By Pater Tenebrarum
Are Commodity-Focused Investment Vehicles Producing Distortions?
Several observers have begun to wonder why gold stocks have so far been unable to consistently outperform the gold price, in light of the fact that input costs have begun to decline noticeably. In line with this, a number of gold producers have reported improvements in mining costs in recent quarters, which have in many cases accelerated of late (in some cases write-offs continue to weigh on net earnings, but these are non-cash events; also, the value of written down reserves will immediately increase again if/when the gold price rises).
Photo via mining-magazine.com
We have previously remarked on the strong correlation between the HUI-gold ratio and the metals and mining stock ETF, the SPDR S&P Metals and Mining ETF (XME). This correlation at times seemed to make little sense. Our conclusion was that it is likely caused by the financialization of commodities during the expired commodity boom. In particular, investment funds focused on resource companies, ETFs, assorted tracker vehicles and indexes serving as benchmarks, occasionally appear to trigger indiscriminate selling and buying across the "metals mining" sector, regardless of what metals the companies concerned actually mine.
Prior to the era during which all sorts of commodities and commodity producing sectors became ETFs (a symptom of the general move to transform commodities into "investable assets" for investors who are not trading futures or are looking to obtain broad sector exposure), the behavior of gold stocks relative to other metal mining stocks as well as trend in all these sub-sectors relative to the prices of the underlying metals seemed to make more sense most of the time.
However, in recent months the relatively tight correlation between the HUI-gold ratio and XME appears to be breaking down to some degree. Nevertheless, the extremely strong downtrend in XME over the past few weeks and days has likely weighed on the gold sector, keeping many producer stocks from advancing as much as one might have expected during the recent rebound in gold. One must also suspect that strength in South African gold stocks due to the big rally in the Rand gold price has contributed a bit to the correlation breaking down.
To illustrate this, here is the HUI gold ratio compared to XME and the gold-copper ratio (we discuss the importance of the latter further below):
Click to enlargeThe HUI-gold ratio and XME have tracked each other very closely until late September 2015. Since then, the correlation has become progressively less pronounced. The gold-copper ratio by contrast has been in a fairly strong uptrend since May 2015 (a long term uptrend persists already since 2007) - click to enlarge.
If we look at a longer term chart of the HUI - XME ratio, and compare it in turn to the gold-copper ratio, we see that there have been times when the trend and momentum in the HUI-XME ratio seemed to make sense and times when it didn't seem to make much sense:
Click to enlargeHUI-XME ratio and gold-copper ratio between 2007 and 2016. During the period in the red rectangle on the left hand side (2007-2012), the trend in the HUI-XME ratio was in line with what one would expect. In 2012 to late 2015 (green rectangle) the HUI-XME ratio seemed to fall far too much relative to the gold-copper ratio. Since then however, the ratio seems to be moving toward a more sensible level again. It is noteworthy that the gold-copper ratio has been in an overarching uptrend since 2007 - click to enlarge.
In order to bring this into context with the more distant past, the chart below shows the following: The HUI-gold ratio, GYX (the industrial metals index as a proxy for XME) as well as the gold-copper ratio in the years surrounding the low in gold stocks in late 2000. GYX is a fairly reasonable proxy for XME, as GYX and XME have correlated fairly closely since 2008, although XME has slightly underperformed GYX over time (the prices of base metal producer stocks have evidently discounted the decline in base metal prices in advance to some extent). Since XME didn't exist in 1999 - 2002, GYX will have to do.
Click to enlarge1999 - 2002: the HUI-gold ratio, GYX and the gold-copper ratio (the red and blue dotted lines illustrate diverging trends or diverging trend momentum between the HUI-gold ratio and GYX - the colors have no significance). This chart makes somewhat more sense, whereby our focus is especially on the HUI-gold ratio and the gold-copper ratio. Note that the HUI-gold ratio essentially doubled before the rally in gold itself really got underway - the undervaluation of gold stocks was so to speak removed in anticipation. One of the causes or symptoms of this anticipation was the trend change in the gold-copper ratio, which happened concurrently with the trend change in the HUI-gold ratio and remained aligned closely with it thereafter. While gold made a secondary low in April of 2001, copper only found its low in 2002, around the time the post-tech mania bear market in stocks bottomed out. The China boom got underway in 2002 as well, although few people recognized it at the time - click to enlarge.
Naturally, fundamental conditions were in many ways different in 1999-2002 compared to today's. However, it certainly does make sense for the HUI-gold ratio to rally concurrently with the gold-copper ratio, even if the latter's rise is largely due to copper declining while gold remains in a sideways range. A rising gold-copper ratio is one of the many symptoms of a deterioration in economic confidence. This is precisely when gold mining margins are usually noticeably improving, as input costs are declining relative to revenues in these periods (or putting it differently: the real price or purchasing power of gold increases).
Panic in Base Metal Stocks
We believe that a lot of trading that is so to speak done on "autopilot" may be influencing the action in gold stocks relative to gold at the moment. This is likely due to this fairly illiquid sector being impacted disproportionately when selling in commodity producing companies takes place in order to align ETF holdings with the dissolution of ETF baskets, or in order to track broad-based mining stock indexes.
However, it also seems to us that this is mainly relevant in the short to medium term. Occasional longer-lasting divergences are likely due to other reasons. For instance, the seemingly exaggerated downtrend in the HUI-XME ratio in 2013 may well have been due to market participants having greater reservations about the profit margins of large gold producers than those of large base metals producers at the time, even though the long term uptrend in the gold-copper ratio remained intact. The margins of the former were probably slimmer to begin with and faith in China's demand for industrial commodities was still strong.
This assessment is evidently changing lately. In recent days the entire resources sector has in addition been jolted by the bankruptcy filing of coal producer ACI, which inter alia inflicted a lot of collateral damage to base metals producers with large debt loads. An example for this is Freeport McMoran (NYSE:FCX), which some genius analyst finally downgraded when it hit $4. At that point it was down 90% from its interim high of a mere 18 months ago. FCX is certainly not without risk given its large debt, but if its 2016 production, capex and cash flow guidance pans out (see pages 7 and 20 of the Q3 2015 presentation for the figures), the timing of the downgrade could well turn out to have been ill-chosen.
As we understand it, the company has undertaken large capital investments in 2014-2015 so as to bring cheaper, more efficient production online that will replace higher cost production that is being idled this year, which should sharply improve its financial ratios. Anyway, the main point is actually that sentiment in certain base metal stocks has turned from merely negative to outright panic:
Even though the HUI gold-ratio remains relatively weak and still diverges greatly from the gold-copper ratio, its trend has clearly begun to decouple from that of XME in October 2015. This could be a first sign of well-worn longer term correlations beginning to reassert themselves. If upcoming quarterly earnings continue to show noteworthy cost improvements, this change in trend may be solidified. The caveat is that all of this is based on the assumption that the gold price doesn't fall out of bed and at least remains fairly stable (we currently see no reason why it shouldn't, but we have been mistaken about this before).
Lastly, traders and investors should keep the short term correlation between XME and the HUI-gold ratio in mind. Once the recent panic conditions in XME subside, it should lend some support to the short term performance of gold stocks. On days when the HUI declines disproportionately relative to the gold price, it pays to take a look at XME to see if it is exceptionally weak as well. If so, then relative weakness in the HUI has little predictive value. In other words, these occasions can at times represent short term opportunities.
Charts by: StockCharts