European Central Bank Attitudes

July 2, 1998

We spent the last week in Europe discussing some of the analysis and conclusions of our first edition of The Gold Book Annual with several central banks in Europe as well as some producers, dealers, and the World Gold Council management in Geneva. Several issues struck us as key.

  1. We have argued that Gold Fields Mineral Services' (GFMS) so- called "official statistics" on the gold market understate global gold demand, borrowed gold flows, and the deficit in the gold market. There appears to be a growing awareness of this. World Gold Council officials do not take issue with our extrapolations of their gold demand data to overall global demand levels that exceed GFMS levels. We were surprised to find that one of the major producers, apparently after having done some research, has concluded that outstanding gold loans may approach 7000 tonnes---a figure that is not far removed from our own estimate of 8000 tonnes. Whereas a year ago our views were widely regarded as "wild" and "irresponsible", we now see growing credibility in our positions. This coincides with some concessions on GFMS' part in revising upward their estimate of outstanding gold loans by 950 tonnes and conceding the existence of fabricator and refinery borrowings.
     
  2. The central banks appear to have taken GFMS' data on the gold market as fact. There may be no awareness on their part that the gold market deficit could be understated. It is widely recognized by central banks that their selling and lending of official gold is depressing the gold price, but, owing to misleading GFMS data, they do not appreciate how severely it has been depressing the gold price.
     
  3. This statistical issues are clearly very relevant to official sector behavior toward gold. We argue that the central bank gold flows were on the order of 1100 tonnes per annum in the period 1994-1996 and were depressing the gold price from almost $600 to an average of $386. If we are correct, and if central bankers believed what we believe, central Banks would be less inclined to lend and sell gold. From this perspective, the World Gold Council's field work has proved critical to the central bank issue: without their Gold Demand Trends survey, there would be no alternative demand survey to that of GFMS.
     
  4. There is another way in which these statistical issues have a bearing on the behavior of central banks toward gold. In The Gold Book Annual we show that, at a constant real gold price, since the early 1800's, global gold fabrication demand has risen at roughly a 4.5% annual rate whereas ex ante mine supply has risen at a 2.8% annual rate. Over the 25 years from 1971 to 1996, at a constant real gold price, ex ante gold demand rose at a 6% rate while ex ante mine supply grew at more than a 1.8% annual rate. In the absence of monetary and Western investment flows, the gold market can clear ex post only if the real gold price rises sufficiently in real terms to ration price elastic demand down to the level of price inelastic mine supply. By our calculations, this implies a trend annual real rate of appreciation of gold of more than 2% plus over almost two centuries and roughly 6% over the last quarter century. Some central bankers sell gold because they believe it is "barren". Almost all central bankers would agree that a 2% real rate of appreciation of gold is a high return for a reserve asset. We demonstrate in The Gold Book Annual that there is a great discrepancy between the growth rates in gold demand in the WGC and GFMS data for the period 1991-1996. The WGC data indicates a rate of growth of gold demand roughly in line with the historical trend of perhaps 5% per annum. The GFMS data shows a trend rate of gold demand relative to a constant real gold price of 2%. The difference between the trend growth rates of demand and mine supply determine the trend in the real price of gold generated by "commodity" dynamics alone. The WGC data suggests a continuing high positive real return to gold; the GFMS data suggests a slightly negative real return. Since the attitudes of many central banks toward gold are greatly influenced by gold's perceived real rate of return, the difference between the GFMS and WGC gold statistics are very relevant to the central bank issue.
     

The WGC, in its work as a marketing arm of the gold mining industry, has developed extensive point of sale information that has produced an alternative more intensive and higher quality survey of global gold demand to that of GFMS. Concerned about central bank lending and selling, the producers who fund the WGC are shifting dwindling WGC resources from gold marketing to central banking issues. The great irony is that the Council's marketing efforts provided a quality demand survey that supports the strongest arguments that can be made to the official sector to stop selling and lending gold.

Only about one third of global gold mine production now supports the WGC. It is often argued that the WGC has spent huge sums on marketing with little in the way of results and that it should have allocated resources to the central banking issue. As we make abundantly clear in The Gold Book Annual, the gold market is now a jewelry market, and it is an increasingly modern market for luxury adornment jewelry. In South East Asia and to an increasing degree in the Middle East demands for luxury adornment jewelry now far overshadow atavistic propensities to hoard gold. For such a market, consumer marketing is productive. What is ironic is that the WGC's field work in consumer marketing has generated the strongest case the gold industry could make to the official sector to stop selling and lending gold. It would be a huge negative to the industry if short-sightedness led the industry to cripple its efforts in its two key markets: the global market for adornment jewelry and the "market" of official gold holders.

A sheet of gold can be made thin enough to be transparent