The previous chapter showed the consequences of the 1967 devaluation for the fall of the Gold Pool. But what were the consequences of the devaluation in Britain? It was intended to give the government breathing space to implement domestic policies and ease international pressure on sterling, but the 1967 devaluation did not ease Britain’s position internationally. Worse still, it led to more instability.

For the British government after the devaluation, it was all about saving face. Wilson had promised that the devaluation was all that was needed even if it quickly appeared that more deflationary measures would be needed.Footnote ¹ This devaluation, unpopular as it was, needed to have some positive effects. And if the positive effects were not apparent in a recovery in reserves, the Bank of England had to be creative in its presentation of the data. On the international front, cooperation started to be questioned. There was a shift towards the United States taking a more self-serving approach. The stability of the international monetary system was no longer a core US policy. And this was well before the Nixon shock of 1971. The Nixon administration was inaugurated in January 1969 and it quickly demonstrated that the United States was no longer willing to cooperate freely in international monetary matters.

Not a Real Solution

The 1967 devaluation meant to resolve the British balance of payments problems and move the economy towards growth and stability. Yet the opposite happened. On the London foreign exchange market, the situation notably worsened. Spreads widened as foreign exchange dealers became increasingly nervous. Volatility expanded significantly. After eighteen years of stable exchange rates, the pound was devalued by almost 14.3 per cent overnight. In the mind of dealers, this could lead to a further devaluation.

After the devaluation, the market became more unstable. It was more pessimistic about the monetary authorities’ ability to maintain the exchange rate at the new $2.40/sterling parity. The 1967 devaluation was more than a simple change in parity. It heightened instability and further decreased sterling’s role in the international monetary system. The 1967 devaluation failed to improve the stability of the pound. Here I look at volatility, bid–ask spreads and forward exchange rates before and after the devaluation. This helps understand the impact of the devaluation.

Figure 11.1 illustrates the ten-day volatility of the three-month sterling–dollar forward rate. As shown previously, the forward is a better reflection of market forces. The figure demonstrates that after the devaluation, volatility was much higher. The ten-day volatility increased on average by more than 126-fold when comparing 1964–66 with 1967–71.Footnote ² This is a substantial increase. It shows that the devaluation made market dealers more nervous.

Figure 11.1. Dollar–sterling three-month forward ten-day volatility

Source: Volatility: author’s calculation; forward data: Accominotti et al., ‘Currency Regimes and the Carry Trade’. Values for the 1967 devaluation are out of the scale of the graph, they reach 11 per cent at peak.

Figure 11.2 depicts the sterling three-month forward bid–ask spreads indexed to 100 for the beginning of the period in 1966. The spread is the difference between the buying and selling price. Here it is normalised to the average for 1966 for each currency to allow for comparison. This is just as in Chapter 3. The spreads inform us about the behaviour of the market-makers: commercial banks. These professionals made a profit from the difference between the buying and selling price. When market conditions worsened, they had to protect their profit by increasing the spread. Spread widening was consistent with the higher volatility presented in Figure 11.1. Both figures show increased uncertainty for dealers.

Figure 11.2. Sterling–dollar bid–ask spread index (whole year 1966 = 100), 1966–69

Source: Spread: author’s calculation: forward data: Accominotti et al., ‘Currency Regimes and the Carry Trade’.

Figures 11.2 and 11.3 show that, in general, the sterling rate became more volatile and dealers became more risk-averse. This does not indicate the relative strength of sterling versus the dollar. The goal of the devaluation was to have a more stable and credible currency, yet at a lower nominal level. How did the devaluation perform in this regard?

Figure 11.3. Sterling three-month forward rate

Note: The scale switches after the 1967 devaluation to allow for continuous reading.

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

Figure 11.3 represents the price of the three-month forward rate compared to the official Bretton Woods band. To allow for comparison before and after the devaluation, the scale is adjusted. On the adjusted scale, both the lower and upper bands before and after the devaluation match: 2.78 matches with 2.38 and 2.82 with 2.42.

Following Bordo, MacDonald and Oliver, the forward rate can be used as a proxy for the credibility of exchange rate bands.Footnote ³ Bordo et al. apply this technique between 1964 and 1967. Here I compare the pre- and post-devaluation period. Figure 11.3 shows that the 1967 devaluation was unsuccessful in restoring the credibility of sterling. After the devaluation, the forward rate was almost constantly breaking through the lower band of the Bretton Woods system. This appears on the right-hand side of the figure. Before the devaluation, the breaks were less frequent. Breaks were linked to sterling crises, such as the 1961 crisis and the crisis in the wake of the 1964 general election. After the devaluation, the forward rate showed that sterling was not credible most of the time. Its falls below the lower band were much more marked.

Disclosure of Reserves

The devaluation failed to restore confidence in the system. How did the Bank deal with this? In a fixed exchange rate system, the exchange rate is credible only if the central bank behind a currency has the means to defend it. The Bank of England needed dollars or gold to be able to buy sterling when the exchange rate was dropping. The level of reserves in itself also has a signalling value for the strength of the currency. The higher the reserves, the more credible a currency appears to investors. During most of its history, the Bank managed to keep its balance sheet obscure enough to make it impossible for investors to understand its true reserve position.Footnote ⁴ This changed in the late 1950s.

The Bank of England was caught between two trends. There was a demand for more transparency following the publication of the 1959 Radcliffe Report. The report was an inquiry into the Bank’s activities. This meant communicating reserves and being vulnerable in case of a reserve drop. At the same time, the Bank witnessed increased international financial flows following convertibility in 1958. This put even more strain on its reserves. The Bank had to satisfy both sets of pressure: more pressure on its reserves, and the need to be more transparent. It started communicating more while manipulating the data it published.

The government pushed the Bank to give an impression of stability and the institution used window dressing to do that. Here I show how window dressing functioned thanks to new data from the EEA account. The Bank worked with the Fed to prevent contradictory information from being shared by both institutions. New archival evidence attests to how the two central banks cooperated to conceal evidence of window dressing.

Window dressing is the manipulation of accounting data before their official publication to make them appear better than they are. It has been a widespread practice among commercial banks throughout history. Roger Hinderliter and Hugh Rockoff show that ante-bellum banks in the United States used window dressing to manage their reserves.Footnote ⁵ Banks under the Bank of England’s jurisdiction also practised window dressing. They moved balances among each other on set weekdays before publishing their reserves.

Reserve publication was a difficult exercise for central bankers during the Bretton Woods period. In a fixed exchange system, reserve information can create a run on the currency. This has been modelled in a second-generation currency crisis model, as first laid out by Maurice Obstfeld.Footnote ⁶ In these models, self-fulfilling dynamics make a run on a stable currency rational for investors as soon as other investors start selling. For this reason, the reserves announcements were well prepared. Credibility was key in currency management. If the central bank was credible enough, it could improve the stability of its currency by exaggerating its reserve position.

The reserves of the Bank were literately made up. An example of this is an internal memorandum from the Bank which read: ‘It will be necessary shortly to decide what figure we are to show for the reserve loss for July.’Footnote ⁷ The wording establishes how the Bank saw reserve publication. It was a guessing game, somewhere between reality and what the Bank thought the market believed.

Disclosure of the reserve position was a communication exercise. High effort went into drafting these communications. An internal memorandum reads: ‘The draft Press Notice is given in two alternative forms. The first alternative is designed to avoid having the fall mentioned at the beginning of the sentence and, therefore, the first thing that meets the eye from the ticker tape. The second alternative is in the conventional form.’Footnote ⁸ The ticker tape transmitted stock price information over telegraph lines. As the text was progressively printed on a thin paper ribbon, the first words mattered. These words could lead to panic selling by traders. The Bank therefore put care into the order of the words in its communication. This problem illustrates how the publication of the reserve position was important to the Bank.

The Bank of England communicated its reserves in press releases before publishing them in the Quarterly Bulletin. The Quarterly Bulletin was first published in December 1960. The Bank launched the Bulletin because of recommendations in the Radcliffe Report. The Bank had anticipated these recommendations. It started internal discussions about a quarterly publication as early as 1958.Footnote ⁹ William Allen analysed the shift in attitude and transcribed the questions to Governor Cobbold in July 1957.Footnote ¹⁰ During his testimony before the Radcliffe committee, Cobbold was not keen to divulge too much information. He argued that ‘it is of some doubt whether it would really clarify the issues for the public if the Bank were continually [issuing] statements with a different slant from similar statements made by Government to the public’.Footnote ¹¹ Cobbold did not see any need for the Bank to communicate its reserves position. The Radcliffe Report, with its 2,294 questions, asked the Bank to change this, as Allen noted.Footnote ¹²

The Fed was ahead in terms of transparency. It had published the Federal Reserve Bulletin since 1914, forty-six years before the Bank started doing the same.Footnote ¹³ The Fed viewed transparency by the Bank of England with amusement. In 1956, the Fed displayed some irony in stating that the Bank of England took ‘a certain pride in pointing out that hardly anything can be inferred by outsiders from their balance sheet’.Footnote ¹⁴ In a memorandum by the Foreign Research Division in 1958, the Fed commented that the governor of the Bank of England ‘for the first time’ publicly considered more transparency. The Fed further commented, ‘it seems clear that the Bank of England is being pushed – by much public criticism – into giving out more information’.Footnote ¹⁵

The Bank had to share more information, and so it did. Starting in 1960, reserve positions were announced to the press. They were then published in the appendix of the Quarterly Bulletin. At first, the Bank only reported a generic reserve figure. Later it started breaking it down into convertible currencies and gold. In an article in May 1963, The Financial Times welcomed this additional transparency. The article also stressed that other countries had been reporting more detailed reserve figures for some time.Footnote ¹⁶

In a note about the meeting of the Court of Directors in October 1964, the governor mentioned that reserve publication would be ‘accompanied by a statement that central bank assistance had been arranged, but that the extent of the assistance used would not be disclosed’.Footnote ¹⁷ It meant that the Bank would disclose that swaps with the Federal Reserve were used to increase reserves. But not the amount drawn on the swaps. This was the Bank’s get-out-of-jail-free card. It could still keep how much it owed the Federal Reserve secret.

In the same meeting, the governor informed the court that he had written to the Chancellor of the Exchequer and to the Prime Minister. The governor was ‘urging that the leader of the Opposition should be made acquainted with the true position in the hope that he would help to discourage irresponsible comment during the election campaign’.Footnote ¹⁸ But Downing Street did not think this was a good idea. The Prime Minister and the Chancellor reminded the governor that disclosure was their decision, not his. The EEA account belonged to the Treasury, not the Bank. And the government was managing the exchange rate. In this sense, it seems that window dressing was a political decision, and not one made by the Bank of England alone. Yet the Bank fully agreed with the decision to implement window dressing.

Swaps became the main tool for window dressing. As seen in Chapter 8, they were the Bank’s favourite source of credit. They were readily available, cheap and discreet. Capie argues that swaps were ‘essentially window-dressing arrangements and allowed a false picture of the reserves to be presented’.Footnote ¹⁹ From their inception, these devices were meant to help cover temporary losses to avoid speculative attacks on the currency. Over time, these temporary measures were made more permanent. Swaps became central to foreign exchange management.

But was it misleading investors and the public? Swaps raised ethical concerns on both sides of the Atlantic. In a November 1968 memorandum, the Bank presented two options for reserves publication. One was ‘consistent with previous practice’.Footnote ²⁰ It would withhold swaps from the public. The other option had ‘been drafted on the assumption that we now decide to come “clean”’. This emphasises that the Bank believed that concealing reserves was not a ‘clean’ business. The Bank was recommending more disclosure in the reserve publication. The Bank argued that once the reserve situation had normalised, they would have to ‘reveal the whole truth’. And then ‘it would be embarrassing if there were then two versions of the “truth”’.Footnote ²¹ The frequent use of quotation marks in this memorandum shows the embarrassing position the Bank was in. It did not want to communicate reserves without the government’s agreement. Yet officials at the Bank knew that they could be blamed for window dressing in the future.

Window Dressing

The Bank of England was window dressing its reserves by publishing only the asset side of the balance sheet of the EEA. It was not disclosing any outstanding loans or swaps. Standard accounting practices require disclosing both the assets and the liabilities. Bordo et al. were the first to highlight the scale of the Bank of England’s window dressing.Footnote ²²

Capie later published more data on window dressing between 1964 and 1967 from a report by Richard Kahn.Footnote ²³ Capie demonstrated that the net reserve position of the Bank of England after December 1967 was negative. The Bank owed more reserves to foreign central banks than it possessed. Capie’s figures are reproduced in Figure 11.4. The figure shows that net reserves continued to decrease after the November 1967 devaluation. Capie’s data stop in 1968 and do not include the daily reserve figures of the EEA. EEA data give a more precise view of the mechanism behind window dressing and its very short-term nature.

Figure 11.4. Bank of England published vs net reserves

Source: Capie, The Bank of England, 231–2.

The Bank of England reported its net position to the Treasury in monthly letters from 1962 to 1972.Footnote ²⁴ These letters gave the net oversold forward position. They also gave the amounts of window dressing. The term window dressing is avoided until 1965. It was too blunt. The more euphemistic term ‘Net short-term aid from central banks’ was used.Footnote ²⁵ From 1965 onwards, the Bank of England explicitly uses the term ‘window dressing’. The cat was out of the bag. The Bank no longer felt the need to use a euphemism when it came to terminology. In May 1968, window dressing reached its peak at $5,000 million.Footnote ²⁶ At this stage, the Bank of England was borrowing up to $5 billion to conceal the fact that it had lost reserves. It amounted to 11 per cent of the country’s gross domestic product.

Next, I use previously unpublished daily accounts of the EEA to show how the Bank used swaps to window-dress its reserve position at the end of each month. The EEA ledgers have not been used in previous literature. They contain daily data on EEA gold, dollar and other currency holdings. These accounts were not published or disclosed at the time. This means they were not window-dressed or manipulated. Rather, they were used for internal and accounting purposes. Previous research has established that the Bank did use window dressing, but here I outline its short-term nature (typically a few days). This is important as it shows that swaps were not used for medium-term reserve management. They were used for investor manipulation. The goal was to convince currency buyers that the Bank was healthier than it really was.

Figure 11.5 illustrates how window dressing worked. The solid line reports the convertible reserves as published in the Quarterly Bulletin. This is the information that was available to market participants. The stacked columns show the actual daily dollar reserves.Footnote ²⁷ Spikes appear at monthly intervals. They are the window dressing, and indicate the short-term borrowing. This borrowing was used to ensure the reserves level was high enough for the days when the reports were prepared.

Figure 11.5. Published EEA convertible currency reserves vs actual dollar reserves held at the EEA

Source: EEA ledgers and Quarterly Bulletin.

The Bank borrowed dollars shortly before the actual reporting day. It drew on swap lines. Swap drawings could be as short as overnight. Table 11.1 illustrates how window dressing worked, using data from the EEA ledgers. On Friday, 31 May 1968, the Bank borrowed over £450 million. This represented an increase in reserves of 171 per cent. The swap operation was then reversed the next working day, and on Tuesday the reserves level was back to where it was before reporting.

The details of these operations emphasise how swap networks were not long-term instruments. They were short-term instruments to manipulate published figures. Another way of hiding the extent of reserves losses was by not publishing the open forward position.Footnote ²⁸ Intervention on the forward market intensified in the late 1960s. The Bank of England was increasingly exposed to a large forward position. This exposure was not published either.

Secrets among Friends

Window dressing was an internal practice. Yet, being the main supplier of funds, the Fed was informed of the use of its funds. More than that, the Fed actively took part in concealing information from the public. This turned into a debate in the Federal Reserve system. The debate was between Coombs, who was inclined to do anything to save the pound in the short run, and some members of the FOMC, who raised ethical concerns. There was direct collaboration between the Bank and the Fed on window dressing, and window dressing worked because of the close cooperation between the two central banks.

Before publishing its Quarterly Bulletin, the Bank of England consulted the Fed on the precise wording of the reserve publication. This was important because the Fed would also communicate periodically on the swap position with the Bank of England. The public statements by the two institutions needed to mesh. Bridge from the Bank called David Bodner from the Fed in October 1966 to discuss strategy. Bodner reported Bridge’s reasoning to his hierarchy: ‘In order to come out in approximately the same position as in the end of September, that is, a slight reserve increase and no net recourse to central bank assistance, Bridge said he would require approximately $500 million.’Footnote ²⁹ At this point Bridge wanted to publish reserves that increased slightly. But this had no relationship with the actual evolution of the Bank’s reserves. It was only a perception that Bridge wanted to create in the market. The goal was either a stable or slightly increasing reserve position. Looking at the line in Figure 11.5 shows this tendency. Despite the true reserves being in decline, ‘imagined’ reserves were stable or increasing. There were some exceptions when the market expected heavy reserve losses and the Bank’s imaginary reserves also dropped slightly.

Collaboration between the Fed and the Bank went much further. Before publishing the minutes of the FOMC, the Fed sent the excerpts of the minutes to the Bank of England. The Bank was to delete anything mentioning window dressing. In December 1971, before publishing the minutes of the FOMC for 1966, Coombs wrote:

Coombs suggested deleting passages in which some FOMC members criticised window dressing. Mitchell of the FOMC did not like window dressing. He suggested that the Bank of England would get better results ‘if they reported their reserve position accurately than if they attempted to conceal their true reserve position’.Footnote ³¹ MacLaury, another FOMC member, was of a different opinion. He stressed that there was a risk of ‘setting off a cycle of speculation against sterling’ if the Bank published a loss of $200 million.Footnote ³² The amount was ‘large for a single month’, in comparison with what was published the previous month. Tension arose between the FOMC, which did not want to support unethical practices, and the New York Fed, which was dealing on the front line of international markets. Coombs held positions at both institutions. He was the link that attempted to convey to the FOMC the reasoning behind this short-term assistance.

The Bank of England understood the FOMC’s reticence. A memorandum read that overnight swaps could not be used for window dressing as ‘the F.O.M.C. regard that as unethical if not immoral’.Footnote ³³ As the logical consequence of this stance, the note continues, ‘this means that any drawing under the swap ought to be left outstanding at least for 32 days’.Footnote ³⁴ The Bank would essentially increase the length of its swaps to make it less obvious that it was manipulation. The only technical difference was that the Bank of England had to pay interest for the period. Interest for a month would be more than for an overnight swap, for which the cost would be negligible. But the fact remained that the Bank used this short-term loan to avoid disclosing the real level of its reserves.