A Relentless Short Term Decline
When we last discussed the gold sector, we noted that with gold approaching its 200 day moving average, a pullback had to be expected soon. In the meantime, a bit more than just a pullback has happened, as a severe sell-off started after the October FOMC announcement.
However, as you will see below, this has most likely merely reset the clock a bit in terms of anticipating a medium term trend change (even if some more near term weakness, or a short term up-down sequence retesting whatever low is put in seems possible, see discussion further below).
Gold and the HUI-gold ratio. At the most recent lows, gold diverged positively from gold stocks. It appears as though the opposite kind of divergence, with gold stocks holding up slightly better then the metal, is about to occur at the upcoming low – click to enlarge.
With regard to recent short term developments, here are a few interesting statistics mentioned by Jason Goepfert of Sentimentrader in an interview at Kingsworld last week. Our comments are interspersed in brackets:
“The sell-off in gold and related stocks is now at historic proportions in terms of persistence and severity, which has led to rallies in the metal and related stocks virtually every time. […] we have been seeing a historic level of focused selling pressure. Never before has gold been down this many days over a three-week period [as of Friday, gold has been down 15 out of 16 trading days – this has apparently never before happened]. With Friday’s down day, this was the 8th straight loss, marking the 7th such streak since 1975. Three months later gold was higher every time.
There have been 9 days when gold stocks either rallied more than 2% on a day that gold fell, or they fell more than 2% on a day that gold rose. This happened three other times over a three-month period: 2001-01-11, 2002- 08-06 and 2008-10-30. Six months later, the HUI Gold Bugs index was higher by 48%, 28%, and 58%, respectively [i.e., this is the fourth such streak in 15 years. As usual, the sample size is small, simply because such things don’t happen very often – and both major lows and highs that are usually accompanied by such extremes are by definition only occurring rarely and at specific points in time].
The biggest caveat here is that money managers are still long gold. They reduced positions quite a bit this week but remain near neutral levels at best. Previous lows in gold occurred when managers were nearly net short” [he doesn’t mention this explicitly, but this has only been true recently, i.e. during the 2012-2015 bear phase. Slight shifts in speculator positioning are not unusual near pivot points – this could e.g. be observed near the 2011 peak as well, only the other way around].
We reproduce one of the charts accompanying the report at Kingsworld here (you can see the other chartsthere), namely the chart showing the number of recent days of closing price divergences between the HUI and gold exceeding 2%.
Number of days of gold stock and gold prices diverging – these divergences tend to become especially frequent in the vicinity of lows of some significance – click to enlarge.
We want to comment on this divergences phenomenon in more detail. We have waited with posting an update because we wanted to see how the HUI Index would act as it approached the 50 day moving average. While the index got closer to the 50-day ma, it initially held up surprisingly well relative to gold. In other words, there seemed to be a decent chance that we would get to see a repetition of one of our “trend change precedents”, which we discussed here. Specifically, it appeared possible that something similar to the late 2000 pattern might occur, when the 50 day-ma held the initial pullback and became support.
With the 50 dma ultimately broken again, the clock has been reset – however, a double divergence may be about to occur here as well (a mirror image of the one with gold shown above) – click to enlarge.
Obviously, this pattern repetition is now off the table again, as the 50 day ma has been broken. However, the question remains, why do such divergences begin to proliferate in the vicinity of turning points? There is actually a fundamental reason for this. Near lows, the gross mining margins of gold miners begin to expand fairly strongly even in the face of a weak nominal gold price. In other words, the decline in input costs becomes noticeable (they may still report headline losses based on non-cash writedowns, but their actual cash flows from mining are turning positive. Recent examples for this are e.g. Goldcorp/GG and Harmony/HMY). There were a number of positive earnings suprises this earnings season, and most negative surprises were of the “headline earnings” sort, belied by a stronger operating performance.
This has led to a great deal of bifurcation within the sector as well, with many individual stocks holding up much better than the indexes (this is evident from the bullish percent chart of the GDM shown below). It also indicates something else (and the medium to long term strength of gold relative to commodities confirms this as well): namely that global economic conditions are weak and getting weaker, as this is the main reason for the decline in input costs.
The bullish percent index shows the percentage of stocks in the GDM (a broad gold mining index) which are on a point & figure buy signal. It is far from overbought (in bull market periods, it often gets up to values between 80 and 90), but it has refused to mirror the decline in the index, which means a number of stocks are actually outperforming – click to enlarge.
Positioning and Sentiment Questions
As Jason Goepfert notes above, as of last Tuesday (the cut-off day for the weekly CoT report), money managers still held a fairly large net long position, or rather, their net position had not returned to the configuration seen near previous lows in the past few years.
However, it is a good bet that a major additional shift in positioning occurred on Friday. Below is a 30 minute chart of the active December gold contract, showing the action over the past week. On Friday, trading volume exploded. It seems that a great many stops were hit on that day. So the positioning data have almost certainly improved considerably further. There is some additional evidence for this, which we show further below. As an aside, CEF’s discount to NAV is closing in on 11% (CEF is a closed-end bullion fund holding physical gold and silver) – not far from the 12% discount seen at the most extreme point so far this year (which in turn represented a long term extreme).
Gold over the past week – last Friday, trading volume spiked enormously. A large positioning shift has likely occurred – click to enlarge.
There could of course be more room for selling of gold futures from money managers, but there is some additional evidence that indicates it should remain fairly limited. In its November issue, the EWI Financial Forecast shows a chart of the 3-day average of the Daily Sentiment Index (DSI) of futures traders (data compiled by trade-futures.com). At a mere 7.3% bulls it is among the lowest readings recorded in recent years, and essentially at the opposite end of the 90% bulls reading at the 2011 peak.
Love and hate in the gold futures market, via Elliott Wave International (see links on the right hand side). The current level does not rule out additional declines, but since mid 2013, similar readings have quite often occured close to short term turning points – click to enlarge.
Almost needless to say, this is also mirrored in terms of anecdotal evidence – it is once again very difficult to find articles in the financial press that are not darkly bearish on gold. Ironically, a handful of bullish/hopeful sounding articles appeared right at the recent peak.
Fundamentals and a Possible Pattern
Finally, we want to briefly comment on the fundamental backdrop and a pattern that could possibly develop from here on out. Obviously, the threat of a rate hike by the Fed once again hangs over the gold market. This threat (they have been talking on and off about finally hiking rates for almost two years) has been a major obstacle to all gold rallies that have developed since the end of QE3.
Previously, the threat of the end of QE3 weighed on the gold market – when it finally happened, a quite sizable rally ensued. It is easily possible that a rate hike will have a similar “sell the rumor, buy the news” effect – in fact, we believe this is quite likely. As Bill Fleckenstein recently remarked, it is quite astonishing that a rate hike is considered bad for gold, but irrelevant for stocks. We believe it could well turn out to be the other way around – even though it should be inconsequential in practice whether the Federal Funds rate is at nothing or at almost nothing, especially as the Fed definitely won’t actively shrink the money supply.
The reason why it might matter after all is however that the economy appears currently weaker than at any time in the past two or three years (lagging indicators like unemployment tell us nothing about future economic trends, and there are moreover good reasons to cast aspersions on Friday’s superficially strong payrolls report). So the Fed is starting a rate hike campaign that seems destined to fizzle out very quickly again – in fact, well before it attains the status of a “campaign”.
Apart from this, the fundamental drivers of the gold price are partly bearish and partly bullish, remaining near neutral overall (this has been the case for quite some time). However, as Keith Weiner points out in today’s update on supply-demand fundamentals, by his methodology gold continues to trade well below its fundamental price as a result of speculative activity in the futures markets. In the past, this has as a rule limited the extent of the remaining short term downside.
We are resetting our “trend change watch” clock, but there are once again a few encouraging signs as the previous lows are approached. There may be more short term downside left (short term momentum is strong after all), and perhaps the previous lows will be undercut, but probably not by much. We think it is very likely that some sort of up-down pattern will be put in prior to the December FOMC meeting. Incidentally, there have been December lows over the past two years as well.
At least three of the recently observable potentially positive signals are new: the persistance of the downdraft and the large number of HUI-gold divergences observed by Jason Goepfert, and the evolving “double divergence” between the metal and the stocks. It is worth noting that such streaks have previously only occurred near pretty significant lows, so this certainly bears watching.