The flagship GLD gold ETF has suffered a radically unprecedented mass exodus this year. The capital fleeing this single vehicle was the primary reason gold plunged so dramatically in 2013's first half. But just this week, money started flowing back into GLD for the first time in months. This likely marks the dawn of the GLD exodus's reversal, which is wildly bullish for gold. Falling stock markets will play a critical role.

The GLD gold ETF is now formally called SPDR Gold Shares. Rising from modest beginnings nearly 9 years ago, it has grown into a dominant force in the global gold markets. This is because GLD acts as a conduit for the vast pools of US stock-market capital to easily and quickly flow into and out of gold bullion. These capital flows can greatly affect overall gold demand over short periods of time, buffeting gold's price.

If GLD operated like a closed-end mutual fund, issuing a fixed number of shares one time at its IPO, this wouldn't be the case. GLD would have zero impact on the gold price after its initial bullion purchase. But GLD isn't structured this way, its primary mission is to track the gold price. And there is only one way to accomplish this. Both excess supply and demand of GLD shares have to be directly shunted into gold.

When stock traders get excited about gold, they buy GLD shares at a faster rate than gold itself is being bought. Thus GLD's price rises faster than gold's, and it threatens to decouple to the upside and fail its mission. In order to prevent this and bring GLD's price back in line with gold's, this ETF's custodians have no choice but to meet this excess share demand. So they sell new GLD shares into the market.

Selling new GLD shares naturally raises cash. And the GLD custodians immediately plow these new funds into physical gold bullion held in trust for its shareholders in its London vaults. This process effectively shunts excess GLD share demand into gold itself, amplifying the underlying gold rally. So when stock traders are exerting differential buying pressure on GLD shares, it is very bullish for gold.

Ever since a gold ETF was conceptualized years before GLD was actually launched, both its advocates and opponents knew this tracking mission was a double-edged sword. Stock traders wouldn't always buy ETF shares faster than gold itself was being bought. GLD would also face differential selling pressure from time to time, when its shares were being dumped faster than gold itself was being sold.

This reverses the bullion-holding process, pulling stock-market capital out of gold or effectively shunting stock-market GLD selling into gold. When traders sell GLD faster than gold is being sold, there is excess supply of GLD shares. So its price threatens to decouple to the downside and fail its tracking mission. In order to avert this, GLD's custodians have to quickly absorb any excess supply of GLD shares offered.

They raise the cash to buy back these shares by selling some of the gold bullion out of their vaults. This of course increases the selling pressure on gold itself. So just like GLD differential buying pressure can amplify gold uplegs, GLD differential selling pressure can amplify gold corrections. Thankfully for gold bulls, we didn't have to contend with a massive differential GLD selloff until this year. Boy was it miserable!

This first chart looks at the total GLD bullion holdings in metric tons superimposed over the gold price. GLD has always been extraordinarily transparent, publishing its total bullion holdings daily right down to the long list of serial-numbered individual gold bars. GLD's holdings offer a fantastic window into stock-trader gold sentiment, showing when American stock-market capital is flowing into or out of gold.

For fully 8 years after its inception, which conspiracy theorists went apoplectic over, GLD was hugely beneficial for the secular gold bull. Stock-market investors and speculators alike bought GLD shares faster than gold was being bought, their capital literally flowing into this ETF's underlying gold bullion as its custodians kept issuing new shares to maintain price tracking. GLD has been a great boon for gold!

In any bull market, its existing investors want as much capital as possible to flow in and drive up the asset prices they are betting on. The more the merrier, the more traders are bidding on an asset the higher its price ultimately goes. GLD provided a wonderfully easy and efficient (very low transaction costs) way for stock-market investors and speculators to participate in the secular gold bull. And participate they did.

There were naturally bouts of differential GLD selling pressure from time to time, as sentiment perpetually flows and ebbs. Sometimes stock traders waxed bearish on gold and sold their GLD shares faster than the metal itself was being sold. So GLD's holdings dropped as its custodians sold gold bullion to absorb this excess supply of shares. For lack of a better term, I call these episodes GLD holdings corrections.

And for the most part they were relatively mild. All of these significant GLD holdings corrections of its entire lifespan are noted in the chart above. The blue number shows the percentage drop of GLD's holdings, the yellow number the tonnage drop, the red number what gold prices did over that span, and the white number how long each correction took in months. GLD holdings actually proved quite "sticky".

This ETF would see big differential buying pressure in gold uplegs, forcing its custodians to shunt this new stock-trader capital directly into gold bullion. But then when the inevitable gold corrections arrived, the differential selling of GLD shares was relatively minor. Typical GLD holdings corrections ran from 4% to 7% of its holdings, and the absolute amounts of gold bullion sold weren't too great for the markets to absorb.

There were occasional exceptions of course. The biggest were during 2008's once-in-a-lifetime stock panic, which also crushed gold through unprecedented US-dollar safe-haven buying. GLD's holdings plunged two separate times, by 12.6% and 13.0%. And gold took big hits over these spans, down 13.3% and 22.0%. This was partially because 83.4t and 91.6t of GLD's bullion was sold back into the gold markets.

But both GLD's holdings and the gold price quickly recovered from the stock-panic carnage in early 2009, when major hedge funds flooded into this ETF to take massive stakes in gold. There was one more large GLD holdings correction starting in mid-2010, and it saw enough differential selling pressure to force GLD's custodians to dump 9.8% of their bullion or 129.1t. That was the most gold it had ever had to sell.

But provocatively gold's price actually surged 20.6% higher over this record period of heavy GLD-share differential selling pressure! Why? Time. That biggest absolute decline in GLD's bullion held in trust for shareholders to that point took 10.6 months to unfold. So there was plenty of time for GLD's bullion sales to be absorbed by normal gold-market demand elsewhere. The duration of holdings corrections is critical.

GLD's holdings reached an initial record high of 1320.4 metric tons of gold bullion in June 2010. Around that time, I was very blessed to enjoy a private lunch with GLD's primary architect. He was rightfully quite proud that his baby had facilitated way over 1000t of gold investment demand that likely wouldn't have existed without GLD. Large stock traders are just not interested in traditional gold-coin investing.

It is too inefficient, with transaction costs in terms of both time and money way too high. A hedge fund that wants sizable gold-market exposure is not going to visit 100+ coin stores to find enough gold bullion to purchase, and then pay enormous premiums of 5%+. GLD revolutionized gold investing for stock traders. There is no doubt the vast majority of stock-market gold investment wouldn't have happened without it.

After the summer of 2010, GLD's holdings consolidated high for the next couple of years with gold prices. There were some periodic draws, but these holdings corrections were minor. Finally in September 2012 (QE3's launch), GLD's holdings started marching up to new record highs again. By early December last year, merely a little over 8 months ago, they had climbed to a new all-time record high of 1353.3 tonnes.

Now on that very day, the gold price was still down 10.1% fully 15.6 months after its August 2011 secular-bull-to-date high. At worst in gold's massive correction out of very-overbought conditions, gold had fallen 18.8% by May 2012. Yet GLD's holdings remained high and stable throughout all of this gold carnage. GLD looked to be held by strong hands, long-term investors who understood gold's bullish fundamentals.

There are some analysts today who claim the GLD liquidation experienced in 2013 was predictable. I call total bullshit on that. Up until February 2013, for GLD's entire existence it had never witnessed any bouts of serious and sustained differential selling pressure other than 2008's stock panic. And GLD's holdings recovered soon after that crazy event on strong differential buying. What 2013 saw was impossible to foretell.

The tenth major holdings correction of GLD's life slowly began in late December, and snowballed from there. In the 8 months since, its holdings have plunged by 32.8%! Even during the stock panic, the most frightening market event most of us alive today will ever witness, GLD's holdings only fell by about 1/8th. Yet this year, it has shed an astounding 1/3rd of its total gold bullion. This is a staggering 444.0 metric tons!

In absolute terms, this was over 3.4x as much bullion as GLD's custodians were forced to sell in its previous biggest correction! It was 5.3x and 4.8x as big as the pair of stock-panic holdings corrections. According to the World Gold Council, total global gold investment demand in 2012 was 1525.8t. So seeing enough differential selling pressure to force GLD's custodians to dump 29% of that was crazy.

This radically unprecedented mass exodus from GLD shares was beyond imagination, so far outside the realms of precedent set in the 8 years of trading before it. I defy anyone to find a reputable gold analyst (not a broken-record perma-bear) that was on the record forecasting such an extreme event back in late 2012. GLD had never experienced anything remotely like this epic 2013 holdings correction before.

So what the heck drove it? I am all but certain the answer is the levitating US stock markets. As far back as February, I was observing and writing about the new and incredibly high inverse correlation between the gold price and the mighty S&P 500 benchmark stock index (SPX). The levitating stock markets seduced traders into believing gold was dead and general stocks would rise forever, so they dumped GLD.

This next chart digs into this thesis, and is pretty compelling. It looks at GLD's holdings over the past year superimposed on the SPX. GLD's draws (differential selling pressure) generally accelerated when the SPX was strong, and then moderated when it was weak. If this causal chain is indeed correct, the long-overdue stock-market selloff is likely to reverse capital flows back into GLD. This is wildly bullish for gold!

Late last year, the gold market and GLD demand were operating normally based on many years of secular-bull precedent. Despite the stock markets being high in their nearly-4-year-old mid-secular-bear cyclical bull, capital was gradually flowing into gold. Stock traders were steadily buying GLD, driving its holdings up to a new record high in early December. But then an extraordinary series of events came to pass.

The SPX had stalled out in early April 2012, ready to roll over into its overdue cyclical bear. For over 5 months the SPX couldn't manage to claw its way to a single new cyclical-bull high, until September when first the ECB launched its own QE and then the Fed birthed its unprecedented open-ended third debt-monetization campaign. The Fed went on to expand QE3 in mid-December to include US Treasury buying.

Stock traders saw the Fed single-handedly reignite the tired cyclical bull, and knew Bernanke had their backs. The brutal fiscal-cliff tax hikes looming scared traders in late 2012, but when a deal was struck to avert the worst of them they started buying stocks. The SPX began levitating to a long series of new cyclical-bull highs in January. As bullishness ballooned, GLD owners started gradually selling to move capital into stocks.

This additional bullion selling began to weigh on gold, and it suffered a capitulation in February. Even then it was readily apparent the levitating SPX was a major factor, as I wrote at the time. That further spooked GLD shareholders, who ended up selling so much faster than gold was falling that this ETF had to liquidate 5.5% of its gold bullion (73.6t) in February alone just to buy back all the excess shares offered.

Though gold prices stabilized in March, averaging $1593, the differential GLD selling continued. Why? The SPX continued to surge dramatically. Gold is and always has been an alternative investment, so when conventional general stocks are thriving interest wanes. Day after day on CNBC and the rest of the financial media, the levitating stocks were euphorically extolled while gold was increasingly ridiculed.

Still, GLD's holdings only fell by 2.6% or 33.2t in March as sentiment steadied. But gold started to fall again in early April, partially on GLD selling. Soon it broke below its critical multi-year support at $1550, spawning an ultra-rare forced liquidation of long futures contracts. This triggered the sharpest gold plunge of our lifetimes, it looked like a gold panic. While GLD selling didn't drive this, it accelerated dramatically afterwards.

As GLD shareholders including elite hedge funds dumped it aggressively after that catastrophic 13.8% 2-day gold plunge, the SPX soon started surging again. Ironically it had pulled back briefly on the fear that plummeting gold had wider implications. The SPX surged dramatically in April to even more new cyclical-bull highs, and GLD ultimately suffered an outlying-record 11.7% draw that month of a staggering 142.7t!

As the SPX inexplicably continued levitating in May on pure euphoria, hitting 11 new record closes in 14 trading days, the differential GLD selling pressure continued. Everyone thought gold was dead, the cool place to be was general stocks. These big GLD draws didn't slow until late May when the SPX finally started pulling back on a Fed QE3-tapering scare. Look what happened to GLD draws during that pullback.

I shaded SPX pullbacks in red above. The differential GLD selling pressure slowed dramatically as the general stock markets retreated. The capital rotation out of GLD shares into general stocks stalled when the SPX started falling. It didn't start again in earnest until a third day of gold-futures forced liquidations was spawned by more Ben Bernanke QE3-tapering comments. That reignited festering gold fears.

Soon after that, the SPX started surging again in July despite a mediocre-at-best Q2 earnings season. As the SPX climbed to new nominal record highs, the differential GLD selling pressure continued. Despite rampant general-stock euphoria though, the rate of GLD selling was slowing. I suspect most of the weak hands had already sold. A larger fraction of the remaining GLD owners are likely strong-handed investors.

Remember how sticky GLD's holdings were in the 8 years before 2013? Long-term investors, predominately institutions, were using this ETF for essential portfolio diversification into gold. They had no reason to sell the metal, and ignored its periodic corrections. Odds are very high that the great majority of GLD's shareholders today fit that normal profile, that the hot money has already been squeezed out.

Then as the SPX started selling off again a couple weeks ago, the GLD draws again slowed and then stopped entirely. We are seeing the same phenomenon that happened during the SPX's last pullback. The capital rotation out of GLD shares into general stocks looks far less attractive when the stock markets aren't levitating day after day after day. 2013's GLD mass exodus was driven by levitating stock markets!

So when the SPX inevitably reverses, which it is again way overdue for, odds are the capital flows out of GLD into general stocks will also reverse. Add on all the other near-term bullish gold drivers like its emerging futures-driven short squeeze, and rallying gold prices are going to look far more attractive than falling stock prices. And indeed we are already starting to see GLD differential buying pressure reemerge!

Both last Friday and this Wednesday, GLD experienced enough differential buying pressure to necessitate builds of 0.2% each day. These are modest, but they are still noteworthy. It is actually the first time in all of 2013 (actually since this massive holdings correction started in early December) that GLD has enjoyed two builds in less than a single week. The GLD exodus reversal looks to be starting!

As of the GLD holdings low the day before that first recent build, 2013 had seen 152 trading days. In a whopping 99 of these, nearly 2/3rds of all of them this year, GLD suffered draws. It had only seen 7 days with enough differential buying pressure to drive holdings builds in all of 2013 to that point! So to see a cluster in the past week is a major change. I suspect the reversal has begun and will soon accelerate.

If you are interested in precious metals at all, you have to stay abreast of these GLD developments. If the radically unprecedented GLD bullion selloffs that crushed gold in the first half of this year turn into GLD buying, the entire precious-metals complex is going to rocket higher. Just this week much new data came out illuminating the GLD holdings correction during Q2 and how it affected global gold demand.

I'll be discussing all this in depth in our acclaimed weekly and monthly subscription newsletters. It is impossible to thrive in the markets if you don't know what's going on. So I draw on our decades of hard-won experience, knowledge, wisdom, and ongoing research to explain what is happening in the markets, why, and how to trade them with specific stocks as opportunities arise. If you've been hiding your head in ostrich mode, now is the time to get plugged back in. Big changes are afoot. Subscribe today!

The bottom line is the incredible differential selling pressure GLD suffered in 2013 was radically unprecedented. Nothing even remotely close had ever happened before, its holdings were actually very sticky before this anomaly. It was spawned by the surreal stock-market levitation, which seduced capital out of GLD to chase general stocks. Whenever the SPX weakens, this capital rotation slows dramatically.

With the SPX starting to sell off again, GLD is already seeing signs of a nascent reversal. As gold climbs for other reasons and general stocks fall, capital is going to come flooding back into GLD. This differential buying pressure will be shunted into underlying gold bullion, amplifying gold's upleg. Stock traders still need to buy back 437.6t worth of GLD shares to unwind 2013's anomaly, which is wildly bullish for gold!

###