The London Gold Pool: Defending $35 Gold, 1961–1968

A Proclamation Signed Before Dawn

Shortly after midnight on 15 March 1968, a handful of Privy Councillors gathered at Buckingham Palace to have Queen Elizabeth II sign a proclamation declaring a bank holiday. There was no banking emergency in Britain that anyone had announced. The proclamation was a legal device, requested by Washington a few hours earlier and drafted in haste, because a bank holiday was the only instrument under English law that could shut the London gold market at a few hours' notice. On the previous day, Thursday 14 March, buyers had taken an estimated 350 to 400 tonnes of gold out of the market in a single session — contemporary estimates of the loss to the central banks standing behind it range from roughly $350 million to $400 million at the official price. The vault of the Bank of England, which had been supplying that gold as agent for eight central banks, could not survive another day like it.

London's gold market stayed closed for two weeks. When it reopened on 1 April, the arrangement that had held the world price of gold at $35 an ounce for six and a half years no longer existed. The London Gold Pool had been the most concrete piece of machinery ever built to defend the Bretton Woods monetary order, and its collapse marked the point after which the dollar's gold anchor was a diplomatic formality rather than a market price. What remained of the formality lasted another forty-one months, until Richard Nixon closed the gold window in August 1971.

Triffin's Arithmetic

Bretton Woods rested on a single promise: the United States would convert dollars into gold at $35 an ounce for foreign official holders, and every other member currency would hold a fixed parity against the dollar. Gold was the system's ultimate settlement asset, and the dollar was its working one. That worked while American gold holdings dwarfed foreign dollar claims, which they did through the late 1940s, when the US Treasury held about $24.6 billion of gold, roughly two-thirds of the world's official monetary stock.

Robert Triffin, a Belgian economist at Yale, set out the problem in testimony to the Joint Economic Committee of Congress in October 1959 and then at length in Gold and the Dollar Crisis the following year. World trade was expanding, and it needed reserves to settle against. Those reserves came overwhelmingly in dollars, which meant they came from American balance-of-payments deficits. Each deficit supplied liquidity the system needed and simultaneously added to the pile of foreign claims on a gold stock that was not growing. Triffin's conclusion was that the gold exchange standard could relieve a shortage of world reserves only to the extent that the reserve-currency country was willing to let its own net reserve position deteriorate — and that at some calculable point, the deterioration would destroy the confidence on which the whole structure ran (Triffin, 1960).

By 1960 the crossover had arrived. Foreign dollar liabilities of the United States passed the market value of the American gold stock for the first time, at somewhere near $19 billion against $17.8 billion of gold. Arithmetic that had been an economist's warning became a trading position: anyone who believed the dollar would eventually be devalued against gold could express that view by buying gold in London, where the metal traded freely and where the price had been quoted since the market reopened in March 1954.

The Spike of October 1960

Speculation found its first real opening during the American presidential campaign. On 20 October 1960, with the election two weeks away and rumours circulating that a Democratic administration would spend heavily and devalue, the London price touched $40.60 an ounce — a premium of roughly 16 per cent over the official parity and a level the market had not seen since before the war. The Bank of England intervened with gold supplied by the US Treasury, and the price came back down within days.

John F. Kennedy responded with a written pledge on 31 October 1960, promising that if elected he would "not devalue the dollar from its present rate" and would instead "defend its present value and its soundness." As president he repeated the commitment in his balance-of-payments message of 6 February 1961. Those words settled the immediate panic without addressing the arithmetic underneath it, and officials on both sides of the Atlantic understood that a second spike would arrive eventually. What the episode demonstrated was that the London market — small, free, and open to anyone with a bank account — could set a price that made the official $35 look like a fiction, and that a fictional official price would invite every foreign central bank to convert its dollars while conversion was still possible.

Eight Central Banks and a Basel Ledger

The remedy was assembled through the Bank for International Settlements in Basel, where central bank governors met monthly, and it was operating by November 1961. Eight central banks agreed to pool gold and sell it into the London market whenever the price threatened to rise much above $35.20 an ounce, a ceiling set by the cost of shipping and insuring gold from New York. The United States took half the burden. The rest was divided among seven European members according to quotas that shaped every settlement for the next six years.

Operationally the Pool was elegant. The Bank of England, which already ran the London fixing as the market's central intermediary, acted as agent for all eight, buying and selling on the group's behalf without disclosing which days it was in the market or in what size. At the end of each month the BIS netted the account and each member settled its share in gold according to its quota. From 1962 the arrangement worked in both directions: when the price sagged toward the $34.9125 level at which the US Treasury bought, the agent purchased for the Pool and the metal was distributed by the same percentages. Gabriele Toniolo's history of the BIS describes the Pool as the high-water mark of an era in which central bankers treated cooperation as a technical craft conducted out of public view, with no treaty, no published rules, and no parliamentary oversight anywhere (Toniolo, 2005).

Years When It Worked

For its first four years the Pool mostly made money. South African mines, which supplied close to three-quarters of non-communist output at around 1,000 tonnes a year, produced more metal than private buyers wanted, and the Soviet Union sold gold heavily in 1963 and 1964 — several hundred million dollars a year — to pay for North American grain after disastrous harvests. Net purchases in those years let members add to their own reserves through the same Basel ledger that was meant to drain them.

Even the Cuban missile crisis of October 1962 produced only a manageable scramble. Charles Coombs, the New York Fed official who ran American foreign exchange operations and later wrote the standard participant account of the period, treated the arrangement in those years as a technical success: a market price pinned to within a few cents of parity, achieved at modest cost, with the speculative community persuaded that betting against the Pool was an expensive way to be right eventually (Coombs, 1976). Barry Eichengreen's account of the dollar's rise places the same judgement in a longer frame, noting that the arrangement bought time without buying a solution, and that time was exactly what the American balance of payments needed and did not get (Eichengreen, 2011).

The deterioration came from Washington. Escalation in Vietnam and the domestic spending of the Great Society pushed American inflation from 1.6 per cent in 1965 to above 4 per cent by 1968, and the payments deficit widened along with it. Rising prices at home meant that $35 an ounce was becoming an ever-deeper subsidy to anyone holding gold, and central bankers in Europe could read a consumer price index as well as anyone.

De Gaulle's Exit

France had never been a comfortable member. Jacques Rueff, economic adviser to Charles de Gaulle and the period's most persistent critic of the gold exchange standard, argued that the system let the United States settle its deficits in its own paper — a privilege no other country enjoyed. De Gaulle delivered the argument publicly at a press conference on 4 February 1965, saying that international exchange needed "an indisputable monetary base" bearing "the mark of no particular country," and that in truth "one does not see how it could really be any standard other than gold."

Words were followed by conversions. France presented dollars to the US Treasury for gold through 1965 and 1966, shipping the metal home, and then in June 1967 withdrew from the Pool entirely. The withdrawal was not announced with a communiqué; it emerged through the Basel accounts and was reported within weeks. The remaining seven quietly took up the French share, raising the American burden to roughly 59 per cent, and continued.

After Sterling Broke

Harold Wilson's government devalued the pound on 18 November 1967, from $2.80 to $2.40, after three years of defending a parity that Britain's reserves could not support. Speculators drew the obvious inference. If the second reserve currency could break, the first could too, and the cheapest way to hold that view was to buy gold in London at a price eight central banks had promised to hold down.

Buying began the following week and did not stop. Through the final six weeks of 1967 and the first ten of 1968 the Pool sold something on the order of $3 billion of gold — more than a fifth of the entire American stock — into a market that treated every intervention as confirmation that the defenders were running out. Demand came from Swiss banks buying for private clients, from Middle Eastern buyers, and from European commercial banks hedging their own dollar positions. By early March the members were meeting almost daily. On 10 March the governors issued a statement from Basel reaffirming their determination to maintain the price; the market read it as weakness, and the following week produced the run that closed the market.

Two Prices for One Metal

Governors of the seven surviving members met at the Federal Reserve Board building in Washington over 16 and 17 March 1968 and produced a communiqué that did not defend the price so much as abandon the market. They agreed that henceforth official gold would circulate only among monetary authorities, at $35 an ounce, and that they "no longer feel it necessary to buy gold from the market" and would "no longer supply gold to the London gold market or any other gold market." The existing stock of monetary gold, the statement added, was sufficient in view of the prospective establishment of the facility for Special Drawing Rights — the synthetic reserve asset agreed in outline at the IMF's Rio de Janeiro meeting six months earlier and first allocated in 1970.

Two prices now existed for the same metal: an official one used for settlements between governments, and a private one that the market set for itself. Congress completed the retreat the next day by repealing the requirement that the Federal Reserve hold gold equal to 25 per cent of its note issue, an act that freed about $10 billion of metal for the defence of external convertibility and removed the last domestic legal tie between American currency and gold.

The two-tier system was an admission dressed as a technical adjustment, and its logic was unstable from the start. Bordo and Eichengreen, surveying the Bretton Woods period, treat March 1968 as the moment the system's credibility passed a point of no return, with everything after it a question of timing rather than direction (Bordo and Eichengreen, 1993). Foreign central banks understood that the official price now had no market behind it, and conversions of dollars into American gold continued through 1970 and accelerated sharply in the summer of 1971. Britain's request for cover on $3 billion of reserves in the second week of August 1971 was one of the immediate triggers for Nixon's announcement on the fifteenth.

Gold's private price behaved as the arithmetic suggested it would. It settled between $38 and $42 through 1968 and 1969, slipped back toward $35 in early 1970 as central banks stayed out, then rose without interruption once convertibility ended — past $100 in 1973, past $800 in January 1980 during the Hunt brothers' attempt to corner silver and the inflation that framed it. Measured against that path, the eight central banks had spent roughly $3 billion of metal in four months to hold a price the market repriced by a factor of twenty within twelve years.

What the Pool proved was narrower and more durable than its failure suggests. Eight institutions with no treaty, no staff, and no public mandate ran a coordinated intervention operation for six years through a monthly accounting entry in Basel, and it worked until the underlying policy it was defending stopped being defensible. Central bankers took the technique forward into the currency interventions of the following decades, including the coordinated dollar sales agreed at the Plaza Hotel in September 1985. The lesson they carried was not that intervention is futile but that it buys time, and that what a government does with the time determines whether the gold in the vault was spent or merely lent to the speculators who came to collect it.