Gold has failed to gain traction over the past couple months, normally a seasonally-strong time. That has really weighed on sentiment, leaving traders increasingly bearish. Gold investment demand has flagged dramatically with lofty stock markets spewing great euphoria. That’s given gold-futures speculators the run of the market, where they have sold aggressively including extreme shorting. But that’s actually very bullish.
Gold price action is driven by the collective trading of both investors and speculators. The former control vast amounts of capital, which dominates gold prices when it is migrating in or out. But investors’ interest in gold withers when stock markets are super-high. When stocks seemingly do nothing but rally, there’s no perceived need to prudently diversify stock-heavy portfolios with counter-moving gold. It falls out of favor.
Extreme stock-market euphoria is gold’s primary problem now, acting like kryptonite for gold investment. This week the flagship US S&P 500 broad-market stock index clawed back to a new all-time record high. That extended its monster rebound rally since late December’s near-bear lows to 24.8%! The farther the stock markets advance, the more gold is forgotten. Investors have relentlessly pulled capital back out of gold.
The best proxy for gold investment demand is the physical gold-bullion holdings of the world’s dominant gold exchange-traded fund, the American GLD SPDR Gold Shares. In early October soon after the S&P 500 peaked but before it started plunging in its severe 19.8% correction, GLD’s holdings slumped to a deep 2.6-year low of 730.2 metric tons. I explained these stock-market and GLD dynamics in depth last week.
Then the very day the stock markets first dropped hard, investors remembered gold. Over the next 3.8 months into late January, GLD’s holdings surged 12.8% to 823.9t on heavy capital inflows from American stock investors. That helped push gold 8.9% higher in that span. But as euphoria came roaring back as the S&P 500 rebounded sharply from its deep selloff, gold’s relative luster again faded in investors’ eyes.
(Click to enlarge)
Between late January and this week, they’ve dumped GLD shares much faster than gold itself was being sold. That has forced GLD’s holdings 9.2% lower in the last 2.8 months to 747.9t, helping push gold’s price down 2.7%. Over 4/5ths of gold’s stock-market-correction-driven investment surge has now been erased, leaving GLD’s holdings just 2.4% above their secular lows of early October before stocks plunged!
The gold-investment selling via GLD in recent months has been relentless, especially in February and now April. During February’s 19 trading days, 13 saw GLD draws averaging 0.4%. And as of the middle of this week, April’s 17 trading days so far have seen 12 GLD-draw days also averaging 0.4%. Gold has faced unyielding selling pressure from American stock investors as the S&P 500 levitated ever higher.
There’s an old proverb stating “when the cat’s away, the mice will play”. That concept perfectly applies to the gold market. When investors are away, the gold-futures speculators will play. Investors’ capital just dwarfs speculators’, so when gold investment demand is robust spec trading is drowned out and usually irrelevant. But when investors aren’t interested, the gold-price impact of gold-futures trading is magnified.
These traders already punch far above their weights, their capital being far more potent than investors’ on a dollar-for-dollar basis. Gold futures allow extreme leverage far beyond anything legal in the stock markets. Each gold-futures contract controls 100 troy ounces of gold, which is worth $127,500 at $1275. But gold-futures speculators are only required to keep $3,400 cash in their accounts for each gold-futures contract.
That gives them absurd maximum leverage up to 37.5x, compared to the decades-old 2.0x limit in stock markets! At 30x leverage, every dollar deployed in gold futures has literally 30x the price impact on gold as another dollar used to buy gold outright. Just $1 of gold-futures capital flows yield the same gold-price result as $30 of investment capital flows. Gold-futures trading’s impact on gold is wildly disproportionate.
Further amplifying gold-futures speculators’ outsized influence, the American gold-futures price is gold’s global reference one. So when heavy gold-futures selling blasts that headline price lower, the resulting negative psychology quickly infects the rest of the world gold markets. Gold-futures trading is effectively the tail that wags the gold-investment dog. This vexing problem shouldn’t be allowed to exist, but it does.
Over the past couple months as mounting stock-market euphoria seduced investment capital out of gold, speculators’ gold-futures selling has soared to extremes at times. That really exacerbated the counter-seasonal downside pressure on gold prices. This heavy selling is evident in the weekly Commitments of Traders reports from the CFTC, which detail speculators’ collective long and short positions in gold futures.
This chart superimposes several years of daily gold prices in blue over the weekly CoT data. Total spec long contracts are shown in greed, and total shorts in red. The falling longs and rising shorts since gold last peaked near $1341 in mid-February are a big reason for its recent weakness. But the lower specs push their longs and the higher they ramp their shorts, the more bullish gold’s near-term outlook grows.
(Click to enlarge)
A couple weeks ago I dug deeper into gold futures’ impact on gold prices in recent years, so I’m going to focus on recent months here. On February 19th when gold surged to $1341, total spec longs and shorts were running 305.0k and 138.5k contracts. While those longs remained way below recent years’ peaks, they were still near the highest levels seen in the past year. I developed a simple metric to quantify that.
This chart shows the general rule on gold-futures trading driving gold price action. When speculators are buying by either adding new longs or covering existing shorts, gold rallies. When they are selling existing longs or adding new shorts, gold retreats. So the lower spec longs, and the higher spec shorts, the more bullish gold’s near-term outlook. The opposite is also true, higher longs and lower shorts are bearish for gold.
Gold’s biggest uplegs in recent years emerged from relatively-low spec longs and/or relatively-high spec shorts. Figuring out how low or high both sides of this trade happen to be can be done by looking at current levels compared to their trading ranges over the past year. When gold peaked at $1341 9 weeks ago, total spec longs were running 96% up into their 52-week trading range. That was certainly relatively high.
That left speculators little room to buy more gold-futures long contracts unless they expanded their total capital allocation back to bigger prior-year levels. If they didn’t, they had a lot more room to sell than to buy. That same CoT week, total spec shorts were running 32% up into their own past-year trading range. Thus the short-side guys had probable remaining room to cover 1/3rd of their shorts, which was relatively low.
If investors had been buying gold, if the mounting stock euphoria hadn’t been sucking capital out of gold, speculators’ gold-futures positioning wouldn’t have mattered much. But with investors missing in action, the gold-futures traders were ruling the roost. And they started selling heavily in the CoT week ending on Tuesday March 5th. Be aware that CoT weeks always run from Tuesday closes to Tuesday closes.
Gold began that CoT week looking great, trading at $1328. But speculators started selling gold futures, pushing gold down towards $1300. That is a hugely-important psychological level for gold, which seems to attract gold-futures stop losses like gravity. So as $1300 neared and failed, gold-futures selling ramped up massively. That CoT week ended with specs dumping 34.0k long contracts while adding 11.9k short ones!
A 20k+ contract change in either spec longs or shorts in a single CoT week is the threshold where huge begins. 20k contracts control the equivalent of 62.2 metric tons of gold, way too much for normal markets to absorb in a single week. That big bout of spec gold-futures long selling that kicked off the last couple months’ gold slump was exceptional. At that point 1053 CoT weeks had passed since early 1999, a long span.
That CoT week’s spec long selling ranked as the 20th largest ever witnessed, a rare event. And in terms of speculators’ total gold-futures selling including both longs and shorts, it was the 11th largest on record! It’s important to realize that gold-futures selling of that magnitude is unusual, unsustainable, and self-limiting. The lower spec longs and the higher spec shorts, the less gold futures these traders have left to sell.
Related: China's Economic Growth Exceeds Analyst Expectations
That extreme selling blitz puking out the equivalent of 142.6t of gold in a single CoT week would probably have been the end of it without the growing stock-market euphoria. Gold usually carves a major seasonal low in mid-March before powering higher in its spring rally. But with the S&P 500 levitating and investors still selling gold on balance, sentiment stayed fairly bearish so gold-futures specs had the run of the market.
Still gold defied the surging stock markets to rally like usual, climbing back to $1322 by March 25th. The gold-futures speculators were responsible, adding 20.4k new long contracts while covering 15.4k short ones in the CoT week ending a day later. That was the equivalent of 111.3t of gold buying. But over the next CoT week, that reversed into heavy selling. That again surrounded gold plunging back under $1300.
For decades now I’ve intensely studied and closely watched the markets in real-time. I get up at 5am and follow the data and news feeds until 4pm or later. Usually, when gold or the stock markets make some big intraday move, it’s explainable by news or data. Neither gold’s 1.7% plunge on March 1st, nor its later 1.4% drop on March 28th, had any apparent catalysts! But both days saw gold break back below $1300.
Running extreme leverage up to 37.5x, gold-futures speculators can’t afford to be wrong for long. A mere 2.7% gold price move against their positions would wipe out 100% of their capital risked at such leverage! So these guys have to maintain an ultra-short-term price-dominated focus, and they have to run tight stop losses or risk quick ruin. Long-side gold-futures traders have long clustered stops near that key $1300 level.
So when gold falls back through $1300 from above, mechanical stop-loss orders start triggering resulting in forced long selling. That quickly pushes gold even lower, tripping more stops to fuel cascading selling. By the time the dust settled in that CoT week ending on April 2nd with gold battered back to $1291, total spec gold-futures longs had plummeted 35.3k contracts! They weren’t short selling then, as shorts fell 2.1k.
That massive long dump was again exceptional, ranking as the 18th largest ever witnessed out of 1057 CoT weeks since early 1999 at that point. Speculators can’t maintain such crazy selling rates for long, as just 7 weeks at that pace would drive their longs to zero which will never happen. For the second time in 4 CoT weeks, extreme spec gold-futures long selling hammered gold from well above $1300 to back below.
But gold soon started recovering even while investors mesmerized by stock euphoria exited. Gold again climbed up over $1300, hitting $1308 on April 10th. This metal really wants to power higher even with investment capital fleeing to chase the lofty stock markets. Yet once again extreme gold-futures selling erupted in the latest CoT week reported before this essay was published, which ended last Tuesday April 16th.
For the third time in 7 weeks, extreme gold-futures selling flared as gold passed back down below $1300. Once again there were no significant data or news catalysts around the world, gold-futures selling just snowballed to a stunning degree. That CoT week total spec longs dropped another 17.5k contracts, close to that 20k+ huge threshold. But total spec shorts exploded an utterly-astounding 36.9k contracts higher!
That single-CoT-week shorting was so crazy it ranked as the 2nd highest ever witnessed out of the 1059 CoT weeks since early 1999! The only bigger shorting week was back in mid-November 2015, soon after the Fed telegraphed its first rate hike of the recent cycle. Yet that record shorting would soon prove very bullish for gold, birthing a major bull market. Gold surged 29.9% higher in 6.7 months in the first half of 2016.
Considered together in that latest reported CoT week ending April 16th, speculators’ total long and short selling rocketed to 54.4k contracts! That is the 5th highest on record, incredibly extreme. The 1st and 4th weighed in at 70.4k and 56.7k, and both occurred in December 2017. That record gold-futures selling also proved very bullish, as gold soon surged sharply to challenge a major bull-market breakout above $1350.
Big gold-futures selling is always bullish for gold, because those bearish bets will soon be unwound with proportional buying. This current episode won’t prove an exception, especially with near-record shorting. While making bullish long-side gold-futures trades is voluntary, short covering is mandatory. Shorting is effectively borrowing gold futures that traders don’t own, those contracts have to be repurchased and paid back.
Between gold’s latest interim high in mid-February to this extreme latest-reported CoT week, total spec longs collapsed 68.5k contracts or 22.5%. That’s a lot in a short span, leaving longs running just 32% up into their past-year trading range. That means specs easily have room to do over 2/3rds of their likely near-term long buying, and much more if higher gold prices excite traders enough to bet at previous years’ scales.
And over the last 8 reported CoT weeks, total spec shorts rose 19.5k contracts. That left them 37% up into their own past-year trading range. That’s not high, but it still leaves a lot more shorts that have to be covered with offsetting buying as gold reverses higher again. Total spec selling since February 19th ran 88.0k contracts, the equivalent of 273.9t of gold. That’s helped force gold 4.8% lower from $1341 to $1276.
The bright side of all this gold-futures selling is it is inherently self-limiting and self-correcting. The more these traders sell, the less they have left to sell. And the higher the odds they will start buying in a big way to mean revert their recent bearish bets back to normal. One of these days some catalyst will arise that will spark major spec gold-futures buying. Gold will surge sharply for weeks as buying normalizes bets.
The biggest casualty of recent months’ extreme near-record gold-futures selling was the gold miners’ stocks, which amplify moves in gold. The major gold miners of the leading GDX VanEck Vectors Gold Miners ETF tend to leverage gold’s action by 2x to 3x. That has weighed on gold-stock prices and psychology since mid-February. GDX slumped while gold-futures speculators battered the gold price lower.
Despite that extreme gold-futures selling nearing records, and incredible stock-market euphoria stunting gold investment demand, the gold stocks have weathered this storm really well. GDX did knife back under its upleg’s support, nearing its 200-day moving average which is much-stronger support. But the major gold stocks have proven impressively resilient overall, largely consolidating high as gold swooned.
Again gold was pounded 4.8% lower over those 8 CoT weeks starting near $1341 and ending way down near $1276. At 2x to 3x normal leverage, the gold stocks would’ve plunged almost 10% to 15%. Yet over that exact span GDX merely slid 5.7%, just 1.2x gold’s loss! And GDX’s leverage was healthy before that as gold rallied, running 2.8x at best by mid-February. The gold stocks have really been holding their own.
Related: What's The Federal Reserve's Next Move?
Gold stocks are set to surge again once gold reverses decisively higher, which is increasingly likely any day now. These lofty euphoric stock markets are going to inevitably encounter some catalyst sparking significant selling, which will snowball after such a massive and long rally steeped in such epic complacency. Gold investment demand will turn on a dime as stock markets roll over, just like back in early October.
And when gold starts moving higher, the hyper-leveraged gold-futures speculators will rush to buy and pile on to its upside momentum. And after slashing their longs and ramping their shorts over the past couple months, they have major buying to do to reestablish bullish positioning relative to gold to ride its next rally. As leveraged gold-futures capital inflows force gold higher, gold stocks will really amplify its gains.
The last time major gold investment buying lined up with major gold-futures buying by the speculators was in roughly the first half of 2016. That catapulted gold 29.9% higher in 6.7 months kicking off this bull. The major gold stocks as measured by GDX soared 151.2% in essentially that same span, amplifying the big gold gains by 5.1x. Gold stocks are the place to be when traders are pouring capital back into gold!
The bottom line is gold has been bludgeoned by extreme gold-futures selling in the past couple months, culminating in near-record shorting. That’s what forced gold lower during its usual spring-rally timeframe. With investors seduced by the lofty euphoric stock markets, gold-futures speculators have been running roughshod over gold prices. But their heavy selling is self-limiting, and will reverse into proportional buying.
Speculators’ big bearish shift in gold-futures positioning will have to be normalized, resulting in big buying that will push gold higher. That upside momentum could really grow, especially when stock markets roll over and again rekindle gold investment demand. The biggest gains as gold mean reverts back higher will come in the stocks of its miners. They’ve proven resilient as gold swooned, and are poised to surge again.
By Adam Hamilton
Adam, you've made the same claim many times already in your past articles. This still heavily contradicts data and even your own charts. In your current and previous charts comparing GLD's holdings with the gold price, you can see numerous periods where the gold price leads GLD's holdings rather than the opposite. So which data leads which data again? In Jan 2017, there was a 7.6% rise in the price of the GLD, which means people have been buying. Yet, since gold bottomed, GLD's holdings have fallen another 4.5%. It fails as both a leading and a lagging indicator. There are countless other examples outside of this January example as well. This indicator is very useless as proven by its track record. Also, the gold exchanges completely dwarfs GLD's movements. GLD's movements are so very insignificant compared to the overall gold market.
"That has forced GLD’s holdings 9.2% lower in the last 2.8 months to 747.9t"
"GLD’s holdings surged 12.8% to 823.9t on heavy capital inflows from American stock investors."
"GLD’s holdings slumped to a deep 2.6-year low of 730.2 metric tons."
Again Adam? I still have yet to see any verifiable evidence to support any of these claims. How reliable are GLD's holding reports? GLD does not give retail investors the right to redeem for any of its mystery physical gold holdings. This fact alone ensures the GLD shares to be nothing more than paper at the end of the day. Daily, even in great detail, doesn't mean much when there is absolutely no way to verify any of it.
GLD also has a glaring audit loophole in their prospectus that states they have no right to audit subcustodial gold holdings. To this day, I have not heard of a single good reason for the existence of this backdoor to the fund. For anyone interested but have not heard, I recommend looking into CNBC's Bob Pisani making a highly publicized visit to GLD's gold vault in a segment called Gold Rush: The Mother Lode. GLD's management organized this visit to show that GLD's gold actually exists. However, the gold bar held up by Mr. Pisani showed a serial number of ZJ6752 which did not show up on the latest bar list during that time. It was later discovered that this "GLD" bar actually belonged to ETF Securities.