JP Morgan to the Rescue

Bullion banks expanded their short position in gold by dramatic proportions in the fourth quarter of 1999. Gold derivatives outstanding increased by a record $24.2 billion to $87.6 billion, the largest quarterly increase ever. These positions, reported by the Office of Comptroller and Currency, do not include activity of large non-US bullion dealers or investment banks. Including those entities, the OCC numbers should be “grossed up” by 50% to 100%.

JP Morgan reported the largest exposure to gold derivatives, $38 billion, over 40% of the total. From June 30, 1999, JP Morgan’s total more than doubled, from $18 billion. It is possible that JP Morgan’s activity was part of a rescue operation for weaker bullion dealers. As the strongest credit among bullion dealers, it might have been called upon (or felt a calling) to shoulder some of the risk of weaker bullion dealer credits rattled by the gold short squeeze of September 1999. Perhaps JPM’s vast derivative expansion was a form of reinsurance, an assumption of a layer of risk to shore up the misadventures of their less competent competitors. The shutdown of the firm’s New York trading desk at year end 1999 and transfer of most trading operations to London is curious in these circumstances. Is it possible there was some motivation to distance these operations from US regulators? Notice that the new address is conveniently near the anti-gold British Exchequer.

The activities of the bullion dealers in general and JP Morgan in particular raise numerous other questions as to the impact these institutions have had on the behavior of the gold price:

Notional Amounts for Gold Contracts
All Commercial Banks
($billions)

Source: Office of Comptroller & Currency

The unprecedented quarterly increase in paper gold took place just subsequent to the late September-early October 1999 short covering rally that took gold from its twenty year low around $250 to an intraday peak of $339. Assuming a gold price of $290, derivative positions expanded by a notional 3600 to 4800 tonnes. At year-end, they totaled 13,350 to nearly 18,000 tonnes. The OCC numbers provide no detail on the underlying positions. These figures undoubtedly contain offsetting long and short positions, lease rate swaps, and numerous other transactions. Still, why was there such a large increase in derivatives when the gold market was threatening to explode? Even more puzzling, the gold price declined about 15% during the quarter. One would expect that the value of outstanding derivatives would have declined more or less by that amount. The 38% increase in nominal terms could hide a larger increase in terms of physical metal. Stranger still, the producers did very little hedging the fourth quarter. In the absence of producer hedging, the predominant and natural source of derivative transactions, what could have initiated this hyperactivity?

The flood of new paper appears to have played a role in snuffing out the short squeeze that began with the announcement of the Washington Agreement. This rally caused extreme discomfort among the bullion dealers. A brief article from the 9/30/99 Financial Times suggests that the bullion dealers sensed great peril:

“A second day of chaos in the gold market left some analysts arguing that European central banks would have to revise the restrictions on gold sales and lending announced on Sunday.”

“This is now a disorderly market,” said Andy Smith of Mitsui, one of the most respected gold analysts. “Gold is still a reserve asset. If you had conditions like this in the bond or foreign exchange markets, it would not be allowed to continue.”

“Over the last three days, gold has been trading like a commodity, not like money. Volatility has shot up; the cost of borrowing has shot up. The situation is untenable.” Mr. Smith called for the European banks to urgently review their strategy.

Andy Smith was and continues to be a leading spokesman for the bearish view and for the bullion trade. His outcry was a fair proxy for the sentiments of the bullion dealers. Never one to complain while gold was steadily declining, his public distress is a most revealing hint of the agony his sponsors must have felt.

We have encountered numerous anecdotal indications that there was much pleading by this beleaguered group to the Bank of England and the US Treasury. The possibility of a linkage between government entities and the subsequent actions of bullion dealers in this time frame is too strong to ignore. In rescuing its weakling counterparts, was JP Morgan acting strictly in its own self-interest or as an agent for the BOE and US Treasury? What were its financial incentives? Did the obvious vulnerability of the bullion trade to a rising gold price throw the dealers en masse into the arms of government saviors, adding them to a long list of bailouts?

The intense three-week rally broke as gold lease rates plummeted, indicating that liquidity had been provided to alleviate the short squeeze. Soon after, the Kuwait central bank announced that it had loaned out its entire 89 tonnes of reserves into what was a shaky and panicked market. One might ask, what was the impetus for Kuwait’s sudden move, an action if taken on its own would have required large measures of decisiveness, boldness and insight that one would not normally associate with the little kingdom.

Shortly after the gold price broke, Ashanti Goldfields announced that its supposedly bulletproof hedge book had failed to survive the stress test of the previous weeks. The company would need infusions of capital, extensive asset restructuring, and lengthy negotiations regarding its troubled hedge book in order to survive. In the workout sessions that followed, the torture experienced by management and shareholders must have paled by comparison to that felt by the syndicate of 17 bullion dealers charged with finding a solution to the credit crisis.

In the process, the dealers undoubtedly came upon some hard realizations. First, their short position was not at all effectively hedged by the long position of Ashanti’s promise to deliver gold in the future as they had once supposed. Despite high leverage and poor cash flow, the dealers had expanded Ashanti’s credit lines to hedge under the doctrine of VAR (value at risk), the totally absurd notion that the lower the gold price went, the more credit worthy Ashanti’s short position became.This thinking ignored the fact that low gold prices threatened Ashanti’s viability as a going concern. The second realization was that hedge instruments were extremely difficult to restructure. After five months of haggling, the total hedge book declined by only 2mm ounces to 9mm. The complicated, high margin instruments the dealers had so willingly purveyed to Ashanti in the prior months proved to be illiquid, impossible to value, and difficult to dispense. Third, despite strict margin provisions, Ashanti’s financial woes quickly became those of the bullion dealers. As intermediaries between central bank lenders and their clients, bullion dealers were then and will once again become ground zero for dislocations in the gold market. A financially weak gold industry is a direct threat to the credit worthiness of the bullion dealer positions themselves.

Our estimate of central bank gold deposits with bullion dealers at the time of the crisis was 6,000 tonnes (see Simple Math and Common Sense). More recently, Dinsa Mehta, head of Chase Manhattan’s gold dealing operations, estimated this figure to be 7,000 tonnes. While gold deposit numbers are not precise and both estimates could be off by 10%-20%, the proportion and trend are accurate. It appears that gold borrowed by dealers has expanded since the crisis. To be safe, let’s say that lending expanded by only 5% to 10%, or 6,600 tonnes at the maximum. Even at the lower end of the range, or 6,300 tonnes, where did this additional liquidity come from besides Kuwait? The record increase in gold derivatives outstanding during the quarter suggests a substantial increase in central bank lending, but aside from Kuwait and the Vatican, few names have surfaced. In all likelihood, their profiles were similar to Kuwait in that they were in some way clients of the US Treasury.

6,300 tonnes of borrowed central bank gold, or fully 19% of world central bank reserves, has been melted down and sold as jewelry. Still called “reserves” by the central banks, the gold exists only as a receivable, an electronic data entry—“due from bullion dealers”. No distinction between gold in the vault and gold in the digital sense is made in the financial accounts of the central banks. This misleading accounting practice obscures the existence of a short position approximately equal to three years of new mine production. At current depressed gold prices, deliveries from mine production are the only way for dealers to repay their obligations to the central banks. Confident that gold prices will “behave”, the bullion dealers have leveraged their short position by creating paper obligations, or derivatives, of up to $150 billion, whose value is affected in some way by the price of the underlying commodity.

Since the September 1999 scare, it has become obvious that any hypothetical future sharp, permanent rise in the gold price would be most harmful to the bullion dealers. With few exceptions, mining companies enjoy lenient margin provisions. The bullion dealers, not the miners, shoulder the risk of their short position, which represents a collective liability of about $60 billion priced at $280 gold. A $100 increase in the gold price would represent a $20 billion increment in their collective liability to the central banks. Such an increase could prove very costly to the dealers in terms of higher lease rate payments caused by the higher dollar liability amount. In all likelihood, a big increase in the gold price would also generate heightened credit concerns and therefore higher lease rates. Such concerns could also lead to withdrawal of bullion deposits.

Gold derivative positions dictate “delta hedge” buying or selling of physical gold according to mathematical formulas. The physical market for gold has become very thin since the September spike. In February of this year, John Henry, the leading CTA hedge fund with presumably thorough knowledge of the gold market, announced that it was downsizing its positions in gold to 60% of normal due to market illiquidity. In light of their near death experience in September, one wonders what measures dealers are taking in this very thin gold market to protect their short positions. Future gold price spikes will be even more troublesome than in 1999. The physical market will be unable to accommodate panic buy orders placed by dealers, leading to price swings almost unimaginable in today’s dispirited market.

Perhaps we are completely wrong in our suspicion and that the 38% fourth quarter expansion in derivatives represented a change of heart to a more bullish stance on gold for this traditionally bearish group. Even if this were the case, we still have to ask who would have been on the other side of the trade? If the dealers did become more bullish, which seems doubtful, which counter parties took the chance that there was money to be made on the other side of these bullish bets?

The September episode exposed the flaws in the bullion trade. The mismatch between paper and physical gold could not tolerate an unexpected, sudden increase in the price of gold. The spike in the borrowing cost of gold (lease rate) reflecting a liquidity squeeze and a credit crisis turned out to be frightening. Only a flood of new liquidity saved the day. However, the flood of liquidity came from an expansion of the short position, not from natural flows in the gold market. It is clear that the fundamental structural flaws have not been amended and that the market is still vulnerable to a new spike in the gold price.

The bullion lending trade is based on the miscalculation that the price of gold will remain locked forever in a trading range comfortably below $350/oz. The low-inflation rationale for anticipating quiescent gold prices is no longer valid, but the derivative positions based on these assumptions cannot be adjusted quickly to the new realities. In the early 1980’s, momentum-driven psychology led banks to lend aggressively against the value and cash flow of hard assets; oil and gas, real estate, and other commodities, on the assumption of never-ending inflation. Ultimately, these attitudes and practices led to the banking crisis of the early 1990’s. In the late 1990’s, banks and other intermediaries aggressively lent gold on the assumption of secular disinflation. Once strong beliefs about the economic climate take hold, they remain in place long after there is considerable change that invalidates the original assumptions. This is the case with the bullion lending trade, which still views every rise in the gold price as an opportunity to expand its short positions.

In a recent piece on the Euro, our partner Francois Sicart discussed the elements of making a contrarian call:

“1. An aging trend. The passage of time is essential. It allows even the most ignorant groupies to acquire the superficial luster of theoretical knowledge that will help them justify decisions really based on momentum investing...

2. A deepening consensus of expert opinions. Since serious academic work requires ample historical and statistical data, well-documented arguments explaining a trend only become available late in the cycle...

3. Straws in the wind. Often, contrary arguments cannot be as solidly documented as those extrapolating a trend because the contrarian looks forward rather than backward.”

The gold market perfectly exemplifies one at the cusp of an opinion change. The weight of institutional thought and scholarly rationalization is heavily invested in the propositions that inflation is dead and that gold, no longer money, has no role to play as an official reserve asset. The repetition of such thinking over a sufficient period of time has sanctioned the establishment of investment positions that have become inherently inflexible. A shift by definition can only come through a cataclysmic market event, which surprises all those with a vested interest in maintaining their no longer valid posture.

Based on an analysis of mining company hedge books which represent a large component of the other side of bullion dealer positions, the tolerance for a gold price that rises and stays above $350/oz is almost non existent. While risk managers among the bullion dealers may regard positions as being suitably squared, such comfort must be blind to the illiquidity of the physical gold market. The physical market will be unable to accommodate the delta hedging required by derivative positions, as was the case in the September rally. What source of liquidity will save the day for the bullion dealers in the next short squeeze? When will the puppeteers run out of ammunition? The producer hedging well has run dry. Producers have forsworn incremental hedging for 2000, and are beginning to realize that any future hedging at current prices would amount to a liquidation of corporate capital. The reasons for this will be discussed in a separate piece “The Folly of Hedging,” to be available shortly on the Tocqueville web site.

It is a pity that George Soros decided to call it a day insofar as macro economic portfolio bets are concerned. The killing to be made by attacking the gold short position fits the pattern of some of his most notable coups. The agony of Ashanti was prologue to what lies in store for the bullion dealers. Once gold trades above its long-term replacement cost of $360/oz or better, the dealers can look forward to a mega LTCM experience. When Alan Greenspan recently warned banks as to the risks of derivatives and in the same breath mentioned that the market should not overly rely on the doctrine of “too big to fail,” could a scenario such as this been what he had in mind?