It has become common to expect at least one country to leave the euro.  Many folk in the City expect the euro to collapse altogether.  There is therefore some considerable interest in how a country could leave with the minimum of disruption.

The actual mechanics of leaving are not especially mysterious.  If the Greeks were to leave the euro, then the Greek legislature would pass a law deeming all contracts under Greek law that featured euros to now feature whatever the new currency of Greece is – let's call it the New Drachma.  The mechanics of paper currency and coins could be slightly trickier, but is largely irrelevant since in any modern economy only a tiny proportion of the total money stock is in the form of notes and coins anyway.  In principle one could simply forget the notes and coin and allow people to keep using euros for petty transactions – as they almost certainly would, for at least a time, anyway.  But if one wanted to go further, then the Slovakian model, following the breakup of Czechoslovakia, of putting a stamp on old notes that converted them into new notes, would be the most obvious.

None of this is especially tricky or interesting.  What is more interesting is that, absent mitigating measures, the above process would be extremely disruptive and quite likely to lead to constitutional overthrow or even civil war.  Why?  Well, amongst the key contracts that would be redenominated into the new currency would be bank deposits.  If people heard that their deposits would be changed into some other currency, they might panic and withdraw them.  (Indeed, that process has been going on in Greece for some time – as per this graph.)  That could lead to bank runs.  When the redenomination actually happened, there would almost certainly be bank closures and restrictions on the right to withdraw deposits.  When that happened in Argentina in the early 2000s, it led to middle class women crowding outside banks, banging on pots, demanding their money back, and eventually a mob storming the Presidential palace, with the President fleeing by helicopter from the roof.

Since no politicians wants to be That Guy, there's a hope that there must be something one could do to avoid it.

Here's my recipe.  Ingredients required: 1 currency union; oodles of cups of political will; about 15 years.  Start by fixing the euro to gold or some similar commodity, with full convertibility.  Then "one euro" simply becomes a way of describing "a claim on x grammes of gold".  Once you've stuck to that for a while – say, five years – then introduce your new currencies, also pegged to gold, in a ratio of 1 to 1 (so no devaluation at the point of redenomination).  Then we have "one lira" equals "one franc" equals "one mark" equals "a claim on [the same] x grammes of gold".  Once we've held that for a while – say, ten years – we can break the link with gold or devalue to a new gold peg.

What's the point of this?  Well, there are a number of historical examples of very straightforward currency union dissolutions that were not disruptive.  A classic such case is the switch in New Zealand at the end of the 1960s from Sterling to the New Zealand Dollar.  The key aspect of these non-disruptive switches is that they were not economically necessary and were not entered into to facilitate some change in economic policy – for example, the currency union was not dissolved so that the exiting country could devalue or inflate.  So because little changed, there was little disruption.

Thus, the key to dissolving the euro with little disruption would be to bring about a situation in which dissolving the euro could be credibly regarded as not involving any change in economic policy.  One way to achieve that would be for the actual dissolution of the currency union not to involve any change in what the currency was – since the euro would simply be a proxy for x grammes of gold and so would the new currency, switching the name of x grammes of gold would not itself facilitate any change of economic policy.

Now I'm sure lots of readers are feeling rather frustrated at this point.  They want to say: "But we don't have 15 years!" and "But the whole point of Greece's exiting the euro would be that it could then devalue!"  I understand that.  But the question asked was how to break up a currency union with relatively little disruption.  If you want to achieve that, then you have to do so over an extended period and without devaluation.  It is, in my view, simply a fantasy to imagine that there must be some low-disruption way to dissolve a currency union quickly and in such a way that after the currency union is dissolved there will be large changes in policy.

If you want to do it quickly and to do it in order to change policy radically, then expect: banking sector collapse; high inflation; civil disorder; authoritarianism or civil war.  Of course, it could well be that not dissolving the currency union quickly would lead to: civil disorder; banking sector collapse; high inflation; authoritarianism or civil war.  There don't have to be any pallatable options – dissolving the currency union quickly might be less unpleasant than the alternative.  But it doesn't have to be able to be done pleasantly at all.