LLCPostings2
Readings from the Book of Barrick: the Sequel by Bob Landis A few weeks ago I was startled out of my afternoon reverie by an inquiry from Melissa Davis, an investigative journalist for TheStreet.com. She was gathering material for a piece on Barrick Gold Corp. and wanted to follow up on my essay, Readings from the Book of Barrick, posted back in May. Sure thing, I said, delighted to receive a little recognition from someone outside the Taliban Wing of the gold bugs. There was just one problem: I hadn't kept up with Barrick ever since it became clear that CEO Randall Oliphant was exactly right when he asserted that the rumors regarding a deal with Gold Fields, which had prompted my musings, had been dreamed up by some guy in his basement. So I had to do a little homework in order to respond to Melissa's thoughtful questions. In the process, I learned some things that may be of interest to fellow gold bugs, but which may be too technical for inclusion in an article geared to a broader audience. Accordingly, Melissa's questions and my responses are set forth below, modified slightly for this format.
We'll post a link here when Melissa's article comes out.
1. Has much changed at Barrick since you published your essay? If so, what?
Three developments strike me as noteworthy.
First, they committed to a serious reduction in their gold hedgebook exposure. This was announced in a press release on September 17, which also laid out a growth plan. The growth plan got top billing but it’s pretty clear the point of the release was the hedge cutback, because the “growth plan” was mostly old news. The announced reduction would chop their gold forward exposure by a third, taking it down from about 18 million ounces to about 12 million ounces by the end of 2003. It would also slash their call option and variable price sales contract position by 1.5 million ounces over the same period, down from about 2.7 million ounces. The pledge is “based on market conditions,” whatever that means, but otherwise it’s unequivocal. This is a big deal if it’s real, but it was quickly overshadowed by the second development.
Second, they wrong-footed the market by lowering earnings guidance out of the blue. This was announced in another press release issued on September 26, just 9 days after the first one. The release attributed the miss to operating issues in Goldco: higher costs and lower output. What was weird about it was the timing. Why not combine the releases rather than stagger them so? It was like Goldco and Hedgebook were no longer on speaking terms. Anyway, the snafu clearly irked some of the bigger players. I happened to catch Randall Oliphant facing the music at the Mining Investment Forum in Denver a week later. I couldn’t help feeling sorry for the guy -- seeing him fidget and glance furtively at his watch under some downright hostile questioning reminded me of Bart Simpson getting reamed by Principal Skinner. I think it’s fair to say that the proffered explanation as to what happened and why was not enthusiastically embraced by a lot of the attendees at Denver.
Third, they have been unusually active with their silver shorts. At the end of the first quarter, they had 26 million ounces subject to spot deferred forward contracts, and about 21 million ounces subject to call options. At the end of the second quarter, they had put another 12 million ounces under the forwards, meaning they had sold short that many more ounces in just one quarter. So at June 30 they were short about 59 million ounces, which is about 10% of the total global annual production of silver. At September 30, they had taken it back down to about 50 million ounces. Now I don’t follow the silver market that closely but their activity here has always struck me as curious. They don’t break out silver separately except to note that Eskay Creek produced 15.5 million ounces of the stuff in 2001, and that they treat it as a byproduct and use it to reduce operating expense. I understand there has been a labor dispute at the Horne Smelter, reducing throughput this year. I suppose this may account for some of the cost increases at Eskay flagged in the September 26 press release. But the bigger issue seems to me to be twofold. First, why bother shorting such small potatoes so aggressively, particularly at historically low prices? Secondly, what provision do their models (and contracts) make for the non-linear event of war, which typically consumes a lot of silver? If there is a war with Iraq that doesn’t end quickly, I guess we’ll see, because there is no more U.S. strategic stockpile, and central banks don’t hold a lot of silver anymore.
2. Did you ever get answers to any of the questions you raised in Readings?
Sort of, indirectly. In August, they trotted out a distinguished consultant/part time academic named Martin Murenbeeld, who authored a document entitled In Defense of Gold Hedging -- The Case of Barrick that was posted at several sympathetic Internet sites in September. Later in that month, a day after the press release announcing their hedging cutback, they posted at their website a presentation entitled “Barrick’s Gold Sales Program / A Detailed Look at Gold Hedging” (www.barrick.com/7_Presentations_Webcasts). Neither one mentions Readings but both touch on some of the topics it covered, and I am flattered to suppose that neither would have been published had I not written Readings.
Dr. Murenbeeld’s report in particular is interesting. It does not disclose the nature of his relationship with Barrick. However, there is no doubt in my mind that Dr. Murenbeeld is more than just a volunteer stirred to action by unkind remarks anent his favorite corporation. This is so for several reasons. First, his report displays a familiarity with Barrick’s hedging contracts that I don’t think could have been acquired simply through review of publicly available information. Second, it dovetails nicely with the Company’s posted presentation. Third, it offers legal opinion, an area with respect to which Dr. Murenbeeld has no disclosed expertise: “A sample of the commentary I have collected serves to highlight the vitriolic nature of the criticism levied at Barrick -- criticism tantamount to slanderous, and which a company smaller than Barrick might otherwise find profitable to litigate”; and “Yet despite an enviable record, Barrick has come under yet more criticism in recent days, criticism bordering on the libelous.” Sounds to me more like not-so-veiled threats from Company counsel than voluntary observations of an impartial observer. Finally, Dr. Murenbeeld lists Barrick as a client at his website (www.murenbeeld.com/clients.htm). So draw your own conclusions.
While Dr. Murenbeeld appears to have read Readings (I recognize some of my language), he seems to have missed the point. In my commentary, I asked two main questions. The first was what, exactly, are the prices on those spot deferreds. I parsed some language in their public documents and concluded that there were in fact no fixed prices, just “projections of future values of existing cash balances using various assumptions on interest rates.” In responding to this question, albeit indirectly, Dr. Murenbleed informs us: “When a gold producer contracts to sell gold forward it is for a specific date at a specific price. For example, a producer may contract to deliver 100 ounces of gold one year from today. The buyer, or counter-party, is typically a bullion bank, and it is legally bound to pay the contracted price upon delivery of the 100 ounces. The price is set today, of course.” I see. That’s enlightening, isn’t it? No wonder the prices reported in their “schedules” are all over the road. The second question was why their prices are so much better than everybody else’s. Again parsing some language in their reports, I concluded that they jazzed up their returns with funky derivative income and other stuff and that this all appears to be history since the banks now hold all the cash. As to this question, Dr. Murenbeeld spares us.
But Dr. Murenbeeld’s report and the Company’s presentation are not total losses. They tell us some things we didn’t know. We learn, for example, that the terms of the spot deferreds range from 10 to 15 years; we had been under the distinct impression that they were all 15 years. We learn also that the contracts “re-price” periodically; I believe this is the first mention of such a provision. This has interesting implications which I’ll touch on in a moment. We learn that the contracts are “evergreen,” that is, they roll over automatically on a year-by-year basis -- unless a counterparty decides not to. (The reader is evidently supposed to derive comfort from the observation that “no counter-party has ever declined to extend an Agreement.”) We learn also that while there are no downgrade provisions (I take this to mean that the contracts don’t have a trigger if Barrick loses its A rating), there are two financial covenants: minimum net worth of $2 billion, and Long Term Debt/Consolidated Net Worth of 1.5:1. We learn also that there are two operating conditions relating to the ability to “roll” the contracts: the existence of a gold pricing market, and the continued ability to produce gold. There are also some matrices in the presentation which suggest that there are even more assumptions and variables that go into the selection of the “prices” reflected in those schedules than I had previously understood.
3. In a nutshell (if any nutshell’s big enough) why is Barrick a risky stock, and what could cause it to blow up?
I don’t know whether Barrick’s a risky stock. I don’t think anyone can have an informed view of this company as an investment, unless and until it discloses in full the terms of those hedgebook contracts. And I don’t mean piecemeal summaries dribbled out occasionally under duress. It’s what we don’t know, rather than what we do know, that’s the problem. Sort of like Donald Rumsfeld’s unknown unknowns. It’s hard to be comfortable with a “just trust me” approach in today’s climate. If I were a shareholder, I would be jumping up and down demanding full disclosure, if for no other reason than to demonstrate once and for all there’s no problem.
As to what could cause it to blow up, again, I don’t know that anything could. The Company’s presentation states explicitly: “There is no price and there is no series of market events that would cause Barrick’s gold hedge program to ‘blow up.’” Maybe they’re right. But it is intriguing to consider three questions implicitly posed (or begged) by the new disclosures in Dr. Murenbeeld’s report and the Company’s presentation.
First, how long have the existing contracts been in effect? Are we in the 10th year, the 12th, the 1st; is there a blended duration; or what? Recall that the Company’s been at the hedging game at least since 1990. So some of these contracts are presumably in evergreen extension mode: they get rolled over every year unless somebody doesn’t want to. Does that mean that some or all of the counter-parties can now say, no thanks, just give us the gold, even though they’ve never before declined to extend? After all, things are getting a little warm over at JP Morgan; who knows, maybe some morning they’ll wake up and decide they’d like their gold back, now that they’ve fired some of their most gold-knowledgeable employees. Do they have the right to demand it? How much is at risk? Is it happening now, i.e., is that the real reason behind the announcement of a dramatic reduction in the hedgebook? It doesn’t take too much imagination to conjure up a scenario more dangerous even than a margin call.
Second, what does it mean that there has to be a “gold pricing market”? Does that mean just Comex? LBMA? Shanghai? Sydney? Hong Kong? Dubai? All or any of the above? It is not beyond the realm of possibility that someday the “paper” gold market, typified by the Comex gold contract, could collapse in a variant of the old fashioned run on the bank, as more claims are presented than can be satisfied in physical gold. As Comex is the principal organ of gold price discovery, does this mean that Barrick’s fate is tied to the continued viability of the Comex gold contract? To other markets? Which, and how so?
Third, how do the reset provisions work, and what risks do they pose? Dr. Murenbeeld and the Company's presentation both note that the contracts feature re-pricing dates at periodic intervals. An example in the presentation suggests that what happens is that at certain points (in the example, the 5th , 6th and 11th anniversaries of the starting date) the existing value of a contract, that is, the proceeds of the original short sale plus the net interest earned, or contango, is substituted for the original contract starting price. This forms the new basis for calculating the contango, and by extension, what the Company will book as revenue when it closes out the short. We are used to thinking of this as a matter of relative positives: that is, the amount they can book as revenue may be less than they could have received by selling at spot, if the spot price goes up, but it’s still bigger than the amount they borrowed in the first place, due to the magic of compounding over long periods of time even at low interest rates.
But Dr. Murenbeeld raises a freshly severed head. What if the net interest rate is negative, which would happen if on a reset date the reinvestment rates on the proceeds are in fact lower than their cost of funds, or lease rate? (The fancy term for this phenomenon is backwardation, which usually relates to inverted price curves in non-monetary commodities.) Under this scenario what he seems to be saying is you could have a situation where the reset contract amount is actually less than the amount borrowed. How low can it go? Below zero, for example? Dr. Murenbeeld pooh-poohs the issue, saying that it’s still a matter of opportunity cost rather than outright loss potential, and that in any event backwardation is rare since 1993. Maybe he’s right on the possibility of loss; I don’t have any way to check the accuracy of his assertion. As to the rarity of backwardation, I’d like to see a slightly bigger data set than the “golden” years of hedging before dismissing the likelihood of a not-so non-linear event. Perhaps they’ll post a sensitivity analysis showing what would happen under a reset provision in today’s low interest rate environment if we were to have a spike in lease rates like the one we had in the wake of the Washington Agreement in September 1999. That’s not so long ago, and what’s striking is the spike in rates coincided with a spike in spot. This scenario would seem to be doubly disagreeable for a heavy hedger in today’s world.
4. What is your reaction to Barrick’s recent changes in its hedging strategy?
I think it’s huge, if it’s real. The market seems to have forgotten about it, maybe because it got lost in the flap surrounding the earnings miss, and was buried in a press release that trumpeted a growth strategy for Goldco. They’re talking reduction equal to an entire year of production. But is it real? One caution flag in their third quarter report caught my eye, a footnote that I haven’t seen before. The footnote relates to spot deferred gold contracts. It states that the current total of 16.9 million ounces subject to forward sale contracts is “net of 300,000 ounces of gold contracts purchased.” This is interesting. What are these purchase contracts? Who is the seller? What is the credit or other risk associated with the contracts, and how real is the offset? How much of the 6 million ounces to be reduced will be cut in this fashion rather than through deliveries into the hedges? Why do it this way, through paper, rather than with metal? Or, if the contracts won’t let you deliver, why not just take your medicine and buy your way out of them? Either of these alternatives would be positive for the gold price. Just strapping on another derivative looks neutral at best. Is that why it’s done? Anyway, let’s see what they do rather than what they say.
5. What do you make of the adjustments to fair value of Barrick’s gold contracts in its latest quarterly report?
I don’t know what to make of them. I must say I was struck by the unheralded appearance of the schedule itself:
Continuity Schedule of the Change in the mark-to-market value of the Gold and Silver Hedge Position The estimated fair value of the gold contracts at September 30, 2002 was approxi- mately $301 million negative, and the fair value of the silver contracts was $19 million positive. These values are based on the net present value of cash flows under the contracts, based on a gold spot price of $324 per ounce, silver spot price of $4.51 per ounce, and market rates for Libor and gold and silver lease rates. The year-to-date change in the fair value of the Company’s gold contracts is detailed as follows: Fair value as at December 31, 2001 $356 Impact of $152 million realized gains in the period to date (152) Impact of change in spot price (from $279 per ounce to $324 per ounce) (883) Impact of contracts added (21) Implied contango period to date 109 Impact of change in valuation inputs other than spot metal prices (e.g. interest rates, lease rates, and volatility) 290 Fair value as at September 30, 2002 $(301)This “Continuity Schedule” is a new one on me, and it demonstrates why trying to track their numbers across reporting periods is like shooting sporting clays: you never know what to expect. Mr. Oliphant told the Boston investment conference hosted by Merrill Lynch back in May that there is about an $18 million negative swing in their mark-to-market with each dollar increase in the spot price of gold. The mark-to-market was negative $127 million at the end of the first quarter, based on a spot price of $302. So at the end of the third quarter, where they based off $324 spot, an investor with the soul of an accountant would look for a mark of negative $523 million, give or take. But no, now we are given a brand new calculation that nets in some stuff we’ve never seen before, derived pursuant to methodologies that remain shrouded in mystery. The result is to pare the mark by 40%, leaving it at the published negative $301 million. Go figure.
November 27, 2002
DISCLAIMER & CONFLICTS THE FOREGOING ESSAY IS FOR YOUR INFORMATION AND AMUSEMENT ONLY. NEITHER MR. LANDIS NOR GOLDEN SEXTANT ADVISORS LLC (“GOLDEN SEXTANT”) IS SOLICITING ANY ACTION BASED UPON IT, NOR IS EITHER MR. LANDIS OR GOLDEN SEXTANT SUGGESTING THAT IT REPRESENTS, UNDER ANY CIRCUMSTANCES, A RECOMMENDATION TO BUY OR SELL ANY SECURITY. THE CONTENT OF THE ESSAY IS DERIVED FROM INFORMATION AND SOURCES BELIEVED TO BE RELIABLE, BUT NEITHER MR. LANDIS NOR GOLDEN SEXTANT MAKES ANY REPRESENTATION THAT IT IS EITHER COMPLETE OR ERROR-FREE, AND IT SHOULD NOT BE RELIED ON AS SUCH. MR. LANDIS AND GOLDEN SEXTANT AND ENTITIES UNDER THEIR RESPECTIVE CONTROL MAY HAVE INVESTMENT POSITIONS, LONG OR SHORT, IN ANY SECURITIES MENTIONED, AND ANY SUCH POSITIONS MAY BE CHANGED AT ANY TIME FOR ANY REASON.
Readings from the Book of Barrick:
A Goldbug Ponders the Unthinkableby Bob Landis A rumor hit the wires last week to the effect that Barrick Gold Corporation (NYSE: ABX) was planning a bid for Gold Fields Limited (NYSE:GFI). The first mention came in the MiningWeb: “Gold Fields raid talk persists.” This was followed closely by the National Post: “Barrick, Anglo seen as suitors for Gold Fields.” Then the Russians picked up the scent with “Gold Fields Rises 6% on Takeover Speculation.” But it was not until Reuters chimed in with “Barrick dismisses rumors of Gold Fields deal” that we knew something was up.
Our first reaction was, let’s be candid, negative. After all, in goldbug demonology, Barrick is Attila the Hedger, a scourge that has ravaged the gold sector for years. Gold Fields, by contrast, is a vestal virgin. The very thought of a coupling is distasteful. At the end of the day, though, this is about money, so we were moved to inquire: from the investment perspective of a Gold Fields shareholder, would Barrick shares be acceptable currency in a bid for our fair sister? We looked into it, and concluded, provisionally: NO. While a close reading of its annual reports confirms that Barrick indeed has a brilliant business model, the exercise raises a number of issues and puts one in mind of other brilliant models that have cracked up. Accordingly, unless Barrick and its accountants were to provide some serious comfort on some fairly basic questions, we’d pass on their paper and opt instead for cash, thanks very much.
A Beautiful Model Barrick’s business model is a very clever piece of financial engineering. In simplest terms, it consists of two parts: a mining business, which produces gold, and a financial business, which earns a spread. The gold producer -- call it “Goldco” -- is a real business: it generates financial information that is captured in published financial statements. Goldco is a collection of very fine mining properties assembled over a period of about 18 years. It provides the raw material for the pictures and much of the commentary in the annual reports; there are physical places you can go, tires you can kick, employees you can talk to.
The financial entity -- call it “Hedgebook” -- is a virtual business. It exists “off balance sheet,” and generates a net yield that is not captured in published financial statements. Hedgebook consists of a file drawer full of contracts no outsider has seen and a giant pool of cash proceeds from short sales of gold over a period of about 14 years. You can’t see it, you can’t touch it, and generally the only mention it gets is a paragraph or two in the annual report.
Goldco and Hedgebook are approximately equal in size and significance. As of yearend 2001, Goldco had $5.2 billion in total assets, while Hedgebook weighed in at $5.5 billion. As stated in the 2000 Annual Report, “Barrick manages this asset as closely as it does its mines to generate the highest returns at the lowest risk.”
Every year, Goldco reaches through a cyber-membrane and pulls some numbers out of Hedgebook. These numbers are thereupon exposed to the light of day, and go directly into Goldco’s reported results. Understanding how Hedgebook generates these numbers, and how they cross over onto Goldco’s books, is the key to understanding Barrick. Kids, don’t try this at home.
A Star Is Born Hedgebook is the spawn of a creative financing technique: the so-called “spot deferred” forward sale. The spot deferred made its public debut in Note 10(c) to Barrick’s 1990 financial statements. For all its future significance, it got a rather low key introduction:
The Company has also entered into contractual arrangements with several major financial institutions to deliver 1,035,000 ounces at prices ranging from $390 to $428 per ounce. Delivery under these contracts can be deferred at the Company’s option for up to 5 to 10 years depending on the individual contract.
The architect of this stunning innovation remains anonymous. One is tempted to speculate that it was CEO Randall Oliphant himself, who was promoted to Treasurer from Assistant Treasurer in January 1991, having joined the company from Coopers & Lybrand, its auditors then and now, in April 1987. Whoever he was, the inventor of the spot deferred should be celebrated in company lore and song, for this was the financial equivalent of discovering the Goldstrike Mine itself. This technique permitted Barrick to monetize that mine’s future revenue streams and harness them to the awesome power of compounding, over a timeframe to be selected by Barrick, and to do all this off the books. The critical feature is not the forward contracts themselves, but rather their funding, which Note 10(c) did not address: each forward contract generated current cash proceeds from a corresponding short sale in the spot market. That cash was to be invested, earn interest and grow into future values which would be booked as revenue only at such time as Barrick decided, in its sole discretion, to settle the account. O brave new world!
Seven years later, in its 1997 Annual Report, Barrick placed proud and proper emphasis on the genius of this structure by describing it thus:
Interest on Gold in the Ground
While Barrick has developed sophisticated tools for its hedging program, the core process is straightforward. The Company sells its gold at the current spot price while the ounces are still in the ground, and earns interest on the proceeds from the spot sale before delivering its production against the contract.
The Company works with a bullion dealer and a central bank, which lends gold from its reserves to the dealer for the sale. When Barrick delivers its gold, which is used to repay the central bank, it receives the net proceeds from the sale: the original spot price, plus accumulated income earned, minus the cost to borrow the gold.
The net interest is the forward premium (“contango”) that each Barrick ounce earns because of hedging. Current contracts will earn over 7% interest on the proceeds from the spot sale while paying 2% interest on the borrowed gold, generating a premium of over 5% a year.
Premium and Protection
Through the use of spot deferred contracts (which it pioneered), the Company may either sell its gold at the hedge price when the contract matures, or - if the spot price is higher - sell its gold at spot instead and roll the contract forward. Barrick, which already had the ability to roll contracts forward up to 10 years, recently negotiated with one of its bullion dealers to extend the roll forward period to 15 years. No other company has this degree of flexibility, which has been granted Barrick because of the unequalled size and quality of its reserves and balance sheet.
How simple. How elegant. How inadequate are mere words to do justice to this great and beautiful idea.
That first amount “on deposit” arising from that first short sale implied by Note 10(c), augmented by over a decade of contango earned on ever bigger sums received on ever more favorable terms, has grown into today’s Hedgebook: a $5.5 billion, throbbing, brooding omnipresence, the product of over 18 million ounces of gold sold short, bubbling, seething, compounding (apparently) tax free; in 15 years this thing could eat Fannie Mae.
Linear Results in a Non-Linear World The reported results of this revolution in finance confirm the brilliance of its author. In the Letter to Shareholders portion of each of Barrick’s annual reports over the last five years, management has prominently and proudly noted Hedgebook’s contribution:
1997:
Hedging has contributed a total of $765 million in additional revenue and $580 million in additional net earnings over the last ten years. That works out to an average premium of $46 per ounce over the spot price, for every ounce sold, during that time.
1998:
Under its Premium Gold Sales Program, Barrick realized $400 per ounce on its gold sales in 1998, compared with $420 in 1997 ($415 in 1996). The Company generated a $106-per-ounce premium over the average spot price of $294 per ounce for the year, resulting in $340 million in additional revenue in 1998. Over the past 10 years, Barrick has realized $56 per ounce above the average spot price of $356 per ounce during the period, or $1.1 billion in additional revenues.
1999:
Our Premium Gold Sales Program contributed $391 million in additional revenue in 1999. Over the past three years, the Program has contributed $1 billion in additional revenue, and averaged a $100 premium over the spot price.
2000:
During 2000, our program realized an average price of $360 per ounce, an $81 premium to the average spot price, which translated into $300 million in additional revenues.
2001:
For the year, we realized a $70 per ounce premium over the average spot price of $271 on the 61 percent of production delivered into our Premium Gold Sales Program. This compares to a realized price of $360 in 2000 and a premium of $84 on the 63 percent of production delivered into the Program during the year. The balance of the ounces sold -- principally Homestake's production -- were sold at an average price of $277 per ounce in 2001. Overall, we realized an average price of $317 per ounce, $46 higher than the average spot price for the year, generating an additional $289 million in revenue.
Riddle Me This When something seems too good to be true, it often is. And so we must wonder, what is really going on here? We have a number of specific questions, but two rather basic ones will suffice to illustrate our puzzlement:
1. What, exactly, are the prices on those spot deferreds?
A casual reader of Barrick’s more recent financial statements and annual reports may be forgiven for concluding that the spot deferreds have specific prices for ounces to be delivered in the future. For example, Note 5B of the 2002 First Quarter Report contains the following comment and schedule:
We have entered into the following spot deferred sales contracts, with various counterparties, that establish selling prices for future gold and silver production, and which act as a hedge against possible price fluctuations in gold and silver [emphasis supplied].
Scheduled for delivery in 2002 2003 2004 2005 2006+ Total Spot deferred gold sales contracts Ounces(thousands) 2,100 2,600 2,800 1,500 9,000 18,000 Average price per ounce $365 $340 $340 $335 $342 $344 Spot deferred silver sales contracts Ounces(thousands) 12,000 8,000 3,000 2,000 1,000 26,000 Average price per ounce $4.75 $5.05 $5.10 $5.10 $5.10 $4.92[Memo to file: Silver? We thought that was just a byproduct used merely to reduce operating expense at Goldco. So what's it doing over here in Hedgebook? And in size! In addition to spot deferreds, Barrick is short another 20,750,000 ounces through written call options, 10,000,000 for 2002 alone. Ay, caramba! Let's make a simplifying assumption: for purposes of this commentary, the silver issue does not exist. We'll let Ted Butler worry about that one.]
Negative, Captain. These contracts, as we know from careful exegesis of other portions of the text in question, do not in fact establish “selling prices;” rather, the “selling prices” are merely projections of the future values of existing cash balances using various assumptions on interest rates. Recall the description of how forward pricing works in the excerpt from the 1997 annual report: “When Barrick delivers its gold, which is used to repay the central bank, it receives the net proceeds from the sale: the original spot price, plus accumulated income earned, minus the cost to borrow the gold” [emphasis supplied]. The sale referred to is the short sale entered into to fund the forward, not the final sale on delivery. All Goldco gets when it closes out a particular spot deferred is the original short sale proceeds plus interest.
So a critical component of the “selling price” is the assumed rate of interest on the proceeds of the short sale. This explains the statement in the 2001 annual report:
Contrary to most businesses, we are adversely affected by lower interest rates rather than higher rates. In higher interest rate environments, we earn higher premiums for our spot deferred sales program because the forward price is primarily a function of US interest rates, as well as higher interest income on our cash balances [emphasis supplied].
Making such a big bet on interest rates involves risk. The cost of funds could spike up, the return on investment could go down; either swing could dramatically affect the size of the amount available on delivery. Barrick’s solution: derivatives.
The Company maintains an interest rate risk-management strategy that uses derivative instruments to mitigate significant unplanned fluctuations in earnings or cash flows that arise from volatility in interest rates.
But what about the rollover option in the spot deferreds? How does a decision to defer delivery affect pricing assumptions? What other costs are incurred in the exercise of this flexibility? Take a hypothetical example. Assume on a day when spot is $250, Goldco enters a contract to deliver 1 ounce on June 30, 2002. But when June 30 rolls around, spot is actually $400, so Goldco does the rational thing and elects to put off delivery for a year, selling its ounce instead into the spot market. Assume one year later spot has fallen to $200, so Goldco elects to deliver into the contract and book a gain. Who bears the loss of the “profit” that the bullion bank/central bank would have enjoyed if it had received its gold on the due date? It could have turned around and sold in the spot market then for $400, but now it’s stuck at $200. Can these contracts be so cleverly drawn that in every scenario, it’s heads Barrick wins, tails the banks lose? Or does each deferral trigger yet another actual or synthetic short sale by the bullion bank to lock in the profit deferred? Does this explain the phenomenal growth in J.P. Morgan’s derivatives position? If so, what is the corresponding adjustment to Hedgebook’s accounts? Or does each deferral just trigger some sort of payment or obligation which need not be disclosed, as it takes place beyond the cyber-membrane?
Grid and Bare It. That spot deferred “prices” are mere projections, changed as circumstances warrant, is demonstrated by the futility of attempting to track them across reporting periods. The following tables show the “selling prices” for gold under spot deferred contracts as presented in Barrick’s annual reports for the last three years and the first quarter of 2002:
2002, First Quarter:
Scheduled delivery 2002 2003 2004 2005 2006+ Total Ounces (thousands) 2,100 2,600 2,800 1,500 9,000 18,000 Avg. price ($/oz/) 365 340 340 335 342 3442001:
Scheduled delivery 2002 2003 2004 2005 2006+ Total Ounces (thousands) 2,800 2,600 2,800 1,400 8,600 18,200 Avg. price ($/oz.) 365 340 340 340 344 $3452000:
Scheduled delivery 2001 2002 2003 2004 2005 2006 2007+ Total Ounces (thousands) 3,800 3,800 2,100 1,600 700 600 2,300 14,900 Avg. price ($/oz.) 340 340 362 364 355 357 360 3501999:
Scheduled delivery 2000 2001 2002 2003 2004 2005 2006+ Total Ounces (thousands) 3,700 3,700 1,800 900 900 500 2,100 13,600 Avg. price ($/oz.) 360 360 360 360 360 361 366 361Try tracking the scheduled amount for delivery in 2002, just for fun (shown in bold in the tables). It starts out life in our series back in 1999 at 1,800,000 ounces, sporting an average price of $360. In 2000, the very next year, some committee or other apparently decided that it would be better for reporting purposes to beef it up to a whopping 3,800,000 ounces, but felt the need to chop the “price” by $20, taking it down to $340. Cooler heads prevailed in 2001, however, and the amount was reduced to 2,800,000 ounces, but what’s this -- the “price” was increased to $365, higher even than 1999’s version. One quarter later, at least the “price” stays the same, but the amount mysteriously declines to 2,100,000 ounces.
Notice a pattern here? Neither do we. These tables breathe new life into the tired term “random.” While we’re on the subject, we wonder precisely what an investor is supposed to make of this information. Whence are the “schedules” derived? Why are they changed so often? Of what utility is a “schedule” which varies as to amount and price so radically period over period?
Is this price derivation and schedule disparity thing a big deal? We don’t know, but the false rigor bothers us.
2. Why are they so much better than the prices everybody else gets?
It is also troubling that no other hedger comes close to matching these results. Fifty, eighty, hundred dollar premiums over spot, year after year? How do you get that at a passbook savings rate, even allowing for the magic of compounding?
Granted, Barrick has an A rated balance sheet, unique in the industry. But is that enough to explain such a huge difference? The financial markets are ferociously competitive, and a great idea has a proprietary shelf life of about a minute. How is it possible that over a 14-year period, swarms of aggressive young investment bankers have not taken this concept and applied it to everyone in the phone book, thereby causing a convergence of results?
Once again, the answer may be found in a synthesis of disparate portions of the annual reports. The MD&A section of the 2001 annual report informs us that Barrick gooses returns in three ways: First, it has until recently (see Holding the Maginot Line, below) put a slug of Hedgebook under the management of gunslingers, we mean, fund managers.
To improve returns, we have diversified [Hedgebook] by investing approximately $1 billion or 17 per cent of the overall Program into an off-balance sheet fixed-income portfolio of corporate securities with a number of top fund managers, with changes in fair value being reflected in the income statement and on the balance sheet.
Second, it takes directional risk on funding rates:
We have locked in gold borrowing costs on approximately two-thirds of the overall Program while maintaining floating lease rates on the balance to maximize the forward premium earned.
Finally, it adds apple juice to orange juice, tossing in premium from the sale of other derivatives:
Third, we sell gold call options to generate additional revenue. The calls are written at prices at which we would be comfortable adding to our forward sales program if we are exercised. We have the ability to convert the call options exercised, at our discretion, including related premium income, into spot deferred contracts, which accrue contango (US$ Interest Rate - Gold Lease Rate) the new delivery date. The call options and the premiums from expired options are recorded on the balance sheet and the fair value adjusted through earnings.
So is this the answer? They take the bread & butter Hedgebook projections, add in the juice from various derivatives, extend or shorten the schedule as needed, divide by some number of ounces which they have sold under this rubric or intend to sell or might sell or could sell unless conditions change, and presto! A super high price-per-ounce calculation, and sure-fire $X-greater-than-spot bragging rights for the annual report. But when the numerator is such a hodgepodge of unlike income categories, and the denominator appears to be a moving target, of what value is the information? What are we to make of this?
We don’t know, we don’t get it, and it bothers us that we have to work so hard to try to figure it out.
The Dance of the Ashanti Goldbugs are paranoid by nature, and we get especially cranky when we’re in the dark on the kinds of basic questions posed above. Our anxieties, as they relate to the prospect of ending up somehow with Barrick paper in a bid for Gold Fields, center on one hellish scenario: a collapse like that of Ashanti Goldfields Ltd. (NYSE:ASL) in the wake of the Washington Agreement in September 1999.
Confession time: we were long Ashanti. Not by much, but still, we’ll never forget the awful sight of hedgebooks blowing up, and the acrid smell of burnt toast wafting out of our portfolio. All because the price of gold rose. (And they wonder why we’re paranoid.)
Now, there are a number of obvious differences between the hapless Ashanti and the swashbuckling Barrick. To begin with, Barrick is not the typical prey of the derivatives sharks described in Frank Partnoy’s FIASCO. Randall Oliphant is a financial maven, not a miner, and Hedgebook is a financial business, not a mine; if anything, it would seem Barrick consistently eats the bankers’ lunch.
Add to this the important differences in the reported terms of the forward contracts themselves. Ashanti’s book, we learned after the fact, consisted of an assortment of amusing little delicacies cooked up by those merry chefs at Goldman Sachs. See, e.g., L. Barber & G. O'Connor, “How Goldman Sachs Helped Ruin and then Dismember Ashanti Gold,” Financial Times (London), Dec. 2, 1999, reprinted at http://groups.yahoo.com/group/gata/message/299. It featured treats like Margin Call Marengo and Death by Chocolate Delivery Requirements.
Not so our spot deferreds, according to Company pronouncements. Barrick’s SVP and CFO, Jamie Sokalsky, put it thus in a letter to the editor published in the November 23, 2001, edition of Canada’s The Globe and Mail:
Critics cite a few high-profile cases of companies running into trouble with their hedging programs when the gold price rose. Barrick’s program is different in kind and quality. For one thing, we have no margin calls to worry about. But the most important difference is that if the gold price rises, Barrick has the unique flexibility to defer its contracts for up to 15 years.
To give an example, if the price of gold shot up to nearly $600 tomorrow and stayed there for 15 years, we would realize every cent of that increase.
Now that’s a striking assertion. Let’s think about its implications for a moment. According to the 2001 annual report, Goldco has proven & probable reserves of about 82 million ounces, and produces at the rate of about 6 million ounces per year. If, in Mr. Sokalsky’s hypothetical, Barrick were to sell all its output in the spot market for all 15 years rather than close out the accounts, at the end of that period it would be out of gold in the ground but the obligation to deliver under those spot deferreds would remain. So unless Barrick succeeded in adding another 80 million ounces of reserves over that 15 year period (how much do new reserves cost in a $600 gold environment, we wonder, and just how plentiful are they?), Mr. Sokalsky’s counterparties would be left holding a rather large empty bag. Now they may not be as smart as the Barrick guys, but it strains credulity that they would just stand by and watch their position disappear altogether. Assuming a normal commercial relationship, at some point they will turn the screws, contract niceties or no.
How Is a Bullion Bank Like an S&L? Readers of a certain age may remember the S&L crisis of the late 1980’s and early 1990’s, a time long ago when greed, fraud and and incompetence stalked the financial landscape, and the federal government ended up “resolving” scores of bust financial institutions through innovative and costly receivership and liquidation techniques. Behind the colorful personalities (Who can forget Charlie Keating?) and the garish irregularities, there lurked a simple problem: the S&Ls had been locked into a classic funding/yield mismatch. They borrowed short, with floating rate funding, and loaned long, in fixed rate mortgages. This worked just fine in the stable, monochromatic world of Leave It to Beaver, but broke down entirely in the gonzo world of Apocalypse Now, as short rates streaked north and the S&Ls’ books went negative.
Consider the parallel in the funding structure of a spot deferred. The bullion bank borrows gold from a central bank on a short-term basis (we understand the typical maturity is under one year), and the loan is routinely rolled over. Hedgebook, by contrast, borrows long from the bullion bank. Not just long, but indeterminate - they pay it back when they want to, as Mr. Sokalsky points out. So the bullion bank is caught in the middle with yet another classic funding/yield mismatch. In the stable world of, say, The Truman Show, this would be fine, because non-linear events do not happen; central banks never change their minds and ask for their gold back, gold prices always stay suppressed, and yield curves never invert. Now it so happens that for the past seven years or so, the gold market has been in a similarly artificial environment [The Greenspan Show?], with events conspiring (oops - there’s the “c” word - it just slipped out, honest!) to produce steady, linear results for players on the short side. Give or take an Ashanti.
But even in The Truman Show, Truman escapes and his world suddenly becomes non-linear. Are we to believe, with Mr. Sokalsky, that if the bullion banks come under pressure from the central banks to return the gold, that the bullion banks will not in turn find a way to force Goldco to hand it over, no matter what the contracts may say? Query, as we say in the law.
Does Moody’s Give an A for Effort? Barrick’s annual reports cite its “A” credit rating as the reason why it gets such favorable terms on the spot deferreds. We wonder, though, whether in fact the rating is a condition rather than merely an explanation: is it possible that the flexibility of the spot deferreds is, explicitly or implicitly, dependent on maintenance of that “A” rating? The question is of more than academic interest.
In Note 16C to its 2001 financial statements, Barrick informed us that, with respect to the spot deferreds:
The favorable fair value of the contracts at December 31, 2001 was $356 million, and is estimated based on the net present value of cash flows under the contracts, based on a gold spot price of $279 and market rates for Libor and gold lease rates.
What a difference a quarter makes. In its 1st Quarter Report 2002, Barrick made the following statement:
Our outstanding gold and silver sales commitments at March 31, 2002 had an unrealized mark-to-market loss of $127 million (calculated at a spot price of $302 per ounce and $4.64 per ounce for gold and silver respectively, prevailing market interest rates and volatilities).
[Memo to file: Silver again. Hmmm.]
That’s some swing on a paltry $23 increase in the price of gold: Four Hundred Eighty-Three Big Ones, or about $21 million for each dollar of spot increase, for the arithmetically challenged. Uncharitable readers will already have noted that the spot price of gold is now around $310, $8 north of the baseline for that first quarter calculation. Randall Oliphant stated at a recent investor conference that Barrick’s mark-to-market sensitivity is about $18 million per dollar increase in the gold price. That suggests we’re looking at another $144 million in negative swing if gold holds $310 through the quarter, which would bring Hedgebook’s total mark-to-market loss to $271 million.
Following the recent unpleasantness at Enron, how long will the rating agencies grant Barrick a pass on an increasingly negative Hedgebook? Forever? Maybe. But if not, what will the counterparties say if that vaunted “A” becomes something less attractive? What consequences may then be set in motion? Gives us the willies just posing the question.
If a Counterparty Falls in the Forest, Does A Hedger Hear It? Maybe Mr. Sokalsky is right, and the spot deferreds are so airtight in favor of Hedgebook that the banks are on the hook for all the risk. Terrific. What happens if the banks, despairing of their desperate lot and staring in the face the grim specter of diminished bonuses, simply decide to exit the business? Like Credit Suisse First Boston did, for instance. Like Morgan is rumored to be doing, for another for instance. (Said rumors are rife out here on the lunatic fringe, at any rate.) Who ends up holding the paper? Will they be gentle? Perhaps the central banks, as the ultimate parties at risk, will step into the shoes of their former agents, and we will be treated to a real time demonstration of top-down disintermediation. Wonder how the central banks would react to a downgrade of their new direct counterparty, if such should come to pass.
Holding the Maginot Line We may soon get some answers to our questions. Last week Barrick announced two changes in Hedgebook’s policy which got our antennae twitching. These were both contained in a press release issued on May 8:
Source: Barrick Gold Corporation
Barrick to Simplify Premium Gold Sales Program
WILL FOCUS ON CORE SPOT DEFERRED CONTRACTS AS CALL POSITION REDUCED; ALL AMOUNTS IN UNITED STATES DOLLARS
TORONTO--(BUSINESS WIRE)--May 8, 2002--[…] Barrick Gold Corporation announced today at its Annual Meeting in Toronto that it is simplifying its Premium Gold Sales Program. The Company said it will not renew its gold call and variable price sales contracts, which should result in a 3-million ounce reduction in the position by the end of the year.
"A simple spot deferred program makes more sense in today's environment," said Jamie Sokalsky Senior Vice-President and Chief Financial Officer. "The overall Program will be simpler, smaller and better positioned to take greater advantage of rising gold prices. At the same time, it will continue to generate significant additional revenues and provide secure and predictable cash flows."
Overall, at the end of the first quarter, the Company had 18 million ounces of spot deferred contracts, representing 22 percent of reserves, and 6 million ounces of call and variable price sales contracts in the Premium Gold Sales Program.
Barrick is simplifying the Program in two ways. First, it will not renew call and variable price sales contracts, and expects this position to decline by at least 50 percent, or 3 million ounces this year. Secondly, the Company will no longer invest a portion of its spot deferred contracts in corporate bond funds, and will instead leave all proceeds invested with its average AA-rated bank counterparties.
These changes are further to Barrick's previously announced decision to sell 50 percent of its production at the spot price, for the first time in 14 years. In prior years, 100 percent of annual production was delivered against the Premium Gold Sales Program.
What is most interesting about this announcement is that the headline should have read: “Barrick Stands Firm on Hedging Policy Despite Heavy Pressure from Shareholders; Premiums to Decline.” There are three reasons why this is so.
First, note that the much ballyhooed sale of 50% of production in the spot market represents an extension of the spot deferred program, not a reduction. You don’t reduce a hedge position by needlessly prolonging the agony. With the addition of the Homestake production, it would seem that Barrick could reduce the hedge at almost twice the rate possible pre merger. Instead, it is delivering into the contracts at a rate which will be only 10% less than that of 2001 (61%) and 2000 (63%). The statement in the release that 100% of production was delivered into the program in prior years is inaccurate (see excerpt above from the 2001 annual report).
Second, elimination of new premium from the sale of calls will eliminate a big swallow of the extra juice on the notional price per ounce calculation, tending to make future such calculations more closely correlated with pure, low octane contango.
Third, it looks like we won't have the corporate bond portfolio to kick around anymore, as all the cash has been pulled right back into the banks. This too should take some of the sizzle out of the price per ounce calculation. But more intriguingly, this has a funny look to it. Who’s driving the bus? A wild guess -- the banks. Come home to Mama. Is this sort of like a margin call in a contract that doesn’t feature margin calls? Just asking.
So Barrick is going to tough it out, despite the prospect of reduced premium and the risk of an increasingly negative mark-to-market. Whose interests are served by such a theological adherence to the strategy that won the last war? We don’t know, but they sure aren’t ours.
Conclusion Call us kooky, but after our little romp through the Book of Barrick, we’re just a wee bit uncomfortable with the prospect of substituting Barrick paper for that of our beloved GFI. Too many issues for our simple taste. So here’s hoping it was all just a misunderstanding. If not, here’s hoping it’s a cash deal.
May 15, 2002
DISCLAIMER & CONFLICTS THE FOREGOING ESSAY IS FOR YOUR INFORMATION AND AMUSEMENT ONLY. NEITHER MR. LANDIS NOR GOLDEN SEXTANT ADVISORS LLC (“GOLDEN SEXTANT”) IS SOLICITING ANY ACTION BASED UPON IT, NOR IS EITHER MR. LANDIS OR GOLDEN SEXTANT SUGGESTING THAT IT REPRESENTS, UNDER ANY CIRCUMSTANCES, A RECOMMENDATION TO BUY OR SELL ANY SECURITY. THE CONTENT OF THE ESSAY IS DERIVED FROM INFORMATION AND SOURCES BELIEVED TO BE RELIABLE, BUT NEITHER MR. LANDIS NOR GOLDEN SEXTANT MAKES ANY REPRESENTATION THAT IT IS EITHER COMPLETE OR ERROR-FREE, AND IT SHOULD NOT BE RELIED ON AS SUCH. MR. LANDIS AND GOLDEN SEXTANT AND ENTITIES UNDER THEIR RESPECTIVE CONTROL MAY HAVE INVESTMENT POSITIONS, LONG OR SHORT, IN ANY SECURITIES MENTIONED, AND ANY SUCH POSITIONS MAY BE CHANGED AT ANY TIME FOR ANY REASON.