HUI (Amex Gold Bug Index)


A moment of truth is quickly approaching for the gold mining sector. I follow and trade this space because the volatility and % returns far exceed anything else on a non-leveraged basis. Add in 3x ETFs and you have incredible trading swings. I personally don’t use this kind of leverage unless I have 100% conviction. And that is lacking at the moment and I’ll show you why.

After the 2016 high, a preferred analysis was presented in October 2016. This focused on using the January 2015 high as the 45 month cycle high. This has been a major uncertainty for me and it changes the view materially if instead, one uses the 2016 summer high as the 45 month cycle high. I’m not going to getting into the pros and cons of either scenario. What I want to show you is that each scenario has quite different outcomes for the remainder of this year.

Here is the preferred view. The projected path from last fall has been tracking well. Even this pullback into current lows is allowed. Daily, weekly, and longer term views as a refresher follow.

So looking at this first analysis presented above, with the 45 month high pinned (yes pinned and is based on my interpretation), the path forward is for the most bullish phase into a 7.5 peak is due to start very soon. A tradeable low is due no later that early next week. Price is expected to attack last year’s highs if this market is truly bullish.

However, if we consider another phasing alternative, and that is simply allowing ST to identify last summer’s peak as the 45 month cycle high, then things look much less bullish for the remainder of the year. I really don’t know which way this goes and will have to trade the coming low and watch price action closely. Price is sitting on several forms of support. The 22 and 45 month FLDs are just below price. There is not much room for lower prices. The real tell will be the coming 16 week cycle after this near term low, which should form by early next week (FED day of course). It needs to be bullish, attack recent highs, plain and simple.

And while it would be grim for the remainder of the year, the long term bull would still stand in front of us.

Another reality check is for the 48 week VTL (green line shown above) to hold on this pullback. I’ll have to update the look on the USD for clues.


John Oestreich is here :sob:

John, is that a hybrid peak and trough analysis? Does it look different with only peak analysis?


Dale Todd


Hi Dale,

That is a hybrid peak and trough view. Having a look at the trough only, the same imminent low is due in the next couple of days. The difference in scenarios and evidence will show at the next high.



Off topic not about cycles, but I use EW and cycles in combination and take a position only when both are saying the same - at the moment this is not the case if you expect important low.
The move lower from February is clear impulse. After an impulse expect more to the downside. It is visible everywhere HUI,GDXJ,single shares. I went trough 20 gold miners and I can not see a way to twist it as bullish.
Yes, bottom is imminent, but only short term. After that there will be one more leg lower. I do not see how it will revers and rally higher… 0 evidence to support it neither EW nor technicals nor market breadth.
I am watching the June-July period the next 20 week cycle low for a possible bottom.

P.S. a little bit confused… in the webinar about gold you were using different model 17w/34w/68w etc. for gold miners which is similar to stocks and now you switched to other model which is similar to gold. Any reason for that?


Hi karni,

I can’t argue with your comments. I have issues with the bullish view. But honestly, I will track and follow my Hurst work until something shows me otherwise. I have concerns with the performance of individual miners, breadth, etc. But lets’ see where this next cycle takes price. I entered long this morning an initial position in Gold miner ETFs (no leverage) as gold is testing $1200 here. TWT.

Will post more on the $USD, other markets over the weekend. Take care.



I’m a big fan of the 20 week wave when it comes to gold and gold miners. We have seen a lower 20 week low (December 2016) followed by lower 20 week high (February 2017) which implies the sum of the waves longer that the 20 week is down. The next 20 week low is still weeks away. I’ve been short from the last 20 week high and just got kicked out today. I’ll be looking to reenter short if I see a lower 5 week high.

As for the overall trend, another lower 20 week low will indicate considerable weakness in my opinion with the last line of support for the bullish argument will be a higher 80 week low. The price waves decompose almost perfectly (using the “mathematical” approach of course) so you can pick just about any wave you want as your trading cycle.



Hi William,

Given the larger cycles involved and the premise that they should have bottomed, one would expect more bullishness. I’m certainly cautious here and stepped out of trading positions yesterday in front of the FED.

This will be an interesting year with the $SPX/$INDU basically going straight up as a backdrop and trying to trade various sectors.




hello i am new here but just an observation i find it interesting that gld and slv have held up

a lot better than the stocks. gld had only a .5 correction off of the feb highs not sure what to make of it

but combined with technical analysis to validate the waves. anyway look forward to learning

more about hurst


The divergence between metals and miners certainly doesn’t look bullish here. However, there was an exception to the rule in January 2016 as stock markets headed to major lows. Cautious and still waiting to see how it plays out.


The slight difference in the gold futures and the miners is the result of underlying trend and amplitude differences in my opinion. They are highly correlated and those differences can disappear very quickly. I just pick my poison ( 20 week wave), damn the torpedoes, and full speed ahead!



Hi Sid.

Nice chart. It is basically the same as mine. I don’t quite understand the logic in placing the 40 week trough at Sept 1 2016. Ordinarily a 40 week trough is also a 20 week trough (at least with respect to gold currently). My filter approach does show a duration fluctuation over that period but the 20 and 40 week wave troughs are closely synchronized.


Hi Willam,

My reasoning for pinning the 40-week to the Sept 1 low is simply because it evens up the spacing of the cycles nicely in comparison with the default ST analysis. Attached are screenshots of ST analysis of HUI starting July 1998 to show what I mean. All analyses utilize the JM Hurst nominal model, troughs only. This first one is a long-term look, to show that the default ST phasing appears to be right on the money with the 9-yr (7.6-yr avg) and 4.5-yr (45.5-mo avg). troughs.


This next screenshot is zoomed in, and shows that the default analysis places the 40-week trough at the Dec 15 low. You can see how very long that 40-week cycle was, if this (default) analysis is correct.


Now here’s the phasing if the 40-week trough is pinned to the Sept 1 low. The phasing is much more normal in its spacing, with a strong 18-month (14.9-mo avg) nest due in about April.


Finally, if the 40-week repin to Sept 1 is correct, what was the most recent swing low on March 8? An 80-day trough (50-day avg)? The HUI analysis is trying to place the 80-day way up on Feb 10, which seems unlikely. However, If I do the very same ST analysis on ASA, which is essentially the same thing as HUI, and do the same repin of the 40-week to the Sept 1 low, if considers the March 8 low an 80-day (58-day avg in ASA) trough, and the composite line suggests that the miners won’t put in their 18-month (13.8-mo avg in ASA) bottom until approximately the first week of May.


Hi Sid/William,
Thank you for posting your thoughts. The current 20 week cycle does not look bullish. The gold mining indices have risen 7 weeks off the Dec. lows and look nowhere near making new highs on this cycle while the metals firmly test 2017 highs. That is typically not bullish.
I have followed this market for a long time and bullish moves or trend changes begin with the miners leading and outperforming the metal. With the shares diverging badly since the Feb. high, caution has been warranted.
When looking a volumes, GDX shows us the biggest weekly volume bar in its history the week of the US election. The price low of that week is likely to be tested a second time after seeing a first test in March. This next test must be on low volume to setup the potential for a spring higher. On the Canadian side I follow/trade the gold miner ETF symbol XGD.TO. This ETF had its largest weekly volume that same week.
The FED early May date would be a potential timeframe for the next 20 week low to occur. William, would love to see your filter work on GDX here. Many trade this instrument globally in the gold space.


Hi John

I was curious as who was usually the leader between the metals and the miners. The chart below is the GDX for the last 8 years or so. The three waves shown are the 20, 40 and 80 week price waves. You will notice there is a small but significant fundamental difference between my analysis and that of most other analysts, including Sid’s excellent analysis. With the exception of the US stock indices, I do not automatically assume a 3 to 1 relationship between the 80 week wave and the next longer wave. I always test any assumption. In the case of gold and the miners I found that a 2 to 1 relationship resulted in a consistent simple harmonic relationship among the waves over many, many years with a minimal amount of modulation. As you can see from the chart the waves are fairly evenly spaced and have consistent amplitudes. Even the waves shorter than the 20 week show this same consistency. This is why I comment on gold being one of the easier instruments to trade from a cyclical perspective.

This results in the average period of the waves being longer than Hurst’s nominal time domain model (which I never use). The 20 week wave has averaged approximately 23.4 weeks over the past 8.3 years. The envelope is the sum of all of the waves twice as long as the 80 week wave shown. It will get real exciting if the 40 week wave puts in a lower high and forms a triangle on the chart.


Nice work as always William. I agree on your phasing and preferred model for BOTH gold and the miners using ~45 month, 22 month, 11 month, 48 week and 24 week for both gold and the miners. I have wrestled with it for many weeks but now prefer this. Your bandpass filter on the GDX makes it very clear. The difference on this new 7.5 year cycle is that the 24 week and 48 week cycles are not making higher lows. Look at the rise out of the late 2008 lows and that is what you see. So the expectation is that this market may not make an important bottom till late this year when the current 48 week and 22 month cycles bottom. Which means another 24 week cycle (23.4 wk) to go after this one bottoms. All of this to say that is why I started this thread and raised the concern in this sector. Thank you for sharing and posting William. Much appreciated.

PS The divergence between gold and miners is typically a short term observation (days /some weeks). It likely would not show up in the cyclic behaviour.