Gold has suffered a rough couple of months, getting pounded below major support. One driver was stock-market capital flowing out of gold again, as evidenced by renewed differential selling pressure seen in gold-ETF shares. But this was minor compared to last year’s, despite extreme bearish sentiment plaguing gold. Gold-ETF selling exhaustion has effectively been hit, paving the way for big rebound buying.

The dominant gold ETF remains the American SPDR Gold Shares, which trades as GLD. This vehicle revolutionized gold trading for stock investors, creating a quick and efficient conduit for the vast pools of stock capital to migrate into and out of gold. And since GLD just celebrated its 10th birthday this week, it’s a great time to take another look at it. Starting from humble beginnings, GLD has matured into a gold juggernaut.

If you weren’t following the precious-metals realm back in the early 2000s, it’s hard to even imagine how different the pre-gold-ETF era was. Before GLD’s introduction in mid-November 2004 kicked it off, stock traders had no easy way to prudently diversify part of their portfolios into gold. Their only options were selling stocks to buy physical gold coins, trading gold futures, or buying gold-miner stocks as a gold proxy.

But for pure stock traders, all these posed real problems. While physical gold is awesome, buying coins is an inefficient and expensive process riddled with high premiums. Gold futures are a highly-leveraged and exceptionally-dangerous game most stock traders avoid like the Black Death. Though gold stocks can be wildly profitable, they are far riskier than gold itself due to an array of serious operational risks.

The gold ETFs led by GLD gave stock traders a cheap and easy way to bypass all these alternatives to gain direct gold-price exposure. GLD held physical gold bullion in trust for its shareholders. And all it took to buy GLD shares was a mouse click and trivial trading commissions. And even with GLD’s 0.4% annual management fee, it is still far cheaper than gold coins given the high premiums they command.

GLD was created specifically for American stock investors by the World Gold Council, the industry group funded by the world’s biggest and best gold miners.  It was never intended to replace physical gold for investors, but to open up gold investing to stock traders who would never or could never (due to legal restrictions like in mutual funds) buy gold coins.  Despite the silly conspiracy theories, GLD has been a great success.

And its resulting big footprint has forever altered the gold landscape.  GLD is a tracking ETF, designed to mirror the gold price.  But GLD’s shares trade independently of gold, leading to constant supply-and-demand mismatches.  If they aren’t immediately addressed, GLD shares will decouple from gold and this ETF will fail its tracking mission.  Literally the only way to maintain tracking is for GLD to act as a capital conduit.

Excess supply and demand of GLD shares from stock traders has to be quickly equalized into physical gold bullion.  The mechanics of this are simple.  When demand for GLD shares exceeds gold’s own, GLD will be bid up faster than gold and decouple to the upside.  GLD’s custodians must step in to supply this excess share demand.  They do this by issuing new GLD shares and immediately selling them.

The proceeds from expanding GLD’s share base are then plowed directly into physical gold bullion in a matter of hours.  So excess GLD-share demand by stock traders effectively shunts their capital directly into gold itself, bidding it up faster and keeping GLD tracking it.  Whenever GLD’s gold-bullion holdings are rising, it always means that GLD-share demand from stock traders exceeds the demand for gold itself.

But conduit ETFs are a double-edged sword.  Sometimes stock traders want out of gold, and sell GLD shares at a faster rate than gold is being sold.  This excess GLD-share supply will hammer GLD’s price below gold’s and cause it to decouple to the downside.  That excess supply has to be quickly absorbed or GLD will fail.  So its custodians immediately start buying back GLD shares to sop up the surplus offered.

They raise the capital to do this by selling some of this ETF’s underlying physical gold bullion.  And that effectively pulls stock-market capital back out of gold, weighing on its price.  Stated another way, excess GLD-share selling pressure is also shunted directly into gold itself.  GLD opened up a major highway for stock-market capital to quickly flow into and out of gold, depending on the prevailing whims of stock traders.

When the gold miners launched GLD via their World Gold Council, they sought to address withering criticism from hardcore physical-gold-or-die conspiracy theorists by being hyper-transparent.  Thus every single trading day, GLD publishes an exhaustive inventory of every single gold bar it holds in trust for its shareholders.  This data includes serial numbers, refiners, weights, and purities.  This week it was 863 pages long!

Having GLD’s holdings available daily is a fantastically-useful dataset, because it effectively shows whether stock-market capital is flowing into or out of gold.  When GLD’s physical-gold-bullion holdings are rising, stock traders are buying gold on balance.  And when GLD’s holdings are falling, stock traders as an aggregate are selling gold.  Stock-market capital flows via GLD can greatly affect gold’s prevailing prices.

Differential GLD-share selling by stock traders indeed helped drive gold’s recent sharp selloff and major support breakdown over the past couple months.  They dumped GLD shares faster than gold was being sold, forcing this world-leading gold ETF’s custodians to sell gold bullion to buy back the excess share supply offered.  But provocatively, excess GLD-share supply was merely a minor factor in gold’s latest selloff.

Before we dig into that, strategic context is essential.  Back in September 2012, gold was trading around $1750.  That’s when the Federal Reserve launched its wildly-unprecedented third quantitative-easing campaign.  QE3 involved the Fed conjuring paper money out of thin air and using it to buy up bonds, monetizing them.  Naturally this was highly inflationary, and should have been great for gold like QE1 and QE2 were.

But curiously, QE3’s open-ended nature along with Fed jawboning about backstopping stock markets instead fomented a monster general-stock levitation.  Stock traders believed the Fed would step up its money printing to arrest any significant stock-market selloff.  And with this effective Fed Put, they aggressively and euphorically bought stocks while ignoring large and mounting fundamental, technical, and sentimental risks.

As the stock markets melted up last year thanks to the Fed, demand for alternative investments led by gold evaporated.  Alternatives only shine brightly when conventional markets are struggling.  This vexing dynamic led to a massive 22.8% gold plunge in 2013’s second quarter.  This happened to be the worst quarter for gold in an astounding 93 years, an extreme once-in-a-century anomaly of epic gold selling!

This unprecedented gold mass exodus primarily came from two fronts.  First was extreme futures selling by American speculators, which I’ve written a lot about.  Second was extreme selling in GLD shares.  Stock traders fled GLD at crazy rates, forcing it to vomit vast torrents of gold supply into the global market.  And naturally heavy selling spawns a vicious circle, where lower prices drive more selling forcing prices even lower.

As the Fed-levitated stock markets melted up, GLD’s selling would increase as the desire for alternative investments and prudent portfolio diversification waned.  Then when they pulled back periodically, the differential selling pressure in GLD would slow dramatically.  The inverse relationship between the benchmark stock-market levels and GLD’s holdings is striking.  And so was GLD’s impact on gold prices.

My chart today looks at the last couple years of GLD’s holdings with the gold price superimposed on top.  Since GLD’s all-time-record holdings high in December 2012, all monthly draws and builds in GLD’s holdings are noted.  And though GLD contributed greatly to gold’s once-in-a-century selling anomaly in 2013, in recent months its impact has been modest.  This is actually a very bullish omen for gold going forward!

Not surprisingly, GLD’s holdings have a strong correlation with gold.  When stock-market capital flows into physical gold bullion via this conduit, GLD’s holdings rise and that buying pushes gold higher.  But when stock traders exit gold, their selling flowing through GLD forces this metal lower.  And that’s the whole story of 2013.  Extreme GLD-share differential selling spawned by the Fed’s stock-market levitation crushed gold.

The World Gold Council does the best research into global gold supply and demand.  According to its latest numbers, gold demand dropped 11.1% in 2013 to 4081 metric tons.  That was 509t less than 2012’s.  But the largest gold-demand categories of jewelry and physical bars and coins actually grew dramatically last year, up 18.0% and 32.0% respectively.  Only one category shrunk, and that was gold ETFs.

Gold demand through ETFs swung from 279t in 2012 to an astounding and probably never-repeatable negative 880t!  Differential gold-ETF-share selling in 2013 added 880 tonnes of gold supply to the world markets, far more than that 509t total drop in global demand!  So truly without the mass exodus of stock traders from gold ETFs, gold’s price would have actually risen last year instead of plunging by 27.9%!

And of that epic global gold-ETF selling, America’s GLD alone accounted for 553t or 5/8ths of the total.  GLD is the dominant gold ETF by far, and can really impact the gold price when there are heavy supply-and-demand mismatches with its shares necessitating gold-bullion buying and selling.  GLD’s holdings liquidation alone in 2013 exceeded the total drop in world gold demand, so it’s effectively solely responsible!

That epic outlying record draw was radically unprecedented.  Remember that gold ETFs were only first introduced in late 2004, and GLD’s periodic draws before 2013’s extreme anomaly were vastly smaller.  So the gold world had never before witnessed American stock traders pulling capital out of this precious metal en masse.  Such an event was never even possible before in history before gold ETFs arrived.

While GLD’s epic draws last year were spread across every month of 2013, the second quarter was the epicenter.  That quarter alone GLD’s holdings plummeted by 252t, or over 45% of 2013’s total.  April 2013 alone, that month gold suffered a panic-like plunge when major support failed, saw a crazy 142.7t draw!  That represented over a quarter of last year’s total GLD liquidations, the pinnacle of popular fear.

The inevitable selling exhaustion was finally hit in January 2014, following a mind-boggling 41.7% draw in GLD’s holdings of 564t over 13.2 months.  Selling is always finite, there are only so many stock traders who own GLD and are susceptible to being scared into selling low.  Thus GLD’s holdings stabilized this year, despite the Fed’s vexing ongoing stock-market levitation.  They started grinding sideways.

Until the last couple of months that is.  After seeing modest GLD holdings’ builds in February, March, June, and July, differential selling pressure resumed in September and October.  The remaining stock traders owning GLD, the strong hands that weathered 2013’s extreme anomaly, were getting scared by gold’s steep selloff.  This metal slumped to support in September, and then broke that key support last month.

While it’s never wise to sell into extreme lows, we can’t blame the stock traders for capitulating on gold given the extreme bearishness in recent months.  As a full-time speculator, investor, researcher, and newsletter writer, I’m as deeply immersed in the precious-metals realm as anyone.  And to me it sure felt like the recent gold bearishness even exceeded that of the spring of 2013 and the end of last year.

In fact fear had grown so crazy-high that gold stocks were recently pummeled to apocalyptic levels.  The leading gold-stock index, the HUI, just fell to fundamentally-absurd 11.3-year lows!  The last time gold stocks had traded so low, gold was merely $350.  But earlier this month it was around $1150, or 3.3x higher.  Gold-stock levels are a reflection of gold sentiment, and hadn’t been worse for over a decade.

So the universal fear infecting gold in the last couple months was the most extreme seen at least since its secular bull was born in April 2001.  If there was ever an event to drive everyone wavering out of GLD shares, gold’s recent $1190 support break was it.  Yet despite this the differential selling pressure on GLD shares remained modest.  September and October only saw relatively-minor GLD draws of 25.1t and 28.7t.

For comparison, in 2013 GLD’s monthly draws averaged 46.0t!  So the recent GLD-share selling was almost trivial relative to that epic extreme.  Between mid-July and early November, gold’s price dropped 14.7% over 3.9 months.  In that entire span, GLD’s holdings merely fell by 67.2t.  There were multiple single months in 2013 that saw comparable or larger draws.  GLD hasn’t been a major factor in gold’s recent selloff!

So why was gold so weak then if American stock traders weren’t to blame?  Extreme selling by American futures speculators.  Every week, their total gold-futures positions are revealed in the Commitments of Traders reports from the CFTC.  And in the CoT-week span that most closely matches that recent gold drop, these guys dumped 36.6k long-side contracts while adding a breathtaking 74.1k short-side ones!

Some perspective is essential on these extreme numbers.  Specs slashed their gold-futures longs by 14.8% in that short span, while ramping their shorts by 90.6%.  Both moves resulted in heavy gold selling, at a total of 110.7k gold-futures contracts.  This is the equivalent of a jaw-dropping 344.3 metric tons of gold supply unleased by American futures speculators alone!  Obviously that dwarfs the 67.2t contributed by GLD.

Extreme futures shorting is the best kind of selling, as every single one of those contracts will soon have to be unwound.  Speculators effectively borrow gold to sell it short in dangerous highly-leveraged bets, and they legally have to rebuy that gold soon to pay it back.  So the near-record gold-futures shorting is super-bullish for gold and portends an imminent sharp short-covering rally.  But back to our GLD focus here.

While American stock traders did capitulate as gold swooned to and through $1190 support, they only contributed less than 1/6th of the identifiable American gold selling.  And that was in the most extreme fear-laden and hyper-bearish gold environment seen in over a decade!  The modest differential selling pressure on GLD shares in light of this reinforces that selling exhaustion has effectively been reached.

Selling low is dumb, there’s no polite way to sugarcoat it.  Smart investors and speculators buy low and sell high, and refuse to succumb to popular fear to do the opposite.  GLD’s remaining shareholders are far stronger and smarter than the crop that abandoned gold last year.  Their holdings are far more sticky, more likely to be permanent portfolio diversification rather than hot money spookable into fleeing on a whim.

Last week GLD’s holdings slumped to 720.6 tonnes, a 6.1-year low!  The last time they were down here was September 2008, heading into that once-in-a-century stock panic.  And gold was trading around $900.  So theoretically ignoring churn, the remaining GLD shareholders are likely still sitting on nice gains in gold.  Since they’ve held on this long, they are highly unlikely to join today’s irrational fear-blinded selling.

And if this proves true, GLD’s holdings have finally decisively bottomed and can only go higher.  That will necessitate new stock-market capital inflows into gold via GLD.  And the catalyst for stock traders returning to alternative investments led by gold will be these vexing Fed-levitated stock markets finally decisively rolling over very soon here.  With gold so incredibly loathed, there is vast room for GLD buying.

As of its recent holdings low, GLD was worth about $27b.  That same day, the total market capitalization of the elite S&P 500 stocks was $19,030b.  If only 1% of that stock-market capital would diversify into gold, GLD’s holdings would soar by 7.1x!  That equates to enough stock capital flowing into GLD to force its custodians to buy about 4400t of gold at today’s cheap prices.  And 1% is really conservative in the grand scheme.

For decades if not centuries, the most prudent portfolio-construction wisdom advocated all investors holding 5% to 10% of their investable capital in gold for diversification and insurance.  So seeing an overall 1% US allocation to gold as the stock markets roll over into what’s almost certain to be a new cyclical bear is a conservative projection.  Alternative investments shine the brightest when conventional ones are weak.

The bottom line is the recent gold selloff was not primarily driven by American stock traders dumping their GLD shares.  Though they did capitulate a bit, the great majority of the selling came from American futures speculators making leveraged downside bets on gold.  The fact that stock traders largely held strong in the most bearish gold environment in at least a decade is a very bullish portent for gold prices.

With GLD’s holdings so incredibly low, there’s vast room for major stock-trader buying as gold inevitably recovers.  As the Fed-manipulated stock-market melt-up starts cracking soon, investment demand for gold among American investors and speculators is going to soar.  And the torrents of capital that will flow back into GLD shares will be shunted directly into underlying physical bullion, catapulting gold higher.