The Golden Pyramid

August 28, 1999

Think about this great business idea for a minute. Let's borrow some surplus stuff and sell it for whatever we can get. We'll buy a futures contract to get it back at some certain future date, so we're covered. Meanwhile, we'll earn an interest spread plus commissions. While we're at it, let's sell puts and calls against the stuff even if we don't have it on hand. Our mathematical models will guarantee that our position is always neutral, and we'll clean up on commissions, interest and other fees on the options too.

The foregoing approximates the rationale of the present day, little-known gold derivatives pyramid. John Exter, a famous gold analyst almost two generations ago, was the first to suggest that gold related to paper assets in the form of a pyramid. He described the relationship of gold to paper assets as an inverse pyramid balanced on trust. Currency at one time was a gold derivative. Government issue was backed by physical gold held by central banks. Because currency was a claim on gold, it was in effect a short position against a physical asset that was relatively easy to calculate. Governments hated the idea because they could never seem to stop issuing new paper. Even the pretense of a link has been long abandoned. Since currencies no longer have gold backing, and the world still appears to function, nouveau central bankers assert that gold is superfluous..

The old currency gold/pyramid has been replaced by a little understood labyrinth of paper claims against gold. Responsible senior officials of mining companies, central banks, and bullion banks cannot begin to understand the internal mechanics in order to make appropriate judgements of risk. There are few published figures, no reserve requirements, no supervision or regulation, and no accountability. It is the private domain of bullion dealers, central banks, and mining companies. The credit worthiness of the old currency/gold pyramid was quantifiable. The credit worthiness of the new pyramid can only be an educated guess. Our guess is that it is bankrupt.

The gold derivatives pyramid is a vigorous free market creature. It cannot be put down with a simple declaration that the paper is no longer redeemable in gold, as governments did with currency. It is a short selling scheme that has become a trap from which few short sellers will escape. Paper claims in the form of derivatives far exceed the underlying physical metal on which they are based. The trust, which balances this new pyramid, is based on false assumptions and lack of information. Paper gold claims have proliferated at a pace rivaling any government printing press. A surfeit of paper gold has driven down the price of the physical on which it is based.

The structure can survive as long as bullion dealers, the mining community and the financial media subscribe to the bearish case. But the position of short sellers is precarious. This is true whether gold stays at current levels, or drops below $200/oz. The point is, they will be unable to realize their paper profits, and stand to lose money on their positions in the aggregate. The compound miscalculations on which the gold market is based rank with the blowup of the yen carry trade in 1998. The yen carry disaster illustrates how over-investment and near unanimity of market opinion can lead to a vicious squeeze. Compared to the yen, gold's liquidity is microscopic. The coming squeeze will lead to a several hundred dollar rally and a permanent change in attitudes towards gold.

Anatomy of a Bear Trap (Part 1)

Following the May 7th announcement that the UK would sell more than half of its reserves, gold and gold shares plummeted. The XAU index of gold shares rose 10% from January 1st through May 6th, just prior to the announcement. In the following weeks, the index dropped by 30%. The post UK selling of gold can only be described as climactic and appeared to discount a never-ending official sector supply. The UK sale came out of the blue and could not have been more perfectly designed to damage investor confidence in gold and gold shares. Still, gold shares have remained above their August, 1998 lows, thereby refusing to confirm the new 20 year lows set by the metal itself. A major bear trap has been set, one that was several years in the making. Mining companies in particular engaged in an orgy of hedging since the UK announcement. The rush to the exits when the bearish case became front-page news is consistent with odd lot behavior and a good omen that a major low is in place.

Gold¹s breakdown was based on two incorrect beliefs: first, that all central banks will sooner or later sell all of their gold; second, that today¹s best of all possible worlds regarding inflation and financial markets will last forever. It was nothing less than capitulation to financial market euphoria.

Central bank and official sector selling represents only a small percentage of the excess supply of gold. Far more meaningful and much less talked about is selling pressure from gold borrowed or leased from central banks and resold for the accounts of mining companies or financial institutions. A full examination of gold leasing will show that the gold market has already absorbed vast quantities of unreported selling that dwarf announced outright sales.

Central bank financial accounts refer to leased gold as "gold receivable," an item lumped together with gold on hand as if it were one and the same. In reality, gold receivable is a dubious asset, and in some instances potential bad debt, an asset class new to central banking. A significant percentage of the borrowed gold has already been melted down and sold into the physical markets. It no longer exists in deliverable form. Aided by poor information and worse governance, physical gold borrowed from the central banks has been sold over and over again in multiple transactions. Through the magic of derivatives, paper claims have multiplied against a shrinking base of physical gold. The short "cover" ratio rivals the most overvalued Internet shares.

How did this happen, especially in the hands of leading financial institutions that should know better? The answer: pervasive bearishness, promoted by bullion traders to build up their lucrative trade, and; lack of good data on over the counter market positions. According to Goldfields Mineral Services, central bank bullion on deposit with bullion banks at year end 1998 was 4,300 tons. Frank Veneroso, a respected and well-known authority on the subject, makes a credible case for 8,000 tons. Since year-end, the volume of this activity has expanded by as much as 30%, according to Deutsche Bank in a first quarter 1999 conference call. Volumes expanded further following the UK announcement.

Using the conservative Goldfield¹s number, it is quite evident that 6,000 tons of central bank gold has already disappeared forever via the leasing route. Using Veneroso¹s, the total exceeds 10,000 tons. Either way, a vast quantity of central bank gold has already been disposed of. It can no longer damage the gold market and could even resurface as a buyback! The gold is now in the hands of jewelry owners, coin buyers, and small investors. The amount far exceeds the high profile outright sales by central banks. It represents 25% to 40% of the total reserves of the 80 or so central banks known to lend gold.

Anatomy of a Bear Trap (Part 2)

Easy to manufacture paper claims overshadow impossible to manufacture underlying gold by several orders of magnitude. There is plentiful evidence as to the mismatch between paper and physical gold. A bullion fund manager who has made a career of and a business strategy based on the chain of custody of the physical metal told me that he recently attempted to buy physical from a large commercial bank. The bank officer tried to persuade him to buy the bank¹s gold certificates instead. It turned out the bank possessed only four bars of gold that were free and clear of various claims and therefore deliverable. The same bank had $452mm of gold certificates outstanding as of 12/31/98. Holders of the certificates do not own gold, even though that may be their impression. What they own is the bank¹s promise to pay. Another bullion trader familiar with the notes said that there was no regulatory requirement for a specific gold backing.

In another instance, a mining CEO mentioned that a small quantity of bullion, a corporate asset, had been deposited with a fiduciary. When the company decided to switch banks, they called for their gold and were told they couldn¹t have it within the notice period specified by the agreement. It was only with great difficulty that delivery was made, highlighted by a last minute appearance of a Brink¹s truck at the depository institution. The delivery was made with borrowed gold.

The high-octane enhancements of the foregoing examples cooked up by the best and the brightest Wall St. minds are based on the same principle: unbridled credit extension based on dubious linkage to physical metal. In a forward sale, a mining company directs the bullion bank to sell a certain quantity of gold for future delivery. The bullion bank borrows gold from a central bank, on a short-term basis and for a very low gold lease rate, and immediately sells the gold into the physical market. At that moment the bullion bank is short. The expected repayment of the gold is from future production of the mining company. Against one spot sale, two claims arise. The central bank has a claim on the bullion bank. The bullion bank has a claim against the mining company¹s future production. Two short positions arise from one physical transaction.

In a forward sale, central bank gold is sold into the physical markets to be paid back from future mine production, an innocuous proposition at first glance. However, substituting yet to be mined gold for gold bars entails the credit risk that the gold will actually be mined. Meticulous research is conducted by bullion dealers to assure that the gold is in place, and that it can and will be mined. There can be a very wide gulf between gold in its two states, as a mine reserve and as refined bullion. If many years elapse between the forward sale and the future delivery, or if a very high percentage of mine reserves and future production are mortgaged in this manner, risks and profit potential for bullion dealers escalate. With high confidence in gold¹s downward price trend, there has been little reason not to push the limits on risk assumption.

In the world of options and delta hedging, the leverage to physical far exceeds the two to one of a simple forward sale. In recent years, mining companies have resorted to writing naked calls to buy put protection. Ever more exotic and indecipherable instruments are being brought forth to clutter the market. The $300mm+ of bullion dealer profits is a voracious money machine demanding more central bank gold and synthetic products to fuel its growth. Simple puts and calls have been supplanted by spot deferreds, knockout puts, and most recently, call options on puts.

Individual strategies too numerous to mention all rest on the central fiction that the underlying physical metal exists in the form of yet-to-be mined reserves. Because of delta hedging, the nominal amounts of gold vary considerably with the spot physical market. These option structures are a time bomb. Overall positions at various strike prices are unknown to the market participants. As spot prices approach the option strike price, mathematical models determine the amount of buying or selling without advance knowledge of market liquidity.

What can go wrong? First, the mining company may be unable to produce the gold. Second, the lease rate charged to the bullion bank could rise sharply, as it has in the past, and wipe out their profitable spread. Third, gold could rise suddenly and sharply, triggering margin calls from the central bank to the bullion dealer. Fourth, under certain extreme market circumstances, the counter-party to the bullion bank¹s call option may be unable to deliver. Fifth, local currency fluctuations could undermine economics of the hedge. Sixth, sovereign risk considerations could substantially alter credit quality.

The 6,000 to 10,000 ton physical short interest has three major components. At year-end 1998, 3,600 tons has been sold short by mining companies against future production. Possibly 1,500 to 2,000 are payables of jewelry fabricators against work in process. The 1,000 to 3,000 ton balance represents speculative positions held by commodity funds, hedge funds, and financial institutions.

Mining company short positions are in theory balanced against their own reserves, which will eventually be produced and delivered against the hedge. The catch words are "eventually produced." A gold forward represents an act of faith on the part of the bullion bank that the reserves are accurately stated, mining plans will be adhered to, and that economic conditions will allow production. Such assumptions would be understandable for periods of a year or two. However, forward sales routinely extend from two to ten years, with some "strong" credits as much as fifteen years.

Ashanti Goldfields, a major West African producer has hedged 11mm ounces as of June 30, up from 8.75mm at March 31st. Ashanti qualifies as an active and sophisticated hedger. They have hedged 50% of their reserves, somewhat higher than other large hedgers, but the rationale for their actions is similar to all other hedging programs, regardless of proportion.

The recent market value of Ashanti¹s hedge book was $290mm, using conservative assumptions. What is interesting is that this major corporate asset would be worth nothing if gold traded at $325. At $350, the company would begin to face margin calls. These numbers assume the company takes no action in a rising market. Management would of course defend the hedge book by buying calls with a higher strike price, close out profitable positions, and other assorted maneuvers. The company is quick to point out that their hedge book would not withstand a spike in the gold price very well. The value of the book could not be realized in a compressed time frame. If the gold price did rally sharply, it is assumed it would settle back after the spike to allow the company to realize the hedge book profits. In theory, a run in the gold price could take place, but management sees nothing imminent. The Ashanti hedge book is a bet that the gold market will remain quiescent and trouble free. They are merely "renting" their gold in the ground to enhance realizations. Ashanti¹s sanguine view is not unusual. Few in the industry are prepared for a spike in the gold price, especially one which does not retrace. We expect this to happen, not only because it would inconvenience so many, but because markets that are far out of balance change in a volcanic, not an evolutionary, manner.

Given the risks inherent in the mining business, the credit worthiness of deliveries the distant years strains rational belief. If gold prices remain low, many of today¹s reserves are uneconomic. 40% to 50% of today¹s production is losing money. Deteriorating finances of mining companies caused by dwindling cash flow and survival tactics such as high grading to buy time add doubt to repayment prospects. In the fourth quarter of 1998, the production of several leading North American companies exceeded the life of mine reserve grade by 39%, and in the first quarter of 1999, by 28%. Some, but not all of this discrepancy reflects high grading. Curtailments of exploration and new project investment complete the picture. The credit outlook for the industry¹s hedge position has weakened dramatically in the last year. Continuing low prices will damage it beyond repair.

Hedging allows uneconomic mines to remain open, and in so doing, depletes firm capital. At the end of the day, the most important asset of some mining companies is their hedge book. The mining business is just an excuse to hob nob with bullion dealers. Without hedging, many mines would have already shut down. But hedging can only prolong the life of marginal assets temporarily. New hedges must follow declining spot prices. Salvation for the industry lies not in hedging, but in its renunciation.

If gold cannot be delivered against hedges, the financial intermediaries (bullion banks) will be caught short. The bullion banker¹s usual retort is that hedges can be bought back in the open market should a mining company founder. However, past examples were isolated incidents. The pattern of red ink, rising debt levels, and shrinking production could trigger multiple simultaneous buybacks that would sharply lift the gold price.

The Coming Melt Up

Bullion banks and miners are vulnerable to rising lease rates. Bullion banks borrow short and lend long. Credit lines extended by bullion banks have terms of several years, but they must roll over their gold borrowings from central banks on a short-term basis. Roll overs have been routine in a benign environment of low interest rates, stable financial markets, and a declining gold price. But concern over the ability of the mining industry or the bullion dealers¹ ability to repay would lead to a sharp rise in the lease rate and calls for repayment. A spike in the gold price through certain chart points would trigger margin calls to dealers or miners as the market value of the hedge book erodes. Almost all participants in this trade assume they will have plenty of time to address these risks. A sharp rise in the gold price or a prolonged rise in the lease rate would do the most damage. Both events are likely to occur simultaneously once confidence erodes in the beliefs on which the rickety market structure is based.

It is interesting to note that Barrick Gold, the icon of all hedging, is largely unprotected on its funding costs. The average duration of Barrick¹s gold lease agreements is six months. The $4 billion off balance sheet asset, a hedge fund in the estimation of some, is invested in long term financial instruments. The mismatch is essential in order to earn the contango, which is the positive carry between the cost of borrowing gold and the returns from investing the proceeds of gold sold short. Barrick, and most other hedgers, face the risk that a prolonged upturn in lease rates will erase all the benefits of hedging. We would not be surprised if increased focus and concern arises over this issue from both shareholders and corporate directors.

In contrast to mine company and jewelry trade short interest, financial shorts have no prospect of covering their positions from natural business flows. In the midst of universal pessimism, thoughts of covering are remote. To date, the shorts have had a field day. Shorting gold has been a one way street to prosperity. Low interest rates of 1%-2% and bullish financial markets (both the case until very recently) have made a compelling case to be short. There has been no cheaper source of funding in recent years.

The gold carry trade is based on a macro view that inflation will remain tame forever, and that the risk of a steady up trend in this cost of funding is non-existent. As with any successful investment strategy that is long in the tooth, latecomers to a party tend to swell the volume of capital at risk with relatively little gain. In the final stages, almost every investor is on the same side of the boat. Negative sentiment readings confirm that the gold market is offside. Market Vane bullish sentiment on gold has stayed under 20% for over two months, an unprecedented stretch of extreme bearishness. The recent hint of a more bearish chill in the financial markets and a rise in lease rates threaten to destroy the rationale for the carry trade. The case to cover speculative short positions is compelling. The four pillars of the carry trade are shaky. Lease rates are up, eroding the spread. Financial markets are turning bearish. Inflation signs are multiplying, possibly awakening investment demand. Finally, the prospect of official sector supply has become far less certain.

How will the financial short sellers cover? The prospective flood of official sales, which triggered intense short selling only a few months ago, is evaporating. The IMF sale is all but cancelled. The now politically-correct Swiss may soft pedal their plans based on social concerns for Black African miners. Similar social concerns seem to be weighing on the deliberations of other central banks. Even the hopelessly stubborn UK posture is under siege, based on recent revelations of strong opposition by Eddie George, governor of the Bank of England. The political sensitivity of these institutions was alluded to in a recent conference call held by Goldman Sachs, which in the same call abandoned its long-held bearish view. JP Morgan, another institution prominent in the bullion trade, has also adjusted its bearish stance.

Gold mining reserves are likely to decline at year-end if current prices prevail. Exploration budgets are being curtailed. Development projects have been put on hold. The raw material for hedging is peaking out. High lease rates undermine the economics of hedging. Panicky hedging by the industry in the last three months was at an unsustainable pace. Anglogold has stated that it will not expand its hedges at these prices. Their stance is equivalent to a 240 ton per year buyback, greater than the recent pace of industry hedging of 200 tons. We believe that certain producer hedgers are considering bypassing bullion dealers and purchasing whatever official sector supply becomes available to close out segments of their hedge books. Perhaps one or two major mining companies that have missed the boat would be foolish or desperate enough to initiate hedges in this depressed market, but the fresh supply would be trifling relative to the short interest. There will be no answer to short seller prayers from the mining industry.

Only gold leased from central banks can provide liquidity in a short covering rally. With physical gold owed to the central banks by bullion dealers no longer in deliverable form, the short sellers¹ only alternative will be to borrow additional gold. The short will remain short, a situation that can only cause great discomfort to the central banks. At current market prices, the aggregate short position represents $40 to $80 billion of capital at risk. A short covering rally of $50 to $100/oz would cost $8 to $16 billion, enough to make the problems of Long Term Capital Management seem insignificant.

The Great Gold Fire Sale

It is axiomatic that the way to create a shortage of a particular asset is to under price it. The US government has proved this beyond doubt across a wide assortment of commodities. There are various ways to set prices such as ceilings or price supports. These methods have been applied to gold with stellar results. Gold¹s special characteristics, which include a large above ground inventory and monetary attributes, subject it to other forms of price fixing. As with paper currency, gold¹s value is affected by the cost of interest, or the lease rate.

The mis-pricing of gold credit has been a central cause for the late stages of the gold bear market. The development of deep forward markets has only occurred in the last decade. The gold pyramid that we have described is nothing more than a conduit for the divestment of central bank gold into the physical markets. Gold in the vault has been replaced by phantom "gold receivables." If the lease rate were comparable to market rates for paper currency, the gold derivative pyramid could not function. The carry trade would blow up and the arbitrage between paper currency and gold, which is the foundation for mine hedging, would not exist.

Pierre Lassonde, President of Franco and Euro Nevada, recently wrote in the Northern Miner: " The single greatest damage caused to the gold price has been indiscriminate leasing, by central banks, of their gold reserves at giveaway interest ratesŠ.These suicidal rates are a gift to the speculators, hedge fund managers and producers who hedge." Low lease rates of 1%/year, he observes, represent an inappropriate 75% discount to US T-bills.

If gold leased by central banks totaled 10,000 tons as of June 30, the interest differential exceeds $2 billion. Since the new age central bankers view their institutions (incorrectly) as profit centers, sub-market lease rates appear to be a glaring departure from their mission. The ill-conceived leasing trade has depressed the value of a major reserve asset, and it has also transferred an embarrassing level of wealth to non- citizens.

What is the correct interest rate for gold? No one can answer the question. Interest rates set by committee lead to distortions and miscalculations. While the interest rate on gold has not been set by a committee, it does reflect the collective negative attitudes of central bankers towards a reserve asset they inherited from the previous generation and a willingness to trust in paper assets, which they did not inherit. In reality, the rate is based on nearly unanimous acceptance by central bankers and mining executives of the bullion dealers¹ sales pitch, in short, a virtual committee.

Who is benefiting from the fire sale? Financial speculators and bullion bankers are high on the list. Even higher up, however, are the world- wide consumers and investors who form the physical markets into which central bank gold is disappearing. Thanks to the gold pyramid, they are able to acquire vast quantities of the precious metal at prices well below the cost of production: present, future, and probably past, if inflation adjusted. Recent dispatches on gold consumption show very positive trends. For example, India, the largest consumer, reported 80 tons of imports in June, on a pace to shatter last year¹s record. Other Asian economies show similar patterns. The US is minting gold eagles at an annual rate of 365 tons, a record pace. Jewelry consumption worldwide is showing strong positive trends. According to the World Gold Council, consumer demand rose 16% in the first half of 1999. Refineries, which melt down central bank gold bars, are as heavily backlogged as any time in history.

Consumer demand for gold is breaking all records. Thanks to the depressed gold price, consumption for jewelry and investment exceeds sustainable sources of supply, mine production and scrap by a wide margin. Without the giveaway engineered by the bullion banks, there would be a shortage and the gold price would be much higher. As with the US dollar, there are twin deficits. The first deficit is the short interest arising from the mismatch between gold derivatives and physical gold. This is a technical market condition, which will be resolved at some point by short covering. The second deficit arises from the growing appetite of world gold consumers, fed by artificially low prices. This chronic, fundamental market deficit will be closed only by much higher gold prices, over a period of years. At some point, frenzied short covering will crowd out consumer demand. Perhaps the grass roots owners of gold will oblige the shorts by melting down their holdings. Issuers of gold derivatives choose to ignore these facts in pursuit of ever greater profits in their risky business, girded in the belief that they have the staunchest of allies in the witless compliance of nouveau central bankers and death-wishing gold mining executives. Few of these players are receptive to a wake up call. Denial is still in high gear.

The recipe for a shortage has been carefully followed. A few finishing touches may be required before a market epiphany. There is no known reconciliation between paper and physical positions, and none will be attempted until after the squeeze. The weakness of credit analysis and supervisory oversight, as well as the many ambiguities in the linkage between paper gold and physical can flourish only if there is supreme confidence in gold¹s permanent downtrend. The trust and confidence essential to balance the gold derivatives pyramid depends on three critical errors: that mine reserves = physical gold; that gold receivables = gold on hand; and that financial markets will enjoy smooth sailing indefinitely. Trust is nothing more than a state of mind. When this levitation is finally exposed and its illusions shattered, it is ludicrous to think the imbalances can be corrected by a small rise in the price and within a comfortable time frame. Expect the resolution to be swift, furious, and uncomfortable for those caught short.

John Hathaway, CFA, Senior Managing Director, Co-Portfolio Manager

Mr. Hathaway is a co-portfolio manager of the Tocqueville Gold Fund, as well as other investment vehicles in the Gold Equity Strategy. Mr. Hathaway also manages separately managed accounts for individual and institutional clients.  He is a member of the Investment Committee and a limited partner of Tocqueville Asset Management (www.tocqueville.com). Mr. Hathaway began his career in 1970 as an Equity Analyst with Spencer Trask & Co. In 1976, he joined investment advisory firm David J. Greene & Co., where he became a partner. In 1986, he founded Hudson Capital Advisors and in 1988 became Chief Investment Officer of Oak Hall Advisors. He joined Tocqueville as a Senior Partner in 1998. Mr. Hathaway has a BA degree from Harvard College and an MBA from the University of Virginia.  

The total world's holdings of gold could be transported by a single solitary oil tanker.