Removing the Cloak from Central Bank Gold Operations

Sale Of UK Gold Reserves

The second period began on 7 May 1999, when Gordon Brown, Chancellor of the Exchequer, announced the sale of more than half of the UK’s gold reserves in a series of auctions to be operated by the Bank of England (BoE). The price of gold was then$282.40. By the time of the first auction, the gold price had dropped by 10% and ultimately reached a low point of $252 on 20 July. This price reaction was actually not so surprising in a market which was dominated by quite substantial short-selling positions. The sale process used by the BoE certainly could not be accused of opacity, but the UK sales had a huge impact in different ways:

• The sales exacerbated the fear that some big gold holders, especially European central banks, would follow and the price of gold would collapse. Indeed, the Swiss National Bank (SNB) had started to modify its legal framework in order to be allowed to sell gold prior to Gordon Brown’s announcement. In June 1999, the SNB decided that half of its 2,590 tonnes of gold reserves were no longer required for monetary purposes. There also were discussions about the IMF selling part of its gold holdings. Furthermore, with the introduction of the euro, intense political debate was spurred as to whether national central banks had to keep on holding so much foreign reserves, both in currencies and in gold. The argument was that the ECB had already been endowed with reserves and the euro was replacing national currencies. I remember having written several memoranda to counter multiple requests from parliament and government members to sell reserves and return the proceeds to the government, which would have been a breach of the principle of non-monetary financing enshrined in most laws of central banks.
• In this context of uncertainty, the first Central Bank Gold Agreement (CBGA) was signed in Washington DC on 26 September 19991. The ECB, comprising 11 national central banks from the euro area, as well as the Swedish, Swiss and British central banks, signed the statement. Even if not part of the Agreement, the Fed and the Bank of Japan implicitly endorsed the aims.
• True, the Agreement took the market by surprise and financially hurt some participants – those who had sold gold short as well as the producers who had initiated quite major hedging programmes in the belief that the gold price could only go one way, meaning downwards.
• The reasons behind the CBGA are, in my view, two-fold. First, the UK sales evidenced that central banks were in a prisoner’s dilemma situation: they would be better off cooperating, not only in their own interests but also by allowing a more efficient functioning of the gold market. However, I remember that central banks’ opinions about the impact of their operations were far from consensual. In 2000, the European System of Central Banks set up a working group to write a report on the gold market. I chaired that panel. The debates we had were very lively: the question of whether gold lending by central banks a negative impact on the spot gold price had divided us. I was among those arguing that, in the context of the gold market at that time, gold lending by central banks did have a depressive impact on the gold price. This concern explains why the CBGA specifically stated that there was to be a block on increasing gold leasing and the usage of futures and options. The second reason behind the Agreement was the recognition of the unique position of central banks in the market as a major player and, more importantly, as a significant holder of gold. Being transparent as to their intentions would stabilise the market by reducing uncertainty. But in no way, as some intimated, was there an intention to manipulate the gold price. The initial CBGA, and the subsequent versions, fully achieved these objectives, in my opinion.

Banque De France Gold Sales

The third milestone in my gold market experience corresponds to the gold sales by the BdF, which I had the privilege to conduct as Director of Market Operations in the mid-2000s. The price of gold was then around $400 per ounce. This programme was determined in full compliance with the second CBGA signed in March 2004. It took the form of an agreement between the BdF and the French Treasury, which was signed on 19 November 2004. This agreement stated in particular that there would be no direct payment of the proceeds (not even the profits) to the French government. Only the return on the reinvested proceeds were to go to the Treasury via the dividend. The main reason behind the sales was to enhance the risk/ reward profile on the whole reserves portfolio.

From an operational perspective, the sales programme was a great opportunity to increase our knowledge of the gold market in all its facets:

• The sale strategy was decided at the highest level. In order to be fully accountable vis-à-vis the stakeholders, the decision of using the London Fixing as a benchmark was made – with the aim of beating the benchmark of course. That is why, gradually, more instruments were used such as options, which required setting up the whole operational framework to process these transactions (legal contracts, front office, risk management and back-office procedures).
• Shipping and refining: part of the BdF gold holdings was London Good Delivery (LGD), and hence easy to sell, but part of it was not. Thus, it was necessary to refine some non-LGD gold.
• Sometimes BdF found it useful to employ location swaps and thus to fully understand the characteristics of that market.

Beating The Benchmark

This was a very exciting period from which I draw two lessons. First, beating the benchmark is not easy; second, the diversification impact can be evaluated by the fact that gold represented 17% of the assets of the balance sheet in 2004, 11% in 2009 (and 8% in 2017, partly because of the resort to unconventional monetary policy operations).

The fourth and last period is more recent. I was no longer involved in the operational aspects of the gold market, but I provided technical assistance in reserve management on behalf of the World Bank Treasury and had contacts with many central banks. I would like to highlight two main observations:

• Central banks still do consider gold as a reserve asset which is useful to hold. I have seen a lot of central banks keeping a very small portion in gold, but they want to keep it even if it does not pay back a high return and contravenes the pure principles of an asset allocation exercise.
• Transparency remains vital for markets. Interest in central banks’ reserve operations in gold surges from time to time, especially when China and Russia are supposed to be active. The various rounds of CBGAs have been quite helpful in promoting an increasing transparency of central banks’ activity in the gold market and in dispelling rumours that central banks are somehow interested in manipulating its price. Furthermore, more and more central banks are subscribing to the IMF Special Data Dissemination Scheme template where they declare their gold holdings (PBOC is one of those).

In conclusion, while central banks may not disclose everything and must preserve some mystique, as evoked by Stefan Zweig, they should be as open and transparent as possible and should not seek to be shrouded in mystery just for the sake of it.