The ESF and Gold: Past as Prologue?

April 10, 2000

Treasury Bulletin (www.fms.treas.gov/bulletin/b10esf.pdf) contains the Exchange Stabilization Fund's balance sheet and profit and loss statement for the quarter and fiscal year ending September 30, 1999. Prior quarterly statements, in mildly varying formats but similarly limited levels of detail, are available in earlier bulletins. These are summarized for five-year intervals from 1935 to 1995 in tables 1 and 2 contained in A. J. Schwartz, "From Obscurity to Notoriety: A Biography of the Exchange Stabilization Fund," Journal of Money, Credit, and Banking (Vol. 29, No. 2, May 1997), pp. 135-153. Although the ESF's current balance sheet shows total assets of about $41 billion, this number must be reduced by at least the $8.2 million of liabilities for SDR certificates to arrive at total resources available to the ESF. For more discussion of the intricacies of the ESF's balance sheet, including "monetizing" SDRs and "warehousing" foreign currencies with the Fed, see W. P. Osterberg et al., "The Exchange Stabilization Fund: How It Works," Federal Reserve Bank of Cleveland, Dec. 1999 (www.clev.frb.org/research/com99/1201.htm).

Created by the Gold Reserve Act of 1934, the ESF began operations in April 1934, financed by $2 billion of the profits realized from the New Deal's gold confiscation and subsequent devaluation. However, only $200 million was made available to the ESF as working capital, of which $20 million was soon invested in gold, held mostly at the U.S. Assay Office, and $30 million in silver. The other $1.8 billion remained in the Treasury's gold account, and ultimately was used to fund part of the original U.S. subscription to the IMF. All SDRs allocated to or otherwise acquired by the U.S. are turned over to the ESF.

A review of the ESF's past quarterly statements, together with other quarterly international financial statistics contained in the same bulletins, reveals some interesting facts: (1) until April 1989, the Treasury's gold stock and the total U.S. gold stock, which also included gold held by the ESF, were reported separately in table IFS-1; (2) from 1974 to April 1989 these two numbers were the same, implying that the ESF held no gold during those years; (3) in 1973 and prior years, the ESF reported gold holdings of varying amounts. According to Schwartz (p. 148), this gold was held in a special ESF account at the New York Fed, and in 1974 was consolidated with the Treasury's other gold in anticipation of its 1975 and 1978-79 gold auctions. Schwartz also reports (p. 138, fn. 6) that at the 1977 IMF gold auctions, the Treasury purchased gold which it sold to the ESF, and that this gold appeared on its balance sheet for the first three quarters of 1977 but not thereafter. Why the ESF purchased this gold and what it did with it remains unclear.

Of more relevance for present purposes is the ESF's income statement, and particularly the first line: "Profit (+) or loss (-) on: Foreign Exchange." In the statement for December 31, 1977, this account read: "Profits on transactions in: Gold and exchange (including profits from handling charges on gold)." Similar language appeared in prior statements, so that until the end of 1977 profits or losses on gold and foreign exchange were lumped together in a single line item. In the statement for September 30, 1978, the language was similar, but there were separate line items for gold and foreign exchange. However, only the line for foreign exchange showed activity during the year. In the statement for December 31, 1978, the separate line item for gold was eliminated, and the current format of a single line item for foreign exchange with no specific reference to gold was adopted. Accordingly, if the ESF has resumed trading in gold or gold derivatives, historical practice indicates that the profits or losses on these activities should be reflected in this line item absent creation of new line item for gold.

The following table shows the ESF's profits or losses (-) on foreign exchange by quarter and fiscal year from fiscal 1981, the first year of the Reagan administration, through fiscal 1999, the most recent data. All amounts are in US$ millions.

   Fiscal          Oct./    Jan./    Apr./    Jul./    Total
    Year            Dec.     Mar.     Jun.     Sep.     FY

    1999            1699     -817     -500*    1257     1637
    1998            -754     -333     -135      576     -646
    1997 (Korea/    -383    -1093      402     -538    -1613
            Asia)
    1996            -449     -547     -419     -214    -1629
    1995 (Mexico)    -38     2623      276    -2054      808
    1994           -1116     1388      883      102     1257
    1993           -1700      965      412      437      114

    1992            1264    -1267     1495     1191     2683
    1991            1020    -1357     -421      739      -19 
    1990             327     -722      944     1752     2301 
    1989             545     -555     -501      471      -39

    1988             994     -236     -414     -133      212 
    1987 (Louvre)     96      589      -51      -15      618
    1986             456      488      478      504     1926
    1985 (Plaza)     -57       50       43      441      477 

    1984             -26      107     -165     -162     -246 
    1983             524      -64       41      112      613
    1982             439     -475      -95       99      -32
    1981            -217     -390     -806      241    -1172

    * Income of $36 million for "Commissions" reported as separate
      line item this quarter; no similar entry before or since.

Because the ESF is self-funding, its earnings on gold and foreign exchange trading (as well as its interest income on investments held) are accumulated in the fund, and it has incentive to operate profitably. As these reports make clear, the ESF does not confine its foreign exchange trading to interventions for purposes of currency stabilization. Indeed, the Treasury has reported to Congress that the ESF did not conduct any "interventions" in 1998 or 1999, yet its activities nevertheless generated foreign exchange profits or losses, as they did in prior years when interventions for the purpose of currency stabilization were few or none. However, since the ESF is sometimes used as a vehicle for providing aid to other countries, e.g., Mexico in 1995-1996, Korea possibly in 1997, some losses in certain years may be attributable to these activities, which really are undertaken to stabilize not the dollar but the currencies of other nations.

For 14 years, from 1982 (the second year of the Reagan administration) through 1995 (the third year of the Clinton administration), the ESF's foreign exchange trading was generally quite profitable, suffering small losses in only three years (1982, 1989 and 1991) and a moderate loss in just one (1984). This creditable record began to fall apart in 1996, with by far the largest loss on foreign exchange trading in the ESF's history, followed by another similarly large loss in 1997, and a significant loss in 1998. Good results in the first quarter of 1999 were halved by losses in the second quarter, and further reduced by a smaller but sizable loss in the third, which is also the quarter in which the British gold sales were announced.

Another noteworthy feature of this quarter is the appearance on the ESF's income account of a separate line item for "commissions." No such entry has appeared before or since. The closest historical analogue are the "handling charges on gold" included in the profits line prior to 1978.

While an examination of the ESF's skeletal financial reports cannot possibly prove that it has engaged in efforts to "stabilize" the gold price in recent years, there is nothing in these reports to suggest that it has not. What is more, there is much to arouse suspicion.

As I have suggested in earlier commentaries, official efforts to cap the gold price probably began in late 1995 or early 1996 as the gold price challenged $400/oz. with the deepening of Japan's economic crisis. These dates also coincide with the ESF's transition from profits to losses. Generally speaking, official bodies like the ESF ought to make profits on foreign exchange trading when profits rather than currency stabilization are their objective since they are likely to have relevant information and intelligence unavailable to others. On the other hand, when they intervene to stabilize a currency in opposition to fundamental market forces, experience suggests that large losses are likely to occur. Accordingly, particularly during 1998 and 1999 when the Clinton administration denies making any interventions and when there were no other obvious activities of the ESF that might explain losses on foreign exchange trading, the losses themselves -- especially large ones -- raise questions.

They become even more suspicious when plotted against gold prices over the past two years, where on a quarterly basis and allowing for shorts lags in realizing profits or losses, firm or rising gold prices tend to correspond with ESF losses and falling prices with ESF profits. This relationship does not hold true for the last calendar quarter of 1997 (first quarter of fiscal 1998 for the ESF), but then declining gold prices reflected concern about lower demand and even dishoarding as a result of the Asian financial crisis, which may have impacted the ESF's results adversely as well. Thereafter, rising gold prices in the first four months of 1998 correspond with ESF losses in the second and third fiscal quarters, and falling gold prices to September with a small recovery that month with ESF profits. From October 1998 through the end of the year gold prices remained weak and in declining mode, and the ESF had one of its most profitable quarters.

For the first two months of 1999, gold remained firm at around 290, spiking upward toward $300 in early March. This rally was contained despite apparently good fundamentals for gold, and the price receded to around $280 by the beginning of April. Then another rally began, pushing the price back above $290 at the end of the month, largely powered by increasing doubts that the proposed IMF gold sales would be approved. In the second fiscal quarter (first calendar quarter) of 1999, the ESF lost almost half of its profits from the prior quarter, and its losses continued into the next quarter on a scale that if continued would have pushed it into a loss position for an unheard of fourth straight year.

In early May, just as gold was threatening a sharp rally that many expected would carry it over $300 and perhaps to much higher levels, the British announced their gold sales. For many knowledgeable about gold, this otherwise inexplicable action was the smoking gun, proof that some official scheme was afoot to cap the gold price. In its wake, the gold price declined within two months to under $260, and stayed at these low levels until announcement of the Washington Agreement at the end of September. At the same time, the ESF had another very profitable quarter, closing out fiscal 1999 with a large profit for both the last quarter and the year.

All these events may be coincidence, but they are also consistent with an ESF program of trying to cap the gold price through a program of selling call options or backstopping calls sold by others. Gold loans and short selling by bullion banks are largely responsible for the weak gold prices of recent years. Both activities require access to a deep, liquid and financially credible market where call options on gold can be purchased. See, e.g., The New Dimension: Running for Cover. Absent such a market or its functional equivalent, these activities simply become too risky, especially for prudent, sophisticated players, because they have no means to hedge their risk. At the same time, as the net short gold position grows, writing or selling calls becomes more risky, and premiums tend to rise, eventually choking off both gold loans and short selling. At that point, gold prices should rally, relieving some of the pressure.

However, given its rather large resources relative to the gold market, the ESF could have enabled gold loans and bullion bank short selling to continue past their normal limits by selling or backstopping calls at reasonable premiums when private parties declined to do so in sufficient quantities or at all. A program of this sort would generally, although perhaps with a slight lag for their actual realization, tend to show losses on rising gold prices and profits on falling ones. Nor would it necessarily require access to official U.S. gold reserves, either for sale or leasing. At first, relatively deep pockets and a high tolerance for risk probably would be sufficient.

But ultimately, as the net short position grew and perceptions of risk increased, central banks and others who loan gold would begin to step back. No central bank wants to be caught with a defaulted gold loan. Then the program to cap the gold price would break down without access of some sort to physical gold, such as a friendly central bank willing to step into the breach -- either with bullion for sale or with a willingness to lease notwithstanding the risk. The British gold sales and Kuwaiti gold loan -- both otherwise without apparent rational basis -- fit this pattern.

Given its long-standing culture of secrecy, any ESF scheme to cap the gold price would almost certainly be carried out without notice to Congress, and quite possibly without notice to the Fed either. As to mechanics, a hidden relationship with one or a very few bullion banks would be sufficient. This relationship would give the favored bullion banks an enormous edge in their own gold trading operations, and one that no Chinese wall could likely neutralize.

From the outset, the constitutionality of the ESF has been open to doubt. Its self-funding operations largely immune from effective congressional oversight seem to contravene both the separation of powers and the exclusive control over appropriations vested in Congress. But any scheme by the ESF today to control the gold price would face further legal and constitutional hurdles.

At the same time that it established the ESF, Congress put the gold value of the dollar at $35/oz. and generally prohibited gold ownership by American citizens. Accordingly, Congress set the gold price for the ESF to target, not the Secretary of the Treasury. What is more, by outlawing most private gold ownership, Congress effectively foreclosed trading of gold. Thus there was little likelihood that the ESF would come into conflict with private investors trading gold or gold futures in New York or anywhere else in the country.

Today, of course, the situation is wholly different. Although Congress has left standing an anachronistic official gold price of $42.22/oz., there is no reasonable argument that this figure remains a legitimate target for ESF stabilization efforts. Congress has also repealed the ban on gold ownership. Substantial trading of gold, including bullion and coins, and gold derivatives, including futures and options, now takes place daily in private transactions, over-the-counter financial markets and public commodities exchanges.

Any act by the executive branch -- through the ESF or otherwise -- to set the dollar gold price today, under these circumstances, would be patently illegal and unconstitutional. The public commodities exchanges are regulated under the authority of Congress precisely to assure that they function honestly and fairly for all participants. Congress has effectively declared by its actions that gold contracts on these exchanges should trade in a free market, not one subject to manipulation by the ESF, the Fed or anyone else, including the bullion banks.

The monetary provisions of the Constitution grant to Congress sole and exclusive power to determine the gold value of the dollar. The Supreme Court, to its everlasting shame, has refused to decide whether Congress may constitutionally sever any meaningful link between the dollar and gold or silver. But if there is to be a link, the Constitution vests in Congress exclusive power to define it.

Were I counsel to a bullion bank, my advice would be to avoid like the plague involvement, or even suspicion of involvement, in any ESF scheme to affect gold prices. Activities of this sort could potentially violate a wide array of federal and state statutes, many of which also confer rights of private action on injured parties. Indeed, the blatant unconstitutionality of any such scheme could well deprive participating government officials of a range of defenses ordinarily available when they perform authorized duties in good faith. But many defenses available to the government or its officials would not be available to the bullion banks. In short, if the scheme blew up, their financial exposure could be enormous.

Seventy-five percent of all gold in circulation has been extracted since 1910