Last year I presented the thesis that the euro would remain weak until its introduction in physical form and would gain strength thereafter (Financial Times of 4 April 2002, p. 15). This is precisely what happened. The euro fell in cycles until February of this year, the date of the official end of the currency changeover, and in March it began to gain strength, recently almost achieving parity with the dollar.

The argument I presented last year was that foreign currency holders, mainly in Turkey and Eastern Europe, moved out of D-mark cash into other currencies. Currency holders outside Germany, who held approximately a third of D-mark currency stocks in 1995, had heard of the planned abolition of the D-mark, but did not know what to think of its replacement, the euro. In any case they were afraid of sustaining losses from the exchange, and did not even know who would carry it out.

I also called attention to cash holders on the European black market who would not want to reveal the source of their funds when exchanging them at the bank counters.

Numerous indicators now support this theory, which initially was only buttressed by the anecdotal evidence of large amounts of cash flowing back to branches of the Bundesbank near the eastern borders. In a survey made by the Austrian central bank in May 2001 of thousands of currency holders in Eastern Europe, of those who had already formed an opinion more than forty percent did not want to shift from the D-mark to the euro. In an expert survey by the Ifo Institute, a majority confirmed that Eastern Europeans did not feel sufficiently informed about the exchange modalities and were worried about the exchange, which later proved to be justified because of the high fees charged by Eastern European banks.

The most important support of the thesis, even last year, was the very sharp decline in the growth of D-mark cash in circulation. The decline had begun in 1997, shortly after the abolition of the D-mark was announced at the Dublin summit, and from that time on the D-mark began to depreciate.

In the meantime the growth decline accelerated, and in the winter of 2000/2001 D-mark cash began to fall even in absolute terms. By February of this year, D-mark currency in circulation had fallen, against the trend, by about €90 billion, from €160 billion to about €60 billion, an unprecedented change in money supply growth.

Parallel to this, total eurozone currency in circulation also began to fall. The decline, against the trend, amounted to about €120 billion by February of this year. This meant, in addition to the drop in the demand for currecny registered by the Bundesbank, another €30 billion in the other countries. If we assume that Germans hold no more illicit money than other Europeans, the only interpretation is that the D-mark cash returning from abroad comprised the lion's share of the money stock reduction. About €75 billion of the decline against the trend falls in this category.

The decline in the demand for cash began much too early to be attributable to technical factors in the currency changeover. Rather it was due to the fact that, in light of the declining liquidity preference, the Bundesbank had to reduce the money supply in order to defend its minimum lending rate and in compensation had to increase the amount of interest-bearing assets in the private sector, which was done in part through discount and repurchase agreements. This passive intervention succeeded in defending interest rates but provided only modest support to the exchange rate. In spite of the passive intervention the euro remained weak, for a time even falling below 85 US cents.

These explanations of the depreciation of the euro were rejected by many actual or putative experts. Initially they disputed the facts as such, and when this was no longer tenable, they claimed that the noted currency effects were too small to be able to explain the depreciation of the euro; indeed, the currency stocks involved were much smaller than even short-term foreign- exchange flows. These assertions miss the point because it is not short-term foreign-exchange flows that are relevant for exchange-rate movements but the long-term demand for currency stocks. Flows go in both directions and explain nothing despite what forex dealers may believe. Fortunately, econometrics offers a clear answer as to the size of the observed effects. The observed reduction in the demand for euro cash of €120 billion implies a long-term depreciation of about 36 cents. That is enough to explain the actual depreciation from Janzuary 1997 to February 2002, and is also far more than the foreign-currency reserves held by the European Central Bank for foreign-exchange market intervention.

A counter-argument that is sometimes put forth is that the money supply M3 should have decreased if indeed there had been a movement out of the D-mark. This argument overlooks that Bundesbank intervention takes place within the broad money aggregate M3. The replacement of cash by assets for the purpose of interest-rate stabilisation, both of which are part of M3, would certainly not have changed this money aggregate.

It is interesting in this connection that the European Central Bank decided last year to change its M3 money supply definition, because in recent years increasingly larger parts of interest-bearing components of this broad money aggregate had moved into the hands of foreigners. According to ECB information and calculations of the Ifo Institute, assets worth about €100 billion were affected by this correction from January 1999 to September 2001. They presumably constitute the other side of the flight cash that was taken back to stabilise interest rates.

In the meantime the flight has ended, and the trend has reversed. Since March the euro has been rising, and at the same time the currency emitted by the ECB has begun to increase again. It is all a mirror-image of what happened before the currency conversion. The euro is well suited for illicit dealings and is finding new friends in Eastern Europe, Turkey and elsewhere in the world. According to a recent information of the ECB, cash of no less than €18 billion has moved to eastern Europe from January to May 2002. Gradually the euro cash stocks are replacing other currencies, which temporarily had found their way into the pockets of the money holders, and euro cash is in short supply. The euro is gaining strength, even though the ECB is now passively intervening by pumping new cash into the economy to defend its interest-rate targets.

Of course there are rival explanations of the strength of the euro and/or the weakness of the dollar. Some have pointed out that the weakening dollar reflects the fading confidence in the American economy. This sounds plausible at first glance, but cannot be supported in light of the confidence surveys among investors and consumers. Not only the surveys of the Ifo Institute in America but also many other surveys clearly show that confidence in the economic upswing in the US is considerably more robust than in Europe.

The often quoted thesis that the weak dollar mirrors the lull on the American stock exchanges is also not very sound. True, the Dow Jones Index has fallen since the beginning of the year, but the DAX and the Euro Stoxx Index have fallen much more. Relative to European equities, American stocks have thus increased in value this year, and that implies a strong, not a weak dollar.

Nowhere is the belief in false and completely untenable theories so wide-spread as in the foreign exchange markets. Superficial plausibility always prevails over economic analysis. The former can frequently win on points because false theories may be true for a time if everyone adjusts his purchasing behaviour accordingly. But in the long term the fundamentals prevail. The falling cash demand among Eastern Europeans and illicit currency holders up to the euro changeover and the rising demand thereafter is one of these fundamentals.

Hans-Werner Sinn
Professor of Economics and Public Finance
President of the Ifo Institut

Published as "Warum der Euro steigt" (Why the euro is gaining strength) in Handelsblatt, No. 129, July 9, 2002, p. 8.