A New Gold War?

As a person who has a long time interest in numismatics and precious metals in general, the information explosion provided by the Internet has excited me. I am able to track, almost to the minute, gold, silver, platinum, and palladium prices around the world, day and night. In the last several months that is exactly what I've been doing.

Stimulated by the large jump in the POG during September of 1999, I watched more closely than ever. Sometime during January I became aware of what seemed to be a trend in the international spot market pricing. It appeared that on many occasions, after the NY closing, the price would rise in the overseas markets overnight, and the London AM fix would exceed the previous NY close. In a parallel fashion, the trading in NY would, at sometime during the day, sell off from the London AM fix and close below it.

Several contributors to the Gold-Eagle Internet Site & Forum have suggested that there is coming, if it's not already under way, a war to see who will be the challenger or the replacement for the U.S. dollar. Others have postulated a manipulation of the gold price, centered in New York, to keep it from rising. If either, or both of these hypotheses are true, then it is likely that during this time period more gold will be accumulated overseas (e.g., Asia and/or Europe), with a tendency for prices to rise. At the same time, there should be a tendency for prices to fall in New York as the U.S. groups try to push the price down. The first part of the effect I have described, rising overnight prices in fact goes along with the accumulation hypothesis. The second part of my observation seems to reflect a manipulation during the trading in New York to reverse or dampen the overseas accumulation.

Since I taught the application of statistical analysis to various sorts of data for over 30 years, it was only natural to turn to a more exact look at the data then I was getting from my visual impressions. I started collecting the London gold fix prices and the NY gold close near the end of January 1999.

The column of data labeled London AM - Previous NY reflects the first part of my original impressions that the price tended to rise between the NY close and the London AM fix. For twenty-three out of thirty-one times, or over 74% of the differences, this result appeared. The column of data labeled NY close - London AM reflects the second part of the cycle, the falling of prices in New York from their London AM close. Twenty-two out of thirty-two observations or about 69% portray this price drop.

If the differences in closing prices were merely a chance phenomenon reflecting unassociated buyers in a market going nowhere, the direction of changes in both columns would vary closely around 50%. If there were a multi-country trend to buy and a concerted effort in New York to suppress the price, then the price differences would vary significantly from the 50% value.

In order to determine if the frequency of the direction of the differences in favor of a higher price in the London AM fix was beyond a reasonable chance expectation, I applied a binomial test with a correction for continuity to that frequency. Since it was my prediction that this direction was expected I applied what is know as a one-tailed test. The results indicate that the probability that the London AM fix exceeds the previous NY close with this frequency, by chance, is less than 1 in 100. In fact 8 out of 9 times that the price change equaled or exceeded $2 / ounce it was up in London over the NY close. Over this period of trading days the total gain per ounce would have been $41.15 from the New York close, through the overseas markets, to the London AM fix.

The NY close - the London AM fix column shows that during this period of time the NY market closed down from the London AM fix twenty-two out of thirty-two times. Applying the same statistical method as before I discovered that this outcome would have occurred less than 5 times in 100 by chance. The concerted effort to push the price down in New York resulted in a net loss of $40.90 over the trading days observed.

Even though we see the consistent rise in the London AM fix for a net gain of $41.15 during the trading days followed, the net loss in New York of $40.90 almost exactly offsets it. The net difference between the total POG gain overseas and the total POG loss in NY is only $.25! Someone who was thinking manipulation might say to the New York people "Come on - At least be a little more subtle".

An additional comparison, the change in price from the London AM to the London PM fix supports the previous conclusions. For twenty-two out of thirty-three trading days the price was down at the PM fix. This is also a rate of one-sided differences that is likely to have occurred less than 5 times in 100 by chance. It appears that the selling pressure in New York had a downward effect on the London price. New York begins its trading between the AM & PM London fixes.

I have located daily London fix data for the last several years and I'm currently searching for NY spot closing data for the same time periods. This analysis should be repeated over a longer period of time to see how far back in time this pattern extends. But meanwhile..

Up to London! Down in NY! Does this represent the start of the war to accumulate gold to back competitor currencies to the U.S. dollar vs. U.S. led manipulation centered in New York?