Untangling The Monster Gold Miner Trade

How leveraged funds are wagging the GDX dog.

Reviewed by: Dave Nadig
Edited by: Dave Nadig

How leveraged funds are wagging the GDX dog.

Yesterday was a great day to be an investor in the Market Vectors Gold Miners ETF (GDX | B-73). GDX was one of my plays of the year, because I thought the sector had been well beaten down. Well, it was a great call until the last three months. The ETF has dropped 21 percent, since midsummer.

However, yesterday was a kind of redemption. GDX gained 7.4 percent when all was said and done, in a single day. It wasn't any particular announcement from a major gold miner that made the move, but a combination of reactions to the easy-money minutes of the Federal Open Market Committee meeting. Everything related to gold had a huge day as a result.

But that's not what's really interesting. What's really interesting is how the trading unfolded in GDX, a fund comprising major gold miners.

Here's my take. First, let's take a look at the chart of GDX yesterday:


Guess when the FOMC minutes came out? Any answer other than "2 p.m. ET" doesn't count. Interest in GDX—already a beleaguered play because of the 20 percent drawdown, and already being played heavily by shorts and options traders—picked up quickly.

My suspicion is that initially short covering and people offsetting options positions caused a quick run-up to about 3 p.m.

Trading Ahead Of Fair Value

And that's where it gets interesting. The red line on this chart is the fair value for GDX, the blue line is what it was it is actually trading at. By late afternoon, GDX starts trading ahead of the fair value—it makes sense: GDX is enormously liquid, and like I said, has a lot of derivative interest.

But it's also the case that GDX is a massive shareholder of all its various constituents—it owns, for instance, nearly 6 percent of Gold Corp, and in fact, it owns roughly that much of all of its holdings. If GDX moves, it's going to bring all those companies along for the ride, and authorized participants will do everything they can to book any discrepancies in price. More on that later.

Notice the enormous closing trade? That's where things get even more interesting. It turns out that 36 percent of GDX is owned not by regular investors, but by ETF issuer Direxion. Direxion runs two funds that invest solely in long and short positions in GDX, the Direxion Daily Gold Miners Bull 3x (NUGT) and the Direxion Daily Gold Miners Bear -3x (DUST). NUGT has about $700 million in assets, offset by the $180 million in DUST. Of course, their exposures are 3X that, so you'd expect a roughly $1.5 billion exposure for Direxion.

Running a fund like NUGT is complicated, because it's not promising 3X the return of GDX over the long haul. It's promising 3X the return of GDX today. And then again, tomorrow. To deliver on that promise, it has to rebalance its exposure every single day.


Understanding The Math

Yesterday, for example, it started the day with $696 million under management, so it needed $2.09 billion in GDX exposure. GDX went up a lot, and at the end of the day, the fund actually had $2.24 billion in GDX exposure (the 7.4 percent gain).

But the actual net asset value of NUGT went up 22 percent (just like you'd expect). That means the assets in NUGT this morning are $852 million, which means NUGT needs to open the day with three times that in exposure—$2.56 billion. The net difference of $311 million is new exposure NUGT needs, and it needs it as close to the end of the day yesterday as possible.

The counterintuitive part is that DUST—which is betting against GDX and had a very, very bad day indeed—also needs to buy on the close. Here's all the math in a table:

Shares out 10/7 (mm)35.275.70
NAV 10/7$19.75$31.64
AUM 10/7 (mm)$696.70$180.32
Exposure for 10/8 (mm)$2,090.09$(540.97)
NAV 10/8$24.16$24.57
NAV Gain/Loss22.32%-22.34%
GDX Gain7.40%
Existing Eposure On Close$2,244.76$(581.01)
AUM 10/8 (mm)$852.21$140.04
Exposure Needed for 10/9 (mm)$2,556.62$(420.11)
"Closing" Trade Needed$311.86$160.90
Total Buy Needed (mm)$472,759,489.43
GDX Price 10/8 3:30 PM$21.90
Shares to Trade (rough, mm)21,587,191












All told, heading into the close, the managers at Direxion know they need almost $500 million in exposure to GDX, and they're going to need it in a hurry. They don't generally go get that exposure directly; rather, they call a bunch of swap counterparties (big banks) who rebalance the leveraged swap positions the funds hold, and those counterparties then go into the market to get the trade done. All told, that means someone has to buy roughly 21 million shares of GDX, on or near the closing price.

So what were the big trades like at the end of the day?


Gee, that looks an awful lot like a 21 million share print during the market's close for the prices reflected (Bloomberg math has everything multiplied by 100, so 211Km as reported is actually 21.1 million.)


What's unusual here is seeing this all in the closing auction. In my discussion with traders, I believe it's far more common for counterparties to actually work these trades over time toward the end of the day. However, given the absolute insanity of how GDX was trading at the end of the day, I imagine discretion was the better part of valor.



The disconnect between GDX's trading price and its fair value is not serious here—at its peak, the deviation was perhaps 20 cents. But that 20 cents is blood in the water for traders trying to arbitrage out the difference through the creation/redemption process. That means there were traders scrambling to sell GDX at these inflated prices, while simultaneously buying the underlying stocks, with the hope of getting a creation unit made through Van Eck.

That creates a lot of chaos for the poor schlub whose only job is to make sure the Direxion funds had their exposure right on this morning's open. Putting in an enormous market-on-close order was probably the prudent call.

So why did the closing price come way, way down when the order was a huge buy? Because every single trader on the Street knows all of the math I just laid out here, and every single one of them is trying to figure out how to get ahead of it. They all bought ahead of the NUGT and DUST rebalance trades, hoping to sell even higher. It's called looking for the greater fool.

In this case, however, the greater fool was them.

How Does This Effect Investors?
Does this matter to you as an individual investor? It can. GDX is a weird-edge case. There aren't really any other ETFs that I know of where there's an enormous position held by just a few leveraged and inverse funds. It's also an odd day when you get this kind of a quick ramp-up in a handful of stocks and the ETFs that track them. I think it's inarguable that the presence of the leveraged and inverse funds added fuel to the trading frenzy in this case.

Does that make it a systemic problem? It's hard to say that. At every step here, GDX did exactly what it's designed to do, and while there was a lot of activity, it's also the case that this all settled out in an orderly fashion.

When we look back again tomorrow, I expect to see a decent chunk of inflows into GDX (probably not exactly 21 million shares, but something close), and the fund trading right back on fair value as it should.

NOTE: I should point out that I noodled on this for a few hours on the train last night while talking with one of the smartest ex-traders I know, a blogger who writes under the name "Kid Dynamite." He also wrote about his first impressions, which you should also read.

At the time of this writing the author held no positions in the securities mentioned. You can reach Dave Nadig at [email protected], or on Twitter @DaveNadig.

Prior to becoming chief investment officer and director of research at ETF Trends, Dave Nadig was managing director of etf.com. Previously, he was director of ETFs at FactSet Research Systems. Before that, as managing director at BGI, Nadig helped design some of the first ETFs. As co-founder of Cerulli Associates, he conducted some of the earliest research on fee-only financial advisors and the rise of indexing.