Similar to many others, we have been waiting for some sort of correction in gold and gold stocks, but obviously, not much has happened in this respect so far. We have written quite a lot about gold and gold stocks between August 2015 and February 2016, because we felt a good opportunity was at hand – a short term trading opportunity at the very least, but one with the potential to become more than that.
Inside the fully mechanized South Deep gold mine in South Africa. Maybe it’s worth to keep digging for more after all?
Photo credit: Themba Hadebe / AP Photo
The strength in gold and gold stocks that could be observed since then certainly suggests that something more than just a short term trading opportunity is at hand. In early February we showed a chart of how the first corrections in the nascent stage of three previous bull markets played out, with the size of the moves adapted to the chart of today’s HUI index.
Obviously, our expectations have been exceeded quite a bit – if the HUI had mimicked the strongest of these previous initial advances (which took place in late 2000/early 2001), it would have put in a short term peak at approximately 167 points. So far, it has made it to nearly 190 points intraday without correcting much as of yet. As far as we know (we stand ready to be corrected), the HUI’s rally after putting in a new bear market low in mid January has been the strongest sector rally in all of history over such a short time period. It has risen by more than 90% in just two months.
The rally in gold has been far weaker, but this is quite normal at this stage. It has nevertheless managed to overcome a number of short and medium term resistance levels, including a major downtrend line. Gold in USD has essentially followed the gold price trend in terms of other major currencies, which hasn’t really surprised us. Readers may recall that we have highlighted this growing divergence for quite some time.
Gold in terms of selected non-dollar currencies: the euro, the yen, the Canadian dollar and the South African rand.
Obviously the extremely strong rally in gold stocks makes a short term correction more rather than less likely. Mainly its timing, extent and duration would normally be open to question. However, overbought conditions have so far been relieved by means of an upwardly skewed sideways move instead of a sharp dip.
One can see this quite well on a 30 minute chart of the HUI that shows the action over the past three months. We haven’t yet given much thought to a possible Elliott wave labeling, but it seems to us that there has to be a “running correction” in there somewhere. This is to say, a corrective structure that exhibits an upside bias and is characterized by continuing the pattern of higher highs and higher lows.
This kind of correction is considered evidence of a very strong underlying trend, as it reflects the urgency of investors to get in. Naturally, a bigger correction could easily begin at any moment. At some point the index should test its 50 day and/or 200 day moving average from above, but it is noteworthy that the sector has so far been strong even on days when gold itself has shown weakness.
A strong rally always invites an increase in bullish sentiment, as sentiment tends to follow prices. In spite of all the recent evidence of technical strength, there is however not as much enthusiasm for gold as one would expect. The severe and long-lasting post 2011 bear market has apparently really worn many market participants down, and the bear market mindset is still quite prevalent.
However, if a new bull market has begun, bear market rules simply no longer apply. This has no bearing on what is going to happen in the short term – with regard to that, we are agnostic at the moment. A fairly sizable correction certainly wouldn’t be a surprise, but as long as short term support levels remain unmolested, it is still possible that a move to new highs will happen first.
Gold with various lateral support levels penciled in. In the short term, the 20 day moving average (blue) and lateral support just below 1230 are probably important. If they were to be violated, a correction to the vicinity of one of the lower support levels would become highly likely. Even though a negative price/RSI divergence and a short term MACD sell signal have recently been recorded, the possibility of a move to a new high remains open as long as these near term support levels are holding (the same holds analogously for the gold stock indexes) – click to enlarge.
An Army of Doubters
In recent weeks there has been quite a bit of anecdotal evidence that the rally remains widely doubted. For one thing, many sell-side analysts continue to be quite stingy with upgrades. Numerous gold stocks are still sporting “sell” ratings.
In fact, lately a number of downgrades have been issued, mainly because gold stocks have allegedly “run ahead” of the gold price. We believe this argument makes actually no sense. The rally mainly reflects the fact that contrary to previously widespread expectations, gold has stopped going down.
Even negative rating changes that superficially seem to make sense, may turn out to be misguided at this juncture. This is so because they are based on assumptions about the future that are almost certain to be wrong, especially if a new bull market has begun.
We can illustrate this with an example from memory lane: when Goldcorp (GG) first hit $5 in September of 2001, it was downgraded to “sell” by a prominent sell-side firm (if memory serves it was CSFB, but don’t hold us to it). The stock had rallied by more than 100% from its late 2000 low, and all assumptions the analyst made about the future suggested there could be no more upside. Many more downgrades were issued over the ensuing years for similar reasons.
About one year after the first noteworthy downgrade to “sell” at $5, the stock hit $10. Another year later it hit $15. Over the next 6 months it suffered a big correction to $8.50, but 18 months later it was trading at $20 and another 6 months later, it was trading at $37. This was followed by another big correction back to $18, but less than two years after the correction the stock peaked at $47 (by the time the stock topped out, i.e., after rising roughly 25-fold, it had become a darling of Wall Street, natch).
We could give you a long list of similar examples, but we mainly want to make the point that it is typical that beginning bull markets are not recognized. What analysts say at the beginning of a bull market is mainly going to help you not make money and what they say when it ends is going to help you lose money. That’s just how it is.
What one needs to know about this is only this: when downgrades are issued and the stocks concerned stubbornly refuse to go down, the probability that a new bull market has begun is very high.
Quite lot of doubt has also been expressed by assorted commodity strategists and the mainstream financial media. Here are a few recent headlines:
Why 2016’s gold rush is already running out of steam (maybe it isn’t?), Gold’s ‘golden cross’ is not a golden ticket for bulls (this is by the way true, but not necessarily if a new bull market has begun…), Why Fed, U.S. Dollar Point to Bearish Gold Market (good question…why?), Goldman No Believer in Gold Rally as Fed to Hike Three Times (this one is from early February and has made the list as an example of really bad timing…), UBS on gold: Sell the rip (as far as we know, these guys are comparable to us, in that they aren’t really bearish, but are thinking it may be time for a correction) and of course…It’s too late to buy gold (not if it’s a new bull market, then it isn’t).
You get the drift – but these are just anecdotes. We could have shown a number of bullish articles as well after all. So are these expressions of doubt actually confirmed by measurable indicators of sentiment? This brings us to something we have discussed in these pages in the past, which we believe is important to consider in this context.
Sentiment and Positioning – Interpreting CoT Data
To begin with, here is a very long term chart of the so-called Optimism Index, or “Optix” for short, published by sentimentrader (the index is an amalgamation of prominent sentiment surveys and options and futures positioning). There is no hard and fast rule as to how to interpret this indicator, as this mainly depends on whether one is in a bear or a bull market as well. However, we find it interesting that it has so far not even moved to “extreme optimism” territory in the face of a technically very strong performance:
Following an extended period of extreme pessimism, the Optix remains relatively subdued in spite of a technically convincing rally. The same happened in 2000-2001, but of course that doesn’t guarantee anything – click to enlarge.
It would actually be best if this indicator were to at least briefly reach the “over-optimism” level, but the fact that it has recently slightly exceeded every previous spike since mid 2013 will have to do for the time being.
Now let us turn to the commitments of traders in the futures market. Many market observers have pinned bearish forecasts on the fact that the aggregate net speculative position (per the legacy CoT report) has recently reached a very high level relative to recent history. Here are two examples of this reasoning from a technical analyst (here and here).
Leaving aside that the “bullish percent” index isn’t what he thinks it is (it merely shows the percentage of stocks on a point & figure buy signal, not the bullish consensus), the question is whether the interpretation of the CoT report is valid. It is certainly true that the aggregate net speculative long position is one of the largest in quite some time (the chart below shows its inverse, the net short position of hedgers):
A position of this size increases the probability of a short term decline , resp. of a shake-out. In an ongoing bear market, it would in fact suggest fairly large downside potential. However, what if the primary trend has actually changed? As to this we should first take a trip back down memory lane to March 6 2013.
At the time, gold was still trading close to the apex of a triangle that had formed after the 2011 peak. Since all previous triangles since 2000 had resolved to the upside, it certainly seemed possible that this one would do the same (given that “QE3” had begun only a few months earlier, it was an especially tempting conclusion). However, important support was very close.
Gold from January to July 2013. a few weeks prior to the spectacular breakdown of mid April 2013, gross speculative short positions in gold futures had begun to increase considerably – click to enlarge.
The net speculative long position reported in early March 2013 was the lowest since December 2008. In the disaggregated CoT report, the “managed money” category had quite significantly added to its gross short position. Superficially, this seemed to be good news, as there was now room for speculative short covering and possibly new buying.
We also thought that the situation harbored bullish potential, not least because gold stocks seemed quite oversold (we obviously had no idea just how oversold they would eventually become). However, we were wary of the change in futures positioning and also noted at the time that one had to be very careful with the interpretation of the seemingly bullish CoT data. Here is what we wrote on the topic:
“One should definitely not assume that just because these short positions have risen, a market rally is necessarily imminent. In fact, it would be a very bad sign if money managers became so bearish as to flip over to a net short position.
It is short term bullish when gross and net long positions are reduced in a correction in an uptrend, but it is far less bullish when the decline in the net long position owes mostly to an increase in shorts.
Let us not forget: big speculators are the group that tends to be right about gold’s major underlying trend. Their growing skepticism is not an unalloyed positive sign as many other writers currently maintain. Quite to the contrary, it is a reason for concern. Bulls would want to see this gross short position reduced as quickly as possible.”
Needless to say, that never happened – not only was the speculative gross short position not decreased, it grew considerably over subsequent weeks and months. The gold price first suffered a mini-crash in April and then declined further – until the speculative net long position reached the lowest point since January 8 2002, after which the first sizable short covering rally began. Incidentally, the gold price stood at a mere $279 on January 8 2002 – but at the time the primary trend had just turned up.
What we are driving at is that one cannot interpret CoT data out of context – least of all nowadays, after we have seen countless examples of what we would term the “persistence of trends” phenomenon. Consider e.g. the acceleration in the yen’s downtrend after Kuroda-san appeared on the scene. The yen was consistently oversold for many months, net speculative short positions were very large, and bounces rarely amounted to much. Similar trend persistence could be observed in many other currencies as well as in industrial commodities in recent years.
It is interesting to see at what point during the uptrend in gold prices that ended in 2011 the greatest net speculative long position was recorded. This is actually quite surprising: it happened in early December of 2009, with gold trading at $1196 (then a new record high). Over the next 20 months the gold price would rise by more than 50% to an intraday high of $1923.
The record high in the speculative net long position in gold futures during the 2000-2011 bull market phase was recorded in December 2009 at $1196. It amounted to more than 308,000 contracts! – click to enlarge.
Obviously, what should be considered a large net position depends on the primary trend as well. As a final remark on futures positioning, it is also noteworthy that small traders continue to be relatively skeptical of the recent rally compared to the enthusiasm they exhibited in 2009-2012. Since 2012 we have seen this group of traders post both a record high net long position at the secondary peak near $1800 in early October 2012, and a record high net short position at the interim low at $1087 in early August 2015. These traders have embraced the recent rally only very reluctantly:
We’re not entirely sure yet how to interpret this, but small traders showed similar caution in late 2000 to early 2002 and again in the early stages of the rally after the 2008 shake-out.
What is the Difference to Previous Rallies?
Since the decline from the 2011 peak in gold began, there have been several bear market rallies, some of which were of roughly similar size as the recent one. So the question is what, if anything, makes the recent rally different? We have already discussed the multitude of bullish divergences around the lows in early February, so there is no need to repeat all this (see “Gold and Gold Stocks, a Meaningful Reversal?” for the details).
As mentioned above, by now the sheer size of the rally in gold stocks suggests that the reversal was in fact meaningful. Not only could a classical false breakdown in XAU and HUI be observed shortly before the rally started, but since then we have seen the largest increase in the HUI-gold ratio as well as the largest increase in the HUI-SPX ratio since late 2008:
We must stress that these large moves make it actually more rather than less likely that a near term correction will begin fairly soon (even if not necessarily right away). However, they also show that the market’s character has changed quite significantly. In addition to the recent move in the ratios shown above, at no point during the bear market have the gold stock indexes managed to rise as far or for as long above their 200 day moving averages as in the recent rally.
The last time short term oscillators like MACD and RSI reached overbought readings similar to those recently seen was in 2012, but at the time the gold stock indexes diverged negatively rather than positively from gold. Maybe we shouldn’t rub this one in too much, but we would note that the bullish percent index has also risen to a much higher level than in any of the preceding bear market rallies.
This is actually a medium to long term bullish signal (and note that in the short term BP can easily attain higher levels than the current reading as well). The BP index essentially represents a sector breadth indicator, as it shows the percentage of GDM component stocks that are considered to be in an uptrend in point & figure charting terms. The long term chart below shows what the difference between bull and bear markets consists of:
While the bullish percent index is somewhat stretched in the short term (which may precipitate a correction), its recent spurt is a positive medium to long term sign. Readings like the recent one are typical of bull markets rather than bear markets – click to enlarge.
There are many signs indicating that the gold market’s character has profoundly changed. We cannot rule out completely that the recent advance isn’t another fluke, as it is still early in the game. Moreover, the fundamental backdrop continues to fluctuate between bullish and neutral. However, the probability that it is a fluke seems quite low given the technical evidence.
If such a change in character has indeed occurred, then the rule book of the past four years needs to be thrown out. This means that henceforth, dips in the sector should be bought, and that different criteria have to be applied when assessing sentiment and positioning and other overbought/oversold indicators.
Readers should always keep the sector’s legendary volatility in mind. Naturally, this volatility is precisely what makes it so interesting to us. Consider the long term chart of GG shown above in this context, which doesn’t even include the crash of 2008, but shows only a nigh uninterrupted bull market period.
Some of the corrections were obviously quite scary and best side-stepped. This just as a reminder that gold and silver stocks are not to be married and that one needs to have some rules or plan in place to deal with the volatility.
Charts by: StockCharts, SentimenTrader