by Matthew Hart
The structural effect of ETFs on price instability arises from the requirement that ETFs “rebalance” themselves at the end of the trading day, buying or selling their underlying assets to match up with the day’s buying or selling of the fund. As rebalancing occurs, a flood of trading orders hits the market at the same time. The market reacts as it does to any large order, moving up if the order is to buy, down if to sell. Since ETF shareholders may have acted in response to a market movement in the first place, their actions intensify that movement.
In a DealBook column in the New York Times, Steven Davidoff, a law professor with expertise in financial regulation, saw in the general buzz about gold the elements of a pricing bubble in which “speculation is aided by the financial revolution. Previously gold could be bought by retail investors only through dealers and street shops. Now anyone can go on the Internet, click and buy gold in the market through exchange traded funds.” In a bubble, wrote Davidoff, television and the Internet would play a large role in spreading the hype. And that is what they were doing. In such an atmosphere, he said, “the marketing of gold to the masses is an ominous sign.”
You can order gold as easily as groceries. Some sellers will deliver to your door. One such company is Goldline International, Inc., of Santa Monica, California. Goldline’s most famous pitchman is Glenn Beck, a popular right-wing talk show host. In 2011 he had a show on the Fox television network. Beck stirred up his pitch with forecasts of ruinous inflation. He advised his viewers to put their faith in “God, gold, and guns.” In a broadcast I listened to online, Beck predicted the collapse of the European Union, and European war.
The fortunes of Goldline soared as the gold price rose. In late August 2011 gold hit $1,917.90 an ounce. The WGC reported that demand for a single category of gold investment—bars—had doubled in the previous two years to 850 metric tons a year. Television business commentators predicted a gold price of $2,500 an ounce, and gold bug websites vibrated with reports that John Paulson, with billions of dollars worth of bullion in his funds, was prophesying $4,000 gold. Leaving aside the holdings of gold coin speculators and private bullion owners, and the uncountable lumps belonging to those who’d melted down their jewelry in home smelters, the amount of investor gold in ETFs alone stood at 2,250 metric tons. Then in September 2011 the gold price reached the end of a ten-year bull run, and fell off a cliff.
In a month gold tumbled $300, landing with a thud at $1,600 an ounce in October 2011. Some investors saw their gains evaporate. On November 3 the Santa Monica, California, city attorney filed a nineteen-count criminal fraud complaint against Goldline, alleging the sale of overpriced gold coins to people who thought they were buying bullion. Goldline agreed to refund as much as $4.5 million to buyers, according to reports, and to pay $800,000 into a fund to settle future claims. ABC News reported that Goldline also had to stop telling customers that the government wanted to confiscate its gold.
As the gold price fell, hedge funds started “unwinding” gold positions to raise cash to cover the increased collateral required by the funds’ bankers when other assets, such as stocks, started falling too. This overall decline of asset values challenged the rule that the gold price rises when equities fall. This time the whole ship was sinking. Analysts call it a “risk-off,” in which investors jettison any risk at all, and head for cash. “Cash” meant only one thing: the U.S. dollar.
You could take the view that selling gold to raise collateral for wilting stock positions demonstrated gold’s utility as a hedge. If, for example, gold had withstood the risk-off better than some other asset, the prudent investor would choose to take a loss on gold in order to escape the greater loss of selling that more depleted asset, an asset that, if held, might recover later. Moreover, the ready market for gold, even as its price declined, showcased one of the metal’s undoubted qualities—liquidity: the ease with which it was turned into money.
As it happened, this quality was crucial to a coup the World Gold Council was attempting to pull off. They were asking European banking regulators to include gold in a special class of top-flight assets. If gold were included in this class, the banks would need to own it. It seemed a strange time to be making the pitch, with the gold price tanking and the markets feeling faint. But they have sturdy hearts at the WGC, and just the metal, they would say, to stabilize the teetering banks of the European Union.
AFTER THE BANKING CRISIS THAT followed the collapse of Lehman Brothers in 2008, the Basel Committee on Banking Supervision, an obscure but powerful caucus of G20 finance ministers, issued new capital requirements for banks. These requirements included bigger liquidity buffers. Such buffers were to be highly liquid assets that could provide cash in an emergency, when normal sources of cash, such as other banks, dry up. Top-quality government debt—German bonds or U.S. Treasury bills, for example—would qualify as highly liquid assets. In the submission of the World Gold Council, so would gold, and on a bright day that heralded a spell of Indian summer for London, where I was living then, I went down to the old financial quarter, known as the City, to hear why.
It was just after noon, and the steps of St. Paul’s Cathedral were crowded with people sitting in the sunshine eating sandwiches and tilting their faces to the sun. I strolled through the gate into Paternoster Square. A well-dressed crowd was flowing out of the London Stock Exchange and into the restaurants around the square. If anguish about sagging markets was troubling the financial heart of London, they had decided to forget it while they ate.
I made my way into Newgate Street. Lawyers wearing wigs smoked on the sidewalk outside the Old Bailey, London’s main criminal court. Down the street from the dark old court stood the entrance to a modern, glass-walled office building, and in a third-floor conference room I sat down with Natalie Dempster.
Dempster, a brisk Scottish economist, worked for the Royal Bank of Scotland and JPMorgan Chase before joining the WGC, where she heads government affairs. When we met, she had just returned from Brussels, where she’d been pitching European banking regulators on gold’s inclusion as a liquidity buffer. She slapped a wad of lobbying material onto the table and began to enumerate the reasons European regulators should like gold, picking her way through the criteria with practiced skill, and explaining how gold met them:
Criterion: The asset cannot be an instrument issued by the institution holding it as a buffer.
Gold: No one “issues” gold. Its value is not tied to anybody’s credit. Therefore it has no credit risk.
Criterion: The asset’s price should be easy to calculate from public information.
Gold: Gold is probably the best known asset price in the world.
Criterion: The asset must be tradable in markets with a large number of participants, a high trading volume, and market breadth and depth.
Gold: The London gold market alone was trading $240 billion a day.
It’s worth repeating that the true size of London bullion trading had surprised even those who’d uncovered it, the London Bullion Market Association. It was the first time in sixteen years they’d polled their members about trading. They’d conducted the exercise in their own interest. If regulators accepted gold as a liquidity buffer, sales to banks would increase. If liquidity was the measure of a buffer, the LBMA survey revealed, gold met it. The liquidity was a function of the intensity of trading. If volume was the standard, Dempster said, it was easier to sell gold than government bonds. There were always buyers—the definition of liquidity.
Because the need for bigger liquidity buffers stemmed from the problems banks had faced when money got tight, Dempster reviewed gold’s performance during the crisis. She contended that the bullion market stayed robust while other markets faltered. A graph showed interest rates spiking as banks stopped lending to each other. The bullion market stayed liquid. Many other markets “assumed to be deep and liquid proved to be the exact opposite,” she wrote in a supporting document, “and assets could only be sold at a large discount. This was even true of so
me AAA-rated assets: credit ratings proved to be no guide to liquidity.”
In Dempster’s view, gold also benefited from a pricing floor. “What happens to gold is that the structural demand of the jewelry market exists under it,” she said. “What happened during Lehman’s failure is that suddenly, as gold fell, industrial and jewelry demand cut in and put a floor under it. So it’s not about gold never falling, but that when you are liquidating it the price is never going to fall out of bed. This is one of the most important points. People from the jewelry and technology sectors have a completely different perspective. When gold goes down in price, they want it.”
Central bank gold buying is another box that gold advocates like to tick, and Dempster ticked it. The Chinese central bank had been expanding gold reserves by about a hundred tons a year, and Mexico, Russia, and South Korea had all bought large amounts. Now European central banks seemed to be changing their position on bullion. They had been sellers for more than a decade, shedding about 400 tons a year. That trend had stopped. They had become net buyers.
In pitching European regulators, Dempster and the WGC were trying to change decisions already made by the Basel Committee of G20 finance ministers. The committee had already considered including gold as a liquidity buffer, but had rejected the idea, mainly, a spokesman for the British Treasury told me in an email, “due to the volatility of its price.” Moreover, he added, the Basel Committee’s recommendations were global. European regulators were supposed to be transposing them into law, not “watering them down” in any way.
In the end they did not water them down. It was the wrong time to be recommending gold as any kind of pillar. Volatility was roiling the bullion market. September 2011 was a chaotic month for the gold price, with sharp swings up and down. Whatever winds were blowing blew the gold price. Gold did not behave like any kind of safe haven. It behaved like what it had become—just another derivative. It was a construct that could be attacked, and someone attacked it.
THE GOLD PRICE WAS VULNERABLE because it was easy to manipulate. Gold traders understand that liquidity is a sea with different depths. When the London market opens, a lot of gold is available to trade. When it’s closed, much less. If you want to sell a lot of gold in an orderly way that will not disrupt the price, you sell when London is open. But what if you do want to move the price? Then it makes sense to pick a time when the trading is thin.
Large miners watch the markets like hawks, because they sell their gold there. They understand to a fine degree the consequences of trading in a thin market. Barrick Gold’s research department once determined, for example, that 100,000 ounces offered during London market hours would move the gold price down $4. That same 100,000 ounces sold when London was closed, would depress it $10 to $15. When half a million ounces dropped like a bomb on September 7, 2011, into the thin market before the London opening, then, it was no accident.
The attack took place during a period of volatility. In the chaotic month of September 2011, the gold price dashed up and down the chart seeking a consensus that it could not find. Into this jittery milieu, on September 7, during lunch-hour trading in Shanghai, hours before the great gold blotter of the London market opened, someone dumped 500,000 ounces. The gold price dropped like a shot crow.
Suspicion turned to Libya, where the regime of Muammar Gaddafi was disintegrating in a bloody rebellion. A former Libyan central bank governor warned that the dictator had possession of the country’s gold reserves. As it turned out, Gaddafi’s gold sale had already come and gone. He had cashed in some thirty tons of bullion the previous April, selling it for a reported $1 billion to dealers who had gone to Libya to transact the deal. Gold traders in Tripoli’s Old City, near the Libyan central bank, said that the regime had started the selling with a trickle of 22-carat coins and then increased to twenty-six-pound bars as its crumbling army demanded pay. But that was five months before the Shanghai sale.
To see what might have caused a bullion seller to make a move so harmful to the price, I spoke to Jim Mavor, an eighteen-year veteran of Barrick’s gold trading operations. Mavor has since become vice president of finance at Detour Gold Corporation, but when we spoke he was still Barrick’s treasurer.
“Well,” he said, “somebody with 500,000 ounces of gold worth [at the time] $900 million, may be assumed to be somebody who knows what’s what. So it’s fair to wonder why he performed a trade guaranteed to drive the gold price down. It fell $50. In trying to understand what happened, it’s useful to construct a scenario in which the person causing the event to happen made money out of it.”
Such a scenario, in Mavor’s view, could work this way: Let’s say the seller of the gold also owned a certain kind of option called a put. The owner of a put option has the right to sell the underlying asset—in this case gold—at a stipulated price. Let’s make the price $1,800 an ounce. If the put owner also happens to own bullion, in addition to owning the right to sell it, then the $1,800 put has placed a floor beneath his possible losses. He knows he can always exercise the put at $1,800 no matter how far the price drops.
But in the scenario we’re envisioning, gold has not dropped, but risen. It has reached $1,825. The owner of the gold has made money on his bullion, but in the process his puts have become useless. There is no value in a right to sell gold for $1,800 when the price is $1,825. The puts, then, represent a loss, because the owner paid money to buy them. The only way his puts can regain value is if the gold price falls. If the put owner decides to drive the price down to revive the value of his options, the best time to do it is in the low-liquidity doldrums before the London opening. The only math required to make his attack on the price appealing is if our hypothetical gold player can make more money from selling the puts into an alarmed market than he will lose from the depressed value of his bullion. We must suppose he can do the math. The likely trader, in Mavor’s estimation, was a hedge fund, where aggressive tactics and large sums come together.
If the September 7 bullion dumpers meant to spook the market, they picked the right time. It was already spooked. Europe was in danger of unraveling in what The Economist called “the greatest crisis to befall the European project in its history”—the dilapidated euro. In the United States, the contest between parties over what to do about the national debt had become a war of religion. In one view the economy would founder without government action, including a tax hike on the rich; in the other, such interference amounted to the destruction of the republic. Meanwhile, American stocks lost $1.1 trillion in a four-day rout. The U.S. government was about to run out of money. Bitter partisans haggled in the Congress. Into this reaper of despair went the gold price. In a single week it lost $200.
Confidence in gold was crumbling in its biggest market, futures. With futures, investors bet on where the price will be at a stipulated forward date. To open the contract, the buyer must deposit with the commodity exchange a certain percentage of the contract’s value. This collateral gives the exchange something to seize if the market turns against the player and he is tempted to abandon his contract. In the face of increased volatility in the gold price at the end of September, CME Group, formerly the Chicago Mercantile Exchange, the world’s biggest commodities market, raised its collateral demand for gold by 21 percent. It was Chicago’s third margin hike for gold in a matter of weeks. The aggregate increase, according to a client note from one large bank, was 90 percent, forcing investors who couldn’t post that much collateral to liquidate their gold positions. The liquidations were depressing the price.
Also preying on the gold price were suspicions about ETFs. The gold market plays in a skeptical arena. Gold bugs are habitually suspicious: mostly of paper money and governments, but anyone will do. You don’t have to search far to uncover doubts about whether gold ETFs actually possess the physical gold they are supposed to. Perhaps it is only natural for a product such as the Spider to attract suspicion, given its spectacular success. Before the market turned against gold, the Spider was the bigge
st ETF of any kind.
In September 2011 the fund owned 1,232 metric tons, or about 40 million ounces. All of it was stored in the London gold vault of the fund’s custodian, HSBC Bank. Since the ETF was ultimately a creature of gold miners, you could say it had simplified the flow of gold from one underground space to another. Or could you? Was the gold at the end of the stream as real as the gold at the beginning? That question troubled some people’s minds.
Online at spdrgoldshares.com, a visitor can navigate to a photo said to show the bars held in the Spider’s account. The Spider says that its shares represent real bullion in a real vault. Such gold is said to be “allocated.” In an allocated account, the asset is not merely produced when the account holder asks for it, like money in a bank account, but stays in its repository all the time, unused by anyone else. It is as if you deposited cash in a bank, and the bank put that actual cash into a separate box and kept it just for you, separate from the cash of every other depositor. No matter what happened to the other deposits, your deposit would be intact. When the Spider fund says that its gold is allocated, it means that the custodian keeps it separate from any other gold. It is always and only the Spider’s gold.
The Spider’s site also gives a list of every bar in the ETF’s account, with each bar’s unique number, its gross weight, and its “fine” weight—the actual weight of gold that it contains. The bars are said to be refined to London Good Delivery standards—a trade definition that specifies a purity of at least 995 parts per 1,000. But rumors persist, about gold ETFs in general, that they do not hold enough gold to redeem all shares. In this view, a surge in redemption demands would collapse the funds, as they would not have the bullion to meet the calls. These rumors, said the client note referred to above, were also helping to push down the gold price.
“The biggest gold and silver funds are now on the defensive,” wrote a commentator in a piece appended to the note, “as they may soon face mass investor exits on the back of heavy discounts to the precious metal spot prices [i.e., a falling market] and doubts about the levels of physical gold they actually hold.”