Shake Out or Fake Out? A Look at the Seasonal Cycle in Gold
By Tim W Wood CPA, April 16, 2004
The seasonal cycle in gold averages 11.45 months. This cycle last bottomed April 8, 2003 at 319.80. Knowing that the seasonal cycle was coming due, I began to be concerned about gold back in January and particularly in February when we saw some of the short-term cycle lows violated. This indicated that the intermediate-term, 18-week cycle, as well as the longer-term seasonal cycle, had topped. It was because of this shorter-term cyclical breakdown combined with the fact that the seasonal cycle top was due and that the monthly oscillators had begun to move down that I found so alarming. The chart below is a monthly gold chart through March 31, 2004. The top oscillator is an RSI derivative. The bottom oscillator is Walter Bressert's 5 double stochastic oscillator. Both have historically done a very good job at helping to identify the seasonal cycle lows. Notice that both of these indicators were moving down in January and February.
With price holding up as well as it did and then the late March rally taking gold back up above the previous trading cycle high, the picture once again turned positive as this all served to confirm that the 18-week cycle low had occurred. Also, because this advance from March turned the monthly double stochastic up, it began to look as if the seasonal cycle low had also occurred. The RSI3M3 is a little slower to turn and did not confirm the March advance. This is expected as it is a slightly slower turning oscillator. Additionally, we saw the short-term trading cycle move above the previous trading cycle high. This served to confirm that the 18-week cycle low had occurred. Then, on April 5, 2004, the trading cycle low bottomed at 414.50. This was above the previous trading cycle low and thus further confirmed that the intermediate-term, 18-week cycle had indeed bottomed on March 3, 2004 at 388.20.
Then came the decline on April 13, 2004. Technically, this was a hard blow to the gold market because it violated the previous trading cycle low, which again occurred on April 5, 2004 at 414.50. Please see the daily gold chart below. I was looking for a trading cycle low to occur around 410. In my weekend update I had told subscribers that it looked like the trading cycle low had occurred and that gold had no business below 414. With the decline below 410, I moved to the sidelines on gold.
The reason that the 410 to 414 level was so important was that it marked the April 5, 2004 trading cycle low. Notice on the chart above that when the intermediate-term trend is moving up, each trading cycle low holds above the previous low. Also notice that at the intermediate-term, 18-week cycle tops that once the previous trading cycle low is violated it marks a trend change as gold then moves into the next 18-week, intermediate-term low.
So, the point here is, Yes, the March low was an important low as it did mark the bottom for the current 18-week intermediate-term cycle. However, with the recent violation of the April 5, 2004 low at 414, all indications are that this cycle has topped out and thereby was a failure. The decline into the seasonal cycle low is likely still at hand. Based on the timing of the next intermediate-term low, we can expect to see the seasonal cycle low occur sometime in the July timeframe.
I also did an analysis of all seasonal cycle declines in Bull market periods and found that the decline into the seasonal cycle low has averaged 3.13 months in duration. So, based on this average, we can expect to see gold continue down into July. It is purely coincidental that July is also the expected timing for the next 18-week cycle low, but the fact that two different measures give me the same answer tends to confirm the prognosis.
In my analysis of the Bullish seasonal cycles I found that the average decline into the next seasonal cycle was 17.94%. A decline of 17.94% would take gold down to 355.32. The largest decline seen in a Bullish seasonal cycle was 48.11%. This was an abnormality as this was the seasonal cycle that advanced into the 1980 top and the seasonal decline that followed was just as severeas the preceding advance. The smallest seasonal cycle decline seen in a Bullish cycle was 6.22%. This too was a bit of an abnormality. A 6.22% decline would only take gold to 406. We have obviously already surpassed this level. As for the time aspect, we also have the remainder of the current 18-week cycle, which must run its course.
Understand, this is not a piece to "bash" gold. I don't bash anything as I have no agenda to do so. I call it like I see it technically and we currently have technical issues to work through before gold becomes a buy again. After all, the current seasonal cycle has seen a heck of a run. History shows that the average advance for a seasonal cycle in a Bull market is 8.5 months. We just experienced a 12-month advance that saw a 35.4% rise. So, history shows that this has been a longer than average run with a very good advance. Expecting more is not supported by the historical norms. Gold has performed well and it’s now due a rest. Expecting more would be just plain greedy and greed will get us hurt. Once the seasonal cycle low occurs, regardless of the level, gold will present its next buying opportunity and guess what, we will be able to buy it cheaper than it is now. What a deal, I can’twait.
The chart below is of the XAU. Notice that we also had an intermediate-term low form in this index back in late January. The trading cycle that followed was positive as it held above the January low, bottoming in mid-March. From the mid-March low the next trading cycle managed to push slightly above the February trading cycle high. However, this gain ran out of steam and has now broken down below the mid-March trading cycle low. This makes the current intermediate-term cycle a failure in that it was not able to better the previous cycle high. This all suggests that we have also seen a turn of intermediate-term degree here as well. We are now moving into the timeframe for the next trading cycle low. I fully expect to see this carry this index back up above the breakdown point at approximately 96. However, we should expect to see this rally contained within the boundary of the upper downward sloping trend line. If by chance this trend line were to be violated on the upside, it could serve to correct the technical damage that currently has been dealt to this sector.
Tim W. Wood
© 2004 Tim Wood