the network just as Bitcoin is: bitcoins have value because other people accept them and namecoins have value because other people are willing to accept the URLs that they represent. The purpose of backing is for there to be a guaranteed minimum value for the currency, but Namecoin does not supply this: Namecoin URLs' cost is based on transaction fees, which means that if a namecoin's value falls by 90% competition will push up the number of namecoins needed to get a URL by a factor of 10, so there is no specific guaranteed minimum value for a namecoin. This may be a good thing - as I argued here, backed currencies have the problem that they push up the price of whatever they are backed by, preventing some people from taking advantage of the normal use of the backing commodity - gold being at $1600 and not $800 prevents people who could have found a use for gold that creates between $800 and $1600 of value from doing so. Namecoin's value going up will not make URLs prohibitively expensive. Namecoin can be considered not backing, but rather bootstrapping: something to encourage initial use, increasing the chance that the currency will pick up and stabilize.

Ultimately, currencies must be backed by someone, making them centralized, or backed by a network good, making their value dependent on the network, a feature that an unbacked currency already has. This does not mean, however, that having a currency that has intrinsic functions is necessarily worthless - some sort of bootstrapping is necessary, but these proposals are not equivalent to gold or commodity-backed bitcoin.

Regardless of which other currency models become available, however, if any of them succeed then Bitcoin is likely to simply because it has the first mover advantage and an established community. Rapid bootstrapping may overcome this, especially with the advent of merged mining, but at this point unless Bitcoin fails it does not seem likely. Unfortunately, just like with all other social sciences we cannot have controlled experiments in economics, so we will never know if the deflationary model is the best one, but it seems to be the one we are stuck with.

Random Articles

blog comments powered by Disqus