The dollar’s rise to a seven month high against the euro in early February provoked sharply lower gold and base metals prices. Yet from its trough on 25 November 2009, the dollar’s latest rally has seen it rise by just 7% - not much by historic moves. Could this be the start of a dollar bull period? If so, much bigger shifts up in the dollar’s strength can be expected over the next few years. Previous such dollar rallies have seen a rise of between 20% and 56%. If that sort of rise happened, what might happen to metal prices?
Using a broad gauge of the US dollar’s strength, the Federal Reserve’s trade weighted major currencies index, it is clear that the US dollar has seen a slow decline in its external value over the past 30 years. However, this decline has been erratic, punctuated by (sometimes-lengthy) periods of dollar strength.
We have identified since 1980 two great dollar bull markets, two significant bear markets, and, more recently, two shorter periods of dollar bull and bear markets.
The first bull market occurred between September 1980 and February 1985, when the dollar rose on its major currencies index from 93.3 to 145.1, an increase of 56%. The second was from June 1995 to January 2002, when it rose from 80.6 to 112.6, an increase of 40%. The two periods of dollar decline followed those booms, and extended from February 1985 to December 1987, when the dollar fell an astonishing 40% in two and a half years (partly due to the September 1985 Plaza Accord, which saw concerted government forex intervention), and then much more recently from January 2002 to April 2008, over which period it fell 38%.
From then there was a short but sharp dollar recovery, which saw it rise by 21% from July 2008 to February 2009, and then another bear market, in which it slipped 15% from March 2009 to November 2009. Since then it has risen by about 7%, suggesting another bull run might be developing.
What will happen to metals’ prices if we see another major dollar rally, similar to those of the recent past? Clearly, the dollar price of metals does badly when the dollar is rising, with only zinc in the first dollar bull period, platinum and palladium in the second, and gold in the third, managing to rise in price. All of the other metals in each period saw substantial price falls.
Similarly, in the dollar bear markets most Of course correlation does not mean causation, and, as always, there are four possibilities. As well as (1) our conjecture that the dollar move is causing the metal prices’ move, the metal prices’ move could be causing the dollar moves (2); another common factor could be causing both to move (3), or it is simply a matter of coincidence (4).
We can probably rule out (2), as the metals’ markets, even gold and copper, are simply not large enough to impact on currency movements, particularly the US dollar. Of the commodities, only oil, given the US’s huge trade deficit, could have such an impact. The role of (4) is always going to have an influence to some degree, but the uniformity of our findings probably rules it out as the main explanation. That leaves (1) and (3) and – at least to some extent – we know (1) has to be the case.
This is because the price of metals is not set in the US, but worldwide, for although producers and consumers in the US and other dollar-linked countries are major players in these markets, they are not the entire market. Thus if for whatever reason the dollar was to fall by 10% in one day against every other currency in the world, then unless the price of each metal also falls 10% when measured in those other currencies, the dollar price of the metals must by definition rise. If the dollar price rises by 10%, then the price in other currencies would be unchanged.
There is a good reason to believe the price of metals in other currencies will fall a bit, because there will be a real impact from a rise in the dollar price of a commodity, and the dollar area is important for producers and consumers of most metals. Put simply, a 10% rise in the dollar price after a10percentage dollar devaluation will see demand from dollar-based consumers decline, and production from dollar-based producers increase. This will have the impact of reducing the world price (for world demand has fallen and world supply has risen) and so the price in other currencies will fall by a certain amount, meaning the dollar price will rise by somewhat less than 10%.
Of course, the fall in price in non-dollar areas will also have an impact on production and consumption there, and so the balance will presumably depend on the patterns of production and consumption between dollar and non-dollar areas. It seems reasonable to assume that the more important the dollar area is to a commodity, the smaller the change in the dollar price will be – after all, if the dollar area accounted for all production and consumption of a commodity, the exchange rate would be largely irrelevant.
Therefore, there is certainly an impact on metals’ prices from dollar moves. Yet in many of the examples above, the move in metals prices has been far in excess of what could have been ‘predicted’ by the dollar move. To put it another way – the swing in either direction, up or down, has been extreme. This could be explained in a number of ways. It might be the case that the dollar did cause the impact expected, but that the other factors were even more important. In this scenario, the commodity prices would have gone up or down, even if the dollar had not moved. This would mean that it is simply a coincidence that they tended to move in the expected direction (and on occasion, when the metal price move was smaller than the dollar move, they did not).
Another explanation is that the dollar move was a catalyst for an even larger move. For example, if the dollar is strengthening and the gold price is falling, investors in the US and other dollar areas take fright and sell their holdings, thus putting more pressure on the gold price, meaning it declines more than the dollar rose. Thus in an explanation the metals’ prices moved in the direction they did because of the dollar move, but in an exaggerated fashion. Finally there is always the possibility that a third factor was influencing both – such as in the rally of 2008-2009, when it was the financial crisis propelling both the dollar higher (on safe-haven demand) and metal prices lower (with the exception of gold, which benefited from its own safe-haven).
Much of this is unpredictable. The most that can be said with confidence is that this ‘mechanical’ effect of a stronger dollar does seem to hit metal prices. Yet for many people around the world who don’t think in dollars this will be irrelevant. A wider issue is whether, for whatever reason, large dollar moves impact on the ‘world price’ of metals – that is, the price in other currencies as well as the US dollar. The concept of a ‘world price’ however is rather nebulous. The most obvious way to calculate it would be to look at the prices of metals using a basket of currencies weighted for each metal’s consumption and/or production profile.
This is a complex undertaking to do for each metal. As a shortcut, we have instead simply looked at the prices of metals in a basket of other OECD currencies, calculated by using the inverse of the dollar index quoted above. In essence, this removes the ‘mechanical effect’ of the dollar moves, and so the price movements give a flavor of how the prices have moved for non-dollar consumers around the world.
The results are rather different than when looking purely in dollar terms. During dollar bull periods, the prices of commodities in other currencies have tended to fall, especially in the first and third periods, but not universally. In the first period aluminium, copper, zinc and tin all gained in price, and in the second period all but copper and nickel gained. In the final period all but gold fell. The dollar bear markets were uniform, with most commodities still gaining in price, although less so in the first period, when silver, palladium, zinc and nickel all fell.
What does this tell us? Essentially, that although a dollar recovery will undoubtedly see the dollar price of metals decline, (which will be bad news for US investors or US producers but good news for US consumers), we cannot say with any certainty what will happen to the price of those metals when measured in other currencies. In the 1995-2002 dollar bull market, metals’ prices measured in a basket of other leading OECD currencies actually tended to increase. In the 2008-2009 one, however, they (except gold) all fell. Clearly one can see reasons why this was – China’s enormous demand for commodities in the early one outweighed the dollar collapse, while the financial and economic collapse in the second reinforced it.
So while the headlines are quick to sound the alert when/if the dollar rises, all tend to focus on the collapse of metals’ prices. To the extent that dollar areas are involved in the production, consumption and investment of metals this will be true but the dollar’s rise does not automatically mean that the price paid for or received by most of the world’s producers and consumers of metals will also fall. That will depend on why both the dollar is rising, and factors entirely unrelated to the dollar’s behavior vis-à-vis other currencies.
