Gold certainly had a rough year in 2015, grinding inexorably lower on Fed-rate-hike fears and investor abandonment. But gold is poised to rebound dramatically in this new year, mean reverting out of its recent deep secular lows. The drivers of gold’s weakness have soared to such extremes that they have to reverse hard. The resulting heavy buying from dominant groups of traders will fuel gold’s mighty 2016 upleg.

Investment demand, or lack thereof, is what overwhelmingly drives the gold price. Investment certainly isn’t the largest component of gold demand, a crown held by jewelry at roughly 4/7ths of the total. But that is somewhat misleading, as gold’s investment merits are the primary reason Asians flock to gold jewelry. But since global jewelry demand is fairly consistent, it’s not what drives the gold price on the margin.

Investment demand is much smaller. According to the World Gold Council, it only accounted for 17.7% of global gold demand in 2013, 19.4% in 2014, and 22.0% in 2015 as of the end of the third quarter. So call investment demand something like 1/5th of total world gold demand. While that isn’t huge, it is a super-volatile demand category. That’s where gold’s biggest demand swings emerge, driving its price.

The reason gold prices plummeted 27.9% in 2013, slipped another 2.0% in 2014, and then fell 9.6% in 2015 by this essay’s data cutoff on the 29th was because investment demand first collapsed and then remained weak. Without strong global investment demand, gold is going to struggle. It is the big swing category of demand, the outlying volatile variable imposed on the steadiness of other demand and supply.

It’s hard to believe after the brutal gold wasteland of recent years, but this unique asset hasn’t always been despised. Between April 2001 and August 2011, gold skyrocketed 638.2% higher in a mighty secular bull earning fortunes for brave contrarians. The flagship S&P 500 stock index actually slipped 1.9% lower over that same span. Even after gold’s summer-2011 peak, its price averaged $1669 in all of 2012.

The collapse of gold’s critical investment demand began in early 2013. And it wasn’t a normal event, but an extreme market anomaly courtesy of the US Federal Reserve. Back in September 2012, gold traded at $1766 the day the Fed launched its third quantitative-easing campaign. QE3 was wildly different than its predecessors in that it was open-ended, its bond monetizations had no predetermined size or end date.

Just a few months later in December 2012 as gold traded at $1712, the Fed more than doubled the size of its brand-new QE3 campaign with massive new US Treasury monetizations. Starting in January 2013 the Fed would conjure up $85b per month out of thin air to buy bonds. The purpose of QE3 was to manipulate long-term interest rates lower, which the Fed openly admitted. QE3 radically distorted the markets.

QE3’s undefined open-ended nature made it a powerful psychological-operations weapon to use on traders to actively manipulate their sentiment. Whenever the stock markets started to sell off, elite Fed officials would run to their podiums to declare their central bank was ready to expand QE3 if necessary. Traders interpreted this just as the Fed intended, believing the Fed was effectively backstopping stock markets!

So traders started pouring increasing amounts of capital into the stock markets, levitating them. With a Fed Put in place, a notion that Fed officials aggressively fostered, traders increasingly ignored all the conventional stock-market indicators. They bought and bought and bought, sentiment, technicals, and fundamentals be damned. This relentless stock buying gradually became a self-fulfilling prophecy.

As the stock markets seemingly magically levitated thanks to the Fed’s deft use of QE3’s undefined nature, they started sucking capital away from other asset classes. Investors love to chase performance, and stock markets were powering relentlessly higher leaving everything else behind. So they started to sell other assets to move that capital into the red-hot stock markets. Gold was collateral damage from this migration.

This mass exodus of investment capital from gold was most evident in the flagship GLD SPDR Gold Shares gold ETF.  After hitting an all-time-record gold-bullion-holdings high of 1353.3 metric tons just 3 trading days before the Fed greatly expanded QE3 in December 2012, stock investors started to dump GLD shares faster than gold in early 2013.  This soon snowballed into a wildly-unprecedented record selloff.

Since GLD’s mission is to track the gold price, it has to act as a direct conduit for stock-market capital to slosh into and out of physical gold bullion.  When GLD shares suffer differential selling beyond what is going on in gold, their price threatens to decouple from gold’s to the downside.  GLD’s managers have to avert this failure by shunting that excess share supply directly into gold itself.  This requires selling gold bullion.

GLD’s managers sell enough of its gold holdings to raise sufficient cash to buy back all the excess GLD shares being offered.  In 2013 as the Fed’s extraordinary QE3-stock-market levitation blasted the S&P 500 29.6% higher, stock investors dumped GLD shares so fast that its holdings plummeted 40.9% or 552.6t that year!  GLD selling alone accounted for over 5/6ths of 2013’s total drop in overall global gold demand.

As GLD was forced to hemorrhage vast record torrents of gold bullion into the markets, another group of traders piled on to ride gold’s downside. The American gold-futures speculators, whose trading has the greatest impact on gold’s price by far, started short selling gold futures at extreme levels. This added to the paper supply of gold, forcing down the benchmark gold-futures price off of which the physical metal is priced.

The more American stock investors jettisoned GLD shares, the faster gold fell. The faster gold fell, the more American futures speculators ramped their short selling. All the resulting gold carnage forced the other futures speculators long gold futures to greatly pare their bets, adding still more selling to the mix. The result was the most devastating vicious circle of selling gold has ever seen, spawning recent years’ wasteland.

As gold fell due to extreme selling driven by the Fed-levitated stock markets sucking investment capital out of it, traders tried to rationalize those losses as fundamentally-righteous. So the futures speculators started to believe first the tapering of the size of QE3’s monthly bond monetizations, then the end of QE3’s new bond buying, then later the Fed’s first rate hike would wreak more havoc on devastated gold.

These rationalizations were always weak though, simply masking self-feeding selling driven by out-of-control bearish sentiment. Remember gold traded at $1766 and $1712 when QE3 was originally born and expanded in late 2012. So gold plunged sharply during QE3. If that largest inflationary event in world history was ludicrously very bearish for gold, then why would the end of QE3 prove bearish as well?

The coming slowing and end of QE3’s epic monetary inflation was the boogeyman used by traders to justify aggressively selling gold in 2013 and much of 2014. QE3 was first tapered in December 2013, and its new bond buying fully ended in October 2014 even though none of those vast monetized bonds have been sold yet to this day. Once QE3’s new bond buying ended, traders shifted their boogeyman to rate hikes.

Fed-rate-hike fears were used to justify futures speculators’ and stock investors’ ongoing gold selling in 2015. Their rationale was simple. Since gold yields nothing, it will be far less attractive in a rising-rate world where yields climb on competing investments. But again this justification was totally emotional, a reflection of extreme popular bearishness that had nothing to do with gold’s actual global fundamentals.

The fatal flaw with this Fed-rate-hikes-are-gold’s-nemesis thesis is that history proves just the opposite.  I’ve extensively studied gold’s performance within the exact spans of every Fed-rate-hike cycle since 1971.  It turns out there have been 11 of them, through all of which gold averaged a stellar gain of 26.9% while the Fed was hiking rates.  In the majority 6 where gold rallied, its average gain was a staggering 61.0%!

In the other 5 Fed-rate-hike cycles where gold lost ground, its average loss was an asymmetrically-small 13.9%.  Gold’s best performance within Fed-rate-hike cycles occurred when it entered them near secular lows and they were gradual.  With gold just off major secular lows today, and the coming Fed-rate-hike cycle promised to have the slowest hiking pace ever witnessed, it’s incredibly bullish for gold’s fortunes.

If higher rates really kill gold, history would be riddled with examples.  The Fed is currently estimating that the federal-funds rate it targets will hit 1.25% to 1.5% in 2016.  While that’s a lot higher than recent years’ 0.0% to 0.25% range, it is still trivial by historical standards.  Between January 1970 and January 1980, gold skyrocketed 2332% higher when the FFR averaged a super-high 7.1% and gold still yielded zero.

During that later 638% secular gold bull between April 2001 and August 2011, the FFR still averaged 2.1% over that span despite the advent of the Fed’s zero-interest-rate policy within it.  Gold has no problem at all rallying mightily during Fed-rate-hike cycles and in much-higher-rate environments as long as global investment demand is strong.  All those rampant gold-to-plunge-due-to-Fed-policy fears are baseless.

Thus as 2016 dawns, the whole premise for selling gold near major secular lows is totally wrong.  Gold only hit these lows because investment demand remained low as extreme bearish psychology choked out all logic and reason.  But the Fed actually hiking rates for the first time in 9.5 years, ending 7 years of ZIRP, will serve as the acid test to shatter these false notions.  Gold hasn’t plunged post-hike as widely forecast!

And that means gold-futures speculators and stock investors alike are going to have to seriously rethink their whole gold thesis.  As they realize that rate hikes won’t slaughter gold, investment demand is going to start returning.  And coming from such epically-extreme anomalous lows, it is going to take a massive amount of gold buying to restore normalcy in the gold market.  That normalization is inevitable in 2016.

Major gold uplegs have three distinct stages of buying, with groups of traders handing off the baton like a relay race.  New gold uplegs are initially ignited and fueled by speculators buying gold futures to cover their risky hyper-leveraged shorts.  That sparking surge of buying lasts several months or so, propelling gold’s price high enough to get speculators interested in redeploying capital in long futures positions again.

Unlike short covering which is mandatory and forced as gold rises to prevent speculators from getting wiped out, long buying is totally voluntary and far less frantic.  So it can take a half-year or more for the speculators to reestablish their upside bets on gold.  That extends gold’s new upleg long enough and high enough to convince investors with their vastly larger pools of capital to start returning, which takes years.

Today gold is in an unprecedented position where the coming speculator gold-futures short covering, speculator gold-futures long buying, and stock investor GLD-gold-ETF buying is all aligned to be utterly huge!  As the extreme anomalies of recent years spawned by the Fed’s stock-market levitation unwind, the vast gold buying necessary to mean revert that market to norms is going to fuel a mighty new gold upleg.

Let’s start with the futures speculators, the early buyers necessary to get gold moving higher again for long enough to motivate investors to return.  This chart shows American futures speculators’ total short-side and long-side bets on gold weekly over the past several years.  These guys are so far out over their skis on the bearish side of this trade it is mind-boggling, and their only way out is extreme gold-futures buying.

American speculators’ aggregate gold-futures positions are released every Friday afternoon current to the preceding Tuesday’s close in the CFTC’s Commitments of Traders reports.  The latest read when this essay was published was December 22nd’s.  That was the week surrounding the Fed’s rate hike which was supposed to obliterate gold.  Yet since gold didn’t plunge as expected, speculators quickly covered.

They bought 15.5k gold-futures contracts that CoT week, cutting their total shorts from near-record levels to 167.5k contracts.  But that is still extremely high by all historical standards, not far from the all-time record of 202.3k in early August.  Even during the recent Fed-distorted years, speculators’ gold-futures short-side bets generally meandered in the trading range between 75k to 150k contracts shown above.

Merely to return near recent years’ 75k-contract support for the fifth time since late 2013, speculators are going to have to buy 92.5k gold-futures contracts to offset and cover their shorts.  And to mean revert to total speculator shorts’ normal-year average levels of 65.4k between 2009 and 2012 before the Fed’s stock levitation started, these traders have to buy a staggering 102.1k contracts.  That’s an incredible amount of gold!

Each gold-futures contract controls 100 troy ounces of the metal, so that equates to total gold buying in speculators’ short covering alone of 317.6 tonnes!  For an idea of how enormous this is, quarterly global gold investment demand in 2015 up to Q3 averaged 228.0t.  So we are talking about overall world investment demand soaring 139% on speculator short covering alone within a condensed several-month span!

Short covering unfolds so rapidly because traders are legally obligated to effectively pay back the gold they effectively borrowed to sell short.  And the leverage in gold futures is so extreme that they can’t afford to wait to cover once gold starts rallying.  A single gold-futures contract controls $107,000 worth of gold at $1070, yet only requires a maintenance margin of $3750.  That makes for extreme leverage of 28.5x!

A mere 3.5% rally in the gold price at that kind of leverage would wipe out 100% of the capital risked by fully-margined gold-futures speculators.  So gold-futures short covering rapidly feeds on itself, with all the covering buying blasting gold’s price rapidly higher which forces additional speculators to cover their own shorts.  The more short covering, the faster gold rallies.  The faster gold rallies, the more shorts are covered.

By the time gold-futures short covering has run its course and fizzled out, speculators buying long-side gold futures start returning.  And their bets are exceedingly low right now, which means they also have huge buying to do to mean revert to normal.  As of that latest CoT report on the 22nd, American futures speculators only held 189.7k long-side gold contracts.  That’s their lowest level since way back in April 2014.

In those last normal years between 2009 to 2012, speculators averaged weekly long-side gold-futures positions of 288.5k contracts.  Merely to mean revert to those normal levels without even overshooting would require 98.8k contracts of buying equivalent to another 307.2t of gold!  In total, American futures speculators alone need to buy the enormous equivalent of 624.8t of gold simply to normalize current extremes!

To put this into perspective, in all of 2013 and 2014 global gold investment demand ran 784.8t and 819.1t per the World Gold Council.  In 2015 current to Q3, that number is running 684.0t.  Annualize the latter and average these years, and you get yearly gold investment demand of 838.6t.  American futures speculators alone are almost certainly going to buy 3/4ths as much gold in 2016 on top of all that normal demand!

And all that mean-reversion gold-futures buying will propel gold high enough for long enough to start to convince investors to return.  And their gold positions are as extreme today as speculators’ gold-futures ones, a guarantee of massive normalization buying coming.  While physical bar-and-coin investment is larger than ETFs, GLD’s highly-transparent daily data is representative of radical underinvestment as a whole.

This chart reveals the total value of GLD’s gold-bullion holdings divided by the market capitalization of that benchmark S&P 500 stock index.  If offers a glimpse into the proportion of stock investors’ portfolios that are deployed in gold.  And thanks to the epic gold bearishness in recent years based on those false notions on the extreme Fed policies’ implications for gold, American stock investors are radically underinvested.

For many centuries, wise investors have recommended every portfolio have at least a 5% allocation in gold.  It is the ultimate insurance policy, a unique asset that moves counter to stock markets.  When something bad happens with the other 95% of one’s investments, that mere 5% gold allocation will often multiply enough to offset most of the other losses.  That old 5% target is probably gold’s full-investment level.

But as of the end of November the last time we calculated the S&P 500’s market capitalization, the ratio of the value of GLD’s holdings to the S&P 500’s collective market cap was just 0.115%.  American stock investors had just over a measly one-tenth of one percent of their capital invested in gold!  That is incredibly low by all historical standards, a Fed-driven anomaly that is as ripe to mean revert as gold-futures bets.

During those last normal years between 2009 to 2012, this GLD/SPX ratio averaged 0.475%.  Seeing stock investors with that nearly 0.5% portfolio allocation to gold is a reasonable conservative baseline.  Just to return to 0.475% would require gold’s portfolio allocation to soar 4.1x from the recent super-depressed levels.  Vast amounts of stock-market capital would have to deluge back into gold to make this happen.

GLD’s holdings averaged 1208.5t between 2009 to 2012 before the Fed’s stock-market levitation sucked so much capital out of other investments including gold.  This week, GLD’s holdings were way down around 643.6t.  So to return to pre-QE3 GLD-investment levels, stock investors would have to buy up enough GLD shares to force this ETF’s managers to purchase another 564.9t of gold bullion in coming years!

And that’s third-stage gold-upleg investment buying on top of first-stage speculator gold-futures short covering and second-stage long buying!  While it will probably take years instead of months to normalize levels of gold portfolio allocations for stock investors, that’s still a tremendous amount of marginal new gold investment demand.  564.9t of GLD buying over 2 years is 282.5t per year, and over 3 years is 188.3t.

The average quarterly gold investment demand in 2015 up until Q3 was 228.0t, so we’re talking about an additional 0.8x to 1.2x a quarter’s gold demand per year on top of all other gold demand.  And don’t forget that GLD is just a window into one aspect of gold investing, ETFs.  Global physical bar-and-coin demand is way larger than ETF demand, and the radical underinvestment there is similar to what GLD has revealed.

So with the gold positions of speculators and investors alike so radically skewed by the Fed’s extreme market distortions of recent years, vast mean-reversion buying is inevitable in 2016 to start to normalize gold investment back to reasonable levels.  Coming off of such an anomalously-low base where virtually everyone loathes gold, all this speculator and investor gold buying is going to fuel a mighty gold upleg.

Gold’s performance will trounce the stock markets’ in 2016, and it can be played via that GLD gold ETF or physical gold bullion.  But the coming gains in the left-for-dead gold stocks will dwarf those in the metal they mine.  With their stock prices recently trading near fundamentally-absurd levels relative to their current profitability, down near extreme 13-year secular lows, gold stocks should be 2016’s top-performing sector.

The bottom line is gold is poised for a mighty upleg in 2016 after being abandoned during the Fed’s surreal stock-market levitation.  That sucked incredible amounts of capital out of other assets including gold, which speculators and investors alike jettisoned with a vengeance.  The resulting bearishness left gold-futures speculators’ bets at epically anomalous levels, and stock investors radically underinvested in gold.

They tried to rationalize their extreme gold selling with Fed-rate-hike fears.  But now that the rate hike has happened and gold refused to collapse as advertised, traders will have to start normalizing all their hyper-bearish gold positions.  This will require vast buying by both speculators and investors, greatly boosting gold investment demand which will fuel a mighty new gold upleg in 2016.  Are you ready to ride it?