# From an economics perspective, what are the ramifications of a currency with fixed money supply?

I'm thinking specifically of bitcoins.

What are the pros and cons of having a fixed number of coins, as opposed to more "normal" currencies? Would the currency have no inflation?

-
The question seems to be about two distinctly different things: zero inflation (the title); and fixed money supply (the body of the question). It would help greatly if you could clarify which you are interested in; and also please do add more substance to the question: we're looking for expert-level questions during this private beta stage. –  EnergyNumbers Nov 21 '14 at 23:49

FooBar is quite right that unless you expect GDP growth to stop, fixed nominal supply currencies will lead to deflation.

A moderate degree of currency inflation serves a number of useful functions in the economy. The most obvious are:

• It induces people to spend their money before it loses its value. In a deflationary environment there is an incentive to put money under your mattress and spend it in a year when it has greater purchasing power. If everbody does this then the lack of demand will lead to a decrease in overall economic activity (i.e. a recession).
• It provides a weapon against downward nominal rigidities. For example, workers are generally reluctant to accept a nominal pay cut, even if market conditions are such that the current wage is above the equilibrium level. Inflation means that their employer can simply increase wages at less than the inflation rate so that the real wage is decreasing.
• It erodes the real value of nominally denominated debt. Now, this is obviously only a pseudo-advantage because (whilst it benefits debtors) it harms creditors. However, this kind of erosion of debt may be desirable if national economic stability is threatened by high debt levels. Also, since debtors are usually poorer on average than creditors, it can reduce inequality, which may be a normative objective for the government.

Without inflation you miss out on these benefits. The first benefit might not seem like a big deal if you think you can simply set the rate of inflation at zero percent. But it is very hard to hold inflation constant at some target level so attempting to hit zero inflation will almost certainly result in occasional lapses into deflation, with the attendant negative economic consequences.

-

It is a fallacy to conclude that a steady number of coins will give you no change on the monetary value (inflation/deflation). The classical quantity theory of money can be used as a first-order approximation here:

$MV = PY$

where $M$ is money, $V$ is the velocity of money, $P$ is the price level and $Y$ is the quantity of real goods. The equation says that the money stock must be sufficient to "buy" all produced goods in the economy, corrected for velocity of money.

Now take $M$ as constant (as in bitcoins). We immediately see that changes in $V$ or $Y$, unless they happen to cancel each other, will affect the price level $P$.

If bitcoins ever took over as the number one real currency, given that we see GDP increasing over time, c.p., they would have to increase in value (forever).

-
(+1) I was reading the question and was thinking "to show the fallacy one needs no more than the classical quantity theory of money" - and then your answer appeared on the screen... –  Alecos Papadopoulos Nov 20 '14 at 16:37
M is money supply, not money base. It is only the money base that is fixed. –  Corone Nov 20 '14 at 21:03
Given that V also has an upper bound, in that transactions are only recorded in the block chain at a fixed rate, MV has an upper bound, so perhaps I should have clarified, but structurally, since MV is bounded and Y is increasing, P must decrease. Perhaps I should have added the cap on V to the question, but I assumed it was known. –  Jason Nichols Nov 21 '14 at 4:17
@jasonnichols MV does not have an upper bound, or at least not a structural one. Not all transactions occur on the block chain. Off block chain payments are already possible. –  Corone Nov 21 '14 at 9:12
If your currency increases in value, that's not inflation, that's deflation (I think you know that, but it's not explicit in your answer). You could get inflation, however, from an increased velocity of money, combined with fixed or declining GDP: but there is a limit there: the speed of a chain of transactions will cap the velocity. –  EnergyNumbers Nov 21 '14 at 23:19

The other answers are correct in respect of what would happen if the money supply of a currency was kept constant, however there is nothing in bitcoins to ensure that money supply will stay fixed

This is such a common misconception so I'll repeat it. Bitcoins limit the money base, but that does not limit the money supply.

The money supply is the money base times the money multiplier, and the money multiplier comes from fractional reserve banking. There is nothing in bitcoins to prevent fractional reserve banking, and we are already starting to see website that allow you to lend or invest bitcoins.

Someone could, if they wished, set up a bitcoin bank. Savers would deposit bitcoins and earn interest, while borrowers could take out bitcoin loans. As soon as you do this the money supply (the $M$ in $MV=PY$) increases.

So, if bitcoins became a global currency, then you would almost certainly find that the money supply was much larger than the base limit - the money multiplier would expand as more banks entered the system. We don't have any concrete evidence to suggest what the limit on the money multiplier would be, but even in "normal" currencies the money multiplier can exceed 10x. The legal limit in the EU is 50x. With no regulatory capital or reserve requirements, who knows what the practical limit could be.

So what does fixed money base do?

Having a fixed money base means that there is no central bank to control the money supply. If the economy heats up, you would normally raise rate, which would reduce the money base and tighten the money supply.

In this case there would be nothing anyone could do - the world would simply rely on the "free banking sector" to do the right thing. This was actually not far off the way the US ran it's currency during the period when it had no central bank (between 1836 and 1913).

What you could do is control the money multiplier directly. This would mean passing laws and regulations forbidding fractional banking above a certain fraction. For bitcoins doing so would be almost impossible, not to mention against the whole "free" concept. You would have reintroduced a central controller for the currency.

The free banking era saw wild swings in money supply (e.g. more than doubling in a year) and hence huge inflation or deflation as banks grew and failed. It is anyone's guess what would be in store if bitcoins became a major currency, but there is no reason to believe they wouldn't have inflation.

-
This isn't quite correct. Even without regulatory frameworks that attempt to directly control the multiplier, the expansion does eventually hit a ceiling that is imposed by the requirements of inter-bank liquidity. It takes decades for this to occur, so while it is happening, it looks like a continuous expansion, but eventually it would flatten off. The American free banking era is also regarded as something of an aberration by European banking specialists. –  Lumi Nov 20 '14 at 22:07
@Lumi can you give an example of hitting that limit for the system? The BoE used a 0% reserve requirement, reserve being limited by liquidity, but the leverage constraint was always the capital requirements. The fact that we view the free banking as an aberration is exactly my point - bitcoins would be a currency without a central bank; exactly the same aberration. –  Corone Nov 20 '14 at 22:11
@Lumi also remember that each tier that you add to a banking system increases the liquidity available in the system, as the bank from the tier above acts as reserve bank to the tier below. –  Corone Nov 20 '14 at 22:13
That depends on the topology of the inter-banking relationships - which I don't think there's been much work on? –  Lumi Nov 20 '14 at 22:22
@Lumi and so which bit is incorrect? I'm saying that fractional reserve banking can increase the money supply. We don't know if there is a limit to how far it can increase from a fixed base, but we do know that it certainly can increase a lot. I will edit to make that uncertainty clearer? –  Corone Nov 20 '14 at 22:24

The old monetary equation is MV=PT, where M is the money supply, V is the velocity of money, P is the price level, and T is the amount of transactions.

The fear in this case is that M will remain fixed. In a growing economy, T, the level of economic activity, will rise. Then, to make the equation hold, one of two things must happen: 1) V must rise to compensate for the rise in T, or 2) the price level must fall to compensate for the rise in T, to hold PT constant with M times V, both of which we're assuming remain constant.

A rise in V can be caused in improvements in money management; banking, ATM cards, smart phone payments etc. That, in fact, is a large part of what has driven economic growth in the past 50 years.

The other possibility, a fall in P, is what scares people. It's called deflation, and represents "sticker shock." Imagine a world in which prices were going down 3% every year, because the economy was growing 3% a year. Your wages would fall 3% a year. You shouldn't worry because your costs are also falling 3% a year, or more (but you will). There will be "reverse" millionaires," people who used to be millionaires, who aren't any longer, but who are better off because prices have fallen faster than their nominal wealth. It would be a strange, funky world, which is why policy makers avoid it, to the extent of making the "opposite" mistake of allowing inflation. Which may be one reason that the Bitcoin experiment is not viewed favorably by central banks.

-