Rethinking Gold Miner ETFs

November 19, 2009

The new junior miners ETF offers investors new possibilities in gold stocks—but the GDX vs. GDXJ choice isn't as simple as it seems.
  • Beyond small-cap vs. large-cap
  • The impact of the Aussie dollar
  • Do we need a better GDX?


As Brad Zigler wrote earlier this week, the new Market Vectors Junior Gold Miners ETF (NYSE Arca: GDXJ) is off to a good start, especially compared with the Market Vectors Gold Miners ETF (NYSE Arca: GDX). But what's really going on under GDXJ's hood?

To understand that, hard-core gold investors must first understand both how GDX and GDXJ differ and where they overlap.

GDX tracks the NYSE Arca Gold Miners Index, an exchange index that understandably limits itself to U.S.-traded equities (including ADRs). GDXJ, on the other hand, is based on a novel index from small player 4asset-management, which exists basically to provide indexes for the Market Vectors ETFs. There's nothing unique about this, as many niche ETFs are based on small, self-defined indexes. But it does mean that two ETFs' selection methodologies are very different, and aren't designed to mesh.


More Than Just Small-Caps

While it's easy to just say "GDX is large-cap, and GDXJ is small-cap," that's not technically true. Although GDX can reach down to companies with just $100 million in market cap, GDXJ actually sets its bottom holding range higher, at $150 million. Indeed, at just $128 million, GDX's smallest-cap holding, Vista Gold Corp (NYSE Arca: VGZ) wouldn't even be eligible for inclusion in GDXJ.

Because both ETFs track cap-weighted indexes, however, GDX ends up being skewed—dramatically—toward large-cap miners. In fact, one criticism of GDX is the overwhelming presence of just three miners, Barrick Gold, Goldcorp and Newmont Mining (NYSE: ABX, GG and NEM, respectively), which together equal nearly 35 percent of the fund. More than that, GDX's top 10 holdings represent almost 70 percent of the fund's assets.

By contrast, the top 10 holdings of GDXJ make up just 46 percent of its assets. Even if two or more of these holdings should merge (not an impossible occurrence in the mining world), the fund's rules state that no one holding may comprise more than 8 percent of the fund.

There is, however, substantial overlap between the two indexes, since GDXJ covers the entire universe of international gold miners, selecting only those in the bottom 10 percent of market capitalization. Consequently, Coeur d'Alene Mines (NYSE: CDE), GDX's 18th-largest holding at 1.19 percent, becomes the largest holding of GDXJ; likewise for the rest of the top five holdings:


GDXJ Weight GDX Weight
Coeur d'Alene Mines Corp. NYSE: CDE 6.45% 1.19%
Silver Standard Resources Inc NASDAQ: SSRI 5.64% 1.09%
Hecla Mining Co NYSE: HL 5.47% 1.01%
New Gold Inc TO: NGD 5.42% 1.17%
Gammon Gold Inc NYSE: GRS 4.90% 0.95%


All told, the overlap is just a bit over 8 percent in terms of total assets. While that's not enough to make holding both a ridiculous proposition, it definitely means that holding both creates a bit of overweighting in the middle of the capitalization spectrum.


The Aussie Connection

After those top five companies, however, GDXJ diverges substantially from GDX, not only due to the larger weightings given to smaller miners, but also because of international exposure. GDX's somewhat arbitrary inclusion of only U.S.-traded companies and ADRs means that the fund ends up with substantial weights in Canada, South Africa and the United States.

In contrast, GDXJ has a much larger position in Australia, and ironically, the U.S.:





It's my guess that more than anything else, this exposure to Australia has had and will continue to have a profound effect on the performance difference between the two ETFs. Take a look at the two-month chart, using the GDXJ index (MVGDXJ) as a proxy for GDXJ (since the fund only started trading last week):


 GDXJ index (MVGDXJ) as proxy for GDXJ


That red line at the bottom is the slow and steady rise of the Australian dollar over the past two months, as represented by the CurrencyShares Australian Dollar Trust (NYSE Arca: FXA). This currency effect implies that even if the market had remained entirely flat, the U.S. dollar value of the Australian holdings inside GDXJ's benchmark index would still be some 8 percent higher than they were two months ago. Given the continued run in gold and gold stocks, that currency effect simply adds extra juice.

The flip side, of course, is that if and when the infamous "cheap dollar carry trade" we keep reading so much about starts unwinding and the dollar begins to strengthen, GDXJ's exposure to the Australian dollar could become a hindrance, rather than a help.


Do We Need A Better GDX?

In the end, I'm struck by one thing: diversification. I like that GDXJ doesn't concentrate so heavily in only a few names like GDX does, and instead reaches further down into the junior miners. But I hate the redundancy; to me, it just feels sloppy. For my money, GDXJ appears to be the better indexing mousetrap, as it reaches internationally for relevant companies regardless of exchange, sensibly weights them and caps any overexposure.

Wouldn't it be nice if GDX did something similar, ditching the NYSE Index and investing in the nonoverlapping 90 percent of the market that GDXJ ignores, using GDXJ's methodology?

In the meantime, GDXJ—or its index, at least—seems to be about as solid a pick-and-shovel ETF play as you can get.


Find your next ETF

Reset All