Bank of France (Herve Ferhani) - GOLD - A VIEW FROM THE OFFICIAL SECTOR 29 May 15:30
LBMA Conference, Istanbul
Ladies and Gentlemen, it is a great pleasure for me to have a chance to visit this beautiful country. I would especially like to thank the organisers of the conference for giving me the opportunity to present our view of the gold market in front of such a knowledgeable audience. Some of you might wonder what a Central Banker is doing in a session devoted to bullion banking. I must confess I myself had a similar reaction at the outset. Nevertheless I think there is a case for a Central Banker to address a bullion banking audience : bullion banks after all make a living out of meeting the supposedly complementary needs of Central Bankers, at least of those who lend Gold, and those of hedgers, who borrow it. As a cautious and rather conservative Central Banker, I thought it might be of interest to express our views, even or maybe mostly if some of you might find these somewhat controversial. Recent developments in the market have been unusual and noteworthy enough for all of us to devote some time to trying to assess the rationality behind the change experienced in market conditions.The market had already displayed some signs of nervousness in the immediate aftermath of the announcement of the Central Bank Gold Agreement in September 1999, but things returned fairly quickly to « normal », if one can use such a word for a market which has been putting in such a lacklustre performance for about two decades. As an introduction to this presentation. I would like to start with a few words on the types of changes we have witnessed before presenting some hypothesis as to what might have generated this change.
Major Changes in the market ? As you notice, I cautiously use a question mark, as it might be too early to tell whether these are long lasting changes or rather, as I am inclined to think, adjustments in response to a constantly evolving environment. The main features of recent developments are the following :
Lease rates volatility ;
A bearish trend (at least in US$ terms, but more on that later)
Some adverse consequences for market participants with the return of credit risk concerns
Market liquidity occasionnally drying up. The spikes in lease rates can mainly be attributed to the difficulty in refinancing borrowings as existing funding expired. : 1 month rates peaked at over 6% earlier this year and even 3 month rates shot up as high as more than 3,1/2%. Conditions have eased since then, but rates remain notably higher than before.
A review of changes in the market must begin with a few words about the Central Bank Gold Agreement. This has been much discussed already but I think it is a testimony of the official sector’s responsible attitude towards the market. It certainly represents a tangible mark of the long term commitment of the signatories to the gold market, irrespective of their individual policy. There is a wide diversity of attitudes regarding the gold market among the signatories of the CBGA. Some are active market participants, others take a more long term view. This diversity shows that this agreement should definitely not be considered as a sellers’ or lenders’ cartel. Each signatory has its own policy on the gold market. What they all have in common is an interest in making sure this market is as efficient as possible. This, in itself, is in the interest of all market participants, whatever their own priorities as sellers or buyers.
Analysing the recent bouts of lease rates volatility, we will try to derive some useful conclusions as far as the structure and functionning of the market are concerned. Reference will be made in particular to the fact that the sudden spikes of lease rates did not spill over into the spot market, where price action remained relatively quiet.
That will lead us to spend some time on price developments. We will try to see why those lease rate increases occured in a bearish market, although this should be qualified somewhat : in US$ terms we are faced with a bearish market, but not in other currencies.
The Central Bank Gold Agreement
The salient features of this agreement are its comprehensive reach, and the elements of transparency and predictability it brings to the market.
The agreement has comprehensive reach in terms of products and participants alike.
In terms of products : The agreement covers all signatories Central Banks activities in the gold market, i .e. selling, lending and derivatives. The aim is to smooth out the activities in all market segments, either at the level already reached (lending and derivatives) or at a level deemed appropriate to preserve orderly market conditions. (400 tons sold per year over the 5 years of the agreement).
In terms of participants : The 11 original EMU members as well as the ECB, the UK, Sweden and Switzerland have signed the agreement. Collectively, they hold 16000 tonnes, i.e. nearly 50% of the world’s official sector holdings. Some major holders such as the U.S. and Japan, without being party to the agreement, are unlikely to behave differently and other countries are considered by some market participants as having ruled out gold sales. Altogether, it has been argued that up to 85% of official gold holdings were more or less affected by the terms of the agreement.
In our view the CBGA has brought much needed transparency and predictability to the marketplace as far as Central Banks’behaviour is concerned.
Transparency : The gold market has traditionally been a rather opaque one where the official sector in particular never commented on its intentions or its policy. Central Banks in general would almost never issue a statement regarding their policy on gold, not until after the fact anyway . The market would find out ex post, if at all, about transactions. The official statement of September 1999 marks a significant departure from this tradition. Before that, the opacity of the market was such that anything was deemed possible. Not anymore.
Predictability : Maybe more than transparency, predictability of behaviour of the official sector brings benefits to all. From this standpoint, the agreement significantly increases market participants visibility.
The agreement has a 5 year duration ;
The yearly amount of sales is public ;
The agreement provides for a stable amount of lending.
But then, if everything is so ideal, why have we experienced such volatility lately ?.
Lease rates volatility was so high indeed recently that it prompted some to question the liquidity of the market. To quote one market participant, « such a squeeze defies logic ». Our view is that what we experienced was some form of adjustment to a constantly changing environment which might indeed be far reaching. Among the major changes that the market had to adjust to, the restructuring of the financial industry stands out as an important factor.
The concentration of market makers has lead to significantly fewer top tier institutions.
To some extent, this has resulted in what could be viewed as market fragmentation with different classes of intermediaries and borrowers. Marginal players can only find funding at a price that fully factors in credit risk considerations, and this might have been one of the contributing factors behind the skyrocketing lease rates lately.
Nevertheless, liquidity is still there : The amounts available for lending have somehow been capped by the September agreement, but definitely not cut. My understanding is that lenders have little or no incentive to restrict their activity in an environment where lending cannot in any case be increased.
Still, the illusion of unlimited liquidity has dissipated. The cap on lending does impact activity : the market probably cannot accomodate the double digit progression of lending of years past. Recent estimates expect a single digit progression in 2000, this activity obviously emanating from non signatories. This is a marked departure from what the market experienced over the last decade : indeed, over that period, official lending grew more than fivefold, starting at 900 tonnes in 1990 to reach 4830 tonnes at the end of last year. In parallel , the overall net short position of the market is estimated to have increased correspondingly, so by 400 tonnes yearly.
Yield curve twists : There has been no parallel shift of the yield curve. The front end came under severe upward pressure at a time where long term rates experienced an opposite move. Investors apparently moved further out on the curve in pursuit of yield. If more long term lending occured, then less short term liquidity became available : it’s a finite world out there ! This is yet another exemple of adjustment to a changing environment. Some hinted that certain lenders increased tenor and reduced amounts. Increased tenor implies higher credit risk concerns, hence a concentration of lending among higher rated institutions and less access to liquidity for second tier institutions.
What was somewhat unusual though was that lease rate volatility coexisted with rather sluggish price action on the spot market.
Price Behavior was in a way remarkable precisely because so little happened on the spot side as lease rates gyrated wildly, somewhat differently from what happened right after the CBGA was published.
A number of factors are at work which could explain that despite the squeeze on the lease market, the spot market remains lacklustre. It has become almost trivial to say that the gold market has been bearish for an extended period of time. Let us have a detailed look at some of the factors behind the price action, focusing on major classes of market participants, namely producers, the official sector and private investors.
On the producer side, it is not self-evident that current market conditions should be viewed as negative. The conjunction of technological progress and exchange rate fluctuations has allowed production to be profitable at previously unattractive US$ price levels.
Increased production has contributed to the flat gold price : production is estimated to have reached 2.000 to 2.500 tons a year for the past 10 years, adding about 20 % to total stock holdings (140.000 tons in 99). Technological progress has significantly boosted production capacity. This obviously contributes to the changing economics of the gold market. Similarly, 80 % of total gold held today was mined during the 20th century and 1/3 of above ground stocks has been produced in the last 30 years.
In addition, exchange rate fluctuations have a tremendous impact on the cost structure and profitability of production facilities. Gold mining is an industry with local currency denominated costs and US$ denominated revenues. From this standpoint, one can say that there is not just one gold price, but many. (Incidentally, this is true for consumers as well : gold was an attractive hedge for rupiah based investors in the aftermath of the Asian crisis.) Therefore, the idea of a bear market in gold has to be qualified. In local currencies, gold has in some cases behaved very differently from the picture we might have in mind.
Another factor explaining the lacklustre performance of gold in dollar terms obviously has to do with demand factors : One of the most important such factor in my view is that we live in a world where positive real interest rates prevail. This is reflected in the sluggish performance of most commodities (with obvious exceptions such as Palladium and Platinum).
Gold is a currency hedge as well : The (admittedly not perfect) negative correlation between gold and the US$ means that part of the demand for Gold has to do with its role as a currency hedge. Hedge against US$ weakness for dollar based investors, hedge against local currency weakness for others as Gold is quoted in US$…(see € based investors). True , this could be qualified somewhat since this source of demand works only if it is not impeded by other factors as stated in this year’s GFMS report. In particular, it has not proven true for yen based investors lately. I think it nevertheless remains an important determinant of demand. It can indeed turn into a determinant of supply : this is what happened this year as far as European investors. There has been a massive private sector disinvestment : 291 T in 2000 (3rd highest level in 10 years, 471 T swing from previous year). It is to a large extent explained by the hedging value of gold for European investors. Since January 1999, gold lost about 10% in $ terms but gained about 20% in € terms. 2000 was not a bad year for European investors to cash in some profits. This explains part of this private sector disinvestment.
So, the question is « how can we reconcile this sluggish price action and lease rates spikes ? ». I think this can only be done if we analyse the conditions under which the lease market has developed over the years and the way in which it traditionally reached equilibrium.
The relationship between Central Bankers and producers can be described to some extent as some kind of Competition/Cooperation. To some extent, this relationship takes a new dimension since the CBGA : up to then, Central Bankers (or some of them…) lent Gold, providing liquidity to hedgers. Case of cooperation with the benefit of forward selling for the miners and a modest yield for the Central Bankers. Essentially Central Bankers were providing cheap funding to the mining industry, 1% or so yearly. This allowed miners to hedge their exposure to gold prices, but also, by taking advantage of the contango, to profitably fund projects which would not have been otherwise considered economically viable.
From this standpoint, it is worth noting that the Gold market is a rare example of a market where the contango situation is almost permanent, ensuring that the forward price remains systematically above the spot price at maturity. (Charts of commodities and currencies). As a result of this, there is a systematic bias in the market to the benefit of short sellers and to the detriment of gold holders. On a long term basis and with only short lived exceptions, short positions have proved profitable. There are two ways to look at this situation. One is to focus on fundamentals and explain the bearish trend by general economic reasons such as low inflation, "commoditization" of Gold which would supposedly lose its monetary appeal etc… This in my view only covers part of the issue. In fact, looking at the various financial assets, it is clear that there are situations of protracted bear markets. But as a general rule, market equilibrium requires that depreciating assets carry high (or relatively high) interest rates. This is obvious for currencies, where the level of interest rates is one of the most potent tools at the disposal of the Central Bank in its task to defend a currency. This is true on securities markets as well, where the existence of a short position on a given security implies that this security becomes expensive to borrow. It has not been the case for Gold, where a persistent bearish trend has indeed coexisted with very low interest rates. The alleged reason for this peculiarity is the existence of above ground stocks, and more specifically Central Banks holdings. This again is only part of the explanation. The full story is that lenders have been ready to accept very low yields on their Gold claims despite the fact that Gold was experiencing a bear market. They accepted these low yields because a) they felt they had no other alternative besides not lending and b) Central Bankers tend to be if not price (or yield) insensitive, at least less sensitive than other market participants.
This situation may have changed since September 99. It might be a case of "No more Mr Nice Guy" : the signatories won't be providing fresh liquidity, but will be selling about 400 tonnes each year. That is 400 tonnes which won't be available on the lease market anymore. They will undoubtedly be replaced by Gold not lent up to then, since I don't think any of the CBGA's signatories was lending its entire holdings. There is therefore no reason to expect a scaleback of official Gold lent. But at the same time, one should definitely not expect the lease market to continue developing at the pace experienced in the last few years. From his standpoint, I think it is fair to say that the illusion of unlimited liquidity has been dispelled. In all cases, it does not seem unreasonable to assume that official Gold sold yearly was in most cases previously lent, since there is a kind of natural progression in the attitude of Central Banks vis à vis Gold, from "diehard holders" to lenders first and to sellers eventually. From this standpoint, one can assume that there must be some relationship-certainly not one to one, but some kind of a relationship nevertheless- between further sales and reduced lending.
So the question now is "How much Gold remains available for fresh lending?" I will rely here on the estimates of our good friends from GFMS as they appear on their "Gold Survey 2001" (Chap.5, pp51 et ss). According to these figures, gold already lent amounts to approximately 4800 tonnes, and of the 34800 tonnes of official holdings, only an equivalent amount to the one already lent (so roughly 4800 tonnes) would still be available. So, maybe Gold bears should not feel overly concerned by the impact of the European Agreement on the lease market after all? Another 4800 tonnes of liquidity available, that's 100% more compared to the current situation, so why worry?
I do think that some caution is mandated here. One can wonder whether fresh liquidity will be as forthcoming as it has been up to now. There is a markedly different attitude as to Gold depending on the main motive from holding it. There is a difference between what I would call Policy held Gold and Portfolio held Gold.
Why do inflation fighters (Central Banks) hold Gold (essentially an inflation hedge)?
Historically, Gold was the 0% asset counterparty of banknotes, a 0% liability. Gold holdings meant that paper money was as good as gold. Today, paper money is only as good as the credit extended to the economy. Indeed, the existence of a 0% liability (Banknotes) explains that Central Banks can afford to hold Gold, which yields nothing, and partly explains their yield/price insensitivity.
They want to be perceived as the anchor of the system (No paper money).
As an insurance policy against a major dysfunctionning of the Global Financial System. Somewhat irreverently, one could say that they hold Gold because they don't believe in themselves…or, rather, as Gold is held as a substitute to reserve currencies, because they don't believe in the reserve currency issuer capacity (or willingness) to fight inflation.
As a war Chest (if the worst comes to the worst).
These are mostly «policy oriented » reasons. These are essentially the reasons why large countries hold gold. Gold held for those reasons is generally covered by the CBGA, or unlikely to be managed differently from « CBGA gold » . I am alluding here to gold held by the U.S., Japan or the IMF. This makes up for the bulk of official gold, roughly 30000 tonnes. There remains the 4800 tonnes presumed to be still available for lending, namely the « portfolio held gold ».
These 4800t of extra liquidity may in our view be more price/yield sensitive than European/Japanese/US/IMF Gold. This liquidity might only come at a price. Should the Bear market persist, these gold holders might react more like other asset holders, i.e. only lend if the yield level compensates for the possible price depreciation. Somewhat differently from policy guided holders, they will have an alternative : bail out and sell! The fact that these holdings are not policy guided implies that the owners have more options.
We used to live in a world where the marginal lenders had the choice between lending or doing nothing. As a result, they tended to lend. Post September 99, the marginal lenders might not be policy oriented anymore, since policy oriented holders are either covered by the September agreement (the signatories) or unlikely to start lending (US/Japan/IMF). These marginal lenders now have the choice between lending and selling. Should this be indeed the case, the end result is that the lease market cannot rely on additional "price/yield insensitive" supply. Lease rates have to move up for the market to reach equilibrium. If this line of reasoning is indeed valid, and I would stress these last words since I offer it as a hypothesis, then it might explain the unusual combination of higher lease rates and lower spot rates which we experienced lately.
Ladies and Gentlemen, as every self respecting Central Banker, I would end with a few words of caution : The Gold market remains driven by the fundamentals and all factors described here are only relevant as far as the way in which the market adjusts to the fundamentals. None of them per se explains the major trend experienced by the market. At the end of the day, we are in a bear market because inflation is absent and real interest rates are positive.
Nevertheless, these factors do have significant consequences as far as the functionning of the market. In particular, if these hypotheses are verified and maginal liquidity now comes only at a price, then the gold market might indeed « normalize » somewhat and lease rates might experience more frequent adjustments in conjunction with market participants price expectations. If that indeed is the case, then maybe the days of the persistant and substantial contango are numbered ? |
