Trading HUI Volume
Adam Hamilton October 22, 2004 3403 Words
While the performance of the gold-stock sector has been nothing short of spectacular in recent years, up 614%, in the grand scheme of things this particular market remains extremely small and largely ignored by mainstream investors.
All 15 elite miner components of the premier gold-stock index, the HUI, had a total market capitalization of less than $54b this week. To put this into perspective, Microsoft alone weighs in at $313b in market cap. The worldís largest gold mining company, Newmont Mining, is worth a comparatively trivial $21b. Regardless of the legendary gains already won in gold stocks in this bull to date, the raw market cap numbers show that serious capital hasnít even yet begun flowing into this sector.
This is great news for those of us investing and speculating in this curious little world, but popular obscurity is a double-edged sword. Due to the lack of general interest so far, the precious-metals stocks sport very few if any well-developed technical trading tools. If you are trading the SPX you have the VIX, VXO, PCR, and a host of other highly specialized sentiment indicators. If you are trading the HUI, you have virtually none.
Thus one of the great analytical challenges in trading an up-and-coming sector is to attempt to develop indicators that help increase our probabilities of timely detecting major interim tops and bottoms. In addition to being a fun and challenging academic exercise, the bottom line goal in all indicator development is reaping big real-world profits. And if history is a valid guide, future HUI profits will utterly dwarf the massive gains with which we have already been blessed so far.
Since precious-metals stocks are my favorite sector in which to invest and speculate due to their staggering potential in this secular gold bull, hardly a day goes by when I am not pondering ideas to better recognize major tradable tops and bottoms. Last week I wrote about the potential for considering the HUI in parabolic terms, my thesis being that when a major HUI parabola shoots vertical the probability for an intermediate trend change is high.
Unfortunately I was stumped on how to elegantly measure these parabolas empirically, so I asked for ideas and I was blown away by your overwhelming response. I received over four dozen detailed e-mails from brilliant minds all over the world, some including gorgeous spreadsheets detailing ways to measure parabolas, so now we have much more parabola research to pursue in the future at Zeal. I am deeply thankful and grateful for all of you who spent your valuable time helping me out.
While this essay isnít about parabolas, I am writing it as a result of the wonderful HUI parabola feedback. A gentleman from Arizona wrote me, a retired United States Air Force Colonel, and prudently suggested that perhaps volume anomalies would help identify when parabolic ascents or descents were reaching terminal tradable turning points. His kind and wise letter unleashed a flurry of thought that solved a problem that has vexed me for years.
Trading volume is a centuries-old and highly respected trading indicator, but unfortunately the HUI has no volume! If you pull up a chart of the HUI on any website, or download HUI data from any data provider, volume data just doesnít exist. The problem is the HUI itself is not actually traded. While we can all trade S&P or NASDAQ futures, and analyze the raw volume on those contracts endlessly, to the best of my knowledge there are no HUI futures anywhere on the planet. So while I have long wanted to analyze HUI volume, there was none to analyze!
I had all but given up on HUI volume, but the good Colonelís letter sparked a deceptively simple insight, one that I feel dimwitted for not considering sooner. Yes, there is no HUI index trading or HUI index volume, but there is trading and volume in the underlying 15 component stocks of the HUI! Thus the solution to analyzing the HUIís volume is simply to add up the individual trading volumes of all 15 HUI stocks and use the resulting sum as a composite HUI volume.
I am very grateful for the Colonelís gracious help with this insight, as it unlocks a whole new avenue of gold-stock analysis. It doesnít matter at all if the HUI index itself trades or not when we can analyze volume at the aggregate component level. HUI volume, while it strictly doesnít exist, really does live in the trades of its individual stocks.
This whole process reminds me of one of my favorite Bible verses as one who traffics in information for a living. In Proverbs 27:17 the incomparable King Solomon said, ďAs iron sharpens iron, so one man sharpens another.Ē By sharing ideas and building off of each otherís insights, I really think the precious-metals community will have no problem developing useful and timely trading indicators to someday rival those of the general markets.
My partner Scott Wright did all the heavy lifting in gathering and aggregating this HUI volume data and building these charts, but I have the privilege of analyzing it. Both of our charts this week graph the HUI over the total daily trading volume of all 15 of its component companies in this gold-stock bull to date. Our first chart shows the raw data, while the second showcases a one-week moving average of HUI volume and offers some trading insights.
From a macro level, HUI volume appears to behave as volume typically does in the general stock indices, no big surprise. As the HUI has marched relentlessly higher since late 2000, its volume has grown accordingly reflecting higher interest in the long-neglected precious-metals stock sector. This is the natural volume progression of all major bulls. In the early 1990s only a tiny minority trafficked in tech stocks, but by early 2000 the NASDAQ volume was staggering as the vast majority chased the bubble.
Now if daily volume was only averaging 5m shares in HUI component companies in 2000 and maybe 25m shares today, it is tempting to think that HUI trading volume has only increased 5x. This is technically correct, but it understates the true volume dynamics. Not only has the raw absolute HUI component volume increased, but all of the share prices of HUI component companies have increased dramatically as well.
The HUI bottomed way back on November 14th, 2000 at a pathetic level just under 36. Its bull-to-date high is just under 257 achieved about a year ago on December 2nd, 2003. In order to really understand the HUIís volume dynamics as reflected by total capital traded, it is useful to consider the HUIís dominant component company Newmont Mining on both pivotal dates as an example.
On the secular HUI low day in November 2000, 0.7m shares of NEM changed hands. On the bull-to-date HUI high day in December 2003, NEM traded 8.1m shares. A simple volume calculation based on these numbers shows an impressive 12x increase in NEM volume from the HUI bottom to the latest HUI top. But, it is crucial to realize that NEM was only trading at a price of $13 in late 2000 but had soared to $50 by late 2003.
True volume considers not only the number of shares changing hands, but the actual amount of capital hitting the markets. At the HUI low NEM did 0.7m shares at $13 or $9m in capital volume. At the latest HUI high NEM trading soared to 8.1m shares at a whopping $50, or $405m in capital volume. So in true capital volume terms the volume in NEM alone has rocketed by 45x in this bull so far, vastly higher than the 12x increase in just the raw number of shares traded.
I bring this up for two important reasons. First, while general raw share volume does indeed rise during a secular bull market, it vastly understates the total capital in play since underlying share prices are also rising at the same time. Second, due to this capital volume dynamic it is not strictly necessary for raw share volume to increase dramatically in future HUI uplegs even though capital volume will continue to explode. This second point is particularly crucial if we are to be able to use volume as a trading indicator.
If vastly more capital can be funneled into the HUI as more and more mainstream investors wake up to this stealth gold bull right under their noses, this capital can manifest itself in HUI component share prices increasing far faster than volume levels. Raw share volumes can grow slowly while at the same time capital volume is soaring. This observation allows for the possibility that absolute volume extremes can be of use for trading clues throughout a secular bull market.
While our second chart below digs into this tantalizing possibility that absolute raw volume levels can remain relevant over years in a bull market, there is one more attribute of the first graph worthy of consideration. If you check out the red volume data, there is a dotted blue-line drawn in. This is not a precise mathematical line, but just a visual center-of-mass line. It is interesting to note that its slope follows the HUIís fortunes fairly well.
When the HUI is charging ahead during a major bull-market upleg, investors and speculators get excited and they increase their trading activity. Volume tends to surge along with the HUI, even when all the excessive day-to-day volume noise is filtered out. Conversely when the HUI has been correcting, slumping, or flatlining in one of its periodic bull-market corrections, investors and speculators grow depressed and weary of trading. This phenomenon is universal across all the markets I have studied, not just gold stocks.
These natural human tendencies based on our psychological reaction to changing short-term market conditions make volume more or less follow prices. When things feel good people love to trade, but when it feels like the markets are kicking them in the teeth people dramatically pull back on trading and want as little to do with the day-to-day markets as possible. This tendency too increases the probability that volume can be a useful trading tool.
OK, to summarize so far, HUI capital volume can soar dramatically as more mainstream investors ďdiscoverĒ our raging bull while raw share volume grows much more slowly. The vast majority of increasing capital volume is the result of rising share prices while the minority is the result of higher raw share volume. On top of this, in general HUI volume tends to follow the fortunes of the HUI due to the inevitable psychological impact of perceived favorable and unfavorable short-term trends on investors and speculators.
The final component to trading HUI volume is to consider adding some kind of filter. As you can see above, the day-to-day volume changes can be utterly colossal. On December 26th, 2003 the HUI only did 12m shares, anemic levels reminiscent of the first year of its bull. Only 13 trading days later on January 15th, 2004, however, HUI volume soared to a spectacular bull-to-date high of 74m shares! This is noisy data!
Ever cognizant of the inherent tradeoffs in noise filters, data smoothness versus timeliness, we ultimately liked a simple 5-day moving average best in our initial HUI volume explorations this week. This is a one-week moving average of volume that remains very responsive to current conditions with little lag yet it still clips anomalous extremes like the 12m and 74m share days mentioned above. Our final chart graphs this HUI volume 5dma with the index itself and offers some intriguing trading insights.
As the good Colonel expected and shared with me, massive volume spikes do indeed usually accompany terminal-stage ascents or descents in HUI parabolas. If you compare last weekís first graph with this one, you will note that all five of the major HUI parabolas reaching maturity from 2002 to today are accompanied by huge volume spikes when they shoot vertical and become unstable and unsustainable. Volume climaxes do mark major interim trend changes!
Most trading indicators, including the various Relativity-based ones we have been developing, follow the oscillator approach. With an oscillating indicator, a low indicated level is always a buy signal and a high level is always a sell signal, or vice versa. Volume doesnít work this way though. If you look closely at this chart, you will notice that huge volume climaxes can happen near both major interim highs and major interim lows.
Now with volume climaxes happening near both highs and lows, what the heck is going on? And if a massive volume spike can mark a high or low, does this mean volume is going to be useless as a trading indicator?
Actually, volume climaxes blooming near major turning points on both the high and low side make great intuitive sense. You, I, and all human traders are hopelessly emotional creatures. Our natural tendency is to grow greedy when prices are soaring and terrified when prices are plummeting. Greed and fear are the most powerful motivators in the financial markets.
They are both the greatest friends and greatest enemies of traders. Greed and fear are wonderful when it is the other guy you are trading against who is under their seductive spell. But when it is your judgment that is guided by your own internal greed and fear you are almost guaranteed to make poor trading decisions. Emotional latitude is lethal at the individual trader level.
When prices are soaring and the HUI is making a new bull-to-date high, most investors and speculators get too excited which breeds greed. Trading activity always increases when people are excited as they sell other investments to free up capital to plow into the surging gold-stock bull as well as churn their own holdings to try and chase the best performers of late. Buy-side volume climaxes marking major interim tops are driven by greed.
Later when prices are plummeting in the inevitable corrections between the uplegs, at some point folks get downright scared. After someone has suffered a loss of a third or so of their PM capital in a matter of weeks or months most people eventually reach a breaking point. They want out now at any price and trading volume soars in a capitulation panic. The last one we witnessed in gold stocks was April 2004, an event that terrified many players. Sell-side volume climaxes are driven by fear.
Thankfully it is amazingly easy to discern whether a volume climax is greed-driven or fear-driven. If the HUI is carving new bull-to-date highs in a spectacular upleg that is shooting vertical, then there is no doubt it is near a greed-laden interim top. But if the HUI is plunging to threatening depths weeks or months after such a major top and people are terrified, then there is no doubt it is near a fear-laden interim bottom. Volume climaxes always make sense in context.
Thus the dual nature of volume climaxes certainly does not preclude them from being useful indicators. When volume is considered with price, it is readily evident whether a volume climax marks a potential top or bottom. To take an initial stab at developing such a volume indicator, we arbitrarily selected levels of above 38m shares and below 16m shares on a 5dma basis as good lines of demarcation marking volume extremes. These are rendered on the graph above.
Each time the 5dma of HUI volume exceeded these boundary conditions, we drew a vertical blue line to compare where the HUI happened to be trading near the volume extreme. If the index was close to a major interim high, we indicated that with an H on the graph. If the volume extreme occurred near a major interim low, naturally an L is drawn. Letís first consider the massive volume climaxes over 38m shares.
Since 2002, there have only been eight times when the HUI volumeís 5dma exceeded 38m shares. Five of these volume climaxes occurred near major interim tops, two near major interim lows, and one in a no manís land between tradable highs and lows. Thus seven of the last eight volume climaxes above 38m shares on a 5dma basis occurred very close to optimal opportunities to trade. In the future when I witness similar events I am definitely going to stop and consider their implications.
If HUI 5dma volume is running above 38m shares and the HUI is carving new bull-to-date highs, then I am going to be considering the high probability that the HUI is short-term overbought and topping, due for an imminent correction. I would also consult all of our other indicators of course to see if anything else corroborated the message of the volume climax. If multiple unrelated indicators concurred in signaling a top, I would either sell my speculative HUI longs, buy synthetic HUI puts, throw short key HUI stocks, are at the very least ratchet up my trailing stops.
Conversely if the HUI has been correcting hard for weeks or months after a major top and I witnessed a 38m+ share spike, I would suspect the HUI had carved a major interim bottom and I would throw long if other indicators confirmed. I would close out any speculative HUI shorts, buy synthetic HUI calls, and buy HUI stocks in preparation for the probable coming upleg. HUI volume climaxes can trigger a welcome alarm for us to carefully consider market conditions, check our other primary indicators, and trade accordingly.
I donít really know what the opposite of a climax is, maybe apathy? In any case, there are also a half-dozen times when the 5dma of HUI volume slumped under 16m shares. Provocatively while these volume wastelands didnít mark major interim tops or bottoms precisely, in five of the six and possibly all six depending on what the HUI does in the months ahead traders would have done well to buy these volume slumps.
Extremely light volume tends to occur when traders are fed up and bored with a market and excitement is low. Naturally as contrarians this is exactly when we want to invest, when most other folks have given up in disgust. Thus it makes sense to consider buying the HUI whenever volume reaches abnormal lows as it likely signals major price gains ahead.
This reminds me of the timeless trading axiom ďnever short a dull marketĒ. During a secular bull when things are getting boring and traders exasperated, like this past summer in gold stocks, it doesnít take much to push the bull higher again once volume returns. If you shouldnít short a dull market, the corollary to this axiom is that you may as well consider buying it. This parallel truth would have paid out handsomely with the HUI.
The bottom line is the HUI volume, calculated by adding the individual volume of all 15 of its components, has some excellent potential to be a trading indicator. We have some more ideas on refining this and looking at the data in different ways which I want to explore in the future, but hopefully you found this initial foray useful. I donít yet know if volume alone can be a primary indicator, but I am fairly confident that it can at least be a great secondary indicator to alert us to potential opportunities approaching, after which we can check primary indicators for confirmation.
We are going to start tracking HUI volume internally and will report on it when appropriate in our acclaimed monthly Zeal Intelligence newsletter for our subscribers. I am going to watch for huge volume climaxes going forward as well as the periodic volume wastelands and try to incorporate these into our decisions on trading and recommending individual gold and silver stocks in the future. If we can develop an indicator, we will also track it in our newsletters.
Just as this idea arose because one gentleman was kind enough to share his thoughts on volume with me, I suspect there are great synergies possible by combining indicators. Different indicators illuminate this phenomenal gold-stock bull from different perspectives, and presumably the more light we can shed on the HUI from any angle the better we will grow at understanding it, trading it, and milking it for enormous profits.
Volume climaxes do indeed mark the crucial moments when the HUI parabolas are shooting vertical and an intermediate trend change is imminent. The good Colonel was right!
Adam Hamilton, CPA October 22, 2004 Subscribe at www.zealllc.com/subscribe.htm