At the same time, though, politicians have competed to attract deep and active financial markets to their own countries. At this very moment, Paris, London and Frankfurt are engaged in a battle to become the dominant financial centre in the European Community. One way in which governments compete is to keep regulatory barriers and capital requirements to a minimum, hoping that this will encourage the global brokers and investment banks to locate on their soil. This encourages the growth of well-capitalised and liquid markets - necessary pre-conditions for the growth of 'speculators'.
Governments enjoy the benefits from financial operators - growth in employment and tax revenue, for example. And they are keen to expand their domestic government bond markets, knowing that a liquid market will help suck in foreign capital to finance their budget deficits. But while governments are all too eager to attract an inflow of foreign capital, they often forget that this automatically increases the potential for currency outflows when things go wrong.
Nevertheless, a sustained run on a currency can only be created when 'real money' - controlled by long- run portfolio managers, corporate treasurers and ordinary citizens - loses confidence in a given exchange rate. It is true that the 'speculators' (defined as the financial intermediaries or market makers, and the hedge funds) can muster billions of dollars to sell a currency in virtually no time, but they are by their very nature quicksilver characters. Unless they are persuaded that 'real money' is following on behind, they are unlikely to maintain a loss-making position for very long.
'REAL MONEY' ATTACK
The truth is that a determined coalition of central banks can always defeat the 'speculators', but can never hope to defeat the 'real money' when it starts to move. A fundamental misalignment in an exchange rate, such as that which probably applied to the lira and pound during 1992, will sooner or later bring an attack from the 'real money', at which point changes in parities become inevitable.
But there is little evidence that the French franc has been misaligned in the traditional sense against the mark. In fact, one of the conundrums facing the EC finance ministers yesterday was that any downward movement in the franc's parity would increase its under-valuation, thus increasing the degree of fundamental misalignment. In these circumstances, there is a strong case for the central banks standing firm against speculative attack. Last week, they chose not to do so.
Several ways have been suggested for improving the hand of the central banks in conditions such as these. The reimposition of capital controls - for example by levying a tax on short-term borrowing of currencies by the 'speculators' - can probably buy some time, but at considerable cost to the long-term viability of the financial centres involved.
An alternative idea, initially proposed by James Tobin, is to levy a worldwide tax on financial transactions, thus throwing sand into the market mechanism. This would be equivalent to increasing the 'spread' between buying and selling rates in the currency markets, thus tilting the balance against short-term transactions. But the problem with this idea is that it would not be sufficient to levy such a tax throughout the EC. This would simply shift business elsewhere, to offshore financial centres, and these centres would certainly not co-operate with the Community in levying the tax.
Other suggestions have focused on the reform of the ERM fluctuation bands themselves, usually involving wider bands and more frequent realignments. But this would transform the ERM into a crawling peg system, with few of the characteristics for which the mechanism was originally designed.
After the upheavals in the ERM last autumn, the EC established a working party to consider the 'fault lines' in the system. All of the above options were no doubt considered, but in the end no significant changes were recommended. This may now seem to have been a mistake, but the fact remains that the ERM system could probably have continued to work perfectly well, provided that all the central banks had been willing to use their powers to the full.
Let us imagine that the Bundesbank and the Bank of France had jointly announced last week that they were determined to maintain the existing parity grid, and they were willing to intervene without limit in the foreign exchange markets to achieve their objective. What would then have happened? A substantial speculative attack would still have built up, but it would have been met by unlimited sales of marks by the Bundesbank, with commensurate purchases of the weaker currencies.
The German money markets would have been flooded with liquidity as the additional marks were placed on deposit. German money market rates would have dropped. Simultaneously, the French markets would have been drained of liquidity, pushing up short-term interest rates there. The interest-rate spread between France and Germany would have widened considerably, making it much more expensive for speculators to sell francs. In both of the previous franc crises, in September 1992 and January-March 1993, this is what happened, and the speculative attack was eventually defeated.
At times last week, there were signs the same tactics were being used again, but this belief was punctured by the Bundesbank's decision on Thursday to leave interest rates unchanged. The crucial decision here was not so much that the discount rate was left at 6.75 per cent. As the graph shows, this rate does not necessarily provide a floor for money market rates, which can drop substantially below it for short periods.
In fact, the Bundesbank's crucial decision last week was to announce that the repo rate this week would be fixed at 6.95 per cent, thus making it clear that money market rates had no further downside. This temporarily breached the rules of the game, and the markets instantly sensed a fissure in Franco-German relations.
I say 'temporarily breached' because the Bundesbank quickly reversed its decision on Friday morning, when it announced that there would in fact be no limit on the amount by which German money market rates could fall. If this announcement had been made on Thursday, even combined with an unchanged discount rate, the currency pressure could well have subsided, as it did on previous occasions.
It is impossible to know why the Bundesbank acted as it did. Was it just a tactical mistake? Or had the Central Council decided the upward pressures on the mark were 'fundamental' rather than 'speculative', in which case they would eventually prevail? And why did the German government not force the Bundesbank to continue intervening, thus in effect forcing them to reduce their money market rates? These are questions the rest of the Community will be asking the Germans for some time.