Central banks bought a net 44.8t of gold in Q4, down 68% y-o-y. This brought annual 2020 purchases to 272.9t, almost 60% lower than the multi-decade record 668t added in 2019. Although 2020 marked the 11th consecutive year of net purchasing by central banks, it was the lowest annual total for central bank purchasing since that trend began in 2010.
The 2020 annual figure masks a year of two halves, with demand dynamics changing as the year progressed. The first half of the year saw a continuation of the longstanding trend of net buying, with central banks accumulating 234.6t. This was despite Russia – the largest gold buyer since 2005 – suspending its gold programme at the end of March. But the overall pace of purchasing began to slow in early H2, coinciding with a sharp pick-up in sales volumes. This resulted in quarterly central bank demand turning negative in Q3 for the first time since Q4 2010, before swinging back into net purchases in Q4.
The number of central bank buyers continued to outweigh the number of sellers over the year, with buying seen from several emerging market countries that mostly have lower ratios of gold-to-total reserves.1 Turkey was once again the largest annual gold buyer, adding 134.5t to its official gold reserves in 2020.2 This brought total gold reserves to 547t (42% of total reserves). India, a regular purchaser since early 2018, added 38t of gold to its reserves during the year, within the 30-40t range it has established over the last three years. And despite its buying being limited to Q1, Russia remained the third largest buyer in 2020, growing gold reserves by 27.4t. The UAE (23.9t), Qatar (14.5t), and Cambodia (5t) also made relatively sizeable additions to their gold reserves.
The increase in sales witnessed in 2020 should not be overlooked. 7 central banks reduced their gold reserves over the year, with the bulk of the sales falling in H2.3 While it is well understood why central banks would want to add gold to their reserves, it is worth also focusing on what drove those larger sales.
Despite being the largest buyer on an annual basis, Turkey was also the largest seller of gold in H2; its gold reserves declined by a net 36.3t after two chunky sales in September and November. These were its largest monthly sales since it resumed regular buying in May 2017.4 But the sales did not signify a strategic shift in policy, rather they were reflective of local gold market dynamics. Higher levels of local gold bar and coin demand in the second half of the year led to increased trading between domestic commercial banks and the central bank, causing this reduction in reserves. This mechanism is part of a range of gold policy tools with which the central bank manages its gold market.
The COVID-19 pandemic was also a driver for some central bank sales. Sales were concentrated among a small number of central banks that buy gold from domestic production, such as Mongolia (12.5t) and Uzbekistan (11.8t). Gold’s performance during the year boosted reserve portfolios, providing central banks with additional firepower when it was needed. And some banks saw this as an opportune time to obtain liquidity to support their struggling economies.
These intermittent sales have resulted in a more complex picture of central bank demand at the end of 2020, having created a small interruption in the pattern of consistent buying since 2010. But we expect continued net buying from central banks in 2021, with purchases expected to remain at a moderate pace although below the record levels seen in previous years. The possibility of capital inflows into emerging markets, resulting in higher reserves, and the continued ultra-low interest rate environment, may lead to central banks adding gold for diversification purposes.