Is it Time to Buy Gold Again?

January 3rd, 2013

by Georg Vrba

Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.


The latest data have a positive message for investors: the gold price could soon move significantly higher. That’s the current finding of my model which determines buy signals for gold using a Coppock indicator.

Figure 1 shows the Coppock indicator and the gold price from 1969 to 2012. According to the description of this indicator a buy signal is generated when it is below zero and turns upwards from a trough. I found that a trough formation usually identified good buying opportunities for gold, irrespective of whether the indicator was below zero or not. Buy signal dates were calculated from the data, not simply found by a visual inspection of the graphs, and are indicated by the green vertical lines on the charts. In the appendix is the information on how to produce the indicator and the buy signals.

Table 1 and 2 gives the dates when the buy signals occurred, together with the price changes after the signal dates over a 1- and 2-year period, respectively. One can see that the 1989 and 1998 signals were least successful in forecasting a gold price increase (although the indicator was well below zero then), while the other nine signals all heralded significant gains for gold within a year or two. It is also evident that an investment in gold at the signal dates never resulted in significant losses over the time periods considered.

Sell signals cannot be determined with this indicator as it was not designed for this, but it is apparent that a negative slope of the indicator coincides with a declining or flat gold price. In the absence of anything better, a good strategy would have been to sell when, in a two year window after a buy signal, the price had appreciated by 40%, but not later than two years after the signal date. This would have resulted in nine gains of about 40% within two years of the signal, and only in two small losses of about 8% after a 2-year investment period. An analysis showed, that if the funds went into a money market account paying interest at the federal funds rate when not invested in gold, this strategy would have provided an average annual return of 11.2% versus 9.5% for a continuous investment in gold from October 1, 1970 to December 31, 2012. Or, put another way, an initial investment of $100 would have grown to $8,750 when implementing this strategy versus $4,600 for a permanent investment in gold. Table 3 lists the rule-based signals with corresponding returns.

It would appear that the indicator is now in the early stages of forming a trough as can be seen in the more detailed figure 2. It is interesting that other models (for an example see the commentary Light at the End of the Tunnel for Gold) are also expecting gains for gold. A chart from this commentary is replicated in the appendix. Once a buy signal is generated, then the odds are good for a significant increase of the gold price to occur.

Appendix

I calculated the indicator for use on a daily time scale. It's the sum of the 294-day rate of change and 231-day rate of change of a 5-day rolling average of the gold price, smoothed by a 210-period weighted moving average (WMA), multiplied by 10.

Indicator = 10 x WMA of (ROC[294] + ROC[231])

Note: Days are trading days (252 per year), not calendar days. The excel function SUMPRODUCT was used to calculate the weighted moving average.

Initial requirement for a buy signal: Before a buy signal can appear the indicator must be at least 1.0 units below its maximum level determined over a preceding 252-day period.

To mathematically find the signal dates, I used an exponential moving average (EMA) of the indicator with a smoothing factor of 0.05. After this EMA moves below the indicator, and when subsequently the difference between the indicator and its EMA is greater than 0.05 while the initial requirement is satisfied, a buy signal is generated. The formula used for calculating the EMA with a smoothing factor α is:

EMAtoday = EMAyesterday + α (valuetoday - EMAyesterday)

Chart from the December 21, 2012 commentary by Frank Holmes, Light at the End of the Tunnel for Gold, with the green vertical buy signal lines from the Coppock indicator superimposed.


Georg Vrba is a professional engineer who has been a consulting engineer for many years. In his opinion, mathematical models provide better guidance to market direction than financial "experts." He has developed financial models for the stock market, the bond market and the yield curve, all published in Advisor Perspectives. The models are updated weekly. If you are interested to receive theses updates at no cost send email request to vrba@snet.net.

Website by the Boston Web Company