国际清算银行-能源转型及其宏观经济影响（英文版）.pdf

BIS Papers No 135 The energy transition and its macroeconomic effects by Alberto Americo, Jesse Johal and Christian Upper Monetary and Economic Department May 2023 JEL classification: Q30, Q40, Q50, O30. Keywords: Energy transition, transition risks, macroeconomics. 2 BIS Papers No 135 The views expressed are those of the authors and not necessarily the views of the BIS. This publication is available on the BIS website (www.bis.org). © Bank for International Settlements 2023. All rights reserved. Brief excerpts may be reproduced or translated provided the source is stated. ISSN 1682-7651 (online) ISBN 978-92-9259-660-6 (online) BIS Papers No 135 3 The energy transition and its macroeconomic effects 1 Alberto Americo, Jesse Johal and Christian Upper Abstract The energy transition will have profound and varying effects across the globe. We assess how clean technologies are evolving – mainly wind, solar and electric vehicles – and the challenges and opportunities the transition poses for fossil fuel and metals and minerals producers in the short and long term. We describe the likely macroeconomic consequences of the energy transition and identify the countries that are most positively and negatively exposed. A small number of fossil fuel-producing countries are likely to be severely hit. Meanwhile, a concentrated group of minerals producers should experience large net benefits. Fuel importers – that is, most of the world – should benefit to varying degrees. Keywords: Energy transition, transition risks, macroeconomics. JEL classification: Q30, Q40, Q50, O30. 1 The authors thank Alexandre Tombini, Benoit Mojon, Jon Frost and Rafael Guerra for comments, and Logan Casey for excellent research support. 4 BIS Papers No 135 1. Introduction The energy transition is well under way and accelerating. Wind and solar energy are the cheapest sources of electricity in most of the world, renewable energy accounts for nearly all global electrical capacity growth, electric vehicle (EV) adoption is rapidly increasing and auto manufacturers are moving towards fully electric fleets (International Energy Agency (IEA) (2023), International Renewable Energy Association (IRENA) (2021)). These trends will be reinforced as technology improves and the impacts of measures such as the US Inflation Reduction Act, the European REPowerEU plan and China’s latest Five-Year Plan for Renewable Energy Development begin to be felt. The increase in fossil energy prices after the Russian invasion of Ukraine has likewise added a further impulse to the transition. This paper seeks to provide a broad, qualitative overview of the energy transition by bringing together research from a variety of disciplines – economics, finance, energy systems and environmental science – to give a directional view of how the transition may evolve in future. That is, it aims both to complement quantitative macroeconomic scenarios, such as those outlined by the Network for Greening the Financial System (NGFS), and to give a more targeted, sector-specific description than is plausible in such large-scale macroeconomic models (NGFS (2022)). We hope this paper can also serve as a launch point for future research that more narrowly focuses on the specific aspects of the energy transition which researchers, policymakers and investors typically focus on (eg quantifying its effects on inflation, trade flows and financial flows). When exploring the macroeconomic implications of the energy transition, we find it convenient to focus on three types of country: (i) fossil fuel exporters, who will see their main source of export and fiscal revenues severely eroded, forcing a shift to a new growth model; (ii) fossil fuel importers, who will spend less on importing fuel due to the abundance and geographical dispersion of clean energy resources; and (iii) exporters of key metals and minerals, who are likely to benefit from a structurally higher demand for their products and possibly from a new metals and minerals supercycle. Of course, these groups are conceptual and actual countries may fall into more than one of these buckets. Moreover, countries could have some regions falling into one category and others falling into a different one. Still, we find this distinction is a good starting point for an exploration of the macroeconomic effects of the energy transition. The transition may follow a variety of paths in the near-term. Constrained fossil fuel investment could push up energy prices sharply for an extended period and bottlenecks in metals and minerals markets could slow or raise the costs of the transition. But the opposite scenario is equally conceivable; rapid adoption or faster than expected improvement of clean technologies could push energy costs and fossil fuel demand down faster than anticipated. Metals and minerals production constraints could create a new commodity supercycle for the handful of countries where accessible reserves and refining capacity are concentrated (eg Argentina, Australia, Bolivia, Chile and Peru for reserves and China for refining). The long-term picture may be as follows. Most of the world, particularly East and South Asia, should benefit from replacing expensive, polluting, imported fossil fuel with cheaper, cleaner, locally sourced energy. For major fossil fuel producers, especially those in the Middle East and North Africa, the economic benefits of clean energy will probably be overshadowed by the decline of existing energy sources. BIS Papers No 135 5 Producers of key metals and minerals (eg copper, lithium, rare earths) should see amplified benefits, but the value of these exports will be substantially smaller than for fossil fuels. Overall, economic activity should shift from fossil fuel producers and towards energy importers and metals/minerals producers. The next section surveys the evolution of and the outlook for clean technology and fossil fuel markets. Section 3 discusses how the transition may manifest itself macroeconomically (eg in growth, inflation, exchange rates and capital flows). Section 4 looks at how exposed different countries are to the transition. We conclude with a discussion on the policy implications of the energy transition. 2. Evolution and outlook for energy markets Renewables are cheap and usage is growing fast Price declines and the adoption of wind and solar energy have far surpassed expectations over the past decade (Graph 1). 2 Measures of learning rates – the rate at which technology costs decline for every doubling of capacity – cluster around 15% for wind energy and 25% for solar energy. 3 The advantages of renewables should thus only grow over the coming decades; applying the International Energy Agency’s (IEA) projections for electricity capacity in its three principal scenarios – the Stated Policies Scenario (SPS), Sustainable Development Scenario (SDS) and Net Zero Scenario – wind energy costs could fall by a further 17–29% by 2030 as compared with 2020, while solar costs could fall by another 40–60%. 2 At a 16% annual price decline, solar energy costs fell by substantially more than assumed in several thousand integrated assessment model (IAMs) projections (Way et al (2022)). The mean projection was for a 2.6% annual decline and the highest forecast was 6%. 3 From 2010 to 2020, learning rates were much faster; 32% for wind energy and 39% for solar energy (IRENA (2021)). Applying the IEA’s scenarios to these data implies that wind energy costs will fall by 35% to 65% by 2030 vis-à-vis 2020, while solar costs will fall by 59% to 79%. Clean energy is growing and improving far faster than expected Graph 1 Projected versus realised wind/solar generation Projected versus realised wind and solar cost declines TWh % Source: IEA. 2,000 1,500 1,000 500 0 202020152010 Actual values Electricity generated from wind and solar: 2010 IEA forecast 2015 IEA forecast 0 –5 –10 –15 –20 SolarWind IEA 2015 projections Actual rate, 2010–20 Annualised cost reduction: 6 BIS Papers No 135 Declines in realised prices have already made renewable energy the cheapest source of electricity in most of the world and thus the preferred choice for new capacity (Graph 2, left-hand panel). Renewables’ share of newly added capacity increased from just under 40% in 2010 to nearly 90% (IRENA (2022)). This share is expected to increase to 95% of new investment by mid-decade (IEA (2021a)). This rapid growth has pushed the share of renewables in global electricity generation from around 20% in 2010 to almost 30% in 2022 (Graph 2, right-hand panel). A lack of transmission capacity could, however, impede the adoption of wind and solar. Intermittency – because of a lack of local wind or sunshine that cannot be offset by sunny or windy conditions elsewhere – may become a more binding constraint. Yet if these issues are resolved, renewables could supply a large share of many countries’ electricity demand; wind and solar could theoretically provide 72% to 91% of major countries’ electricity needs (Tong et al (2021)). 4 Transmission and distribution also create a floor on how low retail electricity costs can go; such costs account for around a respective 13% and 31% of electricity costs in the United States (US Energy Information Administration (EIA) (2021)), Beyond transmission, other technologies (eg hydroelectricity, geothermal, nuclear, batteries, hydrogen) and strategies (eg demand reduction and shifting) will be needed to support variable sources of energy to ensure around-the-clock reliability. In any case, as wind and solar power are still far from hitting even their near-term saturation points, their rapid growth should increasingly eat away at fossil fuel demand in the power sector. 4 It should be noted that, as these figures do not reflect siting constraints, nor do they necessarily reflect the lowest-cost zero emissions grid, actual usage is likely to be lower. Nevertheless, these issues highlight that a lack of sun or wind in a given country are not the main constraints facing the adoption of wind and solar. Renewables’ cheapness and abundance is increasing their usage Graph 2 Cost of electricity by fuel source Share of electricity generation by fuel source USD/kWh Share of total electricity generation, % 1 Levelised cost of energy. Sources: IRENA; BP; EMBER; authors’ calculations. 0.4 0.3 0.2 0.1 0.0 202120192017201520132011 Solar photovoltaic cost range Fossil fuel power generation Global weighted-average LCOE 1 : Onshore wind Offshore wind 60 50 40 30 20 10 0 2022201720122007200219971992 Fossil fuels Market share by type of generation technology: Renewables Nuclear BIS Papers No 135 7 Fossil fuel power is facing a massive challenge from renewables While the use of coal and natural gas in electricity generation – still the two largest sources of electricity supply globally – each hit record levels in 2021 and 2022, the rapid advance of renewables should put huge pressure on them over the next decade and beyond. Growth in fossil fuel electricity generation is slowing while renewable generation continues to grow rapidly (Graph 3, left-hand panel). Investment in fossil production and generation capacity has likewise been in a steady decline (Graph 3, right-hand panel), implying that the existing stock of fossil fuel assets will overwhelmingly be replaced by cleaner sources. Tightening policy around fossil fuel use, strong policy support for clean technologies and the cost advantages of renewables over existing capacity mean that large downside risks to fossil fuel demand in the electricity sector cannot be discounted (IRENA (2021)). Renewables pose a more immediate threat to coal than to gas. Indeed, electricity accounts for around two thirds of global coal demand and demand challenges are already apparent. After decades of steady growth, global coal consumption has been flat since 2014. Coal’s share of global electricity production has fallen from its 2011 peak of 41% to 36% in 2022 and would probably have been even lower if not for the unexpectedly strong rebound in energy demand in the aftermath of the pandemic and the effects of the war in Ukraine on energy prices (Ember (2023)). Global coal capacity additions are also on a steady downward trend (Global Energy Monitor (2022)). And coal financing costs have increased sharply, pointing to growing stigmatisation and poor growth prospects (Zhou et al (2021)). Natural gas has more diversified range of uses, which should help to sustain demand over a longer period. For instance, electricity generation accounts for only 40% of natural gas demand, compared with 65% for coal. Moreover, gas is less polluting, its capital stock is newer and it has quicker start-up times. The latter factor Despite a post-pandemic rebound, the outlook for fossil fuel electricity is poor Graph 3 Electricity consumption growth by fuel type 1 Coal and gas supply and generation investment 2 TWh USD bn % 1 2022 data are on a trailing 12-month basis. Fossil fuels data are a five-year moving average. 2 AEs = OECD regional grouping and Bulgaria, Croatia, Cyprus, Malta and Romania; EMEs = All other countries not included in the advanced economies regional grouping, China included. Sources: IEA; BP; EMBER; Global Energy Monitor; authors’ calculations. 600 450 300 150 0 202220172012200720021997 RenewablesElectricity demand: Fossil fuels 320 280 240 200 160 120 25 20 15 10 5 0 20222021202020192018201720162015 AEs EMEs investment (lhs): Coal and gas AEs EMEs investment (rhs): Share of coal and gas to total energy 8 BIS Papers No 135 makes it suitable for meeting peaks in electricity demand. It is also the largest source of heating across the world (IEA (2022)). Meanwhile, viable alternatives (eg heat pumps) are only just being rolled out at scale or are still years from being commercially competitive (eg hydrogen). Natural gas also remains a major input in some hard-to-abate industrial processes (eg production of glass, fertiliser and other petrochemicals). However, high natural gas prices and aggressive moves by European countries to wean themselves off Russian supplies could accelerate the transition. The electrification of transport is accelerating The EV market, although less advanced than the market for renewable energy, is also growing rapidly (Graph 4, left-hand panel). The EV share in new sales has grown from around 1% in 2015 to more than 13% in 2022 as major hindrances to widespread adoption such as driving range and input costs have declined (Graph 4, centre and right-hand panels). And this rapid adoption is not limited to the conventional passenger market; two/three-wheel vehicles, bus, van and semi-truck markets are also seeing sharp growth. 5 These trends should continue as new manufacturers expand and legacy auto manufacturers shift to the mass production of EVs, which should in turn eventually eliminate the upfront cost competitiveness of internal combustion engine (IC