David Marczis
Gold has several functions: from being widely used in jewelry for thousands of years, to being an important component in modern industry. In addition, it serves as a store of value and is thus an investment asset (Figure 1). According to current estimates, approximately 216,265 tonnes of gold have been mined over the course of history, and about two-thirds of that has been extracted since 1950. Unmined underground reserves are estimated at 54,770 tonnes.
Figure 1: Distribution of the Stock of Gold (As at End of 2024)
Source: Compiled from data by WGC
Gold has long been a central element of financial systems — first it was used as commodity money, then individual countries tied the value of their currencies to gold under the gold standard, and later many countries did so in the Bretton Wood system.
While its significance has decreased over recent decades, it still plays a prominent role, as seen in Figure 2, which shows the amount of gold held on central banks’ balance sheets.
Figure 2: The Largest Official Gold Reserves (Tonnes, As at End of 2024)
Source: Compiled from data by World Official Gold Holdings and International Financial Statistics
Indeed, after the U.S. dollar and the euro, gold is the third most important reserve element.
Experience shows that the co-movement of the prices of many investment assets increases as market uncertainty intensifies. However, gold is weakly correlated with other financial assets (especially in turbulent times), which supports the idea that gold can have a diversifying role in investment portfolios in the medium or long term (Figure 3).
Figure 3: Correlation of Gold With Other Assets
Note: The figure shows average monthly returns for the periods 31 December 2014 – 31 December 2024 and 31 December 2019 – 31 December 2024. Source: Compiled from data by WGC.
The flight-to-quality status of gold has been explored in several empirical studies. Some analyze the connection between the price of gold and geopolitical risks or the pressure on global supply chains, while others point out that gold may be useful as a hedge against unexpected inflation. (For a full list of references, please see the original paper on which this article is based, cited at the end).
Gold has special properties: it is an investment independent of the issuer — that is, it is a store of value with practically no credit risk — and, additionally, its price typically goes up in times of market stress. Considering stress events in the past decades (such as Lehman crisis, the European sovereign debt crisis, Brexit, Covid and the Russia-Ukraine war), these properties may outweigh other detrimental considerations (e.g., gold pays no interest, it has relatively high exchange rate volatility compared with bond investments, and it has potentially lower liquidity). This might explain the recently renewed interest of central banks — as shown by the blue bars in Figure 4.
Figure 4: Annual Demand for Gold and Its Price
Source: Compiled from data by WGC
Investors increasingly combine conventional risk-return goals with other, non-financial objectives, as reflected in the growing interest in sustainable and responsible investments (SRI).
Demand for gold is generally determined by strategic and financial considerations; also, central banks’ purchases are often driven by its favorable safe-haven asset/portfolio diversification properties. Considering many climate scenarios incorporate increased geopolitical tensions, inflation risks, disruptions to agricultural production, declines in GDP and potential threats to the financial system, gold might become relatively more attractive for investors.
In the context of climate change, analyses of financial assets usually apply the double materiality principle: on one hand, exploring the extent to which the operation of the issuer of a security contributes to climate change and how this is financed by investors (impact perspective); on the other hand, looking at how climate change may affect the issuers’ operations and the value of their securities (risk perspective).
Considering gold investments’ global scale and significant role in the reserve portfolios, it seems appropriate to try to apply this approach to gold as well.
Emissions From a Production Point of View
The gold production process has four main stages: mining, milling, smelting, and refining (Figure 5). Each of these stages produces greenhouse gas (GHG) emissions.
Figure 5: GHG Emissions of the Gold Industry’s Value Chain
Source: WGC
The emission profiles of mines are heterogeneous and are determined by several factors, but the separation of gold from ore (that is, milling) accounts for the largest part of the value chain's energy demand.
Looking across all parts of the process, the annual emission of the gold industry is approximately 120-130 million tCO2e. Around 80 percent of the production comes from listed mining companies using large-scale technology, with the rest coming from artisanal and small-scale mining.
Countries with the largest gold production — China, Russia and Australia — accounts for around one third of total output (see Figure 6) and polluting energy sources still play a significant role in their energy mix. However, unlike some hard-to-abate industries (such as steel, cement, and fertilizer production), decarbonization of the gold industry is relatively easy from a technological point of view, for example with the installation of renewable energy sources. Therefore, we expect emissions to decrease in the future as energy mix changes to lower emitting sources.
Figure 6: Production of the Top 30 Gold Mining Countries (2023)
Source: Compiled from data by the WGC
To understand the impact of gold on the climate, one can look at emissions intensity ratios, either expressed as a percentage of ore or of gold itself:
Estimates on GHG per tonne of gold vary widely — from 11,500–55,000 tCo2e/tAu — mainly due to differences in the energy mix of the regional electricity grids. Some studies have also analyzed the environmental impact of gold production through a life cycle assessment, while others have researched additional environmental impacts, such as the use of water and impacts of solid waste created when the gold is processed.
Emissions From an Investor Point of View
Investors usually assess conventional financial assets (such as equities and bonds) using carbon intensity ratios, calculated as the ratio of the annual GHG emissions arising from production to the corresponding annual revenue or the value added generated by the given sector or company. This indicator shows how efficiently the issuer generates value added, from an environmental perspective. Additionally, it is used as a proxy for transition risks, for example as the basis for estimating how a potential carbon tax may affect the operations of a given company or industry.
This logic can be applied to gold as well: the ratio of total emissions to the annual value added of the gold market is around 400–600 tCO2e/million euro value added. On this measure, the carbon intensity of the gold industry is relatively favorable relative to other industries, primarily due to the high value of gold.
A common way to measure climate impacts and risks at a portfolio level is the weighted average carbon intensity (WACI). This takes the carbon intensity (CO2 emissions per unit of revenue or value added) of individual companies within a portfolio and weights them according to the proportion of the portfolio invested in each company.
In the case of central banks’ reserves, gold is held alongside other financial assets, with a WACI ranging between 150-350 tCO2e/million EUR GDP. However, gold is different from equities and bonds, since the latter finance current and future operations and the related emissions of the issuer, so their WACI represents a continuous, annualized emission and value add. In contrast, with gold bars purchased, there is a one-off GHG impact during their production, but no significant recurring emissions while they are held. This raises the question of how the emission aspects of physical gold can be integrated into the current analytical framework used for bonds and shares.
Some studies suggest that instead of intensity ratios, the focus should be on total carbon emissions. For companies, the annual nominal GHG emissions (in tonnes) can be determined. Similarly, the amount of GHG emissions during gold production can be estimated. But to ensure comparability between different assets, the one-time emissions of gold can then be divided by the expected holding period of the investment. The greater the time horizon over which the single emission can be spread, the lower the annualized value of the intensity ratio. Consequently, the climate impact of gold is more favorable in a life cycle assessment if the investment period is long enough.
Transition Risks
When analyzing transition risks, regulatory changes introduced to promote the green transition will typically have a negative impact on companies engaged in carbon-intensive activities. For example, production is made more expensive, impacting profitability, making it more difficult to access funding and giving a competitive advantage to other companies with greener technologies. As a result, the value of the investor’s shares or bonds usually decreases.
However, in the case of gold, this interrelation is not obvious. There are no direct effects (e.g., carbon tax) for physical gold already held in the vaults, meaning that the primary impacts of the transition are probably marginal.
But indirect effects may also play a role. The value of existing gold investments is also determined by the price shaped by the prevailing market supply and demand conditions. To the extent that the safe-haven value of gold increases in turbulent times, gold prices might rise. Alternatively, investors might be pushed to prefer gold mined with more environmentally stringent rules, impacting negatively on the price of gold produced with high greenhouse gas emissions. Estimating these effects is, however, very challenging.
Physical Risks
Physical risks are often analyzed as supply-side shocks, for example, a production facility in a specific location is damaged due to extreme weather events. This logic could be applied to investments in shares of specific gold mining companies.
However, investment in physical gold (coins, bars) is practically free from such risks, as gold is practically indestructible, and it is not damaged even during (climate) disasters. This investment is therefore like taking a position in an entire industry, not a specific company, whose operations may be seriously impacted.
While climate change may negatively affect certain participants involved in gold production, climate-related shocks are not likely to affect the entire market. This is because geographical concentration of mining operations is rather low (Figure 5), with no region accounting for more than a fifth of the global production, which means relatively stable supply conditions.
Figure 5: Geographical Location of Gold Mines With the Largest Annual Production Volumes
Note: Mines with annual production exceeding 10 tonnes in at least one year between 2018-2021
Source: Based on data compiled by Schütte (2023)
An increasing number of central banks publish sustainability and climate risk analyses, focusing mostly on their investments in bonds and equities. Despite its significant role in central bank reserves globally, gold has rarely been discussed, even though climate change can have negative socio-economic consequences that may increase the significance of such safe-haven assets.
Investor demand for gold is mostly determined by strategic and financial considerations, but the sustainability and climate risks of gold should also be explored. Before making new investments, several ESG screening factors can be examined, including emissions.
However, this kind of analysis is not only relevant for central banks that intend to buy new gold, but also for those that hold legacy reserves, as the fair value of their investments is determined by the price of gold on the secondary market, which, could be also influenced by climate-related factors.
Disclaimer: The thoughts above contain the views of the author which are not necessarily the same as the official views of Central Bank of Hungary. A longer version of this article was published in the Financial and Economic Review.
Dávid Marczis is a senior member of the Heller Farkas College of Advanced Studies. He completed his studies in investment analysis and risk management at the Corvinus University of Budapest and at the University of Hull in England. After assuming various risk management roles, in 2022, he joined the Risk Management Department of the Central Bank of Hungary, where his tasks include, among others, the analysis of financial risks of the foreign exchange reserves, including climate risks.