The Gold Lending Market


Diploma Thesis, 2010

76 Pages, Grade: 1,3


Excerpt

Table of Contents

List of Figures

List of Tables

Abbreviations

Symbols and Variables

1 Introduction

2 The Gold Lending Market
2.1 The bullion market
2.2 The main transactions types on the lending market
2.3 Central banks and the gold lending policy
2.4 The development of the gold leasing rates

3 Literature Review
3.1 Gold and the gold price
3.1.1 Central banks and the gold market
3.1.2 The gold price
3.1.3 Hedging
3.1.4 Inflation
3.1.5 Shocks and safe investment
3.2 The gold lending market and the lease rate
3.2.1 The commodity lease rate
3.2.2 Convenience yield and the gold lease rate

4 The Gold Lease Rate Model
4.1 Central bank supply
4.2 Gold producer demand
4.3 Bullion bank demand
4.4 Derivation of the model

5 Data

6 Model Analysis and Results
6.1 The gold leasing rate
6.1.1 Leasing rates before
6.1.2 Expected inflation and economic shocks
6.1.3 Counterparty and investor influence
6.2 Gold and the gold leasing rate
6.3 Speculative influence
6.4 The silver leasing rate

7 Conclusion

References

Appendix

List of Figures

Figure 1 : The Gold Carry Trade

Figure 2: The Gold Loan

Figure 3: The Gold Swap

Figure 4: Swiss National Bank: Relative Gold Lending 1997-2009

Figure 5: Banco de Portugal: Relative Gold Swap Engagement 1997-2009 11 Figure 6: Daily Twelve Months LIBOR, GOFO and the Gold Lease Rate (1990-2009)

Figure 7: Daily Gold Lease Rates and the Nominal Gold Price (1990-2009)14 Figure 8: Quarterly Aggregated Central Bank Gold Lending (Q4 1994-Q4 2009)

Figure 9: Quarterly Aggregated North American (1992-1999) and World Producer Hedging Positions (2001-2009)

Figure 10: Quarterly Speculative Pressure in the COMEX Gold Futures Market (1990-2009)

Figure 11: Quarterly ETF Gold Holdings (2002-2009)

List of Tables

Table 1: World Official Gold Holdings, December 2009

Table 2: Summary Statistics

Table 3: Regression Results based on First Differences

Table 4: Regression Results: Gold Lease Rate and Speculative Pressure (Weekly 1990-2009)

Table 5: Correlation Matrix among Explanatory Variables of the Basic Regression Model

Table 6: Correlation Matrix of the 12 Months Gold Lease Rate and Speculation Variables

Table 7: Granger Causality Test Results: Gold and the Gold Lease Rate...

Table 8: Silver Summary Statistics

Table 9: Silver Regression Results

Table 10: Granger Causality Test Results: Silver and the Silver Lease Rates

Table 11: Regression Results: Silver Lease Rate and Speculative Pressure (Weekly 1997-2009)

Abbreviations

illustration not visible in this excerpt

Symbols and Variables

illustration not visible in this excerpt

1 Introduction

Gold is unique among precious metals with respect to its characteristics. These attributes are durability, divisibility, portability, and storability.[1] Moreover, gold supply is finite and only a fraction of annual production is used for industrial usage. Thus, most of the gold extracted worldwide still exists above-ground. For centuries, gold was used as a store of value and the basis of monetary systems. As a consequence, countries respectively their central banks accumulated substantial gold reserves. With the end of the Bretton Woods agreement, the value of major currencies was no longer tied to the value of a fixed quantity of gold. Consequently, gold lost its importance as a reserve asset. Nevertheless, official institutions kept their large reserves. In 1971, central banks were holding 36,575 tonnes of gold compared to 30,117 tonnes in 2009.[2]

These circumstances set the stage for an active gold lending market that has evolved within the last 20 years. To date, academic literature has primarily focused on analytical and empirical analyses of the physical gold market and the gold price. This study will contribute to the existing gold market literature from a different perspective: the lending market.

Therefore this thesis will initially investigate how the gold lending market functions in general. This will include an analytical overview of the main transaction types, the market participants and their intentions, the gold lease rate, and the impact of central bank lending policy. Secondly, it is unclear as to what the determinants of demand and supply are for the leasing market. Linking theoretical and empirical findings from the gold market and the gold price, this thesis seeks to set up a gold lending market equilibrium. With that, the gold lease rate is derived and empirically tested. The resulting determinants of the gold lease rate are developed with quarterly data from 1990 to 2009.

Overall, the primary long-run determinant of the gold lease rate is central bank lending activity, defined by the amount of gold used in lending transactions. The more gold central banks supply to the lending market, the lower the resulting lease rate. Contrary to that, factors influencing the lending demand have no significant influence. Thus, the gold lending market seems to be mainly driven by central bank lending policy.

The rest of this thesis is organized as follows. Chapter 2 examines the qualitative evidence from the gold lending market. This includes major transaction types, market participants, the gold lease rate, and information about central bank lending policy. Section 3 summarizes the gold related academic literature. Of that, the first part primarily presents determinants of the gold market and the second part reviews commodity lending market fundamentals. Section 4 sets up a lending market equilibrium based on the gold market theory. A gold lease rate model is derived from this approach. After the description of the sample data in Section 5, the empirical analysis of the lease rate determinants is conducted and interpreted in Section 6. Finally, the study concludes by summarizing the basic results in Section 7.

2 The Gold Lending Market

2.1 The bullion market

London has historically served as the market for trading precious metals, primarily gold and silver. Nowadays, it is also the center for precious metal lending. Since 1987, the London Bullion Market Association (LBMA) serves as a trade organization that represents the London bullion market.[3] The London precious metal market is an Over-the-Counter (OTC) wholesale
market and the minimum traded amounts are 1,000 ounces of gold and 50,000 ounces of silver. Thus, only wholesale participants are engaged.[4]

Four different groups are active on the market. Producers of precious metals bear exposure to a hedgeable risk, the commodity price. One possibility to reduce the gold price exposure is hedging via forward transactions. As will be shown later in the theoretical section of this thesis, gold producers are basically concerned with the downside potential. If gold prices are low or declining, a higher level of hedging activity can be observed. A possible explanation might be the influence of gold clientele investors, who hold these investments to benefit from the gold price exposure. As holders of large inventories of gold, central banks try to earn a return on their gold portfolio and therefore lend gold to the market.

This fact will be further confirmed by representative evidence from the German and Swiss central banks. Speculative investors and financial institutions (bullion banks) are, on the one hand, engaged in the market­making process as intermediaries between gold producers and lending institutions. On the other hand, they attempt to benefit from the price developments by speculating on their own account. A minor group on the lending market consists of fabricators who manage their inventories of precious metals.[5]

Gold loan transactions are mainly conducted by central banks that provide the needed liquidity for hedging purposes. In contrast to gold, central banks do not own sizeable amounts of silver. Thus, the silver lease market needs to be supplied from all the other market participants in the forward swap market.

2.2 The main transactions types on the lending market

Major central banks still hold large gold inventories. Due to the reduced need as a reserve asset, these inventories have no primary function. This leads to

the question, as to whether central banks should keep their inventories. The European Central Bank (ECB) points out four main arguments: First, gold shows economic security since the valuation cannot be influenced by the owner and gold has proven to keep its real purchasing power. Second, gold is always usable as a means of payment or collateral. Third, gold can increase confidence in a currency, even if reserves are no longer backed with precious metals. Four, gold reduces portfolio risk through diversification.[6]

However, the demand to borrow gold for different purposes does exist. Thus, there is the possibility to earn returns on these gold portfolios and to employ the reserves on the lending market. Therefore, central banks have a variety of options. They can either engage in a gold lease transaction via the gold carry trade or the gold loan, or they can enter into a gold swap.

illustration not visible in this excerpt

Figure 1 : The Gold Carry Trade

First, the classical gold lease transaction is the gold carry trade shown in Figure 1. Central banks lease their gold in order to generate returns to a bullion bank. Consequently, the investor is then long in gold and has the liability to repay the gold, plus the accumulated leasing fees, in the future. Instead of holding the gold price risk, the investor hedges his own position by
selling the gold in the spot market. The proceeds of this transaction can be invested in a risk-free asset at, for instance, the London Interbank Offered Rate (LIBOR). Accordingly, the bullion bank is able to earn the difference between LIBOR and the gold lease rate (GLR). This is the Gold Forward Offered (GOFO) rate, which is the compensation at which gold is lent on swap against U.S. dollar. The derived GLR is the difference between LIBOR and GOFO.

GLR = LIBOR - GOFO (1 )

At maturity, the bullion bank buys gold on the spot market and repays the borrowed gold including lease payments. Such a transaction is only rational for the bullion bank if the gold price is decreasing or at least stable.

illustration not visible in this excerpt

Figure 2: The Gold Loan

Second, the gold loan is shown in Figure 2. Gold producers are faced with gold price risk, since they have a long position in the commodity. As seen in the second half of the 1990s, gold prices can be subjected to considerable declines. Until July of 1999, the price per troy ounce fell steadily from over 400 to 252.85 U.S. dollar.[7] During the 1990s, gold producers suffered from declining profits and anticipated a potentially declining future gold price. That is the reason why they started to sell their gold forward to lock in a certain level of income and to hedge their gold price exposures. Such forward contracts are offered by bullion banks. Since such banks also have a need for hedging their exposures, they lend gold from the inventories of central banks and engage in gold carry trades. At maturity, when the gold has been produced, the commodity is delivered via bullion banks to the central banks and the positions are closed. Here, the gold price development has no impact, because producers have the duty to deliver the gold.

illustration not visible in this excerpt

Figure 3: The Gold Swap

Third, central banks can enter into a gold swap (Figure 3). This is similar to a sale and repurchase agreement in exchange for liquidity. Central banks sell their gold to financial institutions under the condition of ultimately repurchasing it at a predetermined point in time. In order to receive cash, central banks pay a fee to financial contractors. This is the GOFO rate that can be seen as a fee for a loan collateralized with gold. Central banks are then able to invest the proceeds theoretically at the risk-free LIBOR rate and thus, according to Equation (1), the resulting return is the gold lease rate as is the case in the gold carry trade. The bullion bank earns the secured GOFO rate and has the option to sell and repurchase the gold in the spot market as it is done in the other transaction types.

2.3 Central banks and the gold lending policy

Central banks hold a total of 30,117 tonnes of gold[8] and of that, the largest inventories are owned by the U.S. Federal Reserve System, the German Bundesbank, and the International Monetary Fund (IMF). Table 1 summarizes the reserves of the ten largest gold holding institutions.

illustration not visible in this excerpt

Table 1: World Official Gold Holdings, December 2009

Source: The World Gold Council.

While central banks still have large gold reserves, in the 1990s there was uncertainty in the market about whether further gold sales will accelerate the decline of gold prices. In fear of adverse impacts on the gold market, 15 European central banks set up the first Central Bank Gold Agreement (CBAG) to limit their gold sales and cap the lending activity. The agreement in 1999 limited combined sales to 2,000 tonnes within 5 years. The second agreement in 2004 capped five year sales at 2,500 tonnes and the actual 2009 treaty has a ceiling of 2,000 tonnes up to 2014.

The five largest gold-owning institutions were requested to respond on their gold lending policy in general and the volume respectively maturity of their gold loan portfolio. Out of this group, with the potentially highest influence on the gold lending market, only the German Bundesbank responded to an inquiry.[9]

The Bundesbank confirmed its engagement in the gold lending market and declared that the maximum amount of gold used for lending activities was below 10%. Its lending policy is based on an adequate risk/return ratio and the proceeds are supposed to cover the storage costs. Furthermore, the Bundesbank demands high standards from its counterparties in the gold lending market. Transactions are processed only with experienced banks that serve high credit rating requirements. These requirements are continuously monitored within the risk management process. Additionally, the institution validates that the CBAG capped the gold lending activity and stress that is not engaged in the silver market. With respect to the negative lease rates following the Lehman collapse, the response points out that gold lending activity is not an instrument of monetary policy in the Eurosystem. Consequently, gold loans were not used to supply bullion banks with liquidity, albeit the borrowing banks could have sold the gold easily on the market.

The IMF is, based on its Articles of Agreement, only able to sell gold in the market or to accept it as a loan repayment of a member country. All other possible gold transactions, including explicitly mentioned “loans, leases, swaps, or use of gold as collateral,” are outside its competence.[10]

There is only little public information available on the lending activity of the remaining major central banks, because the reporting is not regulated. The only institution that provides detailed public information on its activity in the gold lending market is the Swiss National Bank as the seventh largest reserve owner. Since an amendment to the Swiss National Bank Law in

illustration not visible in this excerpt

Figure 4: Swiss National Bank: Relative Gold Lending 1997-2009

Source: Swiss National Bank: Annual Reports (1997-2009).

The y-coordinate determines the relative gold lending activity of the Swiss National Bank. The quotient is calculated by dividing the published amount of lent gold in tonnes by the amount of gold that is hold by the central bank. The lending activity is further divided into uncollateralized (blue) and collateralized shares.

In November of 1997, the first 99.2 tonnes were lent to high-class domestic and foreign financial institutions with an average maturity of eight months. The resulting returns were 2.2% p.a. Risk was managed by credit restraints and the maturity structure.[11] In 1999, the central bank started to lend gold against securities as collateral. Initially, 23% of gold loans were secured. Furthermore, the CBAG from the same year restricted further gold sales and also set an upper limit for gold lending of 328 tonnes.[12] This level was kept until 2002 when lease rates started to deteriorate and the earned returns dropped to only 1.2% p.a. As a consequence, the Swiss National Bank reduced its unsecured lending portfolio and entered into secured transactions with maturities up to five years. This resulted in an increase of the average maturity of the portfolio to 12.3 months.[13] Between 1999 and 2003, the size of relative lending remained stable between 12% and 15% and its absolute lending volume was reduced to 232.9 tonnes in 2003. Meanwhile, until 2005, the central bank sold 1,300 tonnes of gold.

Apart from that, the amount of securitization increased steadily over time. In 2004, the absolute lending dropped to approximately 130 tonnes, a level that was kept until 2007. The relative lending still accounted at that time for 10% of the overall gold portfolio. Moreover, the majority of these transactions were collateralized and since 2006, no unsecured gold has been lent. The central bank’s annual report 2008 states that there is low credit risk involved, because the transactions are secured with investment grade bonds.[14] Within the second CBAG, the Swiss National Bank sold an additional 250 tonnes of gold and reduced its lending activity to 111.1 tonnes (10.7%) in 2008. At the end of 2009, the lending activity was further reduced to 92.7 tonnes (8.9%) and the annual return on the lending portfolio was 8.8 million Swiss Francs.[15]

The Banco de Portugal publishes an aggregated amount of their total engagement in gold swaps in its annual reports. These are transactions where gold is exchanged for cash and these operations are part of the bank’s repurchase agreements.[16] Portugal’s central bank already experienced losses in the lending market in 1990. When the investment bank Drexel Burnham Lambert collapsed at that time, the Banco de Portugal lost nine tonnes of gold.[17] Nevertheless, the central bank has continued to be active in
the gold lending market. Holding 382.5 tonnes at the end of 2009, Portugal bought 125 tonnes of gold in 1998 and then sold nearly 250 tonnes by 2006.

illustration not visible in this excerpt

Figure 5 presents an overview on the Portuguese relative swap engagements. From initially 25% of its reserves, the Banco de Portugal extended its swap activity until 1999 to levels around 60%. This level was sustained until 2002, when a continuous cutback of its activities started. In 2006, the relative swap activity accounted for only 2% of the country’s total gold holdings and since 2007, Portugal has not been active in this market.

illustration not visible in this excerpt

Figure 5: Banco de Portugal: Relative Gold Swap Engagement 1997­2009

Source: Banco de Portugal: Annual Reports (1997-2009).

Figure 5 summarizes the Portuguese activity in the gold swap market from 1997 to 2009. The x-coordinate measures time in years and the y-coordinate displays the ratio of annual published gold swap activity divided by the total Portuguese gold holdings, both measured in tonnes.

The Bank of England (BoE) has confirmed that the bank is actively managing its gold reserves of 310.3 tonnes in Q4 2009. To create additional returns, the BoE treasury department has defined a target return for the gold lending portfolio.[18] Furthermore, Young (2003) states that the BoE uses its

expertise and its proximity to the London market to offer custodial lending services for gold deposits of other central banks. Regarding its own activities, the BoE published monthly statistics on the U.S. dollar denominated market value of outstanding gold swapped or loaned. Data is available from July 1999 on and shows that the lending activity peaked at approximately one billion U.S. dollar in 2000. Thereafter, the value of the lending activity declined steadily, and since June 2008, the BoE is no longer engaged in those activities on their own account. Moreover, the BoE reduced its gold holdings to the 2009 level of approximately 310 tonnes by selling 395 tonnes of gold between July 1999 and March 2002.

2.4 The development of the gold leasing rates

According to the LBMA definition, GOFO is the rate at which gold is lent by bullion banks against U.S. dollar. Additionally, the LBMA provides LIBOR data for different maturities and calculates the derived lease rates as difference between LIBOR and GOFO. The GOFO mean is a daily published smoothed average of quotations from the market makers. Meanwhile, the resulting market lease rates are slightly smaller. The derived lease rates include neither the market spread of 20 basis points (bps) nor the required currency deposit factor of 12.5 bps. The published gold lease mid-rates are adjusted by 16 bps. This is the midpoint sum of the two variables and the resulting formula is:[19]

GLRmid_rate = LIBOR - (GOFO + 16bps) (2)

Figure 6 shows the relation between the twelve months LIBOR, GOFO and GLR. Historically, LIBOR exceeds the GOFO and thus, the resulting lease rate is positive. During 1990s, the gap was around 2% and later the difference was converging close to zero. In addition, the usually positive GOFO rate represents that the gold market is in contango. Only under rare circumstances, as observed in 1999, is the GOFO negative and the market turns into backwardation. Figure 7 includes daily gold lease rates with various maturities and the gold price from 1990 to 2009.

illustration not visible in this excerpt

LIBOR GOFO 12 Months GLR

Figure 6: Daily Twelve Months LIBOR, GOFO and the Gold Lease Rate (1990-2009)

Source: LBMA data.

The three variables are measured in percent p.a. and the resulting daily lease rate is the difference between LIBOR and GOFO.

[...]


[1] See Bordo (2003), p. 19.

[2] Official gold reserves are obtained from the World Gold Council based on IMF data and national sources.

[3] Market-making members include the following nine financial institutions: The Bank of Nova Scotia - ScotiaMocatta, Barclays Bank PLC, Deutsche Bank AG, Goldman Sachs International, HSBC Bank USA National Association, JP Morgan Chase Bank, Mitsui & Co Precious Metals Inc, Société Générale, and UBS AG.

[4] See LBMA (2008), pp. 1-5.

[5] See LBMA (2008), p. 13.

[6] These arguments were presented by Paul Mercier, Principal Adviser in Market Operation at the ECB, at the LBMA Conference 2009. See Mercier (2010) for further details.

[7] According to Bloomberg data, the lowest quotation was seen on July 20, 1999.

[8] Data is taken from the World Gold Council.

[9] The following information is provided by the Bundesbank, see questionnaire Bundesbank (2010) for the official responses.

[10] See IMF (2010).

[11] See Swiss National Bank (1998), pp. 41-42 and p. 75.

[12] See Swiss National Bank (2000), p. 50.

[13] See Swiss National Bank (2003), pp. 54-55.

[14] See Swiss National Bank (2009), p. 72.

[15] See Swiss National Bank (2010), p. 136 and p. 144.

[16] See annual reports of Banco de Portugal (2008), p. 348 and Banco de Portugal (2007), p. 322 for exact definitions.

[17] See O’Callaghan (1991), p. 34.

[18] See Young (2003).

[19] Calculations for silver lease rates are based on the same principles see LBMA (2008), p. 21.

Excerpt out of 76 pages

Details

Title
The Gold Lending Market
College
Humboldt-University of Berlin
Grade
1,3
Author
Year
2010
Pages
76
Catalog Number
V164552
ISBN (eBook)
9783640797875
ISBN (Book)
9783640797943
File size
976 KB
Language
English
Tags
Gold, Gold Lease Rate, Gold Price Determinants, Gold Lending Market, Bullion Banks, Central Banks, Gold Producer, Gold Price Model, Determinants, Carry Trade, Gold loan, Gold Swap, Lending Statistics, Hedging Statistics, Speculative Influence, ETF, Central Bank Gold Agreement, LBMA, GOFO, Convenience Yield, Inflation, Economic Shock, Safe Haven Investment, Gold Price, Gold Holdings, Inflation Hedge, Gold Market Model, Counterparty Risk
Quote paper
Diplom-Kaufmann Christian Bohm (Author), 2010, The Gold Lending Market, Munich, GRIN Verlag, https://www.grin.com/document/164552

Comments

  • No comments yet.
Read the ebook
Title: The Gold Lending Market



Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free