Gold’s buoyancy tied to sagging greenback

This article is the first in a series devoted to precious and base metals markets and was culled from Barclays Capital’s monthly report The Commodity Refiner. This week’s instalment focuses on gold. Parts 2 and 3 will be devoted to base metals.

According to classical economic theory, a current account deficit should ultimately result in a weaker currency. Although in the past this idea has been an extremely poor short-term indicator of the U.S. dollar’s value, it remains an article of faith among orthodox economists that a large U.S. current account deficit will, by necessity, result in a weaker dollar. Therefore, many see the dollar’s recent slide as a necessary adjustment to the explosion in the trade deficit (balance of payment on goods, services and income) that occurred in the 1990s.

In the most recent issue of World Economic Outlook (September 2002), the International Monetary Fund noted: “The question is not whether the U.S. deficit will be sustained at present levels forever — it will not — but when and how the eventual adjustment will take place. While history is modestly reassuring, the overvaluation of the dollar has not yet been corrected, and an abrupt and disruptive adjustment remains a significant risk.”

Historically, there has been a relationship between the value of the trade-weighted dollar and the U.S. current account position, albeit not a strong one: between 1986 and 1995, the correlation averaged 0.32. However, that correlation began to break down in 1996. From 1996 to 2000, it actually became negative, at minus 0.32, which means that as the U.S. trade deficit continued to expand, the dollar began an accelerated rally. In the first quarter of 1996, the dollar averaged 86.24 (U.S. Dollar Index) and the quarterly deficit was US$14 billion. By the end of 2000, the dollar averaged more than 115 while the trade deficit had expanded to more than US$94.5 billion.

What caused this change in the deficit/dollar relationship? Perhaps it was money. Beginning in 1997, foreign flows into U.S. equities began to rise dramatically. Monthly foreign equity flows into the U.S. averaged US$5.8 billion in 1997, up from only US$1 billion in 1996. Total foreign capital flows into U.S. equities in 1997 exceeded the total from 1988 to 1996. And the pace continued to accelerate, hitting a monthly average of US$14.6 billion in 2000. Since 2000 monthly flows have subsided, falling back to US$3.9 billion in 2002, and the relationship between the dollar and the U.S. current account appears to be re-establishing itself — the correlation between the two during 2001 and 2002 was 0.74. It appears, therefore, that strong U.S. equity markets drew in capital, which strengthened the dollar, temporarily reversing the traditional relationship between it and the U.S. current account.

So what, one might ask, does this have to do with gold? Gold has traditionally been negatively correlated with the dollar. Barring market events specific to gold, it recently has been trading in step with the euro/dollar exchange rate. Therefore, it is perhaps unsurprising to see that the above relationships apply to gold and the trade balance as well. Between 1986 and 1995, gold actually had a stronger correlation with the trade balance than the dollar (0.43 versus 0.32, as above). Again, the pattern started to break down in 1996, with the correlation turning negative, at minus 0.51, between 1996 and 2000. Notably, the swing in the gold/trade gap correlation was larger than that in the dollar/trade gap. To explain this, we must look at microeconomic events in the gold market during the period.

Central bank gold sales, which started in the early 1980s, began to accelerate in 1996. Of the 3,800 tonnes of identified central bank sales since 1980, 32% (1,952 tonnes) occurred between 1996 and 2000. The fact that these proceeds largely remained in dollars also tended to support the currency during this period.

Similarly, as central banks began to divest, and as investment demand collapsed, producers sought protection by dramatically increasing the size of their hedge books. Of the 2,700 tonnes of identified producer hedging since 1980, 47% or 1,234 tonnes occurred between 1996 and 2000. Moreover, gold was hit hard by the Asian crisis, because its traditional role as an “asset of last resort” was called upon.

In 1998, scrap supply rose to more than 1,080 tonnes, or 47% of mine supply, from an average of 418 tonnes since 1980, less than 25% of mine production. At the same time, investment demand collapsed while gold lost out to exploding equity markets. Annual investment demand fell from more than 418 tonnes per year, or 36% of jewelry demand, to 211 tonnes, or 7% of jewelry demand, between 1996 and 2000. Investment demand in 2000 was actually negative.

So the gold rally that began in 2001 appears to have re-established its traditional relationship with the current account deficit. While models of the gold price based solely on the U.S. current account do not have a high degree of accuracy, the average variance was only US$24 between 1985 and 1996. Such a model would project a “fair value” gold price basis for the current level of the U.S. trade deficit of more than US$690 per oz.

So is the bull market still in its early phases? Is the gold price set to double from current levels? Closer to home, is Barclays’ estimate of US$315 for 2003 and US$275 for 2004 far too pessimistic?

We think not.

In the short-term, foreign exchange market focus is likely to remain fixated on the theme of the twin deficits. While further deterioration in the dollar’s fundamentals looks likely in the short term, higher gold prices are possible.

Nevertheless, once the uncertainty associated with Iraq fades, we expect stronger U.S. fundamentals relative to the Eurozone — notably higher productivity and better growth prospects — to become evident in a dollar recovery.

Put simply, the world is not coming to an end. We believe reflationary policies in the U.S. will result in higher growth and better corporate profits later this year. This, in turn, will boost equities markets (at the expense of bonds), which will, in turn, draw in foreign capital and strengthen the dollar. Capital inflows of this kind will, by necessity, come hand-in-hand with a widening of the trade gap. However, as we have seen, a widening trade gap is not necessarily incompatible with a stronger dollar — between 1996 and 2000, the two coincided. Were such an environment to return, we could reasonably assume that gold would suffer.

Unfortunately for gold, this is likely to coincide with a turn for the worse in gold’s fundamentals. As the economic outlook in the U.S. improves, the “hot money” that has flown to gold will move into higher-yielding assets. As the world economy becomes more confident, the derisory yields offered by gold will pale in comparison to stronger equities and higher bond yields. Moreover, higher interest rates will increase the “cost” of holding gold, removing one of the factors behind the current speculative long position.

At the same time, a rise in interest rates is likely to result in a widening of contangos, pushing forward prices higher relative to spot, removing one disincentive to hedging in the current environment. This incentive will change and producers will be hard-pressed to resist 5-year contangos of around 5% — adding 25% on top of the current spot price.

Moreover, should gold’s promoters be successful with replacing currently collapsing jewelry demand with investment demand, they will be switching gold from “safer” hands to speculative ones, potentially resulting in selling as the price falls. By focusing on investment, gold’s promoters are running the risk that the collapse in physical demand will accelerate a gradual trend away from gold and into other financial assets in gold’s core markets. The risk is that the shift could become permanent, and that jewelry demand will not recover as quickly as it collapsed on the price rise.

Lastly, it is worth noting that a renewal of the Central Bank Gold Agreement (Washington Agreement) is being
discussed now, when the price is strong. Should such price strength result in a larger sales quota (as is likely), this will “kick in” just as the price begins to weaken.

Gold may therefore face more than “just” the pressure of a breakdown in the current account/gold relationship — there may be negative microeconomic shocks as well. This suggests there could be a sharp correction in gold in 2003-2004, something for which a recent poll of gold analysts suggests the market is ill-prepared.

Next week: Part 1 of our detailed outlook for base metals.

— The opinions presented are the author’s and do not necessarily represent those of the Barclays group. For access to all of Barclays’ economic, foreign-exchange and fixed-income research, go to the web site at barclayscapital.com. Queries may be submitted to Kevin Norrish, head of commodities research and energy at Barclays, at kevin.norrish@barcap.com