‘This book is the record of a Government all but a year of whose life was dominated by an inherited balance of payments problem which was nearing a crisis at the moment we took office; we lived and governed during a period when that problem made a frenetic speculative attack on Britain both easy and profitable.’Footnote ¹ This is how Harold Wilson (UK Prime Minister 1964–70) opened his autobiography. Simultaneously with the election of the Labour Party to power in 1964 (which itself added to pressure on sterling), a string of recurring currency crises began that would not be resolved until well after the 1967 devaluation.

After the 1964 general election, the fate of sterling and gold were increasingly intertwined. At this stage, the link between the two reserve currencies (sterling and the dollar) became apparent. US policymakers started to turn their attention to protecting sterling. The 1964 sterling crisis highlighted the role sterling still played in the stability of the international monetary system. Aware of the systemic importance of the British currency, the United States dedicated significant resources towards supporting the currency until 1967. This support would eventually prove insufficient.

From 1958 to the 1964 election, the influence of sterling was not visible on the gold market. Sterling was still relatively stable, bar a minor sterling crisis in 1961. Other international events related to the Cold War (the Cuban missile crisis and the Berlin Wall, among others) took centre stage. This put pressure on both the dollar and gold. The 1964 sterling crisis would reveal that sterling still played a role in global currency markets. Starting in autumn 1964 and continuing through to 1971, pressure on gold did not abate. The pressure was reinforced by the problem of rising inflation in the United States from 1965 onwards.Footnote ²

The 1964 Crisis

The Labour Party won the general election in October 1964. The new government was faced with fears over the devaluation of sterling. The previous administration had already been struggling with balance of payments deficits. The victory of Labour, which was ‘not known for its friendliness towards the markets’, made things worse.Footnote ³ Wilson himself was aware of market animosity, as he later wrote: ‘[W]e had always underestimated the power of the speculators against a Government of whose politics, policies and even personalities they did not approve.’Footnote ⁴ The literature is unanimous in stating that Labour did not want to be the party of devaluation.Footnote ⁵ The party was held responsible for the 1949 devaluation. Wilson did not want the electorate to ‘permanently associate economic incompetence with his beloved Labour Party’.Footnote ⁶ Even if, as the then Chancellor of the Exchequer, James Callaghan, wrote, there was a reason for markets to expect it: ‘We had been out of office for thirteen years, and there would be speculation that our first step might be to devalue sterling.’Footnote ⁷ However, Callaghan, like Wilson, was against devaluation because the ‘Conservatives would have crucified’ the Labour Party.Footnote ⁸

The United States supported the decision not to devalue sterling. This meant that the United Kingdom had a strong hand in negotiating financial aid. As Schenk puts it: ‘[T]he key role of sterling in the international monetary system did allow Wilson and his Chancellors of the Exchequer to garner repeated large doses of international support for the sterling exchange rate both before and after the devaluation of 1967.’Footnote ⁹ This assistance was vital. When Callaghan was appointed Shadow Chancellor, he visited the New York Fed. During his visit, he heard Hayes from the Fed speak ‘very frankly about the strains on the dollar’.Footnote ¹⁰ Hayes also ‘repeated more than once his belief that the best prospect for effective action in monetary matters depended upon Britain and America working together’. This was the Labour government’s guarantee of US support.

There is a debate in the literature on whether 1964 witnessed one or two sterling crises. The interest of dividing the sterling crisis into sub-crises is limited. All these events were closely interlinked. Here I review the debate in the literature in the light of new evidence. What becomes clear from the data is that the Labour government not only inherited a balance of payments crisis from the previous government but was also naive in its approach to the currency market. It had been out of power for a long time. The policy goal was to avoid a devaluation. The government struggled to improve the balance of payments, despite the National Plan which was aiming to curb spending abroad and improve productivity. In this context, US support is the only reason for sterling to have avoided devaluation from 1964 to 1967.

On one side of the debate, Newton suggests that there were two sterling crises in the autumn of 1964 and that Labour handled the first one ‘efficiently’.Footnote ¹¹ The second crisis, Newton argues, was ‘provoked by speculation stemming from market expectations’.Footnote ¹² According to Newton, there was one crisis ‘which coincided with the election result and another one which started three weeks later’. This would place the first crisis around 15 October and the second around 5 November. I analyse these dates against new data below.

On the other side of the debate, Michael Oliver wrote a reply to Newton’s article.Footnote ¹³ Oliver posits two critiques. First, there was only one sterling crisis in the autumn of 1964. Second, the new Labour government did not display a ‘textbook reaction’ to the crisis but failed to react appropriately to stem speculation.Footnote ¹⁴ Using forward rates from The Times, Oliver argues that ‘The behaviour of the 90-day forward rate suggests that sterling was not credible from September 1964’.Footnote ¹⁵ I agree with this timing. In fact, in September 1964 sterling had hit a three-year low and this was reported in the press at the time.

However, Oliver then argues that ‘the daily dealers’ reports from the Bank can be used to reconstruct changes in the reserves and allow a more accurate picture of reserve movements’.Footnote ¹⁶ This is problematic, as Bank of England reserves were not affected by intervention recorded in the dealers’ reports only, but also by many other factors, as argued in previous chapters. But intervention figures from the dealers’ reports, if not fully informative on reserve amounts, are an excellent tool to understand the Bank’s operations. Further, I set out the actual reserve figures from the EEA. These were not available to Oliver.

When looking at the sterling exchange rate, it appears that the currency was beginning to decline in 1962 (see Figure 9.1).Footnote ¹⁷ This coincides with a period when the government accelerated its efforts to stimulate the economy by its fiscal and monetary policy.Footnote ¹⁸ 1964 marked an acceleration of this decline, and sterling hit its lowest point in six years (1960–66) on 26 November at $2.7806. This rate was dangerously close to the Bretton Woods official lower band ($2.78).

Figure 9.1.

Spot exchange rate, 1960–66

Source: Accominotti et al., ‘Currency Regimes and the Carry Trade’, 2017.

The underlying issue was the balance of payments. It turned negative from the last quarter of 1963 and continued to worsen (see Figure 9.2). Yeager argues that at the beginning of 1964, ‘the general picture was one of a booming home economy and weakening balance of payments’.Footnote ¹⁹ In January 1964, Yeager continues, the publication of an ‘all-time record monthly deficit in merchandise trade’ started to increase pressure on the pound. This pushed the government to increase the Bank Rate from 4 to 5 per cent. Coombs attributed the balance of payments problems to the ‘overstimulative budget introduced by Chancellor Maudling’.Footnote ²⁰

Figure 9.2.

UK current account balance of payments (£ million)

Source: Office for National Statistics.

Articles in the Economist in February 1964 give a better understanding of the prevailing mood. The weekly paper expected a current account deficit of less than £100 million for the whole of the second half of 1964. The actual balance of payments deficit was much higher than that only for the last quarter of 1964, when the sterling crisis started.

Well before the 1964 election, the current and capital accounts were already deteriorating. Investors had reason to expect a further decline. Figure 9.3 highlights sterling’s downward trend starting in May 1964. The prospect of a Labour victory five months ahead was most probably not the only reason for the fall. As early as late May 1964, however, the Economist seemed to be arguing that the prospect of a Labour win could have weighed on investors’ expectations:

Figure 9.3.

Sterling spot and one- and three-month forward rate, 1964

Source: Accominotti et al., ‘Currency Regimes and the Carry Trade’, 2017.

Even if there was little coverage of the British election in the US press, some reports in the financial press can be found. In July 1964, the Wall Street Journal reported that the ‘worst’ could happen in an article entitled ‘Capitalism under Fire’: ‘nationalization can be expected to become an increasingly important issue. And one with highly significant overtones for the American businessman and investor’, and further that ‘unions have settled on an approach that strongly attacks free enterprise’.Footnote ²² There are good reasons to believe that the worsening of sterling starting at the end of May might have been due in part to fears of a Labour victory in the context of weak fundamentals.

Sterling weakened both because of the disappointing balance of payments figures presented in Figure 9.2 and the prospect of a Labour win. Press reports from August to November 1964 give a better understanding of what information investors could receive. On 19 August, forecasts from the National Institute of Economic and Social Research announced deficit figures. The current account deficit ‘could well be around £500 million’, a disappointing figure.Footnote ²³ The same day, the New York Times reported that sterling had hit a three-year low.Footnote ²⁴ Sterling would again hit a new three-year low a few days later, as reported in the Wall Street Journal.Footnote ²⁵ At this point, sterling was under stress, with negative press reports published on most days. Most simply mention pressure on sterling but rarely link this with the election campaign.

On 18 September, more negative trade figures were published, though these did not trigger a noticeable market reaction.Footnote ²⁶ The weak state of the economy started to put pressure on Labour. The party would have to deal with the consequences if elected (which was the most likely outcome). Wilson had to ‘present the country with a more grim financial picture’ to avoid taking full blame for a possible currency crisis if elected.Footnote ²⁷

When Labour did win on 15 October, it did not have much of an impact on the foreign exchange market. The market had already factored in the Labour win. The Bank of England dealers reported dollar sales of $10.24 million on the Friday following the election.Footnote ²⁸ This is around 2.5 times the average daily intervention of previous years. A few days later, on 23 October, dealers at the Bank had to intervene for $31.5 million, eight times the average intervention. Pressure had increased. Overall, however, there seems not to have been a major crisis. The election of Labour was indeed factored into the market price of sterling.

The period between the election and early November (the point when Newton sees a second sterling crisis) was a quieter period. This lends some support to his thesis of two distinct sterling crises. The exchange stayed steady, albeit close to the lower band. However, when looking at reserve figures from the EEA (see Figure 9.4), British reserves were continuously worsening. This reveals constant pressure on the Bank of England. The Bank lost reserves from June 1964 onward. Sterling was effectively in crisis from late August 1964 to the 1967 devaluation, and even beyond that. Wilson echoed this when he later wrote: ‘It is difficult to describe what it meant to live against a background of this persistent speculation, speculation in the main made possible only by the balance of payments deficit. Indeed it virtually disappeared as a threat once we moved into strong surplus some five years later.’Footnote ²⁹ Sterling was consistently close to the $2.78 lower band, with the government lurching from one rescue package to another.

Figure 9.4.

EEA gold and dollar reserves

Note: Scale starts at $1,200 million.

Source: EEA ledgers.

What happened in Newton’s ‘second crisis’? Wilson mentions a ‘run on sterling which began, following the Chancellor’s Ways and Means statement on 11th November’.Footnote ³⁰ The run, according to Wilson, was ‘easy to explain’: international companies in London feared devaluation because of the balance of payments deficit and therefore decided to move their sterling into safer currencies (mainly the dollar). This led to a run on sterling, leading to additional pressure. Wilson argued that markets did not like a ‘Government concerned, even at a difficult time, with payments of the old-age pensioners and others in need, concerned to provide charitable largesse which our foreign critics felt Britain could not afford’.Footnote ³¹ Wilson’s explanation describes the channel speculators took and not why they feared for their assets, other than the state of the balance of payments. By pursuing an expansionary domestic policy while still wanting to maintain a fixed exchange rate, the government was violating the macroeconomic trilemma. Beyond the deficit, it appears that the market did not trust the Chancellor’s plans expressed on 11 November. This matches the timing of the ‘second crisis’ posited by Newton.

The press was critical of Chancellor Callaghan’s budget. The Wall Street Journal reported the ‘initial reaction to the special budget from businessmen was that it isn’t likely to help much in solving the nation’s most pressing problem, the serious deficit in the international balance of payments’.Footnote ³² This fell short of the currency market’s expectations, which hoped more drastic changes would be introduced to redress the situation.

Callaghan himself admitted that his handling of the City was a learning process: ‘I did not learn the ways of the City until I had held the post for some time, and consequently made mistakes.’Footnote ³³ One of his first mistakes was made on 11 November. The City was sceptical of his ability. Callaghan did nothing to alleviate their doubts in his Ways and Means speech. His emergency budget fell short of expectations. The budget was meant to calm the markets. But it contained few fundamental reforms for this purpose, apart from a new tax on petrol, which would raise £93 million. Most measures proposed were to take effect much later. The Economist noted: ‘A disturbing point about Mr Callaghan’s first emergency budget is that it contains an unduly large proportion of just such post-dated measures.’Footnote ³⁴

To sum up, the 1964 crisis started before Labour took office. It cannot only be blamed on Labour. The crisis became strongly accentuated in the run-up to the general election by fears of the party’s victory. Once elected, the government failed to solve the crisis until well after the 1967 devaluation. Dividing the crisis into sub-crises does not help us make sense of the events. This long currency crisis put a strain on the economy.

The crisis also put pressure on the dollar, as we will see. In Callaghan’s words: ‘Sterling’s devaluation would add to the difficulties of the dollar and might dislodge the Bretton Woods system, as in fact happened when sterling was eventually devalued in 1967.’Footnote ³⁵

The Gold Crisis

1964 marked the beginning of a four-year crisis for the London gold market.Footnote ³⁶ The crisis only ended with the creation of a two-tier gold market in March 1968. The crisis started in September 1964. At that point, the Gold Pool was no longer able to accumulate gold and the gold price started to rise (see Figure 9.5). Three factors played a major role: first, the US election, followed by fears of inflation in the United States; second, the French attempts to discredit the international monetary system; and third, the crisis in the secondary reserve currency, sterling.

Figure 9.5.

Average monthly gold prices and net monthly intervention

Source: Dealers’ reports (C8) and author’s calculations.

The start of the gold crisis coincides with both the US and UK elections. It is difficult to disentangle the effect of the 15 October 1964 UK general election from the influence of the 3 November US presidential election. Both played a role in the worsening of the London gold price. The market feared a sterling devaluation and Lyndon Johnson’s campaign for a ‘Great Society’ could have caused inflation which might have led to pressure on gold–dollar parity. To make matters worse, in February 1965 French President Charles de Gaulle delivered a speech attacking the foundation of the international monetary system.

The 1964 sterling crisis exacerbated the contagion from sterling to the dollar. Sterling is an important explanatory factor in the gold crisis. This does not mean that the US election (and later US inflation in the context of the Vietnam War) and French calls for reform played no role. The timing of the events presented here seems to give more importance to the role of sterling than previously assumed. The sterling crisis needs to be reconsidered as an important explanation for the increase in the gold price starting in 1964. The literature has focused on both US and French influences, but the role of sterling has not yet been examined by economic historians in detail. This is what I do here.

1961–64 witnessed crises for the dollar emanating from the Cold War. The trend changed in 1964. 1964–67 were crisis years for sterling and this indirectly affected the dollar. Here I review the timing of the contagion from sterling to gold and determine the impact of the sterling crises on the Gold Pool. Normally, a depreciation of sterling, the second most important reserve currency, should have led to an appreciation of the dollar. Investors would normally move from one reserve currency to another when faced with a fall in sterling. However, on this occasion they did not. Econometric evidence shows the opposite. When sterling depreciated, the dollar lost value against gold. More pressure on sterling meant that the gold price rose against the dollar. In response, the Gold Pool had to invest more resources to defend it. This is interpreted as contagion from sterling, the secondary reserve currency, to the dollar, the main reserve currency.

The argument is not new. In 1964, Coombs warned the Federal Open Market Committee (FOMC) of the disastrous consequences of a sterling devaluation. He used the fear of a global collapse to persuade the FOMC to lend more funds to the Bank of England. He argued: ‘[T]he British might decide to devalue sterling. This would probably precipitate an international financial crisis of the first magnitude. He [Coombs] would expect to see a major speculative drive on the London gold market and sooner or later an even more dangerous attack on the US dollar.’Footnote ³⁷

Despite Britain not devaluing sterling, as Coombs had feared, the situation of the London gold market worsened in parallel to the run-up to the 1964 sterling crisis. After the initial crisis of September–October 1964, sterling experienced almost continuous downward pressure until the 1967 devaluation. This had an impact on the international monetary system. The Gold Pool’s losses started in autumn 1964. The losses caused fundamental disagreements among Gold Pool members on the goals of the syndicate. Toniolo and Clement note that at the March 1965 gold experts’ meeting at the BIS the French and Belgian delegates were unhappy. They ‘did not join their colleagues in giving solemn assurances as to the medium-term continuation of Gold Pool operations’.Footnote ³⁸ At the same venue in November 1964, Coombs reported that central bankers across Europe and the United States worried about the impact of sterling on the international monetary system. At the November meeting Coombs reported: ‘In private conversations all the foreign department men from European central banks I met that weekend felt that we faced an explosive situation in both the gold and foreign exchange markets. They were particularly fearful that a massive speculative attack on sterling would react back on the dollar as well.’Footnote ³⁹

The pressure on the gold market was visible in two indicators: the price of gold and the cost of intervention. The more the Bank of England had to intervene to keep the price of gold under control, the more pressure there was. The Bank could either let the price rise, or waste Gold Pool reserves and intervene to keep it in check. These two indicators can be combined to form a rudimentary Exchange Market Pressure index for gold.Footnote ⁴⁰ EMPs help give a more detailed idea of the market pressure. They can be sensitive to calibration and yield a picture that can be biased by the choice of variables. Instead of using daily intervention figures, the index is computed by using the cumulative intervention of the amount of gold bought or sold by the Gold Pool. This shows trends rather than noisy daily changes and helps identify trends and breaks in trends. The index is plotted in Figure 9.6. The average values for 1961 are set at 100. Higher values indicate more pressure on the gold market.

Figure 9.6.

The Exchange Market Pressure index for gold

Source: Dealers’ reports (C8) for the gold price and gold interventions.

The index shows declining pressure at the end of 1962. The lower the index, the less the pressure. A downward trend indicates both that the price of gold was not rising and that the Gold Pool was able to accumulate gold for its members. The Gold Pool was successful in keeping the gold market under control during the early years of the syndicate. There is a reversal of the trend in September 1964. At this stage, the Gold Pool started selling more gold than it was accumulating and the gold price tended to rise as pressure on the market was mounting. What prompted this sudden reversal? Both the US election and speculation against sterling increased pressure on the London gold market. These factors increased in 1964–65. They progressively put pressure on the gold price. This threatened the stability of the international monetary system.

It is uncontroversial to assume that investors were influenced by the press. Therefore, press reports at the beginning of the gold crisis provide partial answers. The trend of the EMP shown in Figure 9.6 reversed on 14 September 1964. The trend for the gold price alone reversed on 25 August.Footnote ⁴¹ On 25 August, the Wall Street Journal published an article entitled ‘Sterling Rate Steadies after Fall to $2.7847, Lowest in Three Years’.Footnote ⁴² This can be seen in Figure 9.7. The Wall Street Journal attributed the fall to ‘the mounting British trade deficit, the prospect of future deterioration in the international balance of payments and uncertainties caused by the upcoming parliamentary election’.Footnote ⁴³

Figure 9.7.

Sterling spot exchange rate, May 1961–November 1964

Source: Accominotti et al., ‘Currency Regimes and Carry Trade’.

On the same day that sterling hit a three-year low, the gold price started its continuous climb from 25 August to September 1969. This analysis does not exclude other explanations. US inflation or French pressure could also be at play. But the negative news about sterling is likely linked to an increase in the gold price. The situation in the United States certainly also played a role. For example, on 15 August the Economist reported that the London gold market was exposed to ‘uncertainties over the Vietnam crisis which had carried the price to a five-month high’.Footnote ⁴⁴ After this announcement, however, the gold price recovered rapidly.

The mounting pressure on the gold market was not noticed by the Bank of England at the time. On several occasions the Bank told the Fed by telephone what it thought caused the gold price increase. One day it was ‘the demand of gold was … still in good volume’ because of ‘the long weekend because of the holiday in New York’ (8 October 1964)Footnote ⁴⁵. Another day it was ‘reflecting the usual Thursday demand and the long weekend because of the holiday here on Monday’ (3 September 1964). The reasons for the increase were not clear to Bridge, who also attributed the ‘gold buying to the situation in Vietnam’ (27 August 1964). The Bank was looking for excuses for the rise in the gold price. With the benefit of hindsight, we can see that it was a trend reversal and not occasional events that moved the gold price. The change of trend in August 1964 was a fundamental change, not a succession of small events as the Bank thought. The Vietnam War certainly put pressure on the gold price, but it was not clear why this started in August 1964.

After taking office, the new Labour government was advised by the Bank to raise interest rates to support the pound. The new government was hesitant. It delayed the rate hike until 23 November and imposed a sharp 2 per cent hike. This increase took place on a Monday instead of the usual Thursday, and informed the market that the situation was serious.Footnote ⁴⁶ Sterling recovered on the morning of the rate rise, but by midday panic had returned.Footnote ⁴⁷ The rate hike had consequences beyond the United Kingdom. It also prompted a rate increase in the United States. According to the Federal Reserve, the rate hike was agreed on ‘to maintain the international strength of the dollar’.Footnote ⁴⁸ Chairman Martin admitted: ‘If it hadn’t been for the British action, the Federal Reserve Board wouldn’t have increased the discount rate at this time.’Footnote ⁴⁹ For the New York Fed, the British move triggered an ‘emergency session’ of its directors. It was the ‘first increase in 30 years that didn’t come at a regularly scheduled directors’ meeting’.Footnote ⁵⁰ If a 2 per cent increase in the United Kingdom had been expected to trigger a reaction in the United States, the last-minute and dramatic nature of the United States’ reaction shows the influence of sterling on the dollar. This certainly did not reassure investors and helps explain their flight out of sterling and out of the dollar into gold.

After showing the impact of sterling, Figure 9.8 offers an overview of cumulative Gold Pool intervention. When the figure increases, it means that the Gold Pool is faced with mild market conditions and is able to buy gold. A decrease means more pressure on the gold market. The figure also highlights key dates in the three possible causes of the gold crisis starting in 1964.

Figure 9.8.

Cumulative Gold Pool interventions in million dollars from the creation of the Pool (6 November 1961) to November 1966

Source: Dealers’ reports (C8)

As the EMP (see Figure 9.6) illustrates, the gold market started to worsen in mid-September. In a first sequence, from around September 1964 to January 1965, interventions were limited while the price was increasing. The cumulative surplus of the Pool stabilised, but was not yet starting to decrease (see Figure 9.8). In other words, the Gold Pool was no longer able to buy gold without upsetting the price. It was not yet forced to sell gold to the market. From the 1964 sterling crisis onwards, the cumulative surplus of the Gold Pool decreased. Gold Pool members increasingly had to contribute to the Gold Pool instead of simply being allowed to buy the excess gold.

De Gaulle’s speech undoubtedly worsened the situation in February 1965. On 4 February, he declared that France would systematically convert excess dollar reserves into gold at the Fed Window. This speech was followed by France’s conversion of dollars into gold at the Fed. French conversions were concentrated from early 1965 to mid-1966.Footnote ⁵¹ After 1966, France had run out of dollars to convert into gold. During this time, France remained a member of the Gold Pool.

The Bank of England dealers were, at best, unimpressed by the speech. Their daily report reads: ‘The statement on the gold exchange standard by General de Gaulle did not create any fresh activity in the gold market; it came after effective dealing hours for Continental operators.’Footnote ⁵² The next day, dealers noted that ‘buying was rather heavy’ on the gold market.Footnote ⁵³ The report continues: ‘General de Gaulle’s discourse had little effect upon the exchange market although there was at first a disposition for dollars to be offered in Switzerland.’

A closer look at the London gold price and the intervention operations by the Bank of England in 1965 lead to a similar conclusion. While the speech was followed by one month of general gold price increases, interventions during this time do not seem to show that the Gold Pool was in distress. It is difficult to disentangle the effects of events in the United States and United Kingdom from de Gaulle’s speech. The French announcement happened on an upward trend in the gold price. One month after the speech, the Bank of England spent $54.6 million on behalf of the Gold Pool. The month before it had only spent $7.7 million. The effect was only short-lived as the net losses of the Pool were $18.2 million during the three months before and $19.2 million during the three months after de Gaulle’s speech. The speech was not a game changer. Later in 1965, Gold Pool operations suffered from the worsening of the US balance of payments and the rise in US inflation.Footnote ⁵⁴

It is not possible to disentangle quantitatively the exact contribution of sterling from the worsening conditions in the United States and the attacks by the French president. All three factors certainly played a role in the gold crisis. Next we will see how to quantify this impact.

Sterling and Gold

There was contagion from sterling to gold.Footnote ⁵⁵ These two prices were linked, but the relationship is different from what might have been expected. A downward shock to the price of sterling tended to make gold appreciate against the dollar. A shock to sterling led to global instability. This led investors to withdraw from the dollar and invest in gold, which was seen to be a safe haven. This in turn put pressure on US gold reserves, adding to the crisis. This link had been described and understood by some, but not all, contemporaries.Footnote ⁵⁶

From the 1964 crisis to the 1967 devaluation, there is a negative correlation between the three-month sterling forward rate and the gold price in dollars. Forward rates show the situation of sterling and offer a better proxy than spot rates. They were less subject to intervention by the Bank of England.Footnote ⁵⁷ They also present more volatility, which helps the interpretation. In Figure 9.9, the relationship seems to show that the lower the three-month forward sterling rate, the higher the London gold price. Troubles for sterling with larger forward discounts seem to affect the gold market negatively. This is most apparent after the election of Labour at the end of 1964. The election marked the beginning of a volatile period for sterling, with sustained pressure on the London gold market. It also appears in 1962 and 1963, although variations then were more modest.Footnote ⁵⁸ Starting in June 1967, the gold price stabilised near its maximum. It reached a ceiling of $35.20, the level at which the Gold Pool intervened constantly. At the same time, the sterling forward rate continued to depreciate.

Figure 9.9.

Gold fixing price at 11 am and London dollar/sterling three-month forward rate

Source: Forward data Accominotti et al., ‘Currency Regimes and the Carry Trade’.

Why is a negative correlation between the two series evidence of contagion between the two reserve currencies? During Bretton Woods, currencies were all subject to one-way speculation. If a currency was under stress, shorting that currency involved practically no risk as it could only be devalued or stay at the existing parity. There was no risk of sudden appreciation of the currency. This made shorting a risk-free bet. Sterling and the dollar were the two most traded currencies at the time. Rumours of devaluation on sterling would lead to a flight out of sterling into the dollar. Equally, when rumours of a dollar devaluation were circulating, investors would sell dollars and seek refuge in the second largest currency, sterling. This is what you would expect. If this does not happen, however, it means that investors running out of sterling do not see the dollar as a safe enough currency. That means there is contagion from sterling to the dollar.

Is there a negative correlation between the gold price and the sterling forward rate? To test this, the London price of gold is regressed on the forward sterling rate with a one-period lag. Data are daily, the gold price comes from the Bank of England archives and the forward rates from Accominotti et al.Footnote ⁵⁹ Using a lag is a way to cope with the fact that the value of the exchange rate is determined at the end of the day, whereas fixing the gold price takes place in the morning. The estimated equation includes a constant. The estimation is corrected for auto-correlation using the Huber–White procedure. Over the full sample (January 1961 to November 1967 or March 1968), the London price of gold is stationary. The data are in levels in the estimation and not as a difference. There is no co-integration relationship between the two variables.

The first estimation sample excludes the 1967 devaluation and stops on 15 November 1967. It yields a coefficient of –0.58. This means that when the forward exchange rate depreciates by one basis point, the gold price increases by approximately 0.6 basis points. Put differently, a decrease in the forward exchange rate from 2.8 to 2.75 is associated with an increase in the price of gold from 35.08 to 35.11. The second column in Table 9.1 shows that the coefficient is smaller when December 1967 to March 1968 is included. During this period, despite large variations in the sterling exchange rate, the gold price was kept constant at the upper band thanks to Gold Pool interventions.

Different sub-samples were analysed to determine whether this relationship was constant over time. This raises an econometric issue. The price of gold was not trend-stationary over 1964–66 when the Gold Pool allowed a continuous increase until it reached the upper bound of $35.20. The Augmented Dickey–Fuller Unit root test confirms that the series is trend-stationary between the October 1964 sterling crisis and the November 1967 sterling devaluation. For this sub-sample, a linear trend is added to the estimation (labelled Trend in Table 9.1). The negative relationship holds. It is still significant over 1964–67 but not over the previous sub-sample (January 1961–October 1964). The period 1961–63 is not uneventful, but sterling troubles in 1961 and 1963 are short-lived. As argued earlier, the period was marked by international political crises. These crises did not foster contagion between dollar and sterling. The sterling effect on the gold–dollar price only really emerges with the 1964 sterling crisis.