Abstract

Gold has extraordinary durability and is never “consumed” in the sense that another good might be. In fact, most gold that has ever been brought above ground has remained there, and analysts believe that at least 80 percent of all gold that has been mined can be accounted for (see Green (1981)). This quality has greatly enhanced gold’s attraction as a store of wealth, but it also means that an enormous stock of gold exists that can potentially return to the market and add substantially to annual supplies from new production, if prices increase sufficiently. This is particularly true of gold bars hoarded for investment or held to back physical gold accounts, as well as jewelry in India, the Middle East, and the Far East that is usually of poor fabrication quality but of high gold content and held for investment purposes.10 Given the magnitude of these potential gold flows from existing stocks, it is clearly not sufficient to consider only the current supply of new gold in the structure of the gold market.

Gold has extraordinary durability and is never “consumed” in the sense that another good might be. In fact, most gold that has ever been brought above ground has remained there, and analysts believe that at least 80 percent of all gold that has been mined can be accounted for (see Green (1981)). This quality has greatly enhanced gold’s attraction as a store of wealth, but it also means that an enormous stock of gold exists that can potentially return to the market and add substantially to annual supplies from new production, if prices increase sufficiently. This is particularly true of gold bars hoarded for investment or held to back physical gold accounts, as well as jewelry in India, the Middle East, and the Far East that is usually of poor fabrication quality but of high gold content and held for investment purposes.10 Given the magnitude of these potential gold flows from existing stocks, it is clearly not sufficient to consider only the current supply of new gold in the structure of the gold market.

Annual Demand and Supply

In theory, the annual supply to bullion markets should be equal to turnover and demand, but in practice it is impossible to estimate turnover and far less possible to estimate all of its sources of supply. For this reason, analysts have confined themselves to estimating sources of supply of new gold (conventionally measured as western world mine production and net sales from nonmarket economies) and have augmented this figure with estimates of major changes in the composition of existing gold stocks. The major estimable sources of stocks that have been retired from a prior use and returned to the free market in bullion form are sales of scrap metal and net sales from official gold holders. These figures have therefore been included in estimates of what will be termed net annual market supply.

The annual supply of new gold bullion from mining and net sales by nonmarket economies has risen steadily during the 1980s. It is estimated to have averaged 1,822 tons per year in 1985–91.11 South Africa still dominates gold production and supplied an average of 33 percent of new gold in 1985–91. The former U.S.S.R. is estimated to have been the second largest producer, supplying 15 percent,12 and the United States supplied 10.3 percent. The inclusion of net official sales by central banks and other official institutions, which actually absorbed an average of 54 tons a year in 1985–91, and the supply of gold scrap from the melting of coins and jewelry, which averaged 407 tons a year (estimates of gold scrap are available only beginning in 1980), brought the total net bullion supply to the private sector during 1985–91 to an average of 2,175 tons a year.13

Net market demand is even more difficult to estimate. Of the 2,175 tons supplied, 2,026 tons are estimated to have been absorbed in fabrication annually during 1985–91,14 and the remainder represents implied net private bar hoarding of 149 tons. Information on these residual flows for hoarding purposes is the most elusive. Estimates must rely on observed cross-border movements, but even these are generally not available for Europe and North America,15 and market stocks in London and Zurich are not revealed.16 Therefore, estimates of the geographical distribution of hoarding are limited to countries outside Europe and North America, with the residual assumed to have been hoarded in Europe or North America. During 1985–91, identified bar hoarding in the Far East averaged 256 tons a year; on the Indian subcontinent, 13 tons a year; in the Middle East, 28 tons a year; and in Latin America, 24 tons a year. This implies that an average of 172 tons a year were dishoarded from Europe and North America over the period.

The breakdown of net market demand and supply given here is used by market analysts to project likely developments in market fundamentals and prices (see Milling-Stanley (1991)). However, behind this estimated breakdown of net market demand and supply from the different categories, there are a multitude of gross flows that have been netted out, in particular with regard to net private hoarding and dishoarding and net official transfers, and these must obviously have had an effect on day-to-day transactions in the market.

Although it is not possible to estimate these transactions, it is possible to reclassify some of the component net gold flows outlined above in order to come closer to an estimate of gross flows and to illustrate better their effect on the market. Since 1989, Consolidated Gold Fields has estimated “gross annual bullion supply and demand” according to whether net investment flows have been negative or positive. This calculation leads to a larger estimate of supply and demand because, for example, net official purchases (if they occur) will be added to demand rather than being subtracted from supply, while net official sales (if they occur) will still be added to supply. More precisely, the approach is to reclassify net flows to or from a particular category as either supply or demand in each year, depending on whether they constituted a net supply to, or demand from, the market. Although the resulting figures are still broad estimates, their recategorization makes for a better representation of the positions adopted by certain sets of market participants in each year. On the basis of such a representation of the structure of the market, annual market demand and supply averaged 2,530 tons in 1985–91 (instead of 2,175 tons, according to the earlier definition). These gross estimates are illustrated in Chart 2 for the period 1968–91, and Table 1 details these estimates of market supply and demand for the period 1985–91.

Chart 2.
Chart 2.

Gold Bullion: Supply and Demand, 1968–91

(In metric tons)

Source: Gold (various years).
Table I.

Annual Bullion Supply and Demand

(In metric tons)

article image
Source: Gold (1992).

Assumed to represent Europe and North America.

Outside Europe and North America.

Composition of Existing Stocks

In 1800, some 4,200 tons of gold were thought to have been above ground, and this stock is estimated to have quadrupled by 1941. By 1968, the stock had risen to almost 70,000 tons, of which 36,192 tons (or 52 percent) belonged to official monetary institutions, and at the end of 1991 estimated cumulative gold production was over 100,000 tons (see CPM Group (1987, pp. 2–16), The Economist (1989), and Gold (1992)). Of this total, about one-third is still held by central banks and other official institutions (see Table 2). Similar proportions are accountable for by jewelry and private holdings of bars and coins. At the end of 1991, the gold stock (excluding that lost or used in industry) was almost 60 times the annual supply of new gold from mines.

Table 2.

Distribution of Holdings of Cumulative World Gold Production, December 31, 1991

article image
Sources: CPM Group (1990, pp. 38–39), Green (1981), IMF, International Financial Statistics (February 1993), Gold (various years), and Russian authorities. Note: Figures are subject to revision if official data are revised.

Most official gold is stored in the United States, with some 10,000 tons placed at the Federal Reserve Bank of New York by foreign and international institutions. The U.S. gold reserves of 8,147 tons are stored at Fort Knox in Kentucky and West Point in New York. At a market price of $350 an ounce, total official gold holdings would be worth $400 billion, but about half of these institutions value their gold holdings at some historic price far below the market price.17 Central bank holdings alone would be worth some $330 billion at $350 an ounce, and, even though gold no longer has an official role in the world’s monetary system, gold reserves are still an important component of international reserves. At the end of 1991, gold reserves at market prices accounted for about 28 percent of the value of total central bank reserves, compared with an interim peak of 58 percent at the end of 1980.18 Clearly, central banks still have a large role to play in the gold market, even if it is a passive one.

The former U.S.S.R. has long maintained absolute secrecy regarding its annual gold production, stocks, and sales to the West.19 In 1935 the Soviet Government announced that gold output for the year totaled 150 tons and that reserves amounted to 745 tons (Euromoney (January 1987, suppl., p. 12)). The next official announcement was made on February 22, 1991, when it was revealed that the U.S.S.R. placed 234 tons of gold on the market in 1990 to pay for food imports.20 Precious Metals: Gold (CPM Group, 1990) provides estimates of Soviet production, sales, and implied reserve accumulation since 1950, and it is on the basis of such estimates that reserves at the end of 1989 were placed at some 2,300 tons. This estimate appears to have been generally acceptable to industry analysts, but the range (including possible error) was put at 2,000–2,500 tons. It is likely that a large part of the total was held outside Gosbank.

These estimates may be compared with recent unofficial information from the former U.S.S.R. Russian economist Grigory Yavlinsky has maintained that the former U.S.S.R. had a total of 390 tons of gold in October 1991, of which 150 tons had been swapped. He has also said that 700 tons had been exported since the end of 1989 (reported by Reuters (October 16, 1991)). Former Prime Minister Nikolai Ryzhkov has stated that the gold reserve was 784 tons at the end of 1989, and that 250 tons was sold during 1990 (Reuters (October 21, 1991)); and a Gosbank official, Alexander Doumnov, has stated that Gosbank had 375 tons in reserve in mid-1991 (reported by AP-DJ (September 24, 1991)). It is not clear whether Yavlinsky was referring to total stocks, or to a second reserve held outside Gosbank, perhaps at the Finance Ministry. Because these recent announcements have not been sufficiently clear about gold held outside Gosbank, a number of market analysts believed that total former U.S.S.R. reserves as of mid-1991 were in the range of 1,000–2,000 tons (see Salomon Brothers (1991)). However, according to recent data provided by Soviet and Russian authorities, the total former U.S.S.R. reserves as of the end of 1991 were only 290 tons. See Table 3 and Box for further analysis and discussion of these data.

Table 3.

Former U.S.S.R.: Gold Production, Sales, Consumption, and Reserves, 1968–91

(In metric tons)

article image
Sources: Data for 1984–1991 are from the former U.S.S.R. Ministry of Finance as reported in Moscow News, November 17–24, 1991. Reserves data for 1986–90 were previously reported in IMF, Economic Review: The Economy of the Former U.S.S.R. in 1991, Table 34. Data for earlier years were compiled as follows: (I) mine production is from market estimates—CPM (1987) and Gold (1992); (2) exports and reserves data were provided by the Russian authorities; and (3) the remaining data for domestic consumption were calculated as a residual by IMF staff.

Data for January–September.

Data from the Russian authorities.

Bar hoarding since 1968 has been estimated by Gold Fields Mineral Services (Gold (various years)), but it is advisable to bear in mind that European and North American accumulation has been compiled as the residual demand in each year and may not be very accurate. Bar hoarding prior to 1968 is not very well identified, but French individuals are generally considered to have accumulated 5,000 tons in the years between World War I and World War II and immediately after. Altogether, some 55 percent of the gold stock is in private hands and (at least theoretically) is available to enter the market in response to price changes. It is, difficult, however, to provide a systematic analysis of the manner in which stocks respond to price increases because, as noted earlier, many gold movements are netted out in the data.

Timothy Green of Gold Fields Mineral Services cites a number of dishoarding episodes, including in 1932 when a sterling devaluation of 50 percent initiated dishoarding from India that was equal to total South African production for that year (Green (1981)). The most recent episode occurred in response to the price increases of late 1979 and 1980. The doubling in the average annual price of gold at that time apparently caused a reduction in hoarding outside of Europe and North America by 168 tons from 1978; in Europe and North America hoarding increased by 154 tons. This effect indicates the prevailing market forces, with speculative demand in the major markets attracting external stocks, but the implied net aggregate dishoarding of only 14 tons completely disguises the extent of the shifts in existing stocks. To a large extent, aggregate stock equilibrium was adjusted through the redirection of annual supply.

Gold use from all sources in the fabrication of carat jewelry fell sharply in 1980, falling to 128 tons from 728 tons in 1979 (and 1,004 tons in 1978). Total fabrication demand in the West fell by 780 tons to 540 tons, but this drop in demand was more than offset by a fall in supply caused by both a reduction in net sales of nonmarket economies and a switch to net official purchases amounting to 883 tons. However, there was a large increase in the supply of gold from scrap. Although exact data are not available, Green (1981) has vividly described the flow of gold from jewelry to speculative holdings through the scrap market after prices peaked on January 19, 1980. Initially, the loco-London price gained a substantial premium (in some cases, 40 percent) over small local markets (principally in the Middle East and Far East). Then, according to Green (1981, p. 28), “within a very few days a wave of reselling of jewelry and coin on an unprecedented scale developed.” Green continues:

The scale of individual dishoarding was often substantial. I was in the Kuwait souk the day gold hit $850. The street of gold shops was under siege. Perhaps two or three thousand people milled around, fighting their way into the overcrowded shops with cigar boxes and biscuit tins crammed with bangles and necklaces. The goldsmiths weighed the ornaments—often two or three kilos from one woman [worth $50,000-$75,000]—and paid cash. Those ornaments were melted down in a nearby basement and the rough gold bars air-freighted to London and Zurich the same night (p. 29).

Green also estimates that some 200 tons of gold returned to the main markets (London and Zurich) in early 1980, while an additional 100 tons returned to subsidiary markets (such as Hong Kong), eliminating the need to buy from Europe to meet investment demand there.21 The main impediments to further flows were a cash shortage in the subsidiary markets and constraints on the capacity to refine scrap to loco-Europe quality once it had arrived there. By late March 1980 the price of gold had fallen to $474 an ounce.

Annual Supply to Bullion Markets

As a rule, the organizers of the physical gold markets do not reveal information on turnover or on the volume that moves through the market on an annual basis, in order to protect the confidentiality of transactions by individual customers. However, some information can be gleaned from the trade figures of the countries involved. Although all physical gold flows through a particular country are not actually transacted on local markets, as pointed out by Schriber (1981), this assumption seems a reasonable approximation. In fact, the reluctance on the part of Switzerland and the United Kingdom to provide detailed trade figures may be an indication that these statistics would be of use in determining the flows of gold to the market.

Table 4 provides estimates of the size of physical gold flows from the gross annual gold supply through the various physical gold markets in 1989. The figures on new gold and scrap, which are very reliable (Gold (1992)), have been combined with estimates of official sales and residual disinvestment in Europe and North America in order to conform to the definition of supply offered in the subsection on annual demand and supply (and used in Table 1).

Table 4.

Estimated Composition of Market Supply and Demand in World Bullion Markets, 1989

(In metric tons)

article image
Sources for columns:

Gold (1992).

The aggregate figure is a residual, representing implied disinvestment in Gold (1992). The breakdown is based on U. K. and Swiss trade data, as supplemented by estimates in Montagu (1990), with “other European” demand as the residual.

For United Kingdom, see Table 7; Switzerland is assumed to accept residual Latin American supplies (after the United States and Canada), and most of other Africa, Europe, Asia, Oceania, and the Middle East; estimates for the Far East assume that supplies from the United Kingdom and Switzerland (col. 5) are equal and distributed according to local market size.

United Kingdom (see Table 7) is assumed to be destined for Far East, and Switzerland assumed to be equal to United Kingdom; Latin America is assumed to export, for refining, everything but local Brazilian demand; others have limited refining capacity and export indigenous supplies. For Australia, see CPM (1990).

For United Kingdom, see Table 7; Switzerland is the residual import figure (from Table 8) after the entry in col. 4.

South Africa and the nonmarket economies export in well-refined form; Canada and the United States are residual estimates; for Australia, see CPM (1990).

Gold (1992); net official sales of all countries are here assumed to have been carried out in London.

Derived as cols. (l)+(2)+(3)+(4)+(6)+(8)–(5)–(7).

Hong Kong and Singapore are residuals but conform to net flows in Montagu (1990); other figures estimated, so that when combined with col. (9), they conform to local market demand.

Equal to cols. (9)+(10)+(11I); also equal to cols. (13)+(14).

Gold (1992).

New gold is defined as newly mined gold and net sales from the former U.S.S.R. and other nonmarket economies.

Unrefined gold is here defined as gold not conforming to loco-London standards of fineness or bar weight.

Loco-London weight and fineness. May be transformed at destination.

Primary market supply signifies gold that has first come to market; secondary market supply signifies gold that is resold on a secondary market.

Net imports differ from secondary market supplies because they are not resold at destination, and refer to net imports (exports if negative) of nonsecondary market participants. The decomposition of the excess local demand over primary market supply between the secondary market supply and net imports is somewhat approximate and thus should be interpreted with caution.

Identified bar hoarding, except in Europe and North America, which is assumed to be the residual supply (in col. 3).

The estimated breakdown of residual disinvestment is based on rough calculations of total demand in the United Kingdom, the United States, and Switzerland (see Samuel Montagu & Co. (1990)), with “Other Europe” treated as the ultimate residual. An attempt is then made to trace the total supply (2,826 tons) through the various bullion markets and ultimately to total demand, which comprises fabrication demand and identified bar hoarding outside Europe and North America.

In Table 4, in columns (3) and (11), estimates of unknown flows have been made so as to conform to the available information; any remaining gaps have been filled on the basis of anecdotal evidence on flow trends or available proxies. Thus, these estimates should be regarded as reflecting some errors and omissions.22 The general approach has been to leave as few residual values as possible and to estimate these on the basis that the total in the relevant column must either conform to the total in another (as with columns (4 and 5) and (6 and 7)) or sum to zero (as in columns (10 and 11)). There is often more than one residual for a particular row or column, but because the data must be consistent for the rows as well as the columns, their number could be reduced significantly. For example, if an item is one of two unknowns in a particular column, but is the only unknown for a row, its identification from the row will also identify the other unknown in the column.

The breakdown of supply has been supplemented with a distinction between refined and unrefined gold (columns 4–7). “Refined gold” is gold conforming to loco-London delivery standards of fineness and weight, and “unrefined gold” refers to gold in other forms. The estimates have been derived to conform to official import data for the United Kingdom in particular, but in most cases they reflect the extent to which gold is thought to have been exported in mine-treated form (80 percent fine) or refined (above 99.9 percent fine) by the producer countries. However, the estimates should not be taken as an indication of the extent of refining operations in recipient countries because imports of refined gold can be further smelted at their destinations.23

Soviet Gold Figures

With the breakup of the former Soviet Union in 1991, there has been much speculation on the size of its gold reserves and its past levels of production and exports. Prior to 1991, the industry consensus was that there were about 2,000 tons in reserve at the end of 1989, but announcements by a number of officials in the autumn of 1991 suggested that the correct figure was much lower, with gold reserves at the end of 1989 officially reported at 784 tons. Table 3 provides information on exports and reserves recently made available to the IMF, along with industry and staff estimates. The data on domestic consumption for 1968–83 have been calculated to conform to the assumption that in each year mine production minus domestic consumption is equal to exports plus change in reserves.

The historical data on the gold transactions of the former U.S.S.R. require some clarification. First, the implied domestic consumption levels are somewhat irregular prior to 1984 and appear to be exaggerated in some years. This could be due to a discrepancy between gold stocks allocated for sale in one period but removed from reserves during the next, or to an over-estimation of production levels. Second, production levels for the period 1984–90 are somewhat higher than previous industry estimates, which suggests that unmined reserves are larger than believed by Western observers. Finally, it is not clear whether there was a significant fall in production in 1991 as suggested by industry sources.1 A former Gosbank official has stated that production was maintained at about 300 tons during 1989–91.2

Reserves data for 1967–91 probably exclude gold that had been swapped, whereas the exports data for later years apparently include swaps in addition to outright sales. Exports (including swaps) and domestic consumption in 1989–91 have significantly exceeded mine production, thereby reducing reserves to 290 tons at the end of 1991.3 Mr. Yevgeny Bychkov, head of Russia’s Precious Metals and Stones Committee, said in March 1993 that the country’s official gold reserves had recovered to 320 tons at that time (Boulton (1993)).

1 See Gold (1992, p. 21). Gold Fields Mineral Services contend that there were increasing problems with production during 1990 and 1991. There may also have been a slowdown of deliveries to Moscow from Russian producers and other republics. Generally, Russia was thought to produce some two-thirds of total output from the former U.S.S.R., with Uzbekistan accounting for a further 25 percent (principally from the Muruntau complex), and Kazakhstan accounting for most of the remainder. Turkmenistan and Armenia also produce gold, while Georgia has discovered gold deposits. See Gold (1992, p. 31). 2 See: “Gold Pause,” Daily Telegraph (April 1, 1992, p. 25) and “Russia’s Gold Output Said Likely to Decline,” American Metal Market (April 3, 1992). This contention has been supported by V. Tarakanovskij, Chairman of the co-operatives (or, “artels”) union in Delovoi Mir (July 23, 1992, p. 10). 3 See “Exports Dent Russia’s Official Gold Holdings,” Financial Times (June 11, 1992, p. 32).

This tenuous distinction is made simply as an aid in tracing overall bullion flows. The distinction between the primary and the secondary market supply should also be treated with caution, because some gold flows to more than two markets, while other flows between markets are netted out. Again, however, this distinction provides a valuable check on the aggregate figures, because some information is available on the extent to which individual secondary markets are supplied by the primary markets. In assessing the total size of the flows through the various markets, the sum of the positive figures in columns 9 and 10 should be used. Thus, for example, U.S. markets are estimated to have handled 485 tons (from column 9), while Hong Kong markets handled 426 tons (the sum of columns 9 and 10). Of course, combined market supply of 3,409 tons, which double counts 583 tons of gold that was sold through two different markets before reaching its final destination, will be greater than ultimate demand and supply of 2,826 tons. “Net imports” are differentiated from secondary market supply, in that they are assumed not to be resold at their destination but are imported directly by the ultimate buyer from a market abroad.

It is likely that most South African gold exports still arrive in Switzerland, with the remainder going to the United Kingdom. Some 50 percent went to the United Kingdom in 1976 and 1977, the last years for which detailed import figures are available for that country, but this proportion is probably somewhat lower now. The remaining gold imports into the United Kingdom probably originated in large part from other African countries, China, the United States, Canada, and Latin America. Switzerland probably shared imports from Latin America (as well as from South Africa and other African countries), but it is thought to have received most sales from the former U.S.S.R. Prior to 1965, gold sales from the former U.S.S.R. were channeled through the Moscow Narodny Bank in London and, to a lesser extent, the Banque Commerciale pour l’Europe du Nord in Paris.24 The Wozchod Handelsbank was established as a dealer in Zurich in 1966. Reflecting the increasing sophistication of its gold policy, this bank both bought and sold gold on behalf of the Soviet Government throughout the 1970s and early 1980s. However, the bank failed in 1985, with forward trading losses estimated from Sw F 400 million to Sw F 700 million, and its gold trading operations reverted to the Bank for Foreign Trade in Moscow.

Switzerland is still thought to be the destination of most gold sales from the former U.S.S.R., although data on these flows are not available.25 The Bank for Foreign Trade, well known for its sophisticated marketing techniques, both buys and sells (in order to camouflage its net position) and has also been involved in forward selling and other hedging techniques. It has been very active in the market for gold swaps in particular, because this practice allows it to temporarily accumulate foreign currency without causing large quantities of gold to come to market and thus depress prices.

It is difficult to estimate the relative importance of various gold trading and conversion activities in the primary markets, but of the estimated 925 tons that flowed through Switzerland in 1989, some 330 tons are estimated by market analysts to have been refined there, whereas 44 tons are known to have been used in fabrication. Of the remaining 551 tons, some 350 tons are considered to have been subject to some form of upgrading, however slight, such as recasting and stamping.

Thus, about 60 percent of gold imports eventually moved out of Switzerland unchanged or slightly upgraded, while 35 percent were refined in Switzerland. Estimates of the value added from refining range from 0.25 percent to 0.50 percent of the total value of the gold involved, depending on the extent of refining and the price of gold. An estimated 50 percent of gold flows through the United Kingdom are upgraded or recast. Switzerland is thought to provide about 39 percent of the supplies to the secondary markets of the Middle East and Far East, while the United Kingdom is estimated to supply 41 percent. Estimates of market size in these regions are based on import and export figures.

Based on this approach to estimating market flows, the combined market supply was 3,409 tons in 1989 (column 9 total plus positive figures in column 10 in Table 4). Of this total, Zurich accounted for 925 tons, and 657 tons moved through London. New York (485 tons) and Hong Kong (426 tons) were the next largest markets, with Singapore (230 tons) and Tokyo (176 tons) also playing a significant role. The remaining 510 tons were shared among the regional markets.

Estimates of the size of the physical gold flows from the gross annual gold supply through various physical gold markets in 1990 are provided in Table 5. Again, an attempt has been made to trace out the total supply of 2,674 tons through various bullion markets and the total size of gold flows through primary and secondary gold markets. The structure of Table 5 follows that of Table 4, in that net gold purchases (of 66 tons) by the official sector in 1990 are added to the supply side as net official sales, -66 tons, instead of being added to the total gold demand. Net disinvestment is added to the supply side as the sum of three supply items from Table 1—“gold loans,” “forward sales,” and “residual disinvestment”—net of two demand items from Table 1—“hedging” and “residual investment.” Therefore, the local gold demand (or supply) in 1990 computed in Table 5 does not coincide with the total supply figure in Table 1.

Table 5.

Estimated Composition of Market Supply and Demand in World Bullion Markets, 1990

(In metric tons)

article image
Sources for columns:

Gold (1992).

The aggregate figure is a residual, representing implied disinvestments in Gold (1992).

For the United Kingdom, see Table 7. Switzerland is assumed to accept residual Latin American supplies (after the United States and Canada), most of other Africa, Europe, Asia, Oceania, and the Middle East; estimates for the Far East assume that supplies from the United Kingdom and Switzerland (col. 5) are equal, and distributed according to local market size.

United Kingdom (see Table 7) is assumed to be destined for Far East, and Switzerland assumed to be equal to United Kingdom; Latin America is assumed to export, for refining, everything but local Brazilian demand; others have limited refining capacity, and export indigenous supplies.

For United Kingdom, see Table 7; Switzerland is the residual import figure (from Table 8) after the entry in col. 4.

South Africa and the nonmarket economies export in well-refined form; Canada and the United States are residual estimates.

Gold (1992). Net official sales of all countries are here assumed to have been carriedout in London.

Derived as cols. (1)+(2)+(3)+(4)+(6)+(8)–(5)–(7).

Hong Kong and Singapore conform to estimates for 1989 in Table 4; other figures estimated, so that when combined with col. (9), they conform to local market demand.

Equal to cols. (9)+(10)+(11); also equal to cols. (13)+(14).

Gold (1992).

New gold is defined as newly mined gold and net sales from the former U.S.S.R. and other nonmarket economies.

Unrefined gold is here defined as gold not conforming to loco-London standard so fineness or bar weight.

Loco-London weight and fineness; may be transformed at destination.

Primary market supply signifies gold that has first come to the market; secondary market supply signifies gold that is resold on a secondary market.

Net imports differ from secondary market supplies because they are not resold at destination, and refer to net imports (exports if negative) of nonsecondary market participants. The decomposition of the excess local demand over primary market supply between the secondary market supply and net imports is somewhat approximate and thus should be interpreted with caution.

Identified bar hoarding, except in Europe and North America, which is assumed to be the residual supply (in col. 3).

Based on this approach, the combined market supply of gold, which double counts 551 tons of gold that was sold through two different markets before reaching its final destination, was 3,225 tons in 1990. Of this total, Zurich accounted for 912 tons; 511 tons and 443 tons flowed through London and New York, respectively, while 396 tons moved through Hong Kong. Singapore (with 215 tons) also played an important role. The remaining 748 tons were allocated among various regional markets. In 1990 Switzerland is estimated to have provided about 65 percent and the United Kingdom only 31 percent of the supplies to the secondary markets of the Middle East and the Far East.

The references made here to “market size” should, of course, be treated with caution. Obviously, they do not refer to the gross volume of market sales but, rather, to net annual International sales through the relevant market. This concept is of more value than might at first be apparent, allowing for inferences about net (dis)investment from the various centers. The market-centered approach adopted here forces one to think in terms of where the flows of bullion must have been derived and sent, which serves as an ultimate check on the reasonableness of the estimates. An examination of Table 6, which is the change between 1989 and 1990 (as derived from Tables 4 and 5), bears this out. For example, consider the slight decrease in Switzerland’s primary market size in 1990 (of 13 tons). On the demand side, a significant decrease of 186 tons in imports by Latin America (which became a net exporter) is somewhat offset by a 44-ton increase in imports by the Middle East. The balance, of about 145 tons, is counted as a decline in direct exports from Switzerland.26 Increased secondary market demand in the rest of Europe (excluding the United Kingdom), the United States, and the Middle East is assumed to have been largely met by an increased supply of 122 tons from Switzerland. (The United Kingdom, however, is assumed to have reduced its secondary market supply (by 76 tons) because of reduced net demand from Asia.) Combining the balance of changes in direct and indirect (secondary market) exports from Switzerland (-23 tons) with increased local fabrication demand of 10 tons, yields 13 tons, which is the reduction in primary market supply to the rest of the world.

Table 6.

Estimated Changes in Composition of World Bullion Market Supply and Demand, 1989–90

(In metric tons)

article image
Sources: Tables 4 and 5.

When this figure is combined with the other supply-side factors that impinged on Switzerland (and produced an increased supply of 248 tons),27 the balancing figure is increased local investment of an estimated 235 tons (this latter figure can only be estimated because of the market centered approach adopted here).28

The relative significance of these markets is to some extent an artifact of their historical evolution, but increasingly results from their geographical significance and the manner in which they operate.

  • Collapse
  • Expand
  • Chart 2.

    Gold Bullion: Supply and Demand, 1968–91

    (In metric tons)

  • Aggarwal, Raj, and Luc, Soenen, The Nature and Efficiency of the Gold Market,Journal of Portfolio Management, Vol. 14 (Spring 1988), pp. 1821.

    • Search Google Scholar
    • Export Citation
  • Baker, James A., Statement Before the Boards of Governors of the World Bank and International Monetary Fund,” Press Release No. 50 (Washington), September 30, 1987.

    • Search Google Scholar
    • Export Citation
  • Bank for International Settlements, Annual Report (Basle: BIS, June 1989).

  • Bank for International Settlements, Annual Report (Basle: BIS, 1990).

  • Bank for International Settlements, Annual Report (Basle: BIS, 1992).

  • Boulton, Leyla, Moscow’s Gold Reserves Now at 320 Tonnes,Financial Times (London), April 1, 1993, p. 7.

  • Brody, Alan J., Gold as a Financial Instrument,Mining Journal, Vol. 311, Supplement (October 21, 1988), pp. 410.

  • Chua, Jess H., Gordon Sick, and Richard S. Woodward, Diversifying with Gold Stocks,Financial Analysts Journal, Vol. 46 (July/August 1990), pp. 7679.

    • Search Google Scholar
    • Export Citation
  • Conger, Harry, Gold: Glitter or Glut?Chief Executive Issues, No. 62 (October 1990).

  • CPM Group, Precious Metals (New York: Christian, Podleska and van Musschenbrock Ltd.), Vol. 1, no. 1 (fourth quarter, 1987).

  • CPM Group, Precious Metals: Gold (New York: Christian, Podleska and van Musschenbrock Ltd.), Vol. 4, no. 2 (second quarter, 1990).

  • Craig, Malcolm, Making Money from Gold: An Essential Guide to Successful Gold Investment (Kingsclere, England: Scope Books, 1982).

  • Craig, Malcolm, Deriving Some Added Value,” in Gold: A Bumpy Ride Ahead (Euromoney supplement, July 1990), p. 24.

  • Craig, Malcolm, The Economist,Fool’s Gold,March 17, 1990, p. 79.

  • Craig, Malcolm, The Real Goldfingers,” August 19, 1989, p. 61.

  • Followill, Richard A., and Billy P. Helms, Put-Call-Futures Parity and Arbitrage Opportunity in the Market for Options on Gold Futures Contracts,Journal of Futures Markets, Vol. 10 (August 1990), pp. 33952.

    • Search Google Scholar
    • Export Citation
  • Gehr, Adam, Undated Futures Markets,Journal of Futures Markets, Vol. 8 (February 1988), pp. 8997.

  • Glynn, Lenny, Doing Business with the Central Banks,Global Finance (September 1991), pp. 4251.

  • Gold Fields Mineral Services Ltd. (also known as Consolidated Gold Fields), Gold (London: various years).

  • Gold, Sidnez, New Products and the Possibilities for More,” in Financial Times World Gold Conference, Venice, June 25 and 26, 1990 (London: Financial Times Conference Organization, 1990), pp. 12.112.5.

    • Search Google Scholar
    • Export Citation
  • Golddealers Luxembourg a.s.b.l., Guide to Precious Metals 1988 (London: Mint, 1988).

  • Gooding, Kenneth, Gold Lending Rate at Record Level,Financial Times (London), December 4, 1990, p. 34.

  • Gooding, Kenneth, (1991a), “SA Bank Sends Bullish Signal to Gold Market,Financial Times (London), January 4, p. 22.

  • Gooding, Kenneth, (1991b), “Forward Selling by Producers Putting Cap on Gold Price,Financial Times (London), January 24, p. 32.

  • Gooding, Kenneth, (1991c), “The Gulf War: Allies Consider Peace Plan,Financial Times (London), February 22, p. 26.

  • Gooding, Kenneth, Gold Continues to Lose its Allure,Financial Times (London), December 19, 1992, p. ii.

  • Green, Timothy S., The Above-Ground Stocks of Gold,” in World Gold Markets, 1981/1982 (London: Proceedings of Conference, May 18–19, 1981, sponsored by Consolidated Gold Fields Ltd. and Government Research Corporation, 1981), pp. 2729.

    • Search Google Scholar
    • Export Citation
  • Green, Timothy S., The New World of Gold: The Inside Story of the Mines, the Markets, the Politics, the Investors (London: Weidenfield and Nicolson, rev. ed., 1985).

    • Search Google Scholar
    • Export Citation
  • Green, Timothy S., The Prospect for Gold: The View to the Year 2000 (London: Resendale Press, 1987).

  • Greenspan, Alan, Statement of the Chairman, Board of Governors, Federal Reserve System, Before the Subcommittees on Domestic Monetary Policy and on International Finance, Trade, and Monetary Policy of the Committee on Banking, Finance and Urban Affairs, U.S. House of Representatives, December 18, 1987,” Federal Reserve Bulletin (Washington), February 1988, pp. 103104.

    • Search Google Scholar
    • Export Citation
  • Gulley, David, Gold and Crises: New Myths for Old?Euromoney (January 1991), pp. 6568.

  • Hok, Gan Tjoen, The Singapore Gold Market,” in World Gold Markets, 1981/1982 (London: Proceedings of Conference, May 18–19, 1981, sponsored by Consolidated Gold Fields Ltd. and Government Research Corporation, 1981), p. 95.

    • Search Google Scholar
    • Export Citation
  • Inoue, Junnosuke, New Products and the Possibilities for More,” in Financial Times World Gold Conference, Venice, June 25 and 26, 1990 (London: Financial Times Conference Organization, 1990), pp. 13.113.2.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF), Balance of Payments Manual (Washington: IMF, 1977).

  • International Monetary Fund (IMF), The Economy of the Former U.S.S.R. in 1991, IMF Economic Reviews (Washington: IMF, April 1992); revised edition forthcoming, June 1993.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF), International Financial Statistics (IFS) (Washington: IMF, February 1993).

  • Jacks, Jessica, Gold Options,Mining Journal, Vol. 314 (May 25, 1990), pp. 1214.

  • Kettell, Brian, Gold (London: Graham & Trotman, 1982).

  • Kuhn, Susan, Is Gold Still the Best Bet for Troubled Times?Fortune, Vol. 122, No. 7 (September 24, 1990).

  • Martin, Michael G., Gold Market Developments, 1975–77,Finance and Development (March 1978), pp. 3135.

  • Martin, Michael G., The Changing Gold Market, 1978–80,Finance and Development (December 1980), pp. 4043.

  • McGanty, Dan, New Products and the Possibilities for More,” in Financial Times World Gold Conference, Venice, June 25 and 26, 1990 (London: Financial Times Conference Organization, 1990), pp. 14.114.3.

    • Search Google Scholar
    • Export Citation
  • McNamee, Mike, Baker’s Plan: No Glitter,Business Week (October 19, 1987), p. 56.

  • Melvin, Michael, and Jahangir Sultan, South African Political Unrest, Oil Prices, and the Time Varying Risk Premium in the Gold Futures Market,Journal of Futures Markets, Vol. 10 (April 1990), pp. 10311.

    • Search Google Scholar
    • Export Citation
  • Milling-Stanley, George, Futures Research: Gold Update,” Shearson Lehman Brothers (New York), August 1991.

  • Milling-Stanley, George, Mining Journal,Soviet Secrets,” Vol. 311 (April 22, 1988), p. 334.

  • Moore, Geoffrey, Gold Prices and a Leading Index of Inflation,Challenge, Vol. 33 (July/August 1990), pp. 5256.

  • Moore, Phillip, Gold in Need of Investor Appeal,” in Gold: A Bumpy Ride Ahead (Euromoney supplement, July 1990).

  • Nathans, Leah, How Drexel Burnham Turned Gold into Junk,Business Week, No. 3162 (June 4, 1990).

  • O’Callaghan, Gary, The Structure and Operation of the World Gold Market,” IMF Working Paper 91/122 (Washington: IMF, December 1991).

    • Search Google Scholar
    • Export Citation
  • Ogden, Joseph P., and Alan L. Tucker, Arbitraging American Gold Spot and Futures Options,Financial Review, Vol. 25 (November 1990).

    • Search Google Scholar
    • Export Citation
  • Poitras, Geoffrey, The Distribution of Gold Futures Spreads,Journal of Futures Markets, Vol. 10, no. 6 (December 1990), pp. 64359.

    • Search Google Scholar
    • Export Citation
  • Potts, David, ed., Gold 1980 (London: Consolidated Gold Fields Ltd., June 1980).

  • Potts, David, Reuters, Russia Exported 98 Tonnes of Gold in 1992,” Mowcow, May 28, 1993.

  • Roethenmund, Otto E., Spot Trading Centres,” in Gold, ed. by Paul Sarnoff (London: Euromoney Publications, 1987), pp. 7380.

  • Salomon Brothers, Precious Metals (New York, September 25, 1991).

  • Samuel Montagu & Co., Annual Bullion Review (London: Samuel Montagu & Co., various years).

  • Sarnoff, Paul, Trading in Gold (Cambridge: Woodhead-Faulkner, 1980).

  • Sarnoff, Paul, ed., Gold (London: Euromoney Publications, 1987).

  • Schriber, R., The Zurich Market,” in World Gold Markets 1981/1982 (London: Proceedings of Conference, May 18–19, 1981, sponsored by Consolidated Gold Fields Ltd. and Government Research Corporation, 1981), pp. 8083.

    • Search Google Scholar
    • Export Citation
  • Shearson Lehman Hutton, Metals Research Unit, Annual Review of the World Gold Industry 1990 (London: Shearson Lehman Hutton, May 1990).

  • Smith, Andrew, Behind the Golden Curtain,Precious Metals, Union Bank of Switzerland (1992), pp. 1213.

  • Smith, Keith, The London Market,” in World Gold Markets 1981/1982 (London: Proceedings of Conference, May 18–19, 1981, sponsored by Consolidated Gold Fields Ltd. and Government Research Corporation, 1981), pp. 7778.

    • Search Google Scholar
    • Export Citation
  • Smith, Keith, Soviet Figures: A Weil-Kept Secret,” in Gold: Refining the Market (Euromoney supplement, January 1987), p. 12.

  • Swiss Bank Corporation, Gold: Myth and Reality (Zurich: Swiss Bank Corporation, 1985).

  • Tan, Ronald H.L., The Gold Market (Singapore: Singapore University Press, 1981).

  • Temple, Peter, Has the Gold Rush Stimulated Renewal?Accountancy, Vol. 105 (May 1990), p. 80.

  • U.K. Central Statistical Office, United Kingdom Balance of Payments (London: Her Majesty’s Stationery Office, 1989).

  • U.K. Department of Trade and Industry, Overseas Trade Statistics of the United Kingdom 1986 (London: Her Majesty’s Stationery Office, 1987).

    • Search Google Scholar
    • Export Citation
  • Williams, David, The Gold Markets, 1968–72,Finance and Development (Washington), December 1972, pp. 916.

  • World Gold Council, World Gold Review (Spring 1992).

  • Yeung, Kenneth B.K., The Hong Kong Gold Market,” in World Gold Markets 1981/1982 (London: Proceedings of Conference, May 18–19, 1981, sponsored by Consolidated Gold Fields Ltd. and Government Research Corporation, 1981), pp. 9294.

    • Search Google Scholar
    • Export Citation