The carbon footprint of the assets in the Riksbank’s foreign exchange reserves

The carbon footprint of the assets in the Riksbank's foreign exchange reserves

To the report's start page
The carbon footprint of the assets in the Riksbank's foreign exchange reserves

Difficulties in calculating the carbon footprint of bonds issued by states and regions

Published: 12 April 2022

Emissions data are published with a time lag and measured in different ways

The Riksbank uses data on greenhouse gas emissions for countries and regions from the national inventory reports reported to UNFCCC. However, these data are calculated and published with a time lag. At the time of publication of this Commentary, the most recently available emissions data are from the year 2019.

Another challenge with emission data for countries and regions is that emissions can be measured in different ways and data are not always available. This Economic Commentary uses only direct greenhouse gas emissions[24] The direct greenhouse gases are carbon dioxide, methane, nitrous oxide, hydrofluorocarbons, perfluoro-carbons, sulphur hexafluoride and nitrogen trifluoride. The indirect greenhouse gases are nitrogen oxides, carbon monoxide, volatile organic substances except methane and sulphur dioxide. converted into tonnes of carbon dioxide equivalents excluding land use, land-use change and forestry[25] A large part of the greenhouse gas emissions caused by man stem from various forms of change in land use. The climate reporting sector "Land use, Land-use change and Forestry" (LULUCF) includes reporting of greenhouse gas emissions and natural sinks for the uptake of greenhouse gases as a result of various forms of land-use change initiated directly by man for purposes such as settlements, commercial activities and forestry. as a measure. Indirect greenhouse gases are excluded because data are not available for all countries and regions represented in the foreign exchange reserves. The reason why land use, land-use change and forestry are excluded is that it is difficult to compare this component between countries since it can be measured in different ways.[26] See Fyson, C. L., & Jeffery, M. L. (2019).

Some countries, such as Australia and Canada, also publish estimates of greenhouse gas emissions and GDP for their regions. If regional data are available, the carbon footprint of the bonds can be calculated more accurately than would have been the case if it had to be estimated based on the country's carbon intensity.

Adjustment for inflation and the effects of exchange rates on GDP

The portfolio-weighted carbon intensity measure recommended by TCFD does not include any adjustment of GDP for fluctuations in exchange rates and inflation. However, the Intergovernmental Panel on Climate Change (IPCC) suggests ways of converting currencies and adjusting for inflation.[27] See Krey et al (2014). De Nederlandsche Bank has carried out an in-depth study of various measures of carbon footprint for asset portfolios in the Dutch pension and insurance sector, where the benefits of adjusting for inflation and exchange rates are clearly illustrated.[28] See De Nederlandsche Bank (2021). The calculations of the carbon footprint of the foreign exchange reserves use data on the countries' GDP from the OECD. These data are adjusted for inflation and exchange rate fluctuations between the years in order to show, as far as possible, the 'real' development of the carbon footprint of the foreign exchange reserves.[29] GDP data from OECD National Accounts Statistics use the year 2015 as the base year for the adjustment of inflation and exchange rates for countries. The GDP deflator is used as a measure of inflation. For the regions, the GDP deflator of each country is used when the GDP of each region is adjusted for inflation. The adjustment of data is performed in two stages:

GDP for each year is inflated/deflated with the GDP deflator from X year’s price level to the desired base year price level.

GDP is converted from the holding’s currency in year X to US dollars using the base year’s exchange rate.

The challenge when it comes to correction for inflation is that there are several measures that can be used to adjust it, for instance the GDP deflator, the consumer price index (CPI) and the producer and import price index (PPI). The GDP deflator and CPI have the best coverage and therefore work best when one wishes to adjust for inflation. In this Commentary, the GDP deflator is used as a measure of inflation. The GDP deflator is a broad measure covering all types of goods and services produced in a country and is therefore a good measure for adjusting inflation in the economy as a whole. It is also a more comprehensive measure of inflation than the CPI, because it is not based on a fixed basket of goods. When consumption patterns change or new goods and services are introduced, they are automatically reflected in the GDP deflator for each year, but not in the CPI. The GDP deflator thus captures changes in a country’s consumption or investment patterns.

One challenge when one wants to correct for fluctuations in the exchange rate is that it is impossible in practice to find a year when the exchange rate for the currency to which it is converted is not strong/weak in relation to other currencies. Assume, for example, that the currency to which the portfolio’s holdings have been converted has appreciated/depreciated significantly for the particular year chosen as the base year. This would result in a higher or lower carbon intensity for the assets, simply due to the exchange rate. This problem would have been smaller if there had been a global standard or practice to consider when choosing a base year. However, the smaller the share of assets in a given currency in relation to the total portfolio value, the less impact the exchange rate will have on the portfolio’s carbon footprint.

In this Commentary, we choose to use GDP converted to US dollars instead of Swedish kronor for two reasons. Firstly, because the majority of the foreign exchange reserve assets are in US dollars, which means that the majority of the reserve assets will not suffer any exchange rate effect. Secondly, because we want to report the carbon footprint in a unit that makes international comparisons easier, and here the dollar is preferable. In cases where comparisons are made over several years, GDP data with the same exchange rate for each year are used to eliminate the exchange rate effect between years.

An alternative to adjusting exchange rates in this way is to use purchasing power parity (PPP), as PPP is often used to compare GDP between countries with different currencies. However, there is no major difference if calculating using purchasing power-adjusted GDP. This is because the difference in relative purchasing power between the countries in the foreign exchange reserve is not so great, given that all are industrialised OECD countries. Nevertheless, we have also made calculations based on purchasing power-adjusted GDP at a constant price level to test the reliability of the results. Our calculations show that the choice of method has little impact on the total carbon footprint of the foreign exchange reserves.

The reason for calculating the carbon footprint of asset portfolios varies, which makes it more difficult for a global standard or practice for adjusting inflation and exchange rates to emerge. One or more standards would make it easier to compare different portfolios either with each other or over time. See the Appendix for a closer description of the data and the calculation method used in this Commentary to calculate the Weighted Average Carbon Intensity.