There are 3 techniques that we have used with varying degrees of success, namely charting, the prediction of moving average cross-overs, and the requirements of exponential smoothing, which are described in this article. These techniques require volatility (movements up and down in the values) to produce meaningful results, they are harder to decipher when the price (of gold or the share) trades sideways and/or poorly.

Charting

I used to think that charting was largely “smoke and mirrors”. However, as part of an MBA thesis in 1983, I devoted a section to charting and found that some of the patterns or lines of coincidence had an accuracy of 60% to 70% when applied to the US$/oz gold price and gold shares, which I regarded as too high to ignore.

Those main patterns were the horizontal lines of support and resistance, right angled triangles to estimate the likelihood of price rises or falls, and scribing half of a base being broken on a uniform (as compared to a log) chart occupying a computer screen into the height as an approximation of point-and-figure target calculations. (It should be noted that while I am not a technical analyst, and have no qualifications in that area, the comments in this paragraph are based on numerous practical observations).

Moving average cross-overs (and their prediction)

Taking a moving average of a series of numbers is a well-known technique. For example, a 10-day moving average may be compiled by adding ten consecutive daily values and dividing by 10. Advancing a day at a time and dropping the first day, a moving average is produced representing the average price of the previous 10 days.

Using two moving averages and observing when they cross over each other, as a tool when to buy or sell commodities has reputedly been used in the 1950s, but was documented by Kaufmann in 1980. It is possible to determine what 2 moving averages give the highest appreciations from either buying or selling between cross-over points, using a computer program. This occurred in 1979 in the gold mining investment division that I worked (using a mainframe overnight), determining that the most optimum days’ averages for gold were 10 and 40 days.

I later double-checked the optimum days for the US$ gold price using my own program over a longer period in 1983, which still produced the 10 and 40-day moving averages as the best appreciation for buying or selling from cross-over points. The classic buy signal (B) is when the 10-day MA passes up through the 40-day MA as shown in figure 1of the XAU gold index (and similarly sell (S) when the 10-day passes down through the 40-day).

Figure 1 XAU Gold Index with 10 and 40 day moving averages (source : BigCharts.com)

However, it is possible to enhance the technique even further, and identify the peak gap between the two moving averages (also shown in figure 1). By producing a chart of the percentage difference between 2 moving averages (say 10 and 40 days for gold), a line oscillating about zero (when the moving averages cross over each other) can be produced. This line is relatively straight, so using least squares regression, it is possible to determine an equation for the line and project it ahead as to what is required for the line to continue to remain straight.

By observation over time it is possible to identify the trend in the gold price (or a gold index) as to whether it appears likely to continue to rise or fall within an accuracy of about 60% to 70%, and what value is required for one moving average to cross-over another. Identifying the cross-over points requires greater experience.

While 10 and 40-day moving average cross-overs do apply for gold and gold indices, based on observations, 5 and 12-day moving average cross-overs appear to work (within reason) for most gold shares as shown in figure 2.

Figure 2 XAU Gold Index with 5 and 12 day moving averages (source : BigCharts.com)

Exponential smoothing

Probably the most easily identifiable technique of when the gold price (and hence gold indices and ergo gold shares) should rise is through using exponential smoothing. I encountered it as a technique that had been used since 1979, and while its buy signals are probably about 70% accurate, its sell signals are far harder to identify.

An exponentially smoothed moving average applies a greater weighting on the last values than the earlier ones. For example the esv (exponentially smoothed value) for 10 values can be determined from the sum of 1 to 10 which = 55, and then the esv is the sum of 1/55 of the 1st value, 2/55 of the 2nd value etc to 10/55 of the 10th value. Again the best correlation for the gold price has been found to be through using 10-day and 40-day exponentially smoothed moving averages.

This time in order to buy gold, both the esv of the 10-day moving average and the esv of the 40-day moving average have been found to need to increase from the previous day’s esv at the same time. It is possible to calculate what value is required on the coming day to produce a “buy” signal (or a more difficult to determine “sell” signal).

The following two examples clearly show “buy” signals being given for gold (using this technique) on 29 August 2002 (see figure 3) and 4 December 2002 (see figure 4).

Figure 3. Exponentially Smoothed Gold Price (US$/oz) (26 Aug to 9 Sep 2002)

Figure 4. Exponentially Smoothed Gold Price (US$/oz) (2 Dec to 16 Dec 2002)

A combination of all three techniques have been used with a success rate of about 60% to 70%. So, where are we now (9 January 2003)……actually overbought (both the gold price and XAU gold index), and ideally in need of correction downwards or stability sideways before moving up again.

Disclosure and Disclaimer : This article has been written by Keith Goode, the Managing Director of Eagle Research Advisory Pty Ltd, who has a Proper Authority with State One Equities, and with his associates, either has or expects to have interests in most of the stocks in this article, This e-mail address is being protected from spambots. You need JavaScript enabled to view it . The opinions expressed in this article should not be taken as investment advice, but are based on observations by the author. The author does not warrant the accuracy or completeness of any information and is not liable for any loss or damage suffered through any reliance on its contents.