The precious metals plunged last week, knifing through key support zones to unleash an explosion of bearish sentiment. This troubling heavy selling wasn’t news-driven, it emerged out of the blue. Who was dumping gold and why? Later data confirmed it was American futures speculators short selling gold and silver at record levels. Extreme shorting is very bullish, as these bets soon have to be covered.

The gold and silver price action has been exceedingly anomalous since early 2013. That’s when the Federal Reserve fomented a melt-up in the general stock markets, through both monetizing debt and jawboning implying it was backstopping stock prices. The levitating stock markets gradually sucked all life away from alternative investments including gold, which was crushed by epic selling of gold-ETF shares.

Western investors all but abandoned gold, leaving it solely at the mercy of American futures speculators. Their collective buying and selling always affects the gold price, but its impact is really amplified in this surreal world of withered investment demand. Since 2013’s extreme record gold-ETF selling petered out in early 2014, gold and silver prices have been utterly dominated by trading in the US futures markets.

Speculating in gold futures is a super-risky hyper-leveraged game. Last month the CME Group cut the margin requirements on gold and silver futures by 7.7% and 8.3%. Now traders only need to keep $6000 in their account to control a single 100-ounce gold contract, and $8250 for a 5000-ounce silver contract. And at $1250 gold and $19 silver, these contracts are worth a whopping $125,000 and $95,000 respectively.

Thus a speculator running minimum margin has maximum leverage of 20.8x in gold futures and 11.5x in silver futures. This is astoundingly high. In the stock markets, the Federal Reserve’s Regulation T has legally limited leverage to just 2.0x since 1974. At 20.8x leverage, a relatively small 4.8% move in gold futures in the wrong direction from any speculator’s bet will wipe out 100% of the capital they risked!

So futures speculators simply can’t afford to be wrong for long, even if they aren’t running minimum margin. On Tuesday May 27th, they greatly expanded their bets that gold and silver prices would just keep drifting lower. So gold and silver plunged 2.1% and 1.9% that day, driving the final nail in the sentiment coffin for many precious-metals bulls. At the time, the source of this heavy selling wasn’t clear yet.

Futures speculators can sell PM futures, effectively adding supply and driving down prices, in two very different ways. They can liquidate longs, the worst kind of selling. Once traders sell their long gold and silver futures positions, they have no obligation to return. They can stay out for a long time, certainly not helping the precious metals’ prices. So paring longs can sometimes be a very bearish omen.

Futures speculators can also short sell PM futures, effectively borrowing the metals from someone else before dumping them. They hope to buy back the futures later at lower prices to repay their debts, pocketing the price drop as profit. But unlike long liquidations, short selling is very bullish because traders are under a legal contractual obligation to buy long contracts in the near future to cover their shorts.

Futures short selling is guaranteed near-future buying, as every single contract sold short has to be bought back which adds demand. This bullish covering occurs soon for two reasons. First, given the extreme leverage inherent in gold and silver futures, speculators can’t risk being heavily short for too long.Second, futures contracts have built-in expiration dates. They have to be covered before those.

Thankfully the serious selling that crushed gold and silver last Tuesday was exclusively American speculators’ heavy futures short selling. We couldn’t know that until late last Friday afternoon, when the Commodity Futures Trading Commission released its weekly Commitments of Traders report on futures positions. Until 3:30pm Friday rolled around, I was worried last Tuesday was a long liquidation.

American stock traders, whose epic GLD-share selling was responsible for gold’s entire plunge last year, played no role in last Tuesday’s precious-metals breakdown. That day GLD’s gold-bullion holdings actually surged higher by 1.1% or 8.4 metric tons. That was their largest daily build in percentage terms since August 2011, and absolute tonnage terms since October 2012! Were stock traders buying gold?

I doubt it. The flagship GLD gold ETF’s mission is to track the gold price. So if stock traders weren’t selling GLD shares as fast as futures traders were dumping gold, GLD’s price would have decoupled to the upside. GLD’s custodians had to add new share supply to keep its price falling fast enough to keep pace with gold. So they issued new shares, and used the proceeds to buy more physical gold bullion.

Since GLD didn’t experience differential selling that breakdown day, American futures speculators had to be the culprit. And Friday’s CoT proved they were,on the short side. This first chart looks at the GLD price superimposed over speculators’ weekly total long and short positions in gold futures, taken from those CoT reports. When I saw this latest CoT data last Friday, my eyes nearly popped out of my skull.

In that CoT week ending on Tuesday May 27th, the day gold and silver plunged, American futures speculators sold short an astounding 30.5k gold contracts! This is a staggering number, the equivalent of 94.8t of gold dumped onto the markets in a single week. Global investment demand, which the World Gold Council says averaged 30.2t per week in 2012 and 14.9t last year, simply couldn’t absorb this burst of supply.

You can see the sharp jump above in the red line, speculators’ total gold-futures shorts. I couldn’t ever remember such a massive shorting spree, so I checked the CoT data back to 1999. In that entire 15.4-year span, encompassing the end of the last secular bear and this entire secular bull, this was the biggest weekly jump in spec gold-futures shorting ever! It may even be an all-time record, it was that extreme.

In fact, there were only two other times since 1999 when specs’ gold-futures shorts soared more than 25k contracts in a single week. Both were last year, in mid-February and mid-November. During those smaller episodes of extreme shorting (30.1k and 27.9k contracts), gold prices plunged by 2.8% and 3.3% in those CoT weeks. Yet this metal merely slipped by 2.3% in this latest new record CoT week.

The fact that extreme futures shorting is having a diminishing impact on the gold price was a pleasant surprise. Given the universal hyper-bearishness plaguing gold these days, and the total abandonment of it by Western investors, you’d think a record shorting assault by futures speculators would ignite a crash-like plummet. But instead gold only retreated around 2%, not a big move even right near lows.

Why? Gold has spent this past year bottoming and basing. It first hit $1200 late last June, and hasn’t traded much lower ever since despite facing the roaring headwinds of the Fed’s ongoing stock-market levitation. You can see that bottoming and basing above clearly in GLD share prices. And each time gold refuses to break to major new lows, fewer American futures speculators are willing to short it again.

Their total gold shorts hit a record 178.9k contracts back in early July after those initial gold lows. And that massive short ramp hammered gold and GLD prices. In fact, over the entire past year the single most important and dominant driver of GLD has been American speculators’ gold-futures short selling and subsequent covering. There’s been a nearly-perfect inverse correlation between their total shorts and GLD prices!

So if you want to know where gold and GLD are going, you have to watch the speculators’ shorting. They covered aggressively after their extreme bets last July didn’t pan out, driving gold sharply higher. Then they started ramping their shorts again late last year to 150.0k contracts in anticipation of the Fed starting to taper its QE3 bond-monetization campaign. They figured both QE3 and the end of QE3 were bearish for gold!

Yet that bet was again spectacularly wrong, so they soon had to cover again which drove the sharp early-2014 rally in gold and GLD shares. Despite all their sophistication, as a herd American futures speculators’ bets on gold are apowerful contrarian indicator. Right exactly when they are most bearish, having the largest total short positions and lowest total longs, gold is on the verge of major new uplegs.

This latest record surge in specs’ gold-futures shorts last week merely boosted their total up to 121.4k contracts, far below the last two peaks. If gold soon rallies sharply from here, which it certainly ought to given these huge new short positions that have to soon be covered, these peaks in spec shorts form a major downtrend. Each time gold drifts to lows, fewer and fewer speculators are willing to bet against it.

And that makes sense, as gold has stabilized in this past year despite enormous ongoing GLD holdings liquidations in the second half of 2013. That means there are big buyers out there absorbing this supply, Eastern investors.While Americans foolishly like to sell low and buy high, hating cheap gold and loving the dangerously overvalued stock markets right now, Asians play this game the opposite smart way.

Interestingly not all American futures speculators are blinded by groupthink and the herd mentality. A majority of traders have been gradually adding to their long-side gold-futures positions all year, shown above by the green line. In fact when their peers were selling short 30.5k gold-futures contracts last week, they actually added 3.1k long-side ones. This was serious contrarian buying given gold’s plunge!

With the legions of bearish futures speculators failing on three major occasions over the past year to push gold to new lows, I suspect their zeal to short this beaten-down loathed metal is going to soon evaporate. And once they start covering, their buying feeds on itself. The more gold-futures contracts bought to cover existing shorts, the faster the gold price rises and the more it pressures other traders to cover shorts too.

And even with specs’ total shorts now nowhere near last July’s record, there is still an utterly huge amount of short covering left to do. During the normal years of 2009 to 2012 before last year’s extreme gold anomaly, American speculators’ total gold-futures shorts averaged just 65.4k contracts. Merely to mean revert from here, they’d have to buy to cover 56.0k contracts which is the equivalent of 174.2 tonnes of gold!

And mean reversions out of extremes rarely stop at averages, they usually overshoot in the opposite direction. So we’re likely to see specs’ total shortswell under 50k contracts after this coming short covering has run its course.On top of that, specs’ total gold-futures longs at 200.0k contracts remain far below their 288.5k average between 2009 to 2012. A mean reversion there represents another 275.2t of gold buying.

So as American gold-futures speculators unwind and mean revert their extreme bearish bets on gold over the coming year or so, we are looking at a staggering 449.4t of buying from them alone. This isn’t far below 2013’s epic record GLD liquidation of 552.6t, which was responsible for nearly 6/7ths of the total drop in global gold demand that hammered its price 27.9% lower last year. Huge gains are coming.

And once gold’s mean reversion gets underway decisively from this speculator futures buying, other investors will pile in. We’ll see big differential buying pressure on GLD shares from American stock traders as they seek to redeploy in gold, and surging Western physical demand. Major gold uplegs start with futures short covering, continue growing on futures long buying, and then accelerate when real investors return.

With gold low and basing this past year, silver has fared worse in some ways.Super-volatile silver is kind of like a gold sentiment gauge. Speculators and investors only get interested in buying it again once gold is already rallying,gold drives silver. And this same CoT chart substituting spec silver-futures positions and the flagship SLV silver ETF’s price highlights even more extremefutures shorting last week.

Last week as gold plunged on futures short selling, speculators also dumped silver. The resulting 6.4k-contract jump in traders’ total silver short contracts was really big, but not a record. It was still quite rare though, as 6k+ contract weekly leaps in spec shorts have only been seen in 34 of the 804 CoT weeks since the beginning of 1999. And nearly half of these occurred before the late 2001 birth of silver’s secular bull.

But the actual total level of American speculators’ silver-futures shorts did surge to a new record last week of 58.2k contracts! This is at least a 15.4-year high, since early 1999 if not all-time. This trounced the last record of 54.3k in early December 2013. The implications of these epic shorts are even more profound for silver than they are for gold, with the potential to launch a moonshot in this hated metal.

Much like gold, silver has bottomed and based over this past year. But unlike gold-futures speculators, silver-futures speculators have continued to press their outsized bearish bets. The peaks of their shorting are actually still rising, leaving them extremely exposed to a sharp rally in this traditionally super-volatile metal. When gold starts running, silver will follow. And the rush to cover will be savage.

Between 2009 and 2012, total spec silver-futures shorts averaged 21.5k contracts. Merely to mean revert to those levels will require traders to buy back 36.7k shorts, the equivalent of 183.5m ounces of silver! This is an enormous amount. According to the Silver Institute, 2013 global mine production averaged 68.3m ounces per month. We are talking about short covering equivalent to nearly 3 months of world mining!

And this silver short covering will likely happen fast, over less than a month.Not only is silver prone to big and fast rallies when gold runs, but silver futures are highly leveraged making shorting extremely risky when silver rallies. So with American speculators’ silver-futures shorts again at record highs, I still believe a massive silver short squeeze is coming. Futures shorts always have to soon be repurchased.

It’s provocative that silver didn’t plunge far in response to last week’s record spec shorts, a sign that there is big buying out there absorbing this burst of supply. Much of this is for physical bullion, where global demand surged to a record high last year despite the wretched prices. But some is futures buying, where contrarian futures traders have been steadily adding to their long silver-futures holdings all year.

So despite the universal bearishness that flared up in response to last week’s heavy gold and silver selling, it was the best kind of selling. It wasn’t dreaded long liquidations, as long-side bets continued to rise in both gold and silver (also up 3.0k contracts). It was extreme short selling, positions that have to soon be bought back to repay those debts. Extreme futures short positions are very bullish omens.

When the great majority of futures speculators have crowded onto one side of a trade, they are always wrong. They are the most bullish, with high total longs and low total shorts, right when prices are topping after long runs higher. And they are the most bearish, with low total longs and high total shorts, right when a price is bottoming before turning higher again. The latter describes gold and silver perfectly today.

With gold-futures and silver-futures shorts just hitting their own records, it virtually guarantees a frenzy of short covering is nearing. Short covering feeds on itself, driving accelerating rallies that propel gold and silver (and therefore GLD and SLV) higher. Are you ready for this? Do you have the courage to make a contrarian bet, to buy low when others are afraid? Or do you sell low and buy high like Wall Street?