With the recent crisis among financial institutions promising further economic weakness and an expansion of fiscal deficits, the decline in the U.S. dollar (blue line above) has gotten the attention of traders and investors. I've been hearing particular talk of hedging this dollar weakness through the purchase of gold as an alternative currency. But is gold an effective dollar hedge?
There's no denying its recent appeal. Over the last three trading sessions, volume in the gold ETF (GLD) has risen to over 20 million shares each day. That's easily twice the average volume over the last few months and the highest three-day total since the Bear Stearns crisis in March.
As the chart above indicates, since 2005 gold (pink line) and the U.S. Dollar Index (blue line) have been traversing opposite paths. The correlation of daily changes in gold and USDX was -.31 from January, 2005 through June, 2007. Since that time, it has soared to -.54. Over a quarter of the daily variance in the U.S. dollar and gold has been shared, suggesting that it has, indeed, been a kind of hedge.
When I went back to November, 1985--the first dates for which I have both U.S. Dollar Index and cash gold data--I found that the correlation of daily price changes between the two has been -.26. That led me to surmise that the current correlation might be historically high.
What I found was that the correlation of daily price changes in gold and the dollar from January, 1986 through December, 2002 was -.20. Since the start of 2003, that correlation has more than doubled to -.43. Interestingly, the period of the correlation's expansion has also been a period in which the U.S. dollar has fallen by more than 28% to modern lows.
The data suggest that, since the marked weakness in the U.S. dollar, gold has increased its role as an effective dollar hedge. Given the surge of interest in gold during March and now in the last few days, it wouldn't surprise me if participation in the gold market also emerges as a kind of sentiment indicator for the dollar.
.
Posts
