Authors: Zhao Wei, Jia Luxi, Li Xinyue (Zhao Wei is the chief economist of Guojin Securities and a director of the China Chief Economist Forum)
+
summary
Recently, gold prices have risen sharply, constantly updating record highs, and central bank gold purchases and demand from Asian retail investors may be important drivers. How to track the "real" central bank gold purchase data and assist in judging the trend of gold prices?
Hot Thoughts: Who are the "marginal pricers" of gold?
Q1: At present, the pricing framework of gold, the central bank and the private sector investment demand together become the "marginal pricing factor" of gold.
The supply of gold is almost rigid, and the high proportion and volatility of private investment and central bank demand are the "marginal pricing factors" of gold. Since 2002, the share of jewellery and industrial demand has declined significantly, replaced by more resilient private investment demand, which was once highly correlated with gold price movements. In the past two years, with the significant increase in central bank gold purchases, the two types of "highly volatile" demand from the central bank and the private sector have jointly become the "marginal pricing factors" for gold.
Under the simplified framework, the traditional center of the gold price is determined by the investment demand that takes into account the real interest rate, but in the past two years, the demand curve brought about by the central bank's gold purchases has shifted outward, resulting in the divergence between the "real interest rate" and the gold price. From 2010 to 2022, the central bank's annual gold purchases averaged 473 tons, which was relatively stable, and the investment demand for inflation, opportunity costs, etc., dominated the formation of gold prices, which was basically in line with the framework of real interest rates. Since 2022, central bank purchases have increased by an average of 587t per year, leading to a widening gap between gold prices and the pivot of real interest rates.
Q2: How to track the "central bank's" gold purchases? Pay attention to the gold exports of Switzerland and the United Kingdom, and the speed of reducing the holdings of US bonds in various countries
In the past two years, the average annual scale of central bank gold purchases has risen from 473 tons in the previous 10 years to 1,059 tons, but the monthly central bank declaration data is "relatively lagging behind". The World Gold Council's quarterly central bank gold purchase data is calculated through cross-corroboration of central bank declared values, confidential unrecorded buying and selling data, and import and export trade flows, or is relatively accurate. There have been several divergences in history, or due to "late filings" by some emerging market central banks. Since 2022, there has been another significant divergence between actual and reported purchases.
This means that we can't simply track countries' "reported purchases"; The "Imports and Exports" and "U.S. Bond Sales" data provide an auxiliary verification perspective. First, central bank gold purchases are usually bought in spot gold, and spot gold trading is concentrated in the London and Zurich markets, which can track exports from the United Kingdom and Switzerland. Second, from the scale of U.S. bond holdings, we can "peek" at the pace of "de-dollarization" of various countries, and then estimate the scale of gold holdings, Russia and Turkey are typical "cases".
Three questions: the possible interpretation of gold prices in the future? There may be twists and turns in the short term, and the central bank is still expected to refresh a new high in the medium term with the support of gold purchases
The recent surge in gold prices is likely to be driven by central bank purchases and demand from Asian retail investors. On the one hand, the dismantling of the gold price time-sharing data shows that since February 14, the gold price has mostly risen sharply in the European independent trading session, and the total gold exports of Switzerland and the United Kingdom in January have also reached a record high of 23.4 billion US dollars. On the other hand, from the perspective of gold ETF flows, the enthusiasm of European and American investors has decreased, but Asian investors are still accelerating their purchases, which is also reflected in the gold premium in the Chinese and Indian markets.
In the medium term, central bank gold purchases may still support gold prices to reach new highs. On the one hand, "de-dollarization" is not a global behavior that is easy to repeat, or the "decoupling" of some countries, given the current low proportion of gold reserves in such countries, as the price of US bonds recovers, the pace of central bank gold purchases is easy to rise and difficult to fall. On the other hand, a timely rate cut in 2024 remains the baseline assumption. Historically, there is a strong certainty that the decline in U.S. Treasury interest rates in the last quarter before the rate cut may support stronger investment demand.
Risk Warning
Geopolitical conflicts have eased significantly, overseas liquidity has tightened sharply, and central bank gold purchases have slowed significantly
+
The main body of the report
Hot Thoughts: Who are the "marginal pricers" of gold?
Recently, gold prices have risen sharply, constantly updating record highs, and central bank gold purchases and demand from Asian retail investors may be important drivers. How to track the "real" central bank gold purchase data to assist in judging the trend of gold prices?
Q1: At present, the pricing framework of gold, the central bank and the private sector investment demand together become the "marginal pricing factor" of gold.
After considering the stock of tradable gold, the supply of gold is almost rigid, and demand plays a decisive role in the formation of prices. Unlike other consumable commodities, gold coins, bars, ETFs and other stocks of gold can be directly circulated in the market. This makes gold pricing more akin to equity products, with new supply likened to an "IPO", with the proportion of incremental gold supply falling further from 12.2% to 9.4% since 2010. The size of the IPO was "stable and limited, making the price of gold more demand-driven." Its gold demand comes from four sources: jewellery demand, industrial demand, investment and central bank demand. By the end of 2023, the relatively rigid industrial demand for gold and the demand for jewelry accounted for 7% and 49%, while private investment and central bank reserve demand accounted for 45%, and there was a close linkage between private investment and central bank purchases and gold prices.
The source of demand for high proportion and high volatility is the "marginal pricer" of gold, which was spot jewelers in the 90s of the 20th century, institutional investors in the first 20 years of the 21st century, and further joined by the central bank in the past 2 years. Prior to 2002, jewellery demand dominated the price of gold, which was mainly determined by spot trading on the London Gold Exchange and the Zurich Gold Treasury, with major participants such as large bullion banks. Since 2002, the share of jewellery and industrial demand has declined markedly, replaced by more resilient private sector investment demand, which was once highly correlated with gold prices. In the past two years, with the significant increase in the scale of central bank gold purchases, the two types of "highly volatile" demand from the central bank and the private sector have jointly become the "marginal pricing factors" of gold.
Under the simplified framework, it can be considered that "investment demand" mainly considers real interest rates and other factors, forming the traditional center of gold prices, but in the past two years, the demand curve brought about by "central bank gold purchases" has shifted outward, resulting in the divergence between "real interest rates" and gold prices. In Gold: Alternative Thinking in the Big Picture, we construct a five-factor model that includes opportunity cost, inflation risk, risk aversion, trading behavior, and central bank gold purchases. If the five-factor model is "simplified", it can be understood that from 2010 to 2022, the central bank's annual gold purchases averaged 473 tons, which was relatively stable, and the investment demand for risk, inflation, opportunity cost and so on dominated the formation of gold prices, which basically conformed to the framework of real interest rates. Since 2022, the central bank's gold purchases have increased by an average of 587 tonnes per year, bringing about an outward shift in the demand curve, which in turn has led to a widening of the gap between the gold price and the real interest rate pivot.
Q2: How to track the "central bank's" gold purchases? Pay attention to the gold exports of Switzerland and the United Kingdom, and the speed of reducing the holdings of US bonds in various countries
In the past two years, according to the data tracked by the World Gold Council, the average annual scale of central bank gold purchases has risen from 473 tons in the previous 10 years to 1,059 tons, but the monthly central bank declaration data seems to be "relatively lagging behind". According to the World Gold Council, quarterly central bank purchases are calculated by cross-corroborating central bank purchases with the IMF, confidential unrecorded buying and selling data, and import and export trade flows, or a relatively accurate "actual purchase". Historically, there have been significant divergences between actual and reported purchases in the first quarter of 2008, the second and fourth quarters of 2009, and the second quarter of 2015, corresponding to several large monthly declarations in Saudi Arabia, India, and China. From 2022 to 2023, the actual gold purchases of the central bank will be 1,082 tons and 1,037 tons respectively, but the total scale of gold purchases declared by the central bank to the IMF is only 395 tons and 325 tons, and the actual gold purchases and the reported gold purchases will deviate sharply again.
This means that we can't simply track countries' "reported purchases"; The "Imports and Exports" and "U.S. Bond Sales" data provide an auxiliary verification perspective. First, central bank gold purchases are usually bought in spot gold, and spot gold trading is concentrated in the London and Zurich markets, which can track exports from the United Kingdom and Switzerland. Since 1995, the UK and Switzerland have been importing gold from gold miners and then selling gold to major bullion banks and central banks, and in 2023, the UK and Switzerland accounted for 18% and 30% of global gold exports, respectively. Compared with 2018-2021, the average annual imports of Turkey, China, and Saudi Arabia from Switzerland and the United Kingdom in the past two years increased by 167 tons, 442 tons, and 72 tons, respectively, but the increase in central bank declaration and consumption was only 62 tons, 150 tons, and 8 tons.
Second, from the scale of U.S. bond holdings, we may be able to "peek" at the pace of "de-dollarization" of various countries, and then estimate the scale of gold holdings. According to a survey by the World Gold Council, "fear of sanctions" and "hedging geopolitical risks" have become important reasons for some central banks to buy gold. Since 2010, Russia, China, Turkey, and Egypt have reduced their holdings of U.S. bonds by $141.8 billion, $97.1 billion, $25 billion, and $18.3 billion, respectively. Historically, the "de-dollarization" process of Turkey and Russia has been accompanied by a rapid increase in the proportion of gold in foreign exchange reserves, with Turkey rising from 28% at the end of 2017 to 57% in January 2024, and Russia also rising from 10% at the end of 2012 to 24% in August 2020.
Three questions: the possible interpretation of gold prices in the future? There may be twists and turns in the short term, and the central bank is still expected to refresh a new high in the medium term with the support of gold purchases
The recent surge in gold prices is likely to be driven by central bank purchases and demand from Asian retail investors. On the one hand, the dismantling of the gold price time-sharing data shows that since February 14, the gold price has mostly risen sharply in the European independent trading session, while the total gold exports of Switzerland and the United Kingdom in January have also reached a record high of 23.4 billion US dollars. This means that investors such as central banks may buy gold significantly in spot markets such as Zurich, which supports gold prices. On the other hand, in terms of gold-backed ETF flows, Asian investors are still accelerating, although the enthusiasm of European and American investors has waned. Asia Gold ETFs are small, and individual investors may buy gold from the local spot market, which is reflected in the gold premiums in China and India: as of April 5, the gold premiums in the local markets of China and India increased by US$34.1 and US$29.6 per ounce respectively from March 8.
In the short term, investor sentiment has become extreme recently, and the risk of a rebound in U.S. Treasury yields may cause gold prices to face "phased twists and turns". Compared to our five-factor model, the current gold price has reached a record high of $288 per ounce. Of course, there are some of the effects of the "central bank purchases" that have been released behind the aforementioned period, and the "spot purchases" of Asian investors, which cannot be accurately observed. However, on the one hand, the historical quantiles of non-commercial longs and net longs in the futures market have reached 89% and 87%, the COMEX gold RSI is as high as 80.1, and the implied volatility of 3-month at-the-money options has also continued to rise to 15.3%, all indicating that market sentiment may be relatively "excited". On the other hand, the stickiness of U.S. inflation has gradually emerged, which may lead to a further revision of the Fed's interest rate cut expectations in the short term, or will exacerbate the "volatility" of gold prices to a certain extent.
In the medium-term analysis of the return framework, the follow-up gold price trend is to see whether the "pricing gap" brought about by the central bank's gold purchases will widen or narrow, and secondly, to look at the interpretation of investment demand under the traditional framework, both of which may have some support for gold prices. On the one hand, from the perspective of SWIFT usage, "de-dollarization" is not a global behavior that is easy to repeat, or the "decoupling" of some countries. On the other hand, although there may be setbacks in the short term, a timely rate cut in 2024 is still the Fed's baseline assumption. Historically, in the transition from the rate hike cycle to the interest rate cut cycle, the yield of the US Treasury at the time of the first rate cut was lower than that of the last rate hike, and the decline in the US Treasury yield in the last quarter before the rate cut was highly certain, or supported the strength of investment demand.
After research, we found that:
1. The supply of gold is almost rigid, and the high proportion and high volatility of private investment and central bank demand are the "marginal pricers" of gold. Under the simplified framework, the traditional center of the gold price is determined by the investment demand that takes into account the real interest rate, but in the past two years, the demand curve brought about by the central bank's gold purchases has shifted outward, resulting in the divergence between the "real interest rate" and the gold price. From 2010 to 2022, the central bank's annual gold purchases averaged 473 tons, which was relatively stable, and the investment demand for inflation, opportunity costs, etc., dominated the formation of gold prices, which was basically in line with the framework of real interest rates. Since 2022, central bank purchases have increased by an average of 587t per year, leading to a widening gap between gold prices and the pivot of real interest rates.
2. In the past two years, the average annual scale of the central bank's gold purchases has risen from 473 tons in the previous 10 years to 1,059 tons, but the monthly frequency of the central bank's declaration data is "relatively lagging behind". This means that we can't simply track countries' "reported purchases"; The "Imports and Exports" and "U.S. Bond Sales" data provide an auxiliary verification perspective. First, central bank gold purchases are usually bought in spot gold, and spot gold trading is concentrated in the London and Zurich markets, which can track exports from the United Kingdom and Switzerland. Second, from the scale of U.S. bond holdings, we can "peek" at the pace of "de-dollarization" of various countries, and then estimate the scale of gold holdings, Russia and Turkey are typical "cases".
3. The recent surge in gold prices, central bank gold purchases and the demand of Asian individual investors may be important driving factors. On the one hand, a breakdown of gold price data shows that since February 14, gold prices have mostly risen sharply in European independent trading sessions, while the total gold exports of Switzerland and the United Kingdom in January have also reached a record high of US$23.4 billion. On the other hand, from the perspective of gold ETF flows, the enthusiasm of European and American investors has decreased, but Asian investors are still accelerating their purchases, which is also reflected in the gold premium in the Chinese and Indian markets.
4. In the medium term, the central bank's gold purchases may still support gold prices to continue to hit new highs. On the one hand, "de-dollarization" is not a global behavior that is easy to repeat, or the "decoupling" of some countries, given the current low proportion of gold reserves in such countries, as the price of US bonds recovers, the pace of central bank gold purchases is easy to rise and difficult to fall. On the other hand, a timely rate cut in 2024 remains the baseline assumption. Historically, there is a strong certainty that the decline in U.S. Treasury interest rates in the last quarter before the rate cut may support stronger investment demand.
Risk Warning
1. The geopolitical conflict has been significantly eased: The Russia-Ukraine conflict has shown "fatigue", and under the pressure of Biden's campaign, he may be willing to promote the reconciliation of the Palestinian-Israeli conflict and the Russia-Ukraine conflict.
2. Sharp tightening of overseas liquidity: The stickiness of the U.S. recovery and inflation is difficult to falsify in the short term, and the expectation of interest rate cuts may be further compressed.
3. The central bank's gold purchase has slowed down significantly: Once the relationship between Russia, Turkey and other countries and the United States is reconciled, there may be a cooling or even reversal of "de-dollarization", and the speed of gold purchase may slow down significantly.