In what seems a lifetime ago, I was the equity index trader at a big bank on Bay Street. Although a lot has changed since then, there are parts of the game that are timeless. So I am putting my old hat back on to analyze VanEck’s recent problems arising from the success of their GDXJ ETF (Junior Gold Miners). And lest you think this will be a boring ETF specific piece, I urge you to suffer through the details as I believe the market is missing the bigger picture message.
For those unaware, VanEck recently announced they were halting the creation of the their GDXJ 3x times Bull shares (JNUG) because they were running into constraints. In fact, the popularity of the GDXJ ETF product has been so overwhelming, VanEck also reported they would be changing the index rules to accommodate the increased demand. From Bloomberg:
The world’s second-largest exchange-traded fund linked to material producers has become the victim of its own success.
With investors piling in amid a precious-metals recovery, VanEck Vectors Junior Gold Miners ETF’s assets jumped 60 percent to $5.54 billion this year. That created a dilemma for the fund that tracks the MVIS Junior Gold Miners Index, as its holdings surge above 10 percent of some of the companies it owns.
On Thursday, MVIS Index Solutions, a VanEck company, announced changes to its equity indexes, widening the criteria for inclusion into the gauge that is tracked by VanEck’s junior gold miner ETF.
“This is the curse of success,” said Sameer Samana, a St. Louis-based global quantitative strategist at Wells Fargo Investment Institute, which oversees $1.8 trillion. “They’re starting to run into issues of how much they own in certain names, how many names qualify for the index and they’re running into issues of how big the fund has gotten.”
Investors poured $1.5 billion into VanEck Vectors’ junior miner ETF this year, the most among global peers tracked by Bloomberg. Assets in the most popular mining ETF, which trades under the ticker GDXJ, climbed amid rising interest in small companies as large producers emerging form the downturn scout for new reserves.
The ETF was growing so fast it was in danger of having the kind of stakes in some junior miners more usually associated with large institutional or activist investors. That, in turn, could be an issue for individual shareholders who might worry such large “passive” holdings could ease pressure on boards around corporate governance, Eric Balchunas, an ETF analyst for Bloomberg Intelligence, said by telephone.
GDXJ invests at least 80 percent of its total assets in securities in the MVIS Junior Gold Miners Index, giving it enough wiggle room to add other companies. Its third-largest holding is VanEck Vectors Gold Miners ETF, which trades under the ticker GDX and tracks the NYSE Arca Gold Miners Index.
ETF.com reported GDXJ’s rapid growth, forcing it to deviate from its underlying index on April 10, before MVIS index announced changes to the gauge.
Some of the more vocal members of the finance community mistakenly believe this is the result of the uncertainty created from Trump’s recent firing of missiles on Syria.
Although there was an increase in GDXJ buying following Trump’s Middle Eastern foray, this problem has been long in the making.
The recent Syria missile strike was not the problem. No, there has been steady increased demand for GDXJ for the past year.
Even though I disagree with my friendly bearish tweeter, he is correct that GDXJ will probably suffer in the coming the weeks. But probably not for the reasons he thinks.
Many will probably look at the increase in GDXJ shares outstanding and think there must be tons of weak longs. Yet, often violent expansions in ETF shares outstanding, are the exact opposite of bearish. These newly created ETF holders might in fact be the ‘smart money,’ not the other way round. Last year’s oil collapse was the perfect example.
So if massive ETF share expansion is a sign of ‘smart money’, why do I think GDXJ might suffer in the coming weeks?
The GDXJ ETF has become so popular, the VanEck ETF holding company is bumping up against 20% ownership in some of these junior Canadian gold miners. From a regulatory perspective, it is difficult to go above the 20% threshold without triggering automatic takeover laws. Also, from a tax perspective, there are some arcane laws pertaining specifically towards ETFs that VanEck was also battling due to the composition of the index.
Faced with all these problems, VanEck chose to dramatically alter the composition of the underlying index that GDXJ is supposed to track. They announced their intention to make these changes on June 17th, but in the meantime, have released their new guidelines for the index.
Over the years I have relied on one firm’s index research more than any other. In fact, even when I was the trader at another bank, I would desperately seek out TD’s Peter Haynes’ research pieces. Peter has been doing this so long, there is a rumour he is grooming his high school aged son to take over his post when he retires to pass along the family business.
Now here comes the important part. Have a look at TD’s index research group’s projected flows arising from this index rebalance. First the adds:
Next up, the deletes:
Look at these numbers! Most importantly, look at the “Days to Trade” column. It represents how many average days’ volume the estimated index flows will take to complete. These are big numbers.
These index flows will affect the index. No two ways about it.
In the coming weeks, the stocks underlying the GDXJ ETF will have the constituent members decline due to the selling, and the members with increased weightings (and new additions) will see their share prices increase. Of course this will all be on a relative basis. But taken together, this large demand for liquidity will result in GDXJ returning much less than would be the case had there been no index rebalance.
To avoid this drag, the obvious no-brainer move would be to swap GDXJ for GDX. It’s not quite that easy as the GDX is comprised of senior gold companies, while GDXJ owns the junior ones. The beta on the junior miners is much higher. Therefore, you might need to volatility adjust the position.
But this swap is the sort of trade needed to avoid the massive GDXJ underperformance that will be experienced due to the upcoming rebalancing.
Let’s step back and really think about this…
Although many are in a tizzy about the “potential” breakdown of the ETF model, and drawing all sorts of ominous conclusions about this hiccup, I take a much different view.
Did investors become too enamoured with GDXJ? Yup, you betcha. But why? And what might this mean for the future?
To answer these questions I refer you back to an absolutely fantastic presentation Grant Williams from RealVision TV fame gave over a year ago. It’s called “Nobody Cares.”.
In the presentation, Grant highlighted how little gold and other precious metal assets are held in pensions. At that time, only 0.15% of pension assets were held in gold and gold equities.
If that were to double, the resulting capital flows would be $100 billion. Grant then went on to demonstrate how much $100 billion would mean to the gold market. For that amount you could buy the following:
The gold market is tiny compared to pension assets. And this example only assumes an increased weighting from 0.15% to 0.30%. Those numbers are still much too low considering gold represents cheap insurance if this whole Central Bank monetary mad science experiment goes off the rails.
Back to GDXJ. While many are viewing this recent episode as an example of why gold investing is best left to the tin foil hat crowd, I wonder if they might be missing the bigger picture.
Maybe this GDXJ problem is the shot across of the bow. Maybe Grant Williams’ theory about how little it will take to move this market is starting to come true.
There is precious little room for everyone in the gold boat. This is made all too clear with the problems we quickly experienced from a little interest in GDXJ.
Can you imagine what will happen if the tide truly turns?
- ← Big Blue Batters Dow To 2-Month Lows As Crude Crashes
- Euro Banks Still Can't Value Their Own Equity Research; Subscriptions Range From $50k - $600k A Year →