To start with the easiest question, it is impossible to see how speculators can disrupt the launch process this year. The final exchange rates for the 11 participating countries will be determined on 3 May, and these will almost certainly be based simply on the exchange rate mechanism (ERM) central rates which are already in place. (Ireland may be an exception, requiring a revaluation of the punt, but this is a detail.)
Europhobes have argued that the May-December period will be exceptionally vulnerable to market attack. But the forward exchange rates for 1 January 1999 have already converged precisely on these ERM central rates, which shows that the financial markets believe that the starting rates are viable. If disruptive speculation were likely to take place after 3 May, then it would already have started. The supposedly vulnerable period this year will therefore pass without mishap.
What about the phase from 1999-2002? This is much more complex, as a recent comprehensive analysis by Martin Brookes of Goldman Sachs amply demonstrates. The Europhobes tend to view this period simply as a glorified version of the old ERM, with exchange rates for the participating national currencies being totally fixed against each other. They therefore argue that the system could be subject to a traditional speculative attack, with markets dumping weak currencies like the lira in favour of strong currencies like the DM. The central banks would have to offset this by purchasing lira and issuing more DM in order to keep the system intact, which in turn implies that Germany would accumulate financial claims on Italy.
Germany might become uncomfortable with this, since in the event of Italy pulling out of the EMU, and defaulting on its liabilities, significant losses could then be incurred. In extremis, the political noise surrounding these developments could lead to further market pressure, forcing an eventual collapse in the system.
It is very hard to say for certain that such an out-turn is impossible. Ultimately, it remains the case that if there were a total loss of confidence in the lira between 1999 and 2002, then there could be a stampede into DM, which would cause a collapse in the Italian banking system and an unwillingness on the part of Germany and others to support the system. It will not be possible entirely to eliminate this risk until the EU becomes a single sovereign entity, with a plain vanilla single currency. And even then - as the case of Quebec demonstrates - fears of a break-up could still develop, which could cause a run on regional banks.
The real question is not whether such an out-turn is theoretically possible - it always is - but whether the design of the system builds in the necessary failsafe mechanisms to make a collapse of confidence almost inconceivable. In a sense, all paper money systems, and all banking systems, are based on a bluff, and they are only viable as long as that bluff instills confidence in the private sector. Fixed exchange rate systems have not proven to be a very successful bluff, so have been open to frequent speculative attack. Currency boards, like that operated in Hong Kong, are more elaborate bluffs, which have proven to be more robust. EMU is a more elaborate bluff still, so a collapse in confidence seems most unlikely.
The Europhobes do not seem to have fully grasped the byzantine nature of this bluff, or why it makes EMU from 1999-2002 fundamentally different from the old ERM. Central bankers will go to every conceivable length to propagate the belief that a break-up of, or an exit from, the single currency, will be technically impossible.
They will have plenty of ammunition. EMU will not be a system of national currencies linked together by fixed exchange rates but will instead be a system in which the euro is legally the single currency of the entire area, albeit with different national currency denominations (lira, marks etc) for an interim period. In consequence, there will be no foreign exchange market in which central banks will have to intervene, and people will be able to switch between the different national currency denominations simply by asking banks to redenominate their deposits.
Contrary to what the Europhobes have often asserted, a stampede into DM would not be a cost-free option. It would lead to a decline in interest rates on DM bank deposits, which would soon deter people from making such switches in large size. And standing behind all this would be an absolute guarantee by all participating central banks that they would automatically exchange banknotes without limit at the fixed EMU rates.
Admittedly, this would still mean that if something did go wrong, it would require Germany to accumulate financial claims on Italy (for example, lira banknotes) in unlimited quantities. In truly extreme circumstances, Germany might become unwilling to hold such claims. But the new system will go to great lengths to disguise this remote possibility, emphasising instead that the treaty requires Germany passively to accumulate Italian paper, and that the workings of the European Central Bank make this automatic. Only if the market comes seriously to believe that either Germany or Italy would renege on treaty obligations would this bluff ever be likely to be called, and that represents a very high obstacle against speculative attack.
Finally, what about the period after 2002? At that point, the bluff becomes even more elaborate, since the euro will replace national banknotes, making it impossible to hold DM instead of lira. And no provision has been made for a country ever to pull out of EMU.
Admittedly, for as long as EU governments remain ultimately sovereign, they can presumably re-establish their own national currencies. Technically, this would be very difficult but not absolutely impossible, so from a nuts-and-bolts point of view, a decision to join EMU would not be irrevocable. But in reality the only reason for exiting the euro would be to introduce, and then rapidly devalue, a new national currency, thus making citizens of the country concerned very much poorer than they were inside EMU. The main mechanism for this would be that debts to foreigners, denominated in the appreciating euro, would be much more burdensome than before. (Asia 1997, writ large!)
A cast-iron rule in a democracy is that electorates do not vote for a devalued currency. In 1990, the Germans opted for the D-Mark, not the Ostmark. If a future political party in Italy wanted to pull out of the euro, it would need to persuade voters in effect to support an Ostmark policy. Only in a 1930s-style recession would that become even the remotest of possibilities.
So while an exit from the monetary union might remain a technical feasibility, a decision to join should, for all practical purposes, be treated as irrevocable. That is why it is so important.