Thursday, December 24, 2015

Was Bretton Woods a real gold standard?

John Maynard Keynes and Harry Dexter White, who contributed to the design of the Bretton Woods system

David Glasner's piece on the gold standard got me thinking about the Bretton Woods system, the monetary system that prevailed after WWII up until the early 1970s.

There are many differences between Bretton Woods and the classical gold standard of the 1800s. My claim is that despite these differences, for a short period of time the Bretton Woods system did everything that the classical gold standard did. I'm using David's definition of a gold standard whereby the monetary unit, the dollar, is tied to a set amount of gold. This linkage ensures that there can never be an excess quantity of monetary liabilities in circulation—unwanted notes will simply reflux back to the issuer in return for gold. When most people criticize Bretton Woods, they say that it lacked such a linkage.

A narrowing redemption mechanism

For a gold standard to be in effect, a central bank's notes and deposits have typically been tied down to gold via some sort of redemption mechanism. In the days of the classical gold standard, central banks didn't discriminate; the right to redeem was universal. Whether black or white, big or small, male or female, you could bring your notes or deposits to a central bank teller and have them be converted into an equivalent quantity of gold coin. Any unwanted notes and deposits quickly found their way back to the issuer.

After 1925, the world adopted a narrower redemption mechanism; a gold bullion standard. Economist and trader David Ricardo had recommended this system a century before as a way to reduce the resource costs of running a gold standard. Gold coins were withdrawn from circulation, and rather than offering to redeem notes and deposits in coin, the monetary authorities would only offer bulky gold bars. This excluded most of the population from redemption since only a tiny minority would ever be wealthy enough to own a bar's worth of notes.

It might seem that this narrowing of the redemption mechanism compromised the gold standard. After all, if too many dollars were created by a central bank, and these fell into the hands of those too poor to redeem for bullion bars, than the excess might remain outstanding rather than refluxing back to the issuer.

But consider what happens if a free secondary market in gold is allowed to operate. Unable to send excess notes to the central bank, less wealthy gold owners can sell them in the free market. This would drive notes to a discount relative to their official price, at which point large dealers will buy them and bring them back to the central bank for redemption in bars, earning arbitrage profits. The free market price is thus kept in line with the central bank's redemption price. So as long as the wealthy and not-so wealthy are joined by a market in which they can exchange together, then a narrower redemption mechanism needn't impinge on the proper functioning of a gold standard.

Anyone who has toyed around with ETFs will know what I'm talking about. Despite the fact that a gold ETF limits direct redemption to a tiny population of investors (so called authorized participants), the price of the ETF will stay locked in line with the market price of gold. Authorized participants earn profits by arbitraging differences between the ETF and the underlying, thus maintaining the peg on behalf of all ETF owners. You don't need many of them; just a few well-heeled ones.

When authorized participants don't do their job

The U.S. narrowed the gold redemption mechanism even further when, in 1934, Roosevelt limited redemption to foreign governments. As long as foreign governments and the public were joined by a market, then excess notes could be sold to these governments and returned to the U.S. for gold at the official price. The free market price and the U.S.'s official price would converge and a gold standard would still be in effect.

Things didn't work that way. After it entered WWII, the U.S. continued to buy and sell gold to governments at $35, but gold traded far above that level in so-called free gold or premium markets in Zurich (see chart below), Paris, Beirut, Macau, Tangiers, Hong Kong, and elsewhere. The existence of premium markets continued through the 1940s and well into the 1950s.

While private dealers have incentives to engage in arbitrage, national governments are driven by political motives. Governments no doubt could have earned large profits by buying gold from the U.S. at $35, shipping it home, and selling it in their domestic free gold markets at $45 or $50, but they chose not to, most likely to avoid raising the ire of American officials.

The monetary system in the 1940s and 1950s was a malfunctioning ETF. For various reason the authorized participants (ie. foreign governments) were not arbitraging differences between the price of the ETF and the underlying, and the ETF was therefore wandering from its appropriate price.

With the redemption mechanism compromised, the supply of U.S. monetary liabilities in circulation could exceed the demand, the result being that the market price of dollars sagged to a discount to their official price. Bretton Woods, with its multiple prices for gold, was not a gold standard, at least not yet.

The London gold market

It was only in 1954 that the so-called "free gold" price in Zurich and elsewhere finally converged with the official price of $35. This happened more by accident than purposeful arbitrage conducted via the redemption mechanism. On the supply side, the Soviets were bringing large supplies of "red gold"  to sell on free markets while the South Africans were diverting more gold away from official buyers in order to earn wider margins. At the same time, the end of the Korean War was reducing safe haven demand.

Source: The Economist

Once parity between free gold prices and the U.S. official price was established, the London gold market reopened for business. London had always been the largest gold market in the world, far eclipsing Paris, Zurich, and the rest. Its re-opening had probably been delayed for face-saving reasons. Given that the Brits and the Americans effectively ran the world's monetary system, they could safely ignore premium markets in Zurich and Paris. But the existence of a British gold price in excess of the official price would have been embarrassing. Upon the market's reopening, British authorities limited the ability of locals to buy on the market (they could freely sell) but put no restrictions on the ability of foreigners to participate in the London gold market.

From the time it opened in 1954 to 1960 the London price was well-behaved, staying locked in line with the official price. In October 1960, however, speculators took over control of the London gold market and sent the price of gold to an intraday high of $40, well above its official price of $35. Rather than arbitraging the market by buying from the U.S. Treasury at $35 and selling in London at $40, foreign central banks stepped aside.

The London gold pool

The authorities' response to the 1960 gold crisis is what finally turned Bretton Woods into a real gold standard, at least in my opinion. While the U.S. had ignored premium markets in the 1940s and early 50s, they couldn't ignore a premium market in their own backyard. At the behest of the U.S., the London gold pool was formed. Under the management of the Bank of England, the pool assembled a gold war chest with contributions from the U.S., U.K., Germany, Holland, France, Italy, Belgium, and Switzerland. Whenever gold rose above $35.20, the pool sold gold on the London market in order to keep the price steady. When it fell to $34.80, it bought in order to support the price. The existence of the pool was never officially declared, but everyone knew it was in operation and had the task of setting the London gold price.

This effectively created a functioning gold standard. Before, the only mechanism connecting the public's excess dollars with the U.S.'s gold was the somewhat unpredictable predilection of foreign governments to buy that excess and exercise the right to return it for redemption. Now the public could deal directly with the U.S. government by selling on the London gold exchange to the U.S.-led London gold pool, which guaranteed a price of at least $35.20.

And as the chart below shows, the pool worked pretty well for the next few years, keeping the price of gold in a narrow range. However, the devaluation of the British pound in 1967 and the departure of the French from the gold pool shook confidence in the $35 peg. The system imploded in March 1968 when a steady jog into gold accelerated into an all out run. Rather than continue to bleed gold to speculators, the London gold pool disbanded and the price of gold in London shot up to over $40, well above the official price of $35.20.

But from 1961 to 1968, the world pretty much had a gold standard. Or, put differently, thanks to the opening of the London gold market and the arming of the London gold pool, the world's monetary system between 1961 and 1968 did pretty much everything that the  gold standard of the 1800s did. After 1968, the U.S. dollar slid back into its earlier Bretton Woods pattern of having more than one price in terms of gold; the $35 official price and the "free" London price. This was no gold standard. When Nixon famously dismantled the already-narrow redemption mechanism in 1971, most of the damage had already been done.


  1. Bretton Woods was not a gold standard, thank God, because there was relatively un-traumatic way for countries to adjust the internal and external value of their currencies. Not as good as floating rates, but better than a "gold standard." Too bad the Eurozone countries gave up this advantage.

    1. In a gold standard with token coinage, a country can simply devalue their currency relative to gold.

      In a gold (and or silver) coin standard, kings re-rated the coinage all the time. Some of the more famous inflations occurred on metallic coin standards.

  2. Excellent post.

    Could I say the London Gold Pool was operating as the world's central bank?

    1. Thanks Nick.

      I think that's right. The London gold pool enforced the U.S. dollar peg, and every other nation pegged to the dollar.

  3. In your (and David's) definition of "a gold standard ... the monetary unit, the dollar, is tied to a set amount of gold." You infer that "This linkage ensures that there can never be an excess quantity of monetary liabilities in circulation—unwanted notes will simply reflux back to the issuer in return for gold"

    It's far from clear that the gold standard as we knew it up to 1914 actually operated in this way at all. Convertibility ensured that an excess circulation would face a forced reduction in a relatively short time period (i.e. counted in months). The Bank of England's commitment to convertibility ensured that it was not rational for a speculator to challenge the peg, even when there were some divergences in prices -- the only outcome would be to speed the pace of the forced reduction that was coming anyhow. The claim that the gold standard implies that "there can never be an excess quantity of monetary liabilities in circulation" is entirely contradicted by the facts as we know them. One of the main effects of 19th century crises was the readjustment of the quantity of "monetary liabilities in circulation," which as you frame the issue could never have been necessary.

    The Bank of England had learned in the crises of 1847 and 1858 that it was indeed possible for an excess of monetary liabilities (i.e. bills of exchange) to circulate and that the forced reduction when it came tended to be ugly. Thus, it was obsessed during the gold standard period with the control of monetary liabilities by controlling the quality of bills that could circulate (and later of the advances that banks issued). This focus on the quality of bank credit lines was designed to support the gold standard by ensuring that the money supply was growing together with GDP and that the money supply was not allowed to grow too fast and generate inflation, which in short order would be forced by the gold standard to be reversed and almost inevitably to cause a crisis.

    I'm not sure how the view that the 19th century gold standard could operate without a central bank actively monitoring the growth of the money supply arose, but it is not at all supported by the behavior of the Bank of England during this period. (I have a paper coming out in the Financial History Review that discusses this in some detail.)