Any deal has two sides. There are buyers and sellers. In the case of earmarking gold in foreign central banks, there are in fact often more than two parties. That is to say, an initial deposit of earmarked gold naturally involves the depositor and the host bank. If the depositor, however, subsequently decides to transfer ownership to another central bank and requests the host bank to do so, three parties become involved in the process. While the old owner of the gold and its new holder may be fully aware of the circumstances prompting the transfer, the host bank will usually know no more than the bare terms of the transfer instruction. For the host bank, it is a simple custodial function. Elsewhere in this report, it has been argued that the advent of war in 1939 obliged the Bank of Canada to modify the conditions under which it would take instruction from its earmark clients, but the relationship still largely remained one of instruction without explication of purpose. Only the Bank of England, with its intimate contacts with the senior officers of the Bank of Canada, usually revealed the pressures behind its needs. Hence, the Bank of Canada was in many instances a blind participant in fiscal transactions that touched the gold in its vaults. As a result, the Bank of Canada's records tell only one side of a story and cast only a partial light on the motivations of its European partners.

The commissioning of this report was prompted by allegations that between 1942 and 1944 the Bank of Canada was party to a series of gold swaps that enabled neutral Switzerland and Portugal to launder "a significant amount of looted gold" through their earmark accounts in Ottawa and at the Federal Reserve Bank of New York. The principal purpose of this report has been to research and to present the Canadian perspective of these events, drawing from historical records over which Canada has complete control and unfettered access. To what degree did the Bank of Canada understand the full implications of the transactions which it facilitated? Did it act with due diligence? A secondary challenge, however, remains -- trying to establish some sense of the European side of these transactions. As this project's mandate remarked: "This project does not have as an objective to document the history of international gold movements during the war, but the secondary results of research by other governments, central banks and others may be useful in providing context for the Canadian story."

GATHERING HISTORICAL INFORMATION IN EUROPE

Creating some sense of the European side of these gold swaps presented formidable challenges. Time was precious and the logistics were daunting. There was no assured access to the archives of European central banks. Language barriers loomed. Records procedures would undoubtedly be unfamiliar. Other countries had their own access-to-information and privacy protocols. There was, however, one very encouraging sign. Since 1995, European central and commercial banks in England, Switzerland, Portugal and Sweden had been prompted by public controversy over the issue of Nazi gold and dormant bank accounts to launch historical investigations into their own roles in wartime financial arrangements. By the summer of 1997, these investigations were well under way and receptive to outside inquiry. They were largely staffed by independent professional historians and accountants who had been given a wide mandate to present as thorough and transparent an analysis of each bank's wartime role as their records made possible. In light of this opportunity, it was decided that the most expedient means of creating some sense of a European perspective was to approach each of these institutions or commissions with a detailed request for information on the 1942-44 gold swaps.

Approaches were therefore made to five historical commissions/teams currently active in Europe. Four are working under the auspices of their central banks: the Bank of England, the Banco de Portugal, Sveriges Riksbank and the Banque Nationale Suisse (BNS). The fifth was the Commonwealth and Foreign Office in London, a group that has been actively reporting its wartime gold research under the title Nazi Gold. The requests from Canada were tailored to the particular role played by each nation in wartime financial arrangements, with a special eye to its contacts with Canada and its central bank. These requests all contained some common elements. Were there records of the specific transactions that involved the Bank of Canada in the 1942-44 period? Were there other references to Canada's role in wartime transatlantic financial arrangements? Did their files contain any memoranda or background papers that shed light on how Canada fit, if at all, into their country's overall wartime financial strategy? Could their research provide any sense of the real motivation that lay behind these swap transactions? In each case, the request contained detailed information, drawn from Bank of Canada records, guiding the recipient to the specific events that involved Canada (e.g., the Anglo-Swiss francs-for-gold swap and the subsequent gold transfer from Switzerland to Portugal in the spring of 1942). These requests were all despatched to Europe in the last week of August 1997.

Without exception, all the requests met with a cooperative response. All were quickly acknowledged and brought some sense of when a detailed response could be expected. Most of the respondents commented on the heavy pressure under which they themselves were operating in trying to fulfil the mandates set by their own central banks. It should be acknowledged that the archives of the Bank of England and the Banque Nationale Suisse were especially forthcoming in responding to our requests. The Bank of England, for instance, arranged for large sections of their files relating to Canada to be microfilmed and sent by courier to Canada. Not one of the respondents set any condition on their willingness to furnish us with information. (In the wake of the recent controversies surrounding the Nazi gold issue in Europe, there now appears to be broad understanding of the high price in public credibility that banks will pay for being unwilling to be open to public attention in these matters.)

The subsequent arrival throughout September of dossiers of information from the five historical groups in Europe allowed a fairly coherent picture of the European perspective on the gold swaps of 1942-44 to be reliably pieced together. As these reports became available, one fairly common trait emerged: there was very little documentary trace or present-day consciousness of Canada or the Bank of Canada in the European historical work on this issue. With the exception of the Bank of England, which reported the existence of ten files relating to Canada, the other respondents all invariably commented that their search for documents and their examination of their own research yielded little on Canada. In all of their minds, Canada seemed to occupy a marginal role in the reconstruction of wartime financial relations that they were undertaking. A good example of this was provided in the prompt and cooperative response to our August inquiry by Dr. Gill Bennett, head of historians, at the Foreign and Commonwealth Office in London. Bennett's group has been busy researching Britain's part in wartime gold for over a year and has produced two studies of their findings under the title Nazi Gold: Information from the British Archives, published in September 1996 and May 1997. Bennett wrote in response to our inquiry:

...I have looked through the documentation we have collected on Nazi gold for any references of gold to Canada before or during the Second World War, but without success. There may be references to such transfers in files at the Public Record Office [Britain's national archive], but the Foreign Office Index for 1942, for example, does not indicate the presence of such information either under headings for Banks and Banking, Canada, or Switzerland. It would be necessary to check each reference to Canada and Switzerland under the broader heading of 'Gold' to be sure....The fact that we have found no references in the large amount of material we already hold here leads me to think, however, that such an effort may well yield disappointing results.⁽¹⁾

It bears reiteration that the massive American study of wartime gold commissioned by President Clinton and headed by Under Secretary of Commerce Stuart Eizenstat -- U.S. and Allied Efforts To Recover and Restore Gold and Other Assets Stolen or Hidden by Germany During World War II (Washington, 1997) -- does not contain a single reference to Canada. The Eizenstat report was based on research among some 15 million pages of documents in the National Archives in Washington. When this trove of documents was reviewed by researchers from the World Jewish Congress, only a single document implicating Canada was uncovered -- the 1944 "From a very Confidential Source" memorandum alleging gold laundering through Ottawa. It was that document that quite appropriately triggered this project.

The scant impression left in Europe of Canada's wartime role in gold in no way deterred our efforts to reconstruct the motivation that lay behind the actions of the European parties to the swaps of 1942-44. In the material supplied to us, three distinct episodes were examined: the initial swap of British gold for Swiss francs between 1942 and 1943, the 1942 swap of part of this gold from Switzerland to Portugal and the 1944 swap of Portuguese gold to Sweden in company with a further transfer of Swiss gold to Portugal. The following reconstructions were therefore based on a mixture of original central bank documentation, documents emerging out of the postwar Allied effort to identify looted gold and present-day historians' opinion.

THE ANGLO-SWISS GOLD-FOR-FRANCS SWAP OF 1942-43

It was the 56 tons of gold that England provided Switzerland with through 1942-43 in exchange for Swiss francs that provided the feedstock of bullion for the subsequent contentious swaps between the Swiss and the Portuguese and, ultimately, between the Portuguese and the Swedes. This gold never physically left Ottawa. It was part of the Bank of England's earmark account at the Bank of Canada, built up by outright purchase and actual transfer from England in the late 1930s and the early war years. There is therefore absolutely no possibility that the gold in question was tainted in the sense of containing gold looted by the Germans in Europe. This was clean gold. What is more appropriately in question is the motivation that first led Britain to begin transferring ownership of the gold to neutral Switzerland in 1942, thereby putting the neutral Swiss in a position to redeploy some of this gold to the earmark accounts of other neutral powers in Ottawa.

Two factors above all else explain the Anglo-Swiss swap of 1942: Britain's desperate need for Swiss francs and Switzerland's need for gold and British trade to stabilize its commercial and financial performance. As has been explained in an excellent paper by Professor Neville Wylie of Cambridge University, "London's acute shortage of Swiss currency" in the early years of the war coloured its entire attitude to neutral Switzerland.⁽²⁾ This shortage was "perennial" and was rooted in Switzerland's prewar balance of trade surplus with Britain. In 1938, for instance, Switzerland exported SFr149.7 million in goods to Britain, but imported only SFr97.2 million. By 1940, the gap had narrowed -- SFr96.6 million in exports to SFr89.6 million in imports.⁽³⁾ What had changed in war was Britain's ability to find the francs to cover its purchases of Swiss goods, many of which were directly related to Britain's war economy and military needs. Even after her ability to trade directly with Switzerland was restricted by the Axis encirclement of the neutral Swiss in 1940, Britain kept trading with the Swiss, even if this necessitated smuggling goods out of the country.⁽⁴⁾ Other invisible expenses, such as the cost of maintaining a legation and possibly supporting covert operations based in Switzerland, also placed demands on Britain's supply of francs. To replenish its supply of francs, Britain sold sterling, Bank of England gold and eventually drew on its own cache of U.S. dollars to procure francs.

By early 1942, francs were becoming harder to procure. The United States had issued injunctions against the use of its dollars in European trade for fear that they would find their way into Axis coffers. For a while, Britain even skirted these regulations by drawing dollars out of commercial banks in New York so as not to alert the U.S. Federal Reserve to its activities. However, Britain's obligation to the United States under the 1941 Lend-Lease Agreement and America's subsequent entry into the war ended this expedient and deceptive means of securing hard currency for trading with Switzerland.

In a parallel subterfuge, the British ignored Swiss requests that the Bank of England transfer part of its store of earmarked gold from London to safer New York. Instead, the Bank of England added some of the Swiss gold to its own transfers of earmarked gold to Ottawa and retained the rest in London. Thus, the British retained control over Swiss gold either in London or in Ottawa, thereby providing themselves with a possible hedge in their commercial relations with the Swiss. Thus, as Professor Wylie has pointed out, by "the winter of 1941/42, the Bank of England was doubly compromised," having gone counter to American currency regulations and at the same time undermined Swiss efforts to find safe haven for their gold in the United States.

This delicate situation became worse when in early 1942 the Swiss, heeding American warnings, refused to take any more U.S. dollars from the British. This coincided with British anxiety that Switzerland was trading too heavily with the Axis. In July 1941, a Swiss-German economic agreement had provided Germany with a credit of SFr850 million that the British viewed with alarm and immediately countered with restrictions on exports of British raw materials to Switzerland. If Switzerland limited its exports to the Axis or increased its shipments of strategic goods to Britain, the British said that they would ease their restrictions on exports to Switzerland. At this awkward juncture, the British hit on the idea of raising francs by selling their gold earmarked in Ottawa or sterling blocked in foreign banks in exchange for Swiss francs. This was exactly the import of Montagu Norman's March 8, 1942, cable to Bank of Canada Governor Graham Towers: "...Difficulties have lately arisen mainly owing to United States freezing policy. Swiss are now reluctant to accept U.S. dollars without special license for their transfer to third party..."⁽⁵⁾ Negotiations were under way to test the Swiss reaction to this offer, Norman reported.

The Swiss were not enthusiastic. As Switzerland's central bank, the Banque Nationale Suisse was tasked with keeping Swiss inflation under control, and in an economy being fanned by wartime demand this meant maintaining a reserve of gold and/or foreign currency within Switzerland's borders. The yardstick for this was a gold or foreign currency reserve equal to 40% of the money in circulation in Switzerland. This operation was already proving difficult because so much -- between 33% and 50% by some estimates -- of Switzerland's gold was already situated outside Switzerland, blocked in London and New York until the end of hostilities and therefore useless for purposes of internal monetary control. As Professor Wylie concludes: "The last thing the National Bank [BNS] wanted was to accumulate further quantities of sterling or gold in London which could not be touched until the end of the war." Adding to the pressure on the franc was the fact that Switzerland had in late 1941 agreed to take responsibility for protecting British assets blocked in Axis territories. This too required francs to flow from Britain to Switzerland and, to the dismay of the BNS, most of these francs went to non-commercial circulation, thereby exerting inflationary pressure on the Swiss economy. Rather than further jeopardize the internal stability of the Swiss economy, the BNS in March 1942 adopted an adamantly vigilant attitude toward any further release of francs for blocked gold beyond Switzerland's borders.

A protracted negotiation thus began in London. In February 1942, Berne appointed a delegation -- composed of a government minister and two academics -- to treat with the English. For both sides the issue was currency liquidity -- the British appetite for francs and the Swiss desire to conserve francs and gold within their borders. Throughout 1942 and 1943, several tentative agreements were drafted but fell apart, usually because of disagreements over what materials were to be allowed to enter Switzerland in the wake of sterling/gold-for-francs swaps.

This impasse opened the way for the swap of British gold held in Ottawa for Swiss francs that began on March 25, 1942, with an initial transfer of 66,815 fine ounces from the Bank of England earmark account in Ottawa to the newly established account of the BNS in Ottawa. Both sides regarded the Ottawa arrangement as a pro tem. measure, designed to feed the British with a minimum supply of francs to cover their basic needs (e.g., legation expenses) until some more durable arrangement could be negotiated in London. As a Bank of England official noted in a letter of April 15, 1943: "...This is a purely temporary arrangement and one which the Swiss, with a declining gold stock at home, accept reluctantly."⁽⁶⁾

The Swiss were indeed extremely reluctant to enter this swap arrangement. As a BNS internal memo noted in mid-January: "While, on the one hand, the National Bank must hand over freely available Swiss francs and pay them out on behalf of the Bank of England in Switzerland and other countries, in return, it receives non-transferable pounds and gold stocks stored in Ottawa that, for all practical purposes, cannot be used by the Swiss economy during the war. The grave danger that this development represents for the Swiss price structure is clear. The question is whether it will be possible to eliminate the inflationary effects for the Federation or prevent the bank of issue from continually issuing new money via taxation and other measures that tie up market resources."⁽⁷⁾ In short, blocked gold in New York or Ottawa could not be fed back into the Swiss economy or used to pay for its imports. Thus, Switzerland was unable to satisfy all of its internal demand for imports and inflation resulted from the fact that wartime demand outstripped supply. A Swiss exchange of francs for earmarked gold in Canada was a way of mopping up Swiss francs and reducing potential inflationary pressures in Switzerland.

The board of the BNS was by early February of 1942 still "not much in favour of a transfer of gold stocks to Canada."⁽⁸⁾ The Swiss were eventually persuaded to accept this temporary gold-for-francs arrangement because, while the gold they were to receive by it was still blocked in the sense that it could not be brought directly to Switzerland, it was nonetheless freer to move than the blocked gold they had in London and New York, which was frozen hard in situ until the end of the war. As the minute book of the BNS board of directors noted on March 26, the gold it received on earmark in Canada "could be exported to other central banks in the Western Hemisphere and transferred to the same central banks, as well as to the central banks of the neutral European countries, Sweden, Spain and Portugal."⁽⁹⁾ To some degree, therefore, the gold in Ottawa was portable and might through some combination of circumstances ultimately be brought home indirectly before the end of the war.

The swap of gold for francs in Ottawa thus satisfied, albeit temporarily and imperfectly, the needs of both Berne and London. The arrangement was the product of London's desperation and ingenuity and was clearly accepted faute de mieux by the Swiss. As Montagu Norman of the Bank of England noted, Swiss agreement "represented a concession by the Swiss National Bank to avoid a breakdown in the financial relations between the sterling area and Switzerland."⁽¹⁰⁾ It provided the rationale for the 155 regular and relatively equal transfers of gold from British to Swiss ownership that followed throughout 1942 and 1943, feeding francs to the British, thereby keeping their financial lifeline to Switzerland in place. There is no documentary evidence, in either the Bank of Canada records or in the Bank of England files sent to us, that the Anglo-Swiss swap was designed to help Britain "assist prisoners of war in Germany," as Graham Towers told the Bank of Canada board of directors in June 1942. It is possible that some of the francs secured by the deal went to humanitarian payments via the Red Cross in Switzerland, but no proof of this has been found. Switzerland was also, for instance, a much-used gateway to freedom for escaped British prisoners of war; such passages to freedom required Swiss francs.

In December 1943 the British negotiators, still facing a chronic shortfall of francs in England, finally broke the deadlock by offering the Swiss what they had wanted all along: the ability to bring gold into neutral Europe. In practice, British gold would now be placed at the disposal of the Swiss, who could use it to satisfy their commercial accounts in Spain and Portugal, thereby conserving their hard currency and gold reserves at home. The agreement also freed hitherto blocked gold acquired in London by the Swiss during 1943 and placed it at Switzerland's disposal in neutral Europe. Significantly, the accord did not free Switzerland's accumulated earmarked gold in Ottawa, but instead reconfirmed the original conditions set upon it in 1942. The Ottawa gold would not be free for patriation to Switzerland until six months after the end of Britain's war with Germany. Not surprisingly, as soon as the Anglo-Swiss accord came into operation on January 1, 1944, the Swiss stopped accepting gold in Ottawa for francs. Ottawa gold was still blocked gold. Gold available in London was now more versatile.

What is striking from the perspective of this report is that both the Bank of England records submitted to us and the lengthy commentary and packet of documentation supplied by the Banque Nationale Suisse were in spontaneous agreement over the above course of events and the motivation behind it. Both saw Anglo-Swiss relations from 1939 to 1943 as essentially an ongoing liquidity crisis. The principal concern was to keep the flow of Anglo-Swiss wartime trade going and, particularly from the Swiss perspective, to allow the BNS to restrain inflationary pressures in the Swiss economy.⁽¹¹⁾ Anxiety over the possibility that looted German gold might find its way into the mainstream of European finance seemed to take a distinctly secondary role in these proceedings. The imposition of conditions that confined Swiss gold earmarked in Ottawa to movement within the Western Hemisphere and paper transfer to European neutrals seemed to provide London with the assurance that the integrity of its dealings with Switzerland was being protected.

THE SWISS-PORTUGESE GOLD SWAP OF 1942

Thus, in the spring of 1942, gold owned by a European neutral power began to accumulate in Ottawa under earmark: 66,815 fine ounces on March 25, another 6,862 ounces on March 28, 13,328 ounces on April 7 -- the payments quickly became regular. By the end of 1943, the transfers would equal 56 tons of gold. Ottawa became one of Switzerland's offshore gold havens, never rivalling New York and London in the magnitude of its holdings, but joining Stockholm and Buenos Aires as secondary locations of earmarked gold. Despite the delicate balance of its internal monetary affairs, Switzerland continued to be a net purchaser of gold from the Allies throughout the war. The United States was the principal seller of gold to the Swiss, selling some SFr1,178.2 million from 1939 to 1945. Britain followed with sales (that would have included the gold earmarked for Switzerland in Ottawa) of SFr580 million, while the French Vichy government sold SFr117 million in gold to the Swiss during the war (although their sales were packed into 1942 and 1945). Even Canada sold Canadian-produced gold to the Swiss, some SFr5.3 million worth, much of it sold after the British and Swiss struck their January 1944 deal to allow Allied gold freer access into Europe.⁽¹²⁾

Only one country sold more gold to the Swiss than the Americans -- Germany. From 1939 to 1945, the Reichsbank sold SFr1,210.3 million worth of gold to the Banque Nationale Suisse. A large portion of this passed through Switzerland to other neutral countries in Europe that were trading with the Axis. Portugal took SFr451.5 million and Spain SFr185.2 million.⁽¹³⁾ Gold transfers into Switzerland were paralleled by a flow of usable foreign currencies into Switzerland. Much of the gold was gold looted from the vaults of European countries that had fallen under Axis occupation; some small amount of it was already probably non-monetary gold stripped from citizens of these same countries.

By late 1941, the Allies, sensing that Germany's meagre prewar reserve of gold was long spent in the initial years of the war, began suspecting that Germany was financing its war effort with looted gold. The British Ministry of Economic Warfare and the newly created Office of Strategic Services in the United States began to monitor gold flows out of Germany. In mid-1942, the British began pressuring the other Allies for some sort of diplomatic injunction against countries accepting looted goods in trade with the Axis. These early efforts were piecemeal and the knowledge of the full extent of Axis bartering of looted gold still vague. But one of the first clear realizations was that the Banco de Portugal was the primary recipient of Banque Nationale Suisse exports of gold, a fact that was quickly linked to the belief that much of this Iberian-bound gold was of Axis origin. At the heart of this early intelligence was the discovery that the Banco de Portugal maintained four accounts at the BNS: Accounts 'A', 'B', 'C' and 'D'. Allied intelligence and postwar account reconstructions have strongly tended to the view that it was accounts B, C and D that the Banco de Portugal employed to carry looted gold from Berne to Lisbon, leaving account A as a legitimate account. For example, a British intelligence report dated August 31, 1944, noted that "intercepts" of coded Portuguese communications indicated that "the gold sold by the Bank of Portugal came from their Account 'B' at the National Bank of Switzerland, i.e., gold with a German taint."⁽¹⁴⁾ Another Allied intelligence document dated September 22, 1944, recapitulated the "estimated gold holdings of the Bank of Portugal at 30/6/44" and indicated that the overwhelming amount of gold "consigned to Portugal" was both received from the BNS and drawn to Lisbon from accounts B and C. Account A registered minimal activity with few transfers to Lisbon and was generally believed to contain only clean (i.e., untainted by Axis looting) gold.⁽¹⁵⁾ While this pattern of moving gold from the Reichsbank through Switzerland to Portugal was only imperfectly understood by Allied intelligence in 1942, its contours were of course familiar to the BNS.

Despite the fact that the Anglo-Swiss gold-for-francs swap was in place in the late spring of 1942, Switzerland still faced the same monetary dilemma. It needed gold and hard currency to secure the internal stability of its wartime economy. Negotiations in London had bogged down. Thought therefore soon turned to ways in which the BNS's mounting store of gold in Ottawa could be directed to Switzerland's monetary advantage. Could the unique hemispheric conditions imposed upon its physical movement and paper transfer be somehow turned to Switzerland's pressing monetary needs? Probably the first test of these conditions came on April 28, 1942 -- just a month after the first swap payments entered the BNS's new Ottawa account -- when the BNS cabled the Bank of Canada to inquire whether a Swiss commercial bank could open a gold account at the Bank of Canada or if the Bank of Canada would permit the BNS to make a "special gold deposit with your bank under our name," such deposit to contain gold owned by a Swiss commercial bank.⁽¹⁶⁾ The request aroused suspicion in Ottawa. Towers cabled Norman: "We are curious to learn what gave rise to the inquiry and before replying should like to know if you have any comments to offer."⁽¹⁷⁾ The Bank of England answered that the purpose of the BNS proposal was all too apparent. Dual BNS-Swiss commercial bank accounts in Ottawa would allow the BNS to transfer gold between these accounts in Ottawa, thereby allowing it to take gold in Switzerland acquired by commercial banks (possibly from dubious sources) in return for legitimate, but blocked, gold in Ottawa. When Graham Towers referred the proposition to Montagu Norman in London, the Bank of England adamantly opposed the idea. "We suspect," Norman told Towers, "therefore that the Swiss National Bank proposes to buy gold in Switzerland irrespective repeat irrespective of origin and ownership either against sales of equivalent gold in Canada or a guarantee to the original holders that the Swiss National Bank holds the equivalent gold in Canada to his order."⁽¹⁸⁾ On May 2, the Bank of Canada shut the door on the Swiss proposal: "...under existing conditions there are obstacles in the way of acceptance of gold deposits unless the gold is the property of a central bank. We regret to advise therefore that it would not be possible for us to enter into the arrangements mentioned in your cable."⁽¹⁹⁾ In the wake of this decision, Governor Towers sent a circular letter to all the general managers of Canada's chartered banks requesting that, in deference to the government's wishes, they not open any new gold earmark accounts for non-residents and to report any such requests to the Bank of Canada.⁽²⁰⁾

The BNS, however, already had an alternative plan. A week before it tested the idea of opening Swiss commercial accounts in Canada, it instructed the Bank of Canada on April 21 to transfer 64,395 fine ounces of its newly acquired Ottawa gold to the Banco de Portugal. There was evident haste in this request, since the Portuguese did not as yet officially have an account in Ottawa. Once that was accomplished on April 30, the BNS cabled that another 64,837 fine ounces be credited to the new Portuguese account.⁽²¹⁾ Thus, within two weeks the Banco de Portugal acquired four tons of earmarked gold in Ottawa. Since the transfer took place entirely within the terms of the original Anglo-Swiss swap -- the gold had been transferred to the central bank of another European neutral power and was not physically leaving the country -- the deal occasioned no undue comment in Ottawa. Ottawa had no idea of what transaction lay behind the BNS transfer of four tons of its gold; the host to an earmark account merely followed the instructions transmitted to it by a client.

What the Canadians did not know was that in releasing four tons of its blocked gold in Ottawa the BNS had gained possession of a similar quantity of gold in Berne out of the account of the Banco de Portugal. Four tons of gold blocked in Ottawa until the end of hostilities had thus been in effect patriated to support the domestic requirements of the BNS. Before striking this deal with the Portuguese, the BNS had offered a similar trade to the Riksbank in Stockholm, but the Swedes had declined.⁽²²⁾ Why would the Banco de Portugal facilitate such a deal, a deal that left it with gold blocked on the other side of the Atlantic? Principally because the Swiss were prepared to pay a hefty commission to Portugal on the transaction. The initial April 22 swap of Berne-for-Ottawa gold cost the BNS a commission of 1.5% and the subsequent swap of May 5 brought a handsome 2.5% commission. The total cost of these commissions was therefore approximately SFr428,000 on a total sale of SFr19.56 million. At its May 7, 1942, meeting, the BNS board of directors decided that such commissions were "too high to speak about" and dropped any further thought of swapping gold in Ottawa with the Portuguese.⁽²³⁾ With negotiations with the British still producing little result in London, the BNS was still sorely tempted to parlay its Canadian earmarked gold to its European advantage. In June of the same year, the BNS learned that the Riksbank in Stockholm had an ample supply of Swiss francs -- francs that if patriated could serve the BNS's domestic needs -- and suggested that it might acquire some of these by selling the Swedes gold in Ottawa. The Swedes declined. Later, in September 1942, the BNS sent a delegation to Lisbon and then Stockholm again to try to swap gold in Ottawa for francs in neutral Europe. Again, the offer was spurned; blocked gold in Ottawa had little allure.⁽²⁴⁾

What was the origin of the four tons of Portuguese gold that the BNS now owned in Berne? Nobody in Ottawa asked. Neither did the Bank of England signal any interest in its cables across the Atlantic, even though it was acting as the Banco de Portugal's intermediary with the Canadians. Others closer to the actual swap did wonder. On June 25, 1942, David Eccles at the British Embassy in Lisbon expressed his concerns to the Ministry of Economic Warfare -- the arm of government that oversaw Britain's economic campaign against the Axis -- in London:

You will have seen the reports of the gold transactions between the Bank of Portugal and the Swiss National Bank, whereby the former has taken over gold belonging to the latter in Canada and released to the Swiss Bank gold belonging to itself deposited in Switzerland. As far as I can see, the result of this transaction is to free gold in Switzerland, which the Germans may now get hold of. I do not know what are the estimated quantities of gold available to the Germans, but if there is any question of their running short for purposes including selling gold to Bank of Portugal, I should have thought we should not allow the Bank of Canada to oblige the Bank of Portugal in this way.

As I am not properly informed on this subject it may well be that such transactions are completely harmless....⁽²⁵⁾

In July, Eccles got a reply from the ministry in London, a reply that reiterated the facts of the original deal struck between the British and the Swiss:

...The Swiss refuse to give us Swiss francs against dollars which they could not use to pay to Portugal .... since they have to meet a deficit with Portugal. It is essential for us to obtain the Swiss francs we need and we reached a provisional arrangement to obtain them against gold in Canada which would be available to be earmarked in Canada for Portugal. As you rightly say, this prevents the Swiss from having to deplete their gold reserves in order to pay Portugal, and naturally we should like Switzerland to have as little gold as possible for fear Germany may get hold of it. However, the Swiss have told us that they do not supply the Germans with gold and we have no reason to believe that up to the present they have done so.⁽²⁶⁾

This interpretation apparently retained its currency in London's corridors of wartime economic power. In August 1944, a memorandum prepared by the Overseas and Foreign Office -- a memorandum that drew on intercepted communications between Lisbon and Berne -- again described the origin of the gold the Portuguese had delivered to the BNS in 1942:

...The Ports obtained gold in Canada from the Swiss, by means of an exchange of Account A * gold against gold in Ottawa.⁽²⁷⁾(* Thought to be without German taint.)

Central to the British reading of the Banco de Portugal's financial dealings with the Banque Nationale Suisse in the mid-war years was the role of its various accounts in Berne. Of the four Portuguese accounts maintained at the BNS -- A, B, C, and D -- British intelligence usually identified account A as a clean account. accounts B and C, however, were usually described as containing "gold with a German taint."⁽²⁸⁾

Historical data released by the Banco de Portugal in April 1997 confirm these patterns. The largest European acquisitions of gold by the Banco de Portugal were made through its accounts B and C at the BNS, the latter account in particular serving as the channel for gold acquired from the Reichsbank. Gold movement figures for 1942 show two entries representing the transfers involving Canada: one for 2,018 kilos of gold in account A and a second for 2,000 kilos of gold in account B. This obviously partially contradicts the opinion in the minds of British commentators that all the 1942 transfer was drawn out of account A. Nonetheless, none of the transactions appears to have been drawn from the principal Reichsbank account, C.⁽³¹⁾ The commentary accompanying these statistics -- commentary ultimately presented to the United States House of Representatives Committee on Banking and Financial Services in June 1997 -- stressed the fact that gold accumulation was Portugal's preferred strategy of shoring up the external value of Portugal's currency, a policy applied to "any trade partner of the country."⁽³²⁾ It is perhaps worth noting that, when the war in Europe drew to a close in mid-1945, Bank of Canada Governor Graham Towers drew up a report on the financial condition of Europe. "Portugal," he reported, "has done well out of the war, accumulating $75 millions and large amounts of gold and dollars."⁽³³⁾

Did Canada unwittingly facilitate the laundering of tainted German gold in the spring of 1942? The swap of Swiss gold earmarked in Ottawa for Portuguese gold deposited in Switzerland was first and foremost driven by Swiss monetary needs, not Portuguese needs. The imperative of bolstering the domestic requirements of the Banque Nationale Suisse in its fight against inflation was the paramount motive in the whole transaction. This need was sufficiently powerful to persuade the Swiss to pay a stiff transfer commission to prod the Banco de Portugal into accepting gold on the other side of the Atlantic that was blocked until the end of the war. The commission was daunting enough to dissuade the BNS from pursuing the strategy of further using its clean gold in Canada as a bargaining chip in its European operations.

It is, however, impossible to say with absolute surety that the four tons of gold that the Banque Nationale Suisse received in Europe from the Banco de Portugal in return for its Ottawa gold was also untainted. Contemporary intelligence reports reflected the belief that the gold was "without German taint." Similarly, the gold appeared to be drawn from accounts maintained by the Banco de Portugal that were not generally perceived as being the principal conduits for tainted Nazi gold on its journey from its looted origins to the mainstream of legitimate finance. If swapping gold in Switzerland for clean, if blocked, gold in Canada was an effective means of laundering gold, why did Portugal not employ the tactic again, especially in light of the handsome commission the Swiss were prepared to pay on the transaction? After all, Switzerland still had gold left in Ottawa after the first four tons had been transferred. But there is no record whatsoever of the Portuguese ever asking to duplicate the transaction. In fact, the Banco de Portugal spurned the BNS's offer of further sales in mid-1942. Other contemporary intelligence reports and a convincing burden of more recent research indicate that Portugal had more expedient ways of moving gold from Germany's borders through Switzerland to Portugal. There was little inclination or incentive to detour through distant Ottawa.

THE PORTUGESE-SWEDISH-SWISS GOLD SWAPS OF 1944-45

During 1942-43, growing recognition that gold looted by the Germans was making its way to the outside world through neutral countries began to be reflected in a tightening of Allied policy around the countries believed to be the middlemen in the trade. In January 1943, 17 nations signed the "Inter-Allied Declaration Against Acts of Dispossession." The signatories reserved the right to declare invalid any transfers of property taken out of countries occupied by the Axis when peace returned to Europe. The declaration did not specifically mention looted gold, but a subsequent February 1944 declaration by the United States emphatically stated that the Americans, who would plainly be the driving force of postwar economic activity, would not recognize title to any gold that had been moved by Germany onto world markets during the war. Britain and the Soviet Union quickly echoed this policy, thereby sending European powers that were alleged to be involved in the trade clear warning that they did so at their own financial peril. In the spring of 1944, the search for Nazi gold became active as the U.S. Treasury Department initiated its Safehaven program, which began to actively seek evidence of the dispersement of looted gold. Despite these early injunctions, the Banco de Portugal's gold reserves continued to rise steadily through the 1943-45 period.⁽³⁴⁾

Paralleling its acquisition of gold, Portugal had been an active participant in European wartime trade, capitalizing on its strength in such strategic minerals as tungsten -- used to harden steel -- while at the same time maintaining its more normal commercial relations with European trading partners. These included trade with its oldest trading partner, Britain, which since signing a 1940 agreement between their two central banks had allowed its trade with Portugal to be conducted in sterling. The settling of Portugal's trade accounts with other countries was, however, less straightforward and more problematic. Once the war had made the reichmark an unconvertible currency, Portuguese trade with the Axis was generally financed by German payment in gold, gold usually transferred through accounts at the BNS in Switzerland. In 1941, the Banco de Portugal and the BNS signed a central bank agreement that regulated payment between the two countries. At the same time, Portugal tried to maintain its commerce with other European trading partners such as Sweden by using gold to secure hard currency to satisfy its accounts with these countries.

Throughout 1944, the turning tide of the war began to have a constricting effect on the financing of inter-European commerce. Looted gold had now acquired a second taint; few would accept it as a form of payment for fear that its true origins might be exposed with the peace, opening the way to Allied confiscation. At the same time, countries with large holdings of such gold grew anxious to rid themselves of the liability they represented. "Source states," a British secret report copied to the Bank of England warned in August 1944, "that the Bank of Portugal is now endeavouring to liquidate the larger part of its gold holdings in Switzerland. These consist almost exclusively of German gold deposited by the Reichsbank. The Bank intends to utilize these deposits to cover imports from Switzerland or Sweden."⁽³⁵⁾ A month later there were even reports of German gold being transported by air via Spain to Portugal, where it was being smelted into Portuguese ingots and hoarded for "a postwar credit for the Germans."⁽³⁶⁾

Amid these pressures and rumours, Canada again found a small role in the disposition of European gold. It began innocuously enough in mid-February 1944, when the Swedish ambassador to Canada paid a visit to the Bank of Canada and inquired into opening an account in Ottawa. The purpose of the account would be to allow the Sveriges Riksbank, Sweden's central bank, to buy newly minted Canadian gold and earmark it in Canada. This the Canadians were prepared to do at six cents below the going New York rate (a discount made possible by the saving of not shipping the gold to New York) as long as gold was available surplus to Canada's own needs. Early in March, the Riksbank exercised this option, buying 1,076 bars of Canadian gold for earmark and agreeing to the usual condition that it could physically export the gold only within the Western Hemisphere or transfer the gold on paper to central banks of Allied or neutral European powers.⁽³⁷⁾ The deputy minister of finance routinely approved the transaction. The Bank of Canada had another wartime earmark client. There is nothing in the correspondence covering the opening of this new account to suggest that there was an ulterior motive behind the Swedish initiative. Sweden appeared to be applying some of its U.S. dollar holdings to the acquisition of gold at a good price.

Events in Europe would soon create a new purpose for the account. By the summer of 1944, Portugal was experiencing real difficulty converting gold into the hard currencies that fuelled its trade with Switzerland and Sweden. The exchange rate in Lisbon on the Swiss franc and the Swedish krona climbed in relation to the Portuguese escudo. In early August, the Banco de Portugal instructed local banks not to sell francs or kronor except for the purpose of "conventional transactions." Under the agreement signed in 1941 between the Banco de Portugal and the BNS, the Portuguese had the right to sell gold to the Swiss in exchange for francs that would then be applied to financing Swiss-Portuguese trade. In August 1944, however, the Swiss learned that the Portuguese were in fact selling them gold for francs, but then using the francs to purchase Swedish kronor to finance their trade with Sweden. The Swiss complained and reminded the Portuguese that francs obtained for gold must be fed back into Swiss trade with Portugal. The Banco de Portugal assured the Swiss that the gold sales were to cover commercial transactions with Sweden only. The BNS did allow one such gold-for-francs-for-kronor swap to take place, but then insisted that the Portuguese respect the 1941 agreement. In desperation, the Portuguese then suggested various schemes to the Swedes, all of which revolved around the Swedes buying some form of Portuguese gold in return for kronor or swapping gold in Stockholm for gold in Berne. The Swedes proved very reluctant to accept any form of Portuguese gold in Europe and in fact were even reluctant to acquire escudos. The Swedes may have sensed what Allied intercept had already revealed: the Portuguese were selling gold out of their B account -- gold with German taint.

Confronted with Swiss intransigence and Swedish hesitancy, the Portuguese then struck upon the idea of offering the Swedes gold -- gold that was undeniably clean -- in their Canadian earmark account in return for kronor in Stockholm. The Swedes replied cautiously, asking whether the Portuguese would promise to spend the kronor they received from the deal solely for trade with Sweden. On August 30, the Banco de Portugal, possibly responding to a Swedish request about the quality of the gold in Ottawa, asked the Bank of Canada to reconfirm that the 321 bars in its earmark were "good delivery New York." They were. With these assurances, the Swedes and the Portuguese struck a deal. On September 6, the Banco de Portugal instructed Ottawa to transfer "about 1000 kilos gold fine" -- 32,174 fine ounces -- into the Riksbank account in Ottawa. On October 2, another 500 kilos -- 16,108 fine ounces -- followed.⁽³⁸⁾ In its secret commentary on this swap, the Overseas and Foreign Office attributed Portugal's desperate need for kronor as the principal motive behind the deal. The second, smaller swap in early October was prompted by the exhaustion of the kronor obtained in the first swap. On October 23, the British intercepted a telegram from Banco de Portugal to the Riksbank in which the Portuguese reported that their reserve of kronor was now healthier and that "they may possibly accumulate more kronor than they need." Would, therefore the Swedes be willing to resell the gold, or some part of it, they had acquired in Canada back to the Portuguese? The Banco de Portugal was willing to pay slightly more for the returned gold than they had received from the Swedes in August and October.⁽³⁹⁾ The Riksbank agreed as long as the kronor it received were derived from "normal commercial transactions." The Swedes agreed to honour this reswap arrangement for six months until April 26, 1945, at which time they would be "prepared to reconsider the matter."

On November 21, the new agreement was activated. A half ton of gold was returned to the Banco de Portugal's account. On November 24, another half ton followed. Another half ton was returned on January 5, 1945, in effect cancelling the original 1944 Portuguese-Swedish swap. But then the tide turned again, and in six separate transfers between May 7 and September 13, 1945, the Portuguese transferred gold totalling two and a half tons back to the Swedes in Ottawa.⁽⁴⁰⁾ Coincidentally, the Riksbank bought more gold -- 1,230 bars -- directly from the mint in Ottawa through the minister of finance.

This tango of Portuguese-Swedish swaps was accompanied by one last Portuguese-Swiss swap. On September 19, the Banque Nationale Suisse instructed the Bank of Canada to transfer two tons -- 64,386 fine ounces -- of gold to the Banco de Portugal's earmark account in Ottawa.⁽⁴¹⁾ Unlike the 1942 transactions, the BNS paid a much lower commission on this transfer, only 3/8%.⁽⁴²⁾ Once again, the Portuguese drew their gold from their A account at the BNS, the account generally believed to be without German taint. Statistics supplied by the Banco de Portugal show a withdrawal of 2,001 kilograms from account A in 1944.⁽⁴³⁾ There is no definitive explanation for this Portuguese-Swiss swap. Coming a week after the initial Portuguese swap with Sweden, it may have been occasioned by a Portuguese desire to replenish its Ottawa stock of clean gold in case its need of kronor necessitated further Swedish swaps, swaps that would take it beyond the two tons it by then had left in Ottawa. Hence Portugal's willingness to accept a much lower commission. Switzerland may well have been ready to facilitate this swap because it promised to ease Portugal's liquidity problems, thereby easing Switzerland's own strained exchange situation with Portugal. This is, however, only speculation. The BNS's own recent interpretation of the swap suggests that it "seems to have been motivated by a desire to avoid the cost of transporting gold from Bern to Portugal."⁽⁴⁴⁾

Thus ended Ottawa's role in the wartime swapping of gold between the accounts of the central banks of neutral Europe at the Bank of Canada. The flurry of late-war swaps between Portugal, Sweden and Switzerland passed with absolutely no undue concern in the Bank of Canada. These swaps followed a pattern that was by 1944 familiar to the Bank's officials and to the deputy minister of finance. There is no written record of any sense of anxiety or alarm over the requests for transfer received from European central banks. Similarly, the record of these events supplied to us by the European banks involved -- a record admittedly viewed at arm's length -- gives every indication that the gold swaps that took place through the auspices of the Bank of Canada from 1942 to 1945 were driven by liquidity concerns, not by a desire to launder dirty gold into clean gold. There were too many other ways to pass off bad gold under the cover of war in Europe for the route through Ottawa to ever have had much appeal as a site for laundering gold. Documents can, however, lie. And it is in the nature of human deviousness that some acts are never captured for posterity in the first place. We can never therefore be absolutely sure that Ottawa did not unwittingly aid the designs of gold launderers, but the burden of evidence would suggest otherwise.

On July 11, 1945, the Banque Nationale Suisse cabled Ottawa to inquire whether the restrictions that had since 1942 conditioned their ability to dispose of their Canadian gold were finally in abeyance. Peace, they suggested, meant that "these restrictions have lost their importance." Ottawa concurred, but reminded the BNS that an export licence was still required and that until peace returned in the Pacific some countries were still considered "enemy." No dramatic withdrawals of gold followed. The Swedes, Portuguese and Swiss all kept their Ottawa accounts after the peace.

In 1955, a ripple of horror spread through the Bank of Canada when it was discovered that deep in its vault was a small holding of what were described as "Prussian gold bars." There was concern that the 597 bars might be "tainted" and had somehow slipped into the Bank's gold reserves as a result of wartime acquisition. A quick check revealed that they had in fact come to the Bank as part of a shipment from the Bank of England and that they were tainted only in that they were not "good delivery London" (i.e., they had been manufactured by a prewar German refiner, whose products were, in the 1950s, no longer acceptable in international gold markets). The "Prussian" bars were subsequently remelted and integrated into the supply of acceptable gold.⁽⁴⁵⁾ In a real way, this was probably as close as the Bank of Canada ever got to actual tainted gold crossing the portal of its vault.