United States: Swaps to the Bank for International Settlements and Deutsche Bundesbank, 1967

The devaluation of British sterling in November 1967 caused major disruptions in currency markets; led to concerns that the US dollar, the lynchpin of the global gold standard system, would also devalue; and threatened the stability of financial markets, particularly the market for US dollars overseas. The Federal Reserve and European central banks used a network of preexisting swap lines in the ensuing weeks to stabilize exchange rates, defend the gold standard, and calm global markets. In most cases, central banks used these swaps to stabilize exchange rates. However, the main purpose of several of these swaps was arguably to ensure US dollar liquidity in markets outside the United States in the wake of the sterling devaluation. For that reason, they are relevant to the New Bagehot series, which seeks to draw lessons about crisis-management tools that promote financial stability. This case focuses on the Federal Reserve’s collaborations with the Bank for International Settlements (BIS) and West Germany’s central bank, the Deutsche Bundesbank, to provide dollar liquidity to the eurodollar market at the end of 1967. In these arrangements, the parties used the swaps to acquire dollars that they could then lend to European banks. The BIS drew USD 346 million on its line with the Fed; in an unusual arrangement, the Fed provided USD 300 million to the Bundesbank by drawing on its swap line. The BIS fully repaid all swap obligations by January 1968. The Fed fully repaid all swap obligations to the Bundesbank by March 1968. The Federal Open Market Committee authorized the expansion of the BIS and German facilities to USD 1 billion on March 17, 1968.

Keywords: Bank for International Settlements, BIS, Bretton Woods, central bank swap line, Deutsche Bundesbank, eurodollar, Federal Reserve

On November 19, 1967, the United Kingdom devalued sterling by 14.3% against gold. The devaluation caused major disruptions in currency markets; led to concerns that the US dollar, the lynchpin of the global gold standard system, would also devalue; and threatened the stability of financial markets, particularly the market for US dollars overseas (McCauley and Schenk 2020).

Global central banks used a network of preexisting swap lines in the ensuing weeks to stabilize exchange rates, defend the gold standard, and calm global markets. In most cases, central banks used these swaps to stabilize exchange rates.

However, the main purpose of several of these swaps was arguably to ensure US dollar liquidity in markets outside of the United States in the wake of the sterling devaluation. For that reason, they are relevant to the New Bagehot series, which seeks to draw lessons about crisis-management tools that promote financial stability.

This case focuses on the Federal Reserve’s collaborations with the Bank for International Settlements (BIS) and West Germany’s central bank, the Deutsche Bundesbank, to provide dollar liquidity to the eurodollar market at the end of 1967. In these arrangements, the parties used the swaps to acquire dollars that they could then lend to European banks. Five other central banks participated in swaps with the Fed in late 1967, but the main purpose of most of these transactions was to defend the dollar (see Key Design Decision No. 2, Part of a Package).

We focus on the Fed’s swaps with the BIS and Bundesbank because their stated purpose was to provide liquidity to eurodollar markets.^FNThe eurodollar market, which now refers to any non-US market for US dollars, came into existence in 1957, and by 1962 it had become an interbank dollar market in London. It was largely an avenue for regulatory arbitrage, initially to avoid Bank of England regulations, but later to avoid US reserve requirements, bank deposit yield cap regulations, and deposit insurance costs. US bank branches paid higher rates in the eurodollar market than in domestic (US) money markets (McCauley and Schenk 2020). Moreover, the Fed described these transactions as having been particularly impactful in lowering eurodollar rates, arguably for financial stability reasons (Coombs 1968a).^FNBecause of the fixed exchange rate system, drawing hard boundaries between the purposes and uses of swap lines during the Bretton Woods era is difficult: pressures in the eurodollar market affected the dollar exchange rate, and liquidity provision to those markets was part of foreign exchange intervention (McCauley and Schenk 2020). Indeed, even in the case of the explicitly eurodollar market–focused BIS swap line, the Fed used the line “particularly when the [eurodollar market] was creating special problems in the exchange market” and with dual consideration for the defense of sterling (to shield sterling from pressures in the eurodollar market) (MacLaury 1969). Inasmuch, the Fed’s swap operations with the BIS and Bundesbank also served to counter feedback into the US market and to safeguard monetary stability. For a discussion of the use of central bank swap lines in foreign exchange (and thereby gold) intervention in the context of fixed exchange rates, see Appendix A.

The strains in eurodollar markets were partly seasonal. Swiss commercial banks, wanting to show Swiss franc assets at year-end, would sell dollars for Swiss francs to the Swiss National Bank (SNB), leaving a glut of dollars at the SNB^FNIn 1962, the Fed established a dollar-for-Swiss franc swap line with the SNB that could be used to absorb the SNB’s excessive amounts of dollars through cover operations. The swap line was for up to USD 100 million to provide dollar liquidity to the SNB in exchange for Swiss francs. The majority of the Fed’s cover operations through 1971 were with the SNB (Bordo, Humpage, and Schwartz 2014). but a shortage in the market, drying up dollar liquidity and thereby pushing rates higher. Foreign exchange–related (and in this case sterling-related) tensions could also result in eurodollar rate spikes, since the dollar was a vehicle currency (see Appendix A). When, for example, an investor wanted to sell sterling for Swiss francs, they would first sell sterling for dollars and then sell dollars for Swiss francs, leaving the SNB with dollars and removing dollar liquidity from the market (Bordo, Humpage, and Schwartz 2014).

Another possible reason for elevated eurodollar rates was a risk premium for holding dollars. In other words, depositors—worried about a possible dollar devaluation—demanded higher returns for holding dollars. Our research was unable to uncover whether the dollar-liquidity-draining currency phenomenon or the dollar risk premium phenomenon was primarily the cause of the 1967 year-end eurodollar market rate spike, and it is entirely possible that the driving force was a combination of both.

The devaluation of the pound sterling in late 1967 significantly aggravated market conditions. Investors feared that speculation against sterling would result in contagion and additional pressure would mount against the dollar (Bordo, Humpage, and Schwartz 2014; Bordo, Monnet, and Naef 2017). The resulting market outflows from the dollar and into continental European currencies pushed eurodollar rates even higher. Elevated eurodollar rates also threatened spillover effects in foreign exchange markets and domestic (US) money markets (Coombs 1968a; FOMC 1967b; McCauley and Schenk 2020).

Figure 1 shows eurodollar yield curves at three inflection points in November and December of 1967. Figure 2 shows a stylized subset of the same data, specifically for the November rate spike that occurred before market tensions eased.

Source: FOMC 1967b.

Figure 2: Eurodollar Rates, November 17 and 28, 1967

Source: FOMC 1967b.

In 1967, the Fed maintained a dollar-for-European-currency swap line with the BIS (the BIS swap line, originally entered into in 1965), which permitted the BIS to request dollars in exchange for equivalent amounts of deutsche marks (DM), Austrian schillings, Belgian francs, French francs, Italian lire, Netherlands guilders, Swedish kronor, or Swiss francs as well as Canadian dollars, Japanese yen, and pounds sterling (Fed 1968). In practice, most of the BIS draws on the swap line involved the BIS drawing dollars against marks (McCauley and Schenk 2020).^FNThe Fed also had in place a first swap line with the BIS that solely permitted the exchange of dollars for Swiss francs. Federal Reserve Bank of New York (FRBNY) Vice President Charles Coombs said in a confidential 1965 memorandum to the Federal Open Market Committee (FOMC) about talks with the BIS over the European-currency swap line that the fact that the (then only) BIS swap line was denominated in Swiss francs was “partly a matter of geographic accident; the BIS happened to be in Switzerland” (Coombs 1965, 2). He said that there was “no inherent reason” that the BIS couldn’t obtain other European currencies and serve to supplement existing swap facilities, such that the Fed could obtain foreign currency liquidity from the BIS when it had already drawn its maximum from bilateral swaps with European central banks (Coombs 1965, 2–3). The BIS swap line was intended to be used for eurodollar market interventions.

In response to the Fed’s request that the BIS reactivate its European swap line, the BIS initially agreed that “no more than a token intervention” in the eurodollar market was appropriate (FOMC 1967a, 6). However, conditions continued to deteriorate, and between November 13 and December 27, the BIS drew a total of USD 346 million from the European swap line in 10 draws, mostly with maturities of one month. Figure 3 shows the BIS’s swap draws with associated repayment periods and interest rates.

Figure 3: BIS Draws on FRBNY Swap Line, November–December 1967

^A Date is value date, all in 1967.

The BIS then lent dollars downstream to eurodollar market participant banks via deposits at rates above the rate it paid to the Fed. While we found no explicit reference to risk premiums in 1967, in similar 1966 year-end eurodollar market interventions, the BIS received market rates on its downstream lending to commercial banks but paid a lower rate to the Fed in recognition that it was taking uncollateralized credit risk in the form of deposits (whereas collateralized downstream lending is typical in central bank swap line arrangements) (McCauley and Schenk 2020).

In the autumn of 1967, as strains in the eurodollar market developed, the Bundesbank entered into dollar swaps, with commercial banks, worth USD 860 million, USD 600 million of which it channeled into the eurodollar market.^FNIn the spring of 1967, Germany maintained “plenty of dollars” but pledged voluntarily to avoid exchanging those excess dollars for gold at the US Gold Window (McKinsey 1967, 14). Half of the Bundesbank’s eurodollar market interventions, USD 300 million, were funded from its swap line with the Fed, which the parties had originally entered into on August 2, 1962, and which in November 1967 was authorized for up to USD 400 million. Unlike most of the Fed’s swap lines, the Fed was the drawing party under the swap line and provided dollars to the Bundesbank that it used for the eurodollar funding operations. These, according to an FRBNY report, were “especially effective in bringing down [eurodollar] rates” (Coombs 1968a, 51; Bundesbank 1968). Since the Fed was the drawing party, it received West German marks from the Bundesbank and deposited dollars as collateral in the Bundesbank’s account at the Fed, which the Bundesbank then moved into West Germany to lend downstream; normally, collateral currency is held by the lending central bank (see Figure 6 at Key Design Decision No. 2, Part of a Package) (Leimone 1967). Our research was unable to uncover whether the Bundesbank action was prompted by a request from the Fed, but the direction of the transaction (Fed drawing) indicates that the Fed paid the Bundesbank for the swap transaction rather than the Bundesbank paying the Fed (which would be expected, given that the Bundesbank used dollars for downstream lending) (see Key Design Decision No. 10, Downstream Use of Borrowed Funds).

From July to November 1967, the Fed increased the sizes of its swap lines with the BIS and Bundesbank by USD 300 million and USD 350 million, respectively, as shown in Figure 4. In March 1968, the Fed authorized expansion of each swap line to USD 1 billion (Coombs 1968b).

By January 1968, the BIS had fully repaid the USD 346 million obligation outstanding as a result of the year-end operations. By March 8, 1968, the Fed had fully repaid its commitments under its drawings on the Bundesbank swap line (Coombs 1968a). The Federal Open Market Committee (FOMC) voted to allow both lines to lapse in its November 17, 1998, meeting, when Board of Governors members described many of the Fed’s various central bank swap lines as “increasingly obsolescent” in light of years of disuse and the advent of the European Central Bank (Fed 1999; FOMC 1998, 36).

The head of foreign exchange operations for the FRBNY reported that the central bank actions helped to “induce reflows . . . of hot money” into continental European currency markets (Coombs 1968a) and that, “in response to [BIS] smoothing operations, the [eurodollar] market continued to function efficiently, with no more than a normal seasonal rise in rates” (Coombs 1968a, 40). The report also said that coordinated actions with the Bundesbank in November were “especially effective in lowering [eurodollar] rates—from nearly 7 percent per annum for three months to about [6.25] percent by early December” (Coombs 1968a, 51). In its 1967 annual report, the Bundesbank said that its dollar-providing operations “contributed to the relaxation of tensions in Euro-currency markets” (Bundesbank 1968, 38). In its 1967 annual report, the BIS said that eurodollar rates fell as a result of the interventions (BIS 1968).

Some scholars have said that the Fed’s legal mandate to partake in foreign-exchange operations in the 1960s and 1970s was unclear (Bordo, Humpage, and Schwartz 2014).

Scholars at the BIS have said that “substantial” BIS dollar lending in November “did not seem to succeed” in signaling forthcoming dollar liquidity from the Fed and the Bundesbank (McCauley and Schenk 2020, 32). Those scholars said that the eurodollar operations “seemed to exert a more muted impact, but also faced a greater challenge” (from the sterling devaluation) as compared with previous Fed-BIS eurodollar interventions (McCauley and Schenk 2020, 32). On the whole, though, the BIS scholars said the eurodollar interventions were largely successful in limiting eurodollar rate spikes, even in the context of “a greater challenge” because of pressure on the dollar resulting from the sterling devaluation (McCauley and Schenk 2020, 32).

Federal Reserve Board Governor Sherman Maisel was “concerned over a seeming conflict” between the Fed’s BIS eurodollar operations, which promoted credit overseas, and the Fed’s domestic goal of limiting reserve expansion to tighten credit (MacLaury 1969, 1).

Figure 5 shows eurodollar rates in the relevant scope of the swap operations.

Source: McCauley and Schenk 2020.

Appendix A

Federal Reserve Swaps and Foreign Exchange Intervention in the Bretton Woods Era

The Gold Standard and Fixed Exchange Rates in Bretton Woods

Under the Bretton Woods system, exchange rates were fixed to one another, ultimately to the US dollar (USD), which itself was pegged to the price of gold. Dollars were convertible on demand for gold at the Treasury at the official peg of USD 35 per ounce. Since the advent of the Bretton Woods system after World War II, the United States ran a widening balance of payments deficit, which, according to US authorities, was at least partially the result of post-war reconstruction and security guarantees. This deficit resulted in a ballooning ratio of external dollar liabilities to gold stock, which put pressure on the gold peg^FNThis is because an increasing ratio of dollar liabilities to US gold stock implied that the US may not have been able to convert all its dollar obligations into gold if needed (Bordo, Humpage, and Schwartz 2012). Some scholars have suggested alternative explanations for the collapse of the Gold Standard and Bretton Woods. For a dissenting view on this narrative, see Bordo and McCauley 2019. (the resolution of which would be a devaluation) (Bordo, Humpage, and Schwartz 2014). Figure 7 shows the US gold stock and external dollar liabilities from 1951 to 1975.

Source: Bordo, Humpage, and Schwartz 2014.

In sum, the US dollar and its gold peg was the foundation of the fixed exchange rate system, but both the dollar and sterling faced increasing pressure to devalue as a result of balance of payments pressures. When fear spread about the ability of nations to maintain their pegs, fund flows moved (1) out of fiat currencies and into gold, and (2) out of troubled currencies and into safe ones (frequently the Swiss franc), often leaving a dollar shortage in the process (and in turn leaving the safe currency-issuing central bank with an excess of dollars).

Federal Reserve Swaps During Bretton Woods

In March 1962, the US set up a network of reciprocal swap lines^FNPrior to the Fed’s network, the Treasury had established swap lines—which, in contrast to the Fed’s, were not all reciprocal—with foreign monetary authorities since 1936 on a case-by-case basis (Bordo, Humpage, and Schwartz 2014; 2015). through the Fed to forestall a run on the US gold stock after the Treasury’s Exchange Stabilization Fund (ESF)^FNThe Treasury historically used the ESF to intervene in foreign exchange markets, but it was limited through the Congressional appropriations process. The Fed, on the other hand, had more “elastic capacity” to obtain foreign exchange (Bordo, Humpage, and Schwartz 2014, 3). had reached its limits. From that point on, the swaps network became the “first line of dollar defense” (Bordo, Humpage, and Schwartz 2014, 3). The Fed used its swap lines^FNThe Fed also, on an ad hoc basis, from time to time engaged in third-currency swaps, wherein it swapped one non-dollar currency to obtain another foreign currency, but these swaps were not a part of the standing network (Leimone 1967). for three distinct (though related) purposes: (1) to provide cover for unwanted dollar inflows at foreign central banks; (2) to intervene in foreign exchange markets; or (3) to provide dollar liquidity to overseas markets (the focus of this case study) (Bordo, Humpage, and Schwartz 2014). (For a discussion of dollar liquidity provision to overseas markets, see Overview.)

1. Cover Operations. When a foreign central bank received unwanted dollar inflows, it could exchange them at the US Gold Window for gold, draining down the US gold stock. In such a case, the Fed defended the US gold stock through “cover operations,” in which the Fed activated a swap to obtain foreign currency with which to absorb temporary unwanted surges in a foreign central bank’s uncovered dollar balances, so as to obviate the need for the foreign central bank to convert those dollars for gold at the Treasury (Bordo, Humpage, and Schwartz 2015, 152). A cover operation took place in two distinct transactions: (1) the Fed swapped dollars for a foreign currency, for example, deutsche marks, receiving deutsche marks and providing dollars; (2) the Fed then sold that foreign exchange back to the foreign central bank for dollars. The foreign central bank ended up with the same amount of dollars as before the transaction, but those dollars were now protected against foreign exchange risk. The risk to the Fed was that if, in this example, the Bundesbank chose not to renew the swap, the Fed would have to come up with the deutsche marks by purchasing them in the market or obtaining them from the Treasury^FNFor the Treasury to obtain foreign currency, it sometimes resorted to the issuance of so-called Roosa Bonds (named after Undersecretary of the Treasury Robert Roosa), which were foreign currency–denominated Treasury bonds. For example, in 1962, the Treasury issued three-month certificates in Italian lira worth USD 150 million. Outside the context of swap line backstops, the Treasury could use them to similar effect in diminishing the desire of foreign banks to convert dollars for gold: for a foreign central bank, holding non-dollar US debt protected them against a dollar devaluation (Bordo, Humpage, and Schwartz 2015). (Hetzel 1996). These operations protected the US gold stock. From the beginning of the Fed’s swap network to 1971 (when the US closed the Gold Window), the Fed drew USD 11.6 billion for cover operations (Bordo, Humpage, and Schwartz 2014).
2. Foreign Exchange Intervention. A foreign central bank could draw dollars from a Fed swap line and use those dollars to buy up its currency—reducing supply and driving up the exchange rate—to defend an exchange rate (which was always a peg in the Bretton Woods context^FNFor example, in June 1962, the Bank of Canada drew down its new swap facility with the Fed and used the dollar proceeds to defend its exchange rate parity (Bordo, Humpage, and Schwartz 2015).) (Bordo, Humpage, and Schwartz 2015; Leimone 1967).

In 1962, the Fed entered a dollar-for-Swiss franc (CHF) swap line (the “CHF swap line”) with the BIS. This line was initially meant to provide dollar liquidity to the BIS, which would in turn provide such liquidity to the eurodollar market (for a background on the eurodollar market, see Appendix B). The CHF swap line was a tool to bypass Swiss regulations limiting the SNB’s ability to lend (including via swaps) to foreign central banks. This swap line was essentially a pass-through to the SNB, wherein the BIS swapped gold with the SNB for Swiss francs and then, in a second step, swapped Swiss francs with the Fed for dollars.

From 1962 to 1971, the BIS was a frequent user of Fed swap lines, accounting for 15% of all draws during this period (Bordo, Humpage, and Schwartz 2014).

Originally, the Fed made a clear distinction that the European currencies swap line should be used for eurodollar market operations, not for foreign exchange interventions. In 1966, the FOMC authorized the BIS to use the European swap line for eurodollar interventions, after which point it became the primary funding source for BIS intervention in the eurodollar market (for background on the eurodollar market, see Appendix B) (Bordo, Humpage, and Schwartz 2015; McCauley and Schenk 2020).^FNIn 1967, the Fed also drew on its CHF swap line with the BIS, but this is treated separately as a related but distinct cover-providing intervention and not part of the effort to manage dollar liquidity and interest rates (see Key Design Decision No. 2, Part of a Package). Figure 8 compares the changing uses of the different BIS and SNB swap lines with the Fed in the 1960s.

Scholars have said that the cover operations via the Fed’s swap network were ultimately ineffective:

Forms of Foreign Exchange Intervention

The Fed and the Treasury used foreign exchange tools to intervene in the foreign exchange markets and defend gold parity, including spot operations, forward operations, IMF drawings, the issuance of foreign currency-denominated Treasury bonds, and the use of the Gold Pool. In 1961, the US Treasury began to use the ESF for foreign exchange intervention in defense of the dollar, but by 1962 it had reached its resource limits and thereafter swaps became the first line of defense for the dollar (the US continued to make use of the Gold Pool) (Bordo, Humpage, and Schwartz 2014).

The Gold Pool

In the London gold market, buyers and sellers could engage in the private sale of gold (in other words, not through redemptions at government gold windows). As a result, deviations of the London gold market price and the official gold pegs created arbitrage opportunities for central banks (one could, for example, buy gold on the London gold market and then redeem it for dollars at the US Treasury’s Gold window). In order to defend gold parity, eight central banks pooled gold reserves (the Gold Pool) in the London gold market in order to defend the 35-dollars-per-ounce peg (and thus their reserves) and safeguard the stability of the international monetary system (Bordo, Monnet, and Naef 2017).

According to scholars, the Gold Pool functioned successfully for years, but ultimately the November 1967 sterling devaluation caused its collapse. Gold Pool interventions accounted for three-quarters of the US gold stock losses during the 1967 year-end operations (Bordo, Monnet, and Naef 2017).

Sterling Devaluation Crisis, 1967

The US had numerous incentives to support sterling and help the UK avoid a devaluation: (1) the US estimated costs of USD 1-2 billion in losses to the US gold reserves; (2) a larger devaluation could force the US to suspend its own gold convertibility, thus ending the Bretton Woods system; and (3) the US linked sterling support with British military presence in Southeast Asia, which it wanted to remain^FNThe relationship between American support for sterling and British military presence in Asia was documented in archives. A US national security adviser said, “George Ball really put it to the British on Singapore and our support of the pound,” and a scholar said that there was a “formal quid pro quo” between the two governments about American support for sterling and British military presence east of the Suez (Schenk 2010, 168). An American government official informally suggested that “a battalion would be worth a billion dollars of sterling support”; he later revised this comment (Schenk 2010, 168). (Schenk 2010, 157–67).

US authorities recognized a relationship between pressure on sterling and further speculation about the strength of the dollar peg. A scholar of sterling’s decline said that “sterling’s fortunes were of crucial importance . . . in Washington because of the potential for contagion from speculation on sterling to spread to the US dollar . . . .” (Schenk 2010, 157).

Appendix B

Note: Negative amounts (in blue) indicate draws by the Fed, while positive amounts (in red) indicate drawings by European counterparties (so that the total amounts indicate the aggregate as an asset or a liability of the Fed).