57874 W O R L D B A N K W O R K I N G P A P E R N O . 2 0 3 A Financing Facility for Low-Carbon Development Christophede Gouvello Ivan Zelenko with Philippe Ambrosi THE WORLD BANK W O R L D B A N K W O R K I N G P A P E R N O . 2 0 3 A Financing Facility for Low Carbon Development Christophe de Gouvello Ivan Zelenko with Philippe Ambrosi Copyright © 2010 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First Printing: April 2010 Printed on recycled paper 1234 13 12 11 10 World Bank Working Papers are published to communicate the results of the Bank's work to the development community with the least possible delay. The manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formally edited texts. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank of the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission promptly to reproduce portions of the work. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, Tel: 978 750 8400, Fax: 978 750 4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, Fax: 202 522 2422, email: pubrights@worldbank.org. ISBN: 978 0 8213 8521 0 eISBN: 978 0 8213 8525 8 ISSN: 1726 5878 DOI: 10.1596/978 0 8213 8521 0 Library of Congress Cataloging in Publication Data has been requested. Contents Acknowledgments .................................................................................................................... v About the Authors....................................................................................................................vi Acronyms and Abbreviations .............................................................................................. vii Executive Summary............................................................................................................... viii 1. Introduction............................................................................................................................ 1 2. Meeting the Mitigation Challenge: The Decisive Participation of Developing Countries ............................................................................................................................. 3 3. Call for New Financing Mechanisms ................................................................................ 5 4. The Proposed Low Carbon Development Facility .......................................................... 9 LCDF Concept: Objective, Design, and Scale.................................................................. 9 Assessing the LCDF Impact............................................................................................. 10 Financial Viability ............................................................................................................. 12 Economic Efficiency.......................................................................................................... 16 5. Building on the Kyoto Instruments: Attracting and Monitoring a Large Portfolio of Low Carbon Development Projects .......................................................... 18 6. Coordination with Carbon Markets: Complementarities of the LCDF and CDM ................................................................................................................................... 22 7. Conclusion............................................................................................................................ 24 References................................................................................................................................. 25 Appendixes............................................................................................................................... 27 Appendix A. Mitigation Potential in Non Annex I Countries.................................... 28 Appendix B. Historical and Projected Evolution of the CDM Pipeline: Rapid Widening of Scope and Quick Growth of the Project Portfolio and Mitigation Potential .................................................................................................. 29 Appendix C. LCDF Financial Model Initial Capital Estimate and Financial Viability ...................................................................................................................... 34 Tables Table 3.1. Estimated Potential Demand for International Offsets by 2020 ........................ 7 Table 4.1. Profitability of the LCDF: Net Income Forecast 2010­22.................................. 14 Table A3.1. Profitability of the LCDF over a 12 Year Period, Net Income and Return on Equity.............................................................................................................. 39 Figures Figure 4.1. Evolution of the LCDF Loans for a 10 GtCO2e Annual Emission Reductions Target in 2030............................................................................................... 15 Figure 5.1. Historical and Projected Growth of the Number of Projects Submitted to the CDM by the End of 2012 ...................................................................................... 20 iii iv Contents Figure 5.2. Projected Mitigation Capacity of the CDM Portfolio by the End of 2012..... 20 Figure A2.1. Pace of Submitting New CDM Methodologies............................................. 29 Figure A2.2. Pace of Approving CDM Methodologies ...................................................... 30 Figure A2.3. Number of CDM Projects Submitted for Validation .................................... 31 Figure A2.4. Mitigation Capacity of Projects That Have Applied to the CDM (Validation Stage) ............................................................................................................ 31 Figure A2.5. Projected Number of Projects Submitted to the CDM by the End of 2012 .................................................................................................................................... 32 Figure A2.6. Projected Mitigation Capacity of the CDM Portfolio by the End of 2012 .................................................................................................................................... 33 Figure A3.1: Number of Loans in Default, Cumulative Distribution and 99.9996% Level .................................................................................................................................. 37 Boxes Box 1.1. Countries' Common but Differentiated Responsibilities ...................................... 2 Box 3.1. Major International Regimes for the Carbon Market............................................. 7 Acknowledgments T he authors would like to thank Michael Toman, Research Manager, Environment and Energy in the World Bank Economic Research Department; Kenneth G. Lay, Vice President and Treasurer for his support; Hennie van Greuning from the Treasury for helping us drive the publication process; Olivier Schmitt from the Quantitative, Risk and Analytics Department of the Treasury for reviewing the LCDF model; and all the experts from the World Bank who reviewed and gave comments on the paper. Of course, all errors are the responsibility of the authors. v About the Authors A ll authors are with the World Bank. Christophe de Gouvello is a Senior Energy Specialist in the Latin American and Caribbean Region, Ivan Zelenko is the Head of Structured Finance at the Treasury Department, and Philippe Ambrosi is an Economist in the Environment Department. vi Acronyms and Abbreviations CDM Clean Development Mechanism CER Certified Emission Reduction COP Conferences of Parties CTF Clean Technology Fund DOE Designated Operational Entities ERTC Emission Reduction Transformation Cost EU European Union EUA European Unit allowance EUETS European Union Emissions Trading Scheme FDI Foreign Direct Investment GEF Global Environment Facility GHG Greenhouse gas IDA International Development Association IPCC International Panel on Climate Change KP Kyoto Protocol LCDF Low Carbon Development Facility MRV Monitoring Reporting and Verification NAMA Nationally Appropriate Mitigation Activities ODA Official Development Aid UNFCCC United Nations Framework Convention for Climate Change vii Executive Summary T he reality of climate change requires a drastic reduction in global emissions of greenhouse gases (GHG) in the coming decades. In its 2007 report, the United Nations Framework Convention on Climate Change (UNFCCC) Secretariat predicted that emissions would reach 61.5 gigatons carbon dioxide equivalent (GtCO2e) in 2030 under a baseline (or "no mitigation" scenario), whereas a strict upper limit of 29 GtCO2e in 2030 is required to stabilize climate degradation. The UNFCCC subsequently established a challenging target for emission reductions of 32.5 GtCO2e. Under the baseline scenario, Annex I countries, which are the countries committed to reducing emissions under the 1997 Kyoto Protocol,1 would emit together 22.1 GtCO2e in 2030. Clearly, even by drastically cutting emissions, these countries cannot alone meet the UNFCCC target of 32.5 GtCO2e. The same basic arithmetic also demonstrates that the offsets based flexibility mechanisms of the Kyoto Protocol will not be sufficient to solve the problem. Developing countries must reduce emissions beyond the important but too limited role of offsetting a share of Annex I country emissions. Recent studies have shown that the cost effective2 emissions mitigation potential in developing countries ranges between 7.7 and 25 GtCO2e per year in 2030. Low carbon development leading to avoided emissions in developing countries near the upper end of this range likely would bring the global emissions target within reach. Moreover, these GHG mitigation opportunities would come from the implementation of low carbon technologies in investment projects that support the economic development of non Annex I countries. However, many low carbon investment projects do not materialize because they have restricted access to financing, even though the projects may offer low or negative GHG abatement costs. In fact, many projects validated under the Clean Development Mechanism (CDM) of the Kyoto Protocol cannot achieve financial closure, even though they are eligible for carbon finance. Carbon finance alone cannot support the full GHG emission abatement potential in non Annex I countries. Therefore, removing the investment financing barrier should be a priority, independent of the evolution of the carbon finance market. This paper proposes an innovative financing mechanism known as the Low Carbon Development Facility (LCDF). The LCDF would bring additional investment financing at concessional rates to unlock economically beneficial, low carbon development projects in non Annex I countries, thus enabling a rapid scaling up of project based emissions avoidance in these countries (potentially up to 10 GtCO2e in 2030). The LCDF could be a modality of the Copenhagen Green Climate Fund to implement the financial pledges made by Annex I countries as a result of Copenhagen and post COP153 negotiations to support projects, programs, policies and/or other activities in developing countries related to Nationally Appropriate Mitigation Actions (NAMAs). The LCDF would not replace the Global Environment Facility (GEF) and the Clean Technology Fund (CTF), but would rather support the scaling up of innovative projects pioneered by these instruments. viii Executive Summary ix The LCDF would function as a AAA lending facility initially endowed with capital by Annex I countries. At inception, the LCDF would receive $68 billion4 in capital from Annex I countries. This amount would grow to $80 billion during a 10 year build up phase, which would be enough to sustain the AAA rating over the long run while lending to low carbon development projects an annual $100 billion. The LCDF would manage in steady state a loan portfolio of $1 trillion with an average BBB rating. The LCDF would invest its capital in a portfolio of liquid and safe securities and would fund its loans by issuing bonds in capital markets. It would have a light cost structure (with operational costs representing 0.15 percent of the loan portfolio) relative to international financial entities. The LCDF would offer loans at the LIBOR rate plus 10 basis points. This rate is very competitive, comparable with the lending rates offered by multilateral development banks, and situated significantly below the borrowing conditions faced by developing countries. The financial viability of the LCDF would mean that its annual revenues would more than cover the default risk on its loans on average to yield a positive net income on average. In the build up phase, profits would be reinvested to reach the $80 billion in capital in year ten. Profits could then be distributed to finance sustainable development projects worldwide. LCDF capital would be enough to weather a worst case operating loss in 99.9996 percent of cases and thereby justify the AAA rating. The projects financed, with an average 21 year lifespan, would together generate estimated emission reductions on the order of 10 GtCO2e per year in 2030. The annual $100 billion of financing to developing countries would compare well to $613 billion foreign direct investment (FDI) to developing countries and $74.3 billion official development assistance (ODA) in 2007. As noted, the purpose of the LCDF is to significantly scale up low carbon development in non Annex I countries--development that is in their own economic interest but cannot be realized due to financial barriers. It is not a primary purpose of the LCDF to generate tradable Certified Emission Reductions (CERs).5 Only projects that remain financially unviable, after enjoying the concessional financing conditions offered by the LCDF, would be awarded CERs, thus promoting a transparent and sound interface with the carbon finance market. Moreover, the proportion of emissions reductions achieved by LCDF financed projects that could be converted into tradable CERs would be capped initially at 20 percent, so that their market price would not be driven so low by an overflow of CERs, which could in turn weaken national policies and actions in Annex I countries to reduce GHGs. Because CERs supplied to the market would only come from low carbon projects with indispensable needs for revenues from CER sales to become profitable, the risk of hot air jeopardizing the environmental integrity of global mitigation efforts would be reduced. To ensure environmental consistency with the existing UNFCCC framework and with the methodology and regulation assets built under the CDM, the LCDF would rely on an enhanced CDM monitoring, reporting, and verification process (MRV) to measure emission reductions. All projects financed by the LCDF would be registered under an enhanced CDM process and apply approved CDM monitoring methodologies, therefore ensuring the reliability of environmental performance in projects deemed eligible for financing under the LCDF. As an additional safeguard to environmental integrity, the concessional rate could be revised and increased if the x Executive Summary project does not comply with the CDM MRV requirements or does not maintain its environmental performance. The identification of projects and assessment of their financial returns would associate private banks and financial intermediaries in a public private partnership framework. As a result, additionality screening--one of the current weaknesses of the CDM--would be strongly improved by the financial due diligences performed during the LCDF financing process. Regarding the origination and the management of the large portfolio of low development projects of the LCDF, two capacities would be deployed: potential candidates for financing would be identified by (i) the CDM, as of a natural focal point for low carbon projects, and (ii) by the private sector banks, to which the LCDF offers a huge business opportunity in terms of originating loans. Regarding the capacity of the CDM, statistical and analytical reviews indicate that this mechanism already has the potential to support thousands of projects. While several limitations have constrained its performance (although these are expected to be relaxed in a renewed and enhanced CDM), the CDM portfolio has in fact grown rather steadily for several years, now with more than 5,000 projects in validation, registration, or implementation stages. Simple projections indicate that this number could increase to about 10,000 projects by 2012 and more than 40,000 projects by 2030, just by maintaining the current annual flow of new projects applying to CDM every year. Private and development banks are also expected to bring a sustained flow of projects. Private sector entities would lend to projects alongside the LCDF, whose conditional rates should considerably enlarge and strengthen their initial borrower base. We estimate the cofinancing ratio to be two thirds brought by the LCDF and one third by other lenders. This cofinancing would create a strong incentive for strict selection of projects from the point of view of their solvency. The cost of this new large scale financing mechanism would be borne by Annex I countries, consistent with both the KP "common but differentiated responsibility" dictum and the pledges of Annex I countries at Copenhagen to increase financing for low carbon development. The likely cost would be far lower than if such effort were supported through carbon finance only. Much if not most of the emissions avoided under the LCDF could be delivered at less than $1 per tCO2e, compared to the market price of CERs of $12 per tCO2e in 2010, expected by analysts to move toward $30­$40 per tCO2e going forward. The LCDF thus would enable Annex I countries to significantly leverage the resources they commit to the global mitigation effort. The estimates of potential abatement volumes given in the paper should be taken for what they are: estimates based on informed judgments of analysts. The actual potential could be more or less than the figures cited here. One key advantage of the LCDF is that it provides a low cost means to address this uncertainty, which has loomed over debates about GHG mitigation from the start. If access to finance for low cost, economically beneficial abatement projects is less than is hoped, the result will be a limited draw on the capital pledged to the LCDF by Annex I countries, and the mechanism can be reworked or abandoned after a trial period with limited financial cost. If the access barrier is even more important than anticipated, demand for LCDF resources will end up exceeding its financing capacity and the program can be enlarged. Executive Summary xi Notes 1 Annex I countries historically have produced the most GHG emissions. 2 Below 60 euros/tCO2e. 3 Fifteenth session of the Conference of the Parties (COP15). 4 All dollar amounts are U.S. unless otherwise noted. 5 Also called "carbon offsets." CHAPTER 1 Introduction T he Climate Change Conference of 2009 saw the heads of state and governments of the largest nations agree on the so called Copenhagen Accord, whose text begins with: "We underline that climate change is one of the greatest challenges of our time. We emphasize our strong political will to urgently combat climate change in accordance with the principle of common but differentiated responsibilities and respective capabilities." The reality of climate change associated with anthropogenic emissions is now widely acknowledged by the scientific community. Its potential devastating future harms are equally well perceived and, as stated in the Copenhagen Accord, major nations agree on the need to jointly and urgently combat climate change. The international community is also quite aware that stabilizing atmospheric concentrations of GHG at supportable levels will require a drastic reduction in GHG emissions within a limited period of time. Undertaking such an enormous effort triggers several interlinked challenges: (i) technically mitigating GHG emissions to the required level; (ii) implementing these solutions in countries where the required amount of emission reduction is most realistically and efficiently achievable, in particular through involving and using in full the large potential of developing countries; and (iii) mobilizing the large amount of financing needed to ensure that the corresponding projects and programs can be effectively implemented. Furthermore, these challenges must be simultaneously addressed in a way that is acceptable to all the parties involved. This means in particular that any arrangement designed to meet the global GHG emission reduction challenge must be consistent with the principle of the common but differentiated responsibilities of developed and developing countries (box 1.1). Daunting as they may be, these challenges also present an extraordinary development opportunity. While helping to meet the most urgent emission reduction challenge, the mobilization of vast amounts of financing may also trigger powerful new growth paths in developing countries based on low carbon and efficient energy. Emission reduction and development are closely intertwined. Arrangements to address one of these two major global issues should ideally address both and benefit from all potential synergies. Another crucial criterion by which to assess new, concrete arrangements is the need to accelerate the fight against climate change. Time is short, and if ambitious GHG concentration targets are to have any prospect of being achieved, new 1 2 World Bank Working Paper arrangements should be built, to the extent possible, on existing assets of international cooperation, in global finance as well as in carbon finance. To be more specific: New arrangements should use and leverage the capacity of multilateral banks, international public agencies, but also and quite importantly private sector banks. New arrangements should make use of the CDM, a powerful tool constructed over the past two decades for verifying and monitoring projects from an environmental angle, and a tool that can be enhanced in response to recent critiques to achieve higher productivity. New arrangements should however be more discriminating in what types of emission reductions require the extra financial benefit of carbon finance through the sale of CERs. Because of the significant uncertainties surrounding these issues, new arrangements should facilitate learning and subsequent adaptation of mechanisms to new information. This paper proposes an international financing instrument that meets the challenges and the conditions set out above. The Low Carbon Development Facility (LCDF), as its name suggests, is designed to foster development together with mitigation. The LCDF could be a modality of the Copenhagen Green Climate Fund to implement the financial pledges made by Annex I countries as a result of Copenhagen and post COP15 negotiations to support projects, programs, policies, and/or other activities in developing countries related to NAMAs. By neither limiting financing to incremental costs nor focusing on transformational projects, the LCDF will not replace the GEF and the CTF, but instead would support the scaling up of the innovative projects pioneered by these instruments. Box 1.1. Countries' Common but Differentiated Responsibilities Acknowledging the global nature of climate change, the preamble of the UNFCCC calls for "the widest possible cooperation by all countries and their participation in an effective and appropriate international response, in accordance with their common but differentiated responsibilities and respective capabilities." The notion of differentiated responsibilities recognizes that "the largest share of historical and current global emissions of greenhouse gases has originated in developed countries." As a result, the UNFCCC, in article 4.5 (p. 8), determines that "[t]he developed country Parties and other developed Parties included in Annex II shall take all practicable steps to promote, facilitate and finance, as appropriate, the transfer of, or access to, environmentally sound technologies and know-how to other Parties, particularly developing country Parties, to enable them to implement the provisions of the Convention." Source: United Nations, text of the UNFCCC (1992). CHAPTER 2 Meeting the Mitigation Challenge: The Decisive Participation of Developing Countries C limate change is a threat to the prosperity all nations and needs to be mitigated by rapid and effective joint actions to reduce emissions. In addition, global efforts to overcome poverty and advance development can no longer ignore the serious and immediate threat that climate change poses to development prospects. At the same time, efforts to mitigate climate change need to creatively leverage development opportunities for less developed countries. To date, the global mean temperature has risen about 0.8°C above preindustrial levels. Among the changes observed in the earth's climate are more frequent and severe extreme weather events, affecting disproportionately the poor and sometimes eroding development gains earned over decades. Drawing on a wide set of business as usual GHG emissions scenarios (that is, with no further mitigation efforts), models project that the global mean temperature could rise 2.5°C to 7°C above preindustrial levels by the end of this century. There is convincing evidence that the capacity of societies and ecosystems to cope with climate change will be tested with increasing severity as warming advances beyond 2°C. Limiting global mean temperature rise to below 2°C requires urgent actions or else the target soon will be out of reach given inertias in the climate system (climate change will progress further even after emissions decline), and inertias in the built environment (which drive up the costs of action once higher carbon capital investments are put in place). Stabilizing GHG atmospheric concentrations at around 450 parts per million CO2e is consistent with the objective of trying to limit warming to 2°C.1 However, this concentration target would require global GHG emissions to peak over the next decade and decrease at least by 50 percent below 2000 levels by 2050 (IPCC 2007).2 By 2030, this could represent absolute emission reductions of about 30­40 GtCO2e. The target of 30 GtCO2e by 2030 is much more than anticipated Annex I baseline emissions without new mitigation action (20 GtCO2e).3 In other words, even if Annex I countries entirely eliminated or offset their emissions by 2030, the world would still fail to reach the global 2030 mitigation target. Therefore, the global mitigation challenge goes well beyond the capacity of Annex I countries alone. Whether such large and rapid global emissions reductions are realistic is a matter of intense debate. However, the mostly untapped mitigation capacity in developing countries looks large enough to 3 4 World Bank Working Paper at least reduce the gap. Several studies have estimated the mitigation potential in non Annex I countries (see Appendix A: Mitigation Potential in Non Annex I Countries). The results of these studies tend to show that, once added to the global effort, the mitigation potential in developing countries could enable sufficient global reductions to stay on track for staying below a 450 parts per million carbon dioxide equivalent. Moreover, these studies show that 50­70 percent of the global mitigation potential would be located in non Annex I countries. Developing countries could yield up to 25 GtCO2e of reduced emissions annually. In sum, developing countries are an indispensible part of the mitigation solution. The global community must find the mechanisms and the proper incentives to fully bring their emission reduction forces into the climate change battle. Ideally this mechanism would also scale up investment in emerging countries and the global development agenda. Notes 1 Using a best estimate of 3°C for climate sensitivity, the amount of global warming associated with a doubling of atmospheric carbon dioxide concentration from preindustrial levels. Taking into account uncertainty on climate sensitivity, limiting global warming below 2°C above preindustrial levels with an estimated likelihood of about 80 percent (or higher) requires stabilizing GHG atmospheric concentrations at 378 parts per million (or below) (IPCC 2007), that is, very close to today's levels. 2 In 2005, global GHG emissions were approximately 13 percent above 2000 levels. 3 See, for instance, UNFCCC (2007), with required emission reductions of about 32 GtCO2e by 2030 or close to 150 percent of Annex I baseline emissions at this time, and McKinsey & Company (2009), with required emission reductions of about 42 GtCO2e by 2030 or close to 220 percent of Annex I baseline emissions at this time. CHAPTER 3 Call for New Financing Mechanisms T he Kyoto Protocol laid the foundation for a global carbon market--an innovative scheme to efficiently manage GHG emissions, primarily led by industrialized countries (see box 3.1). Several countries, alone or jointly, have chosen carbon trading (or plan to do so) as a key component of their climate change policies, with The European Union (EU) through its EU Emissions Trading Scheme (EU ETS) a leader (see box 3.1). An entity seeking to reduce its GHG emissions at least cost can either opt for internal abatement measures or acquire CERs (also called carbon offsets) from the carbon market, depending on the relative costs (abatement cost versus CER price). CERs are generated by emission avoiding projects conducted in non Annex I countries. Such projects can be registered under the CDM of the Kyoto Protocol. An official UN certification process, the CDM validates applying mitigation projects, assesses the expected emission reductions of such projects, imposes monitoring procedures, and certifies actual reductions. The corresponding CERs are then issued and delivered to the sponsor of the project (the primary market). These CERs are tradable. They can be used for compliance (that is, to meet specific abatement targets) by the project sponsor or sold on the secondary market for CERs. The volume of CERs issued has been steadily increasing since 2005. From 2002 to 2008, CERs totaled about 1,900 MtC02e with an approximate value of US$23 billion, mainly sponsored by private sector. There are about 6,000 projects presently being processed at the CDM, which should yield an additional 1,035 MtC02e in carbon offsets over 2008­12.1 In 2009, the total value of transactions in carbon markets was around 90 billion, with its three main components being the EUETS (around 73 billion), the primary CER market (2.5 billion), and the secondary CER market (15 billion).2 But despite its formidable achievements--creating a large market with a single price for carbon, catalyzing and implementing CDM projects--carbon finance has three main limitations. First, carbon finance provides only limited means to overcome the investment financing barrier faced by many low carbon projects. Second, carbon finance only incentivizes generation of carbon offsets; consequently, emission reductions achieved in non Annex I countries can only compensate for insufficient reduction effort in Annex I countries (the buyers of carbon offsets). Third, Annex 1 countries' future demand for CDM carbon offsets may fall off, partly as a consequence 5 6 World Bank Working Paper of the second limitation. The EU ETS limits the volume of CERs that can be used for meeting emission targets. Let us now look in more details at these three limitations. Several recent studies have shown that one of the main barriers for low carbon projects to be implemented in developing countries is limited access to investment financing.3 Successive annual World Bank carbon market studies have underscored that achieving financial closure is one of the main difficulties that mitigation projects face, leading to delays or failure to deliver the expected emissions reductions.4 Such a barrier is not specific to low carbon projects and affects generally investment projects in many developing countries. But CDM emission reduction projects present a number of specific challenges: Because they would mostly use new and uncommon practices and technology, low carbon projects seem to face even more difficulties than other projects when it comes to financing. The revenues generated by selling the CERs will typically be much smaller that the size of the investment. Although they are necessary to make the investment viable, CER sales cannot by themselves guarantee the financing of the project. Payments for carbon credits generally occur upon delivery, once the project is already operational. So far, carbon finance has not addressed the need to convert future revenues from CERs into upfront financing of mitigation investments. So far, and given the risks to the underlying project, there have been few attempts by financial institutions to monetize forward carbon revenue streams to contribute to upfront investment. In short, the sale of CDM offsets bring strong benefits in improving the profitability of low carbon projects in developing countries, but does not help much to overcome limited access to finance. Looking forward, carbon finance will likely continue to play a role in catalyzing low carbon investments in developing countries. However, the market of carbon offsets is intrinsically limited. It is of course mechanically capped by the size of the projected emissions in the Annex I countries. More specifically, the EU ETS limits the size of carbon offsets accepted for compliance to $1.4 billion until 2020. Likewise, current draft legislation for a cap and trade regime in the United States limits the volume of CERs acceptable for compliance. A few studies have produced estimates of the future demand for offsets. A series of bottom up estimates of the annual potential demand for offsets range from 400­600 MtCO2e per year--roughly the size of today's project market--to about 1,700 MtCO2e per year, representing a doubling or tripling over current market levels (table 3.1). In his review of Carbon Energy Economy integrated models--top down models as opposed to the bottom up approaches reviewed here--Haites (2007) estimates that the annual demand for offsets will reach about 3,150 MtCO2e in 2020.5 These numbers clearly remain below the size of the overall mitigation effort and below the contribution expected from developing countries. A Financing Facility for Low-Carbon Development 7 Table 3.1. Estimated Potential Demand for International Offsets by 2020 Source Potential demand (MtCO2e) New Carbon Finance 1,000­1,500a IDEAcarbon 500­1,200 Point Carbon 1,700 Barclays Capital 600­1,100b Sources: Barclays Capital (2008); IDEAcarbon (2008); New Carbon Finance (2008); Point Carbon (2008). a. Double or triple current levels. b. Annualized estimate of the maximum potential demand. The shortcomings of the market for CDM offsets discussed above call for complementary instruments for decisively scaling up emissions reductions in non Annex I countries. There is ample room for other instruments focused on investment financing rather than purchase of carbon offsets, to unleash the immense available potential in developing countries. Box 3.1. Major International Regimes for the Carbon Market The United Nations Framework Convention on Climate Change (UNFCCC). Treaty produced at the United Nations Conference on Environment and Development (Earth Summit) in 1992, which serves as the overall framework for governments' efforts to mitigate climate change, following "common but differentiated responsibilities," based on "respective capabilities" (box 1.1). The Kyoto Protocol (KP). Principal update to the UNFCCC treaty adopted in 1997 and entered into force in February 2005. Under the Kyoto Protocol, 38 industrialized countries (with the exception of the United States, which has not ratified the Protocol) have committed to country- specific targets that reduce their collective GHG emissions over the 2008­12 period by an average of at least 5.2 percent below 1990 levels; developing countries (with no obligations) can take no-regrets actions and participate in carbon markets. To comply with their Kyoto targets, industrialized countries can (i) take domestic actions (such as carbon tax, carbon trading, standards, or subsidies); (ii) trade allowances (Assigned Amount Units) among governments; and (iii) purchase emission reductions from projects in developing countries (via the CDM, generating CERs) or economies in transition (via Joint Implementation, generating ERUs). EU Emission Trading Scheme (ETS). World's largest multinational emission trading scheme, initiated in 2005. Under the ETS, EU member states allocate part of their efforts toward meeting their Kyoto Protocol targets to private sector emission sources (mainly utilities). In 2008­12, emissions from mandated installations (about 40 percent of EU emissions) are capped at an average of 6 percent below 2005 levels. Participants can reduce emissions via (i) domestic actions, (ii) purchase of EU Allowances, or (iii) acquiring CERs and ERUs (within a 13.4 percent average limit of their allocation over 2008­12). In January 2008, as part of its climate change strategy, and looking toward 2020 and beyond, the European Commission proposed continuing the EU ETS beyond 2012 and cutting emissions from mandated installations by 21 percent below 2005 levels in 2020 or beyond, depending on whether a satisfactory international successor agreement on climate change materializes. This is a major step in establishing the continuity of the carbon market beyond 2012; similar initiatives are emerging in Australia, New Zealand, and other major countries. Source: Authors. 8 World Bank Working Paper Notes 1 Source: UNEP, CDM pipeline, estimate as of April 2010. 2 Source: Point Carbon (2008). 3 See, in particular, World Bank (2008a); World Bank (2009a, 2009b). 4 World Bank (2007, 2008b, 2009c). 5 Median estimate by 2020 for Annex B Parties, should they wish to reduce GHG emissions by 20 percent below 1990 levels by that date. CHAPTER 4 The Proposed Low Carbon Development Facility LCDF Concept: Objective, Design, and Scale We propose to establish a LCDF with the aim to: (i) overcome the financing barrier currently faced by development of low carbon projects in developing countries; (ii) build on the capacity created by the Kyoto Mechanisms to screen, monitor, and verify emission reduction projects and push to considerably enhance this capacity; and (iii) decisively scale up investment flows toward low carbon development projects in non Annex I countries. The LCDF would be capitalized and guaranteed by Annex I countries so as to maintain the AAA rating on a sustainable basis. It would borrow in capital markets and lend at affordable rates to mitigation projects located in non Annex I countries. By leveraging the capital provided by Annex I governments, releasing the financing constraint, and offering highly competitive terms, the LCDF would thus combine (i) support to development and (ii) scaling up of GHG reductions to the magnitude required to meet the GHG concentration stabilization challenge. Specifically, the LCDF would channel $100 billion per year from private investors in capital markets to low carbon development projects. The LCDF would complement and partner with multilateral development banks and other public sector national or international entities combating climate change. Financial management of the LCDF could lead to the creation of a new international entity or be placed under the fiduciary responsibility of an existing international financial institution. In a public private partnership arrangement, the LCDF would also rely on the capacity of the global banking sector to screen, select, and monitor low carbon projects for profitability and financial soundness. For appraisal and the monitoring of projects' mitigation and environmental impacts, as detailed further in chapter 5, the LCDF would rely on an enhanced CDM. The monitoring could be associated with financial sanctions for projects that failed to deliver on their initial emission reduction commitments or to comply with MRV requirements. In such a case, the rate serviced on the LCDF loan would revert, in proportion to the failing, from its initial low spread to LIBOR plus 10 basis points towards the actual borrowing spread of the considered project, which would be charged by a commercial bank. The LCDF would significantly increase existing and future climate change financing. Based on our modelisation below, the LCDF could generate an additional 9 10 World Bank Working Paper annual $100 billion investments and annual GHG emission reductions of 10 GtCO2e to be reached in 2030. If delivered in 2030, 10 GtCO2e would represent almost a third of the 32.5 GtCO2e needed by 2030 stabilize climate change. The LCDF also contains two other interesting features, related to development impact and the cost of the mitigation effort. First, the investment flows channeled by the LCDF from international capital markets would significantly increase net foreign direct investment (FDI) to non Annex I countries, currently forecasted at around $613.3 billion in 2007.1 In addition, compared to the current level of official development assistance (ODA, $74.3 billion in 20072), the LCDF would be a new, effective financial development tool. Through its financing power the LCDF can help be the locus of a powerful North South collaboration toward sustainable, low carbon economic growth. Second, because emission reduction projects typically are small in size, the industrial risk of these investments can be particularly well managed through diversification. As a result, the unit cost of abatement (or the cost of abatement of one ton of CO2) associated with the LCDF, and measured in financial terms (interest and credit risk premium), is also a source of efficiency in the mitigation effort. At the core of the LCDF concept are three sources of value. One is to scale up decisively the financing of projects in non Annex I countries, where considerable sources of emission reduction can be exploited for the benefit of all. A second is to increase significantly the flow of investments to developing countries and to provide support for strong, sustainable growth path. The third is the LCDF's comparative advantage in terms of abatement costs. We will analyze these three sources of value in the assessment sections below while, at the same time, checking that the LCDF structure is financially sound. Assessing the LCDF Impact We start with a more detailed description of the LCDF. The LCDF Model presented in Appendix C translates this discussion into a more formal and modeled presentation. The LCDF would provide project lending at highly competitive terms to non Annex I countries and would refinance itself through issuance of AAA rated bonds in international capital markets. Lending costs would thus reflect a strong AAA borrower rating plus the operational cost margin and the average default rate. Through its capacity to diversify away its credit risk, the LCDF would be able to charge a stable credit risk premium that strictly reflected the statistical average rate of default on its portfolio. Governments from Annex I countries would guarantee the repayment of these bonds by providing the LCDF the amount of capital required to maintain the AAA rating on a sustainable basis. Shareholding would be aligned with the UNFCCC principle of "common but differentiated responsibilities," while Annex I countries would be responsible for the LCDF's financial viability. The LCDF would originate, select, manage, and monitor a large and diversified portfolio of loan financing projects with high emission reduction impact in non Annex I countries. At the outset, the LCDF would receive paid in capital to statistically ensure full repayment of the bonds it issued in capital markets and to secure its AAA rating. Through efficient portfolio diversification and careful selection of projects, the worst A Financing Facility for Low-Carbon Development 11 case loss from default on loans would remain strictly below the capital pledged by Annex I countries. The computation of the worst case loss would be derived from the distribution of portfolio losses and would result from a probability threshold, chosen to provide the LCDF its AAA rating. The LCDF would receive paid in capital to firmly protect the AAA level in the initial years of operation and in steady state (see Appendix C for more details on this critical aspect). Sound financial management would further firm up the AAA rating. Ultimately, Annex I countries would act as guarantors if the LCDF needed additional capital to sustain its AAA rating. However, the initial capitalization is designed to allow the LCDF to keep its AAA status over the long run without calling the guarantee from its Annex I government shareholders. This is a major element of the LCDF concept: it both limits and makes less uncertain to Annex I countries the cost of their contribution to the mitigation effort in developing countries. The LCDF would lend $100 billion annually and raise the same amount on international capital markets. This is a large program compared to multilateral development banks or national agencies: it is larger than the World Bank, larger but still comparable to the funding programs of EIB and KfW, but smaller than the annual funding programs of Fannie Mae and Freddie Mac (before the financial crisis). If the $100 billion borrowed every year were all lent out, 1,000 new loans (assuming an average loan size of $100 million) could be granted each year. After 10 years, the balance sheet could total $1 trillion (without the securities portfolios) and 10,000 loans. Loans to emission reduction projects would have a $100 million average size and a 10 year average maturity. These two characteristics are in line with the observed features of the projects having gone through or presently in the pipeline of the CDM. Another key element in the LCDF concept is its ability to leverage a relatively small amount of capital brought by the Annex I countries by cofinancing projects with other public or private financial partners. For instance, in countries with a low income per capita, the LCDF could blend its loans with IDA financing. In middle income countries, LCDF financing could be complemented by commercial bank loans or project finance. At the core of the LCDF lies the connection between the abatement potential in a project and its initial financing need. Let us introduce a parameter we refer to as the emission reduction transformation cost, or the ERTC. This is the amount of (metric) tons of carbon dioxide equivalent per year (tCO2e yr) that is abated by a dollar of investment in a mitigation project. Based on data on projects taken from the CDM pipeline, we estimate the intrinsic ERTC at about $300 per tCO2e yr; that is, projects requiring initial capital of $300 can each abate one tCO2e annually over their lifespans (10 years, 20 years or more in certain cases).3 In fact, the $300 investment will do more than abate one tCO2e per year over, say, 20 years. This capital primarily supports projects with economic development, industrial growth, environmental, and energy efficiency objectives. It is likely that most of their value will come from their impact on sustainable growth. Financing a low carbon project yields a wealth of economic benefits extending far beyond the incremental amount of abated emissions. Considering only the emission reduction impact, we assume that the LCDF can further leverage its lending capacity by cofinancing projects with partners from both 12 World Bank Working Paper public and private sectors, including development banks, funds, governments, private sector investors, and banks. If the LCDF were financing about two thirds (66.67 percent) of total capital, then the ERTC would go down to US$200 per tCO2e. We further consider that the LCDF, with its 10 year maturity loans, would reach a portfolio size of $1 trillion after 10 years (in 2020). Assuming that all LCDF loans represent only two thirds of the financing, the actual financing of low carbon projects generated by the LCDF would amount to $1.5 trillion. The total volume of emissions abated would then represent $1.5 trillion divided by $300 (the ERTC factor), that is, 5 GtCO2e. But another aspect has to be factored in. The life span of the emission reduction projects (the time during which they can generate emission reductions) goes beyond the 10 year financing period. If we assume a 20 year horizon, we see that the impact of the LCDF will reach its steady state only after 20 years (when the whole portfolio of projects sees its first generation of projects coming to an end). In sum, the LCDF has the scale to achieve the desired impact: after 20 years, in 2030, the total annual amount of emission reduction achieved would total 10 GtCO2e. Financial Viability This section focuses on future net income forecasts of the LCDF. Taking the balance sheet of the facility as a starting point, we analyze the conditions under which the LCDF can sustain its AAA rating over time and yield a stable net income. The detailed analysis--credit portfolio model, simulations, and derivation of key parameters--can be found in Appendix C. We introduce the following notations for the LCDF balance sheet, schematically represented as assets (L and P) and liabilities (C and B): C (capital and accumulated reserves) B (borrowings) L (loan portfolio) P (liquid portfolio) We model the facility as having bullet loans with final 10 year maturities. Within 10 years, the LCDF loan portfolio would grow to the $1 trillion (not including capital and reserves) and would maintain this size in steady state thereafter. The paid in capital and accumulated reserves would be invested in a low risk liquid portfolio. Annual revenues from this portfolio and from the spread charged on loans, net of operation costs, would exceed the average annual loss resulting from defaults. Therefore, the LCDF would have a steady stream of profits and a natural capital build up process to reinforce its capacity to absorb extreme losses. Loans would not necessarily require any guarantee from the governments of the countries where the projects are developed, but we suppose that the loans selected by the LCDF would be of a good average credit quality, which is assumed in the model to be around BBB. Rating agencies publish values for the number of defaults of a cohort of loans followed over time. From these data we estimate the expected default rate as the arithmetic average of the default rates of a BBB cohort taken in successive years A Financing Facility for Low-Carbon Development 13 (one year to ten years). In the simulation performed in Appendix C, the loan portfolio is composed of loans of similar maturities and ratings (all BBB/Bbb). The expected annual loss on the loan portfolio is 0.52 percent. The correlation is factored in the cumulative probability distribution formula developed by Vasicek (2002) with a value of 10 percent, which is quite conservative (see Appendix C). We simulate the evolution of the profitability of the LCDF and its capital in its first 12 years, during which it builds up and moves to a steady state phase. The amount of capital and accumulated reserves (C) results from the AAA target. For a given size of loan portfolio (L), C is determined by the facility's need for AAA rating. For a $100 billion portfolio of 1,000 loans (the facility in its first year), we want to estimate the initial paid in capital allowing the LCDF to sustain its AAA rating over 10 years. Contributing countries pay the initial capital, which will then be invested, along with the reserves, in safe securities. The LCDF's average annual profit can be expressed as follows: rP L( s e (1 )d) (1) with the following additional notation: r = rate of return on the liquid portfolio (P) s = spread over LIBOR charged on the loans. This is the single credit spread applicable to all loans, which guarantees, along with the initial paid in capital, the financial viability of the facility: the margin spread earned on loans matches the operating costs and the average annual loss on loans. The margin spread embedded in the lending rate achieves the preservation of the capital, and the AAA rating e = expenses, operating costs of the facility as a multiplier of the outstanding loans = recovery rate upon default 1 ­ = loss given default d = expected annual default rate on the portfolio of BBB loans. The expected annual default rate (d) is for a diversified portfolio of several thousand loans to projects distributed among non Annex I countries, where the main sources of emission reduction projects are located. We can infer from the current distribution of projects in the CDM that China, India, Brazil, Mexico, and other comparable developing countries with a good credit status are likely to capture a larger share of the financing. Using the annual default rate published by the rating agencies, averaged over 10 years, we obtain an expected annual default rate of 0.52 percent (more specifically, we took the arithmetic average of default rates estimated by the rating agencies over 10 years for a cohort of loans of initial rating BBB. The recovery rate is assumed to be 30 percent. We assume a 4 percent annual rate of return on the liquid portfolio and operational costs at 15 basis points of the entire portfolio. The LCDF Model is presented, simulated, and solved in Appendix C: see in particular the modeled correlation between loans and the impact of this correlation on the cumulative probability distribution of losses. Given the assumptions made on the cost structure and the lending rate, the AAA objective will determine two key parameters: the lending spread and the initial paid in capital. The financial structure and the pricing of loans will provide the LCDF with an expected positive net profit and 14 World Bank Working Paper a level of capital that can withstand an annual loss happening in less than 0.0004 percent of cases. The AAA rating is obtained in models made public by rating agencies when the considered entity remains solvent in 99.9996 percent of all cases. In line with the studies published by rating agencies, this should guarantee the LCDF's AAA rating. We find that, after 10 years, with a loan portfolio of 10,000 loans and an outstanding amount of $1 trillion, the capital needed to be given the AAA rating should be $78.8 billion. With an initial paid in capital of $68 billion, and assuming that the net income is fully retained as reserves in the first 10 years, the capital grows to the critical value of $80 billion (table 4.1). To guarantee a positive net income we allow the LCDF to charge a 30 basis point margin above its cost of funding, on its loans. In a steady state reached after 10 years the LCDF will hold a portfolio of 10,000 loans amounting to $1 trillion. It will have an expected net income of about $300 million (return on equity (ROE) of 0.4 percent on capital of $80 billion). The LCDF is a nonprofit entity so this average level of net income can be used for various actions related to the environment and mitigation. When above or below this average level, the net income replenishes reserves (to sustain the AAA rating) or takes from reserves to absorb a shock with larger than expected losses Table 4.1. Profitability of the LCDF: Net Income Forecast 2010­22 Year 1 2 3 4 5 6 7 8 9 10 11 12 Lending spread (bp) 30 30 30 30 30 30 30 30 30 30 30 30 Capital (in $B) 68.0 70.7 72.6 74.5 76.0 77.4 78.4 79.3 80.0 80.0 80.0 80.0 Loans (in $B) 100 200 300 400 500 600 700 800 900 1,000 1,000 1,000 Losses from defaults (168) (476) (819) (1,253) (1,701) (2,142) (2,513) (2,884) (3,241) (3,612) (3,612) (3,612) ($M) Portfolio income ($M) 2,720 2,121 2,179 2,234 2,281 2,321 2,353 2,380 2,401 2,401 2,401 2,401 Operating costs ($M) (150) (300) (450) (600) (750) (900) (1,050) (1,200) (1,350) (1,500) (1,500) (1,500) Loan margin ($M) 300 600 900 1,200 1,500 1,800 2,100 2,400 2,700 3,000 3,000 3,000 Net income ($M) 2,702 1,945 1,810 1,581 1,330 1,079 890 696 510 289 289 289 ROE 4% 3% 2% 2% 2% 1% 1% 1% 1% 0.4% 0.4% 0.4% Source: LCDF simulation model (see Annex C for details). The lending spread (or lending margin), which results in 30 basis points in steady state, does not reflect the actual credit risk of projects but the capacity of the LCDF to offer competitive lending rates associated with a AAA rating. The lending spread remains limited to what is necessary to guarantee financial viability, cover operational costs, and cover average losses on loans. The supporting rationale of the LCDF is not to eliminate the financial cost of emissions reduction to developing countries, which would take the form of a market based credit risk spreads, but to reduce it through diversification of credit risk and appropriate initial capitalization by Annex I countries. Note that the simulation does not factor in any preferred creditor status that the LCDF would be given, because facility would lend to projects and sovereign default risk would not be taken into account. Yet it is likely that project loans would be concentrated in large developing economies such as China, India, and Brazil) and could benefit from some form of sovereign guarantee, accompanied by a de facto preferred creditor status. ancing Facility for Lo A Fina pment ow-Carbon Develop 15 volution of the LCDF Loans for a 10 GtCO2e A Figure 4.1. Ev L r n Annual Emission Reductions T Target in 2030 ght he Source: Calculation of the authors. The gray line (rig scale) tracks th size of the LCDF portfolio over time. The black line (left scale) ind on ided by projects dicates the volume of annual emissio reductions provi finance by the L LCDF. How competitive wou uld LCDF lend ng ding rates be to developin countries? Assuming a c cost of funds of LIBOR minus 20 basis points on average (a estimate in f s an line with the funding costs experienced by multilateral d y development ba anks over the past 10 years) the LCDF wo ), cing at a rate of LIBOR plus 10 basis points. ould offer financ f 0 The same ave erage cost of fuunds can be exp ry pressed relative to U.S. Treasur Note rates. 2010 LIBOR min 20 would be about U.S. Tre As of April 2 nus e easury plus 30 basis points.4 Adding its margin, the LCDF average lendin rate would b around U.S. T F ng be Treasury plus 40 basis poin By compari nts. d ison, the spread of emerging m market bonds o over the U.S. e Treasury rate has been 350 basis points on a e average over the past 10 years ((which would be 300 basis points over LIB BOR). In April 2010 this sprea stands at about 250 basis ad points. Where do this leave th Annex I gove oes he at t ernments? Wha is the net cost of the LCDF e supposing it is going to function as per the model in App pendix 6? The a answer is that Annex I coun ely ntries collective have to com n mmit $68 billion in initial paid in capital. In comparison w with the commi y itments recently reaffirmed by large industria alized nations en uite he d in Copenhage this seems qu possible. Th LCDF would need no furth support in her the form of l get nts. rs long term budg commitmen Shareholder will guarante the LCDF, ee g is ery be but the strong financial structure makes thi guarantee ve unlikely to b called (see he by analogy th callable capital structure used by the World Bank). 16 World Bank Working Paper If the LCDF was not to meet the expected demand for financing, the liquid portfolio in which the initial paid in capital has been invested can easily be liquidated and repaid to the contributing countries. There is considerable reversibility in the initial capital commitments depending on the success of the facility. Economic Efficiency In this section we analyze a little further the economic benefits of the LCDF in the perspective of the overall mitigation effort. Some analysts have focused on the marginal cost of abatement as the principal indicator to measure economic costs. McKinsey & Company (2009) in particular has published curves showing the increase in the marginal cost of abating a ton of CO2e with the volume of emission reduction targeted. Well functioning carbon markets typically would reveal this marginal abatement cost as the market price of carbon. The approach taken in designing the LCDF focuses on the cost of financing an abatement effort of a certain size ($1 trillion equivalent to 10 GtCO2e emission reductions per year in steady state). However, we can still try to analyze how the LCDF helps to deal with the cost of reducing 10 GtCO2e a year from 2030. In essence, the LCDF supports the cost of financing the mitigation projects, including the cost of financing nonperforming projects that will default. Seen from the LCDF then, the economic cost of abating one ton of carbon in this LCDF financed 10 GtCO2e mitigation effort is measured, on average, by: l(1 )d (2) where l is the ERTC introduced above, that is, the dollar amount of investment needed to generate an emission reduction of 1 ton CO2e. We estimated the ERTC to be $300 per ton (assuming that the LCDF is financing alone the mitigation projects). If we replace the parameters by the values we estimated above we obtain: $300 x 70%.0.52% = $1.1 An immediate comment is that $1 per ton of CO2 abated is very low compared to the marginal cost of abatement revealed by the EU ETS market. The EU ETS is today, by far, the largest and most liquid market for carbon. The price of the carbon unit, the European Unit Allowance (EUA), drives the price of the secondary CER, which trades at a slight discount to the EUA. The EUA price has been quite volatile. As of April 2010, against the backdrop of an economic slowdown in Europe, the EUA is trading slightly above 12 per ton. But some market analysts estimate that for carbon capture storage technology to be developed in Europe, the EUA price should rise above 30 per ton. In any event, the EUA price stands well above the cost of abatement resulting from the LCDF, as measured by the expected loss given default in (2) above. In our analysis, emission reduction projects are reduced to a credit risk return profile and their cost is measured accordingly, that is, by the cost of default. It may be argued that the LCDF does not incorporate the entire cost structure of the project it finances. This is true; but on the other hand, the entrepreneurs undertaking these projects necessarily create profits in excess of these costs if, on A Financing Facility for Low-Carbon Development 17 average, the lending provided by the LCDF can be serviced. And, of course, the cost of nonviable projects is supported by the LCDF. A second comment is that the cost of abatement embedded in the LCDF is also much below the market price of offsets. The CER price is driven and closely correlated to the price of the EUA. The CER secondary market grew markedly starting in 2008. As of April 2010, there is enough liquidity to consider the CER market price a reliable indicator. Both the EUA and CER prices can be publicly observed, including on public carbon exchanges like the ECX or Bluenext. CERs trade at a slight discount (one or two euros per ton) relative to the EUAs. This comparison between (i) the financial marginal cost of abatement associated with guaranteeing the LCDF and (ii) the price of the EUA (marginal cost of abatement in the EU) and the price of the CER (price of carbon offsets) tends to support the following argument: By capitalizing the LCDF with an initial paid in capital of $68 billion, Annex I countries, consistent with their historical responsibilities, can help finance a mitigation potential situated in developing countries. The size of this potential (10 GtCO2e a year) is commensurate with the current global environmental challenge and creates a decisive incentive for developing countries to join forces with Annex I countries. Notes 1 Source: World Bank datafinder (http://datafinder.worldbank.org/). 2 Source: World Bank datafinder (http://datafinder.worldbank.org/). 3 Looking at the World Bank carbon finance portfolio (excluding projects targeting HFCs, PFCs, and SF6), we find a weighted average leverage factor of about US$270 per tCO2e yr. Looking at the entire CDM pipeline using data from Seres and Haites (2008), we find similarly a weighted average leverage factor of about US$250 per tCO2e yr. Such values are of course sensitive to the technology breakdown of the sample but confirm that our order of magnitude is correct. 4 The swap spread (that is, the difference between the Treasury reference and the LIBOR reference) has been 50 basis points at the 10 year maturity over the past 10 years (source: Bloomberg). CHAPTER 5 Building on the Kyoto Instruments: Attracting and Monitoring a Large Portfolio of Low Carbon Development Projects T o match the magnitude of the required global GHG mitigation effort, the proposed LCDF would need to support several tens of thousands of projects. This reality triggers a triple challenge: (i) to identify such a large number of mitigation projects; (ii) to technically screen these projects from a global environmental perspective to ensure the quality and measurability of mitigation benefits (volume of emission reductions); and (iii) to collect related financial information with which to assess, select, and monitor project loans. The latter activity would require establishing links with an extended network of multilateral development banks and commercial banks; in turn, these banks would benefit from an expanded borrowers' base resulting from the facility's concessional funding terms. They would take on a significant share of the credit risk for each project sourced to create a direct incentive for carefully assessing project viability. Although the organizational task of technically screening the mitigation performance of so many projects may appear daunting, the CDM of the UNFCCC, or a future enhanced CDM, is well positioned to meet the challenge. The CDM is also well positioned to identify and channel many projects, even if the financial analysis of the project would be handled by the LCDF and/or a commercial bank. Although the CDM bureaucratic process could certainly be improved, the existing institutions have successfully expanded the regulatory framework at a steady pace. Since the CDM was officially created at the Conference of the Parties 7 (COP 7) in Marrakesh, Morocco in 2001, an intensive learning process has allowed the mechanism's regulatory bodies (the Executive Board and MethPanel) to examine some 300 methodologies, more than 120 of which have been approved.1 Once approved, a methodology can be used by any similar project worldwide, thereby unleashing a new segment of GHG mitigation projects. In addition, a growing flow of revisions and consolidations is widening the scope of already approved methodologies, allowing for even more projects to be developed (see Appendix B). 18 A Financing Facility for Low-Carbon Development 19 As a result, the range of clean development projects eligible to the CDM has grown rapidly. In only five years, more than 5,000 projects in 70 countries have applied for validation (figure 5.1).2 Over the past two years, the pace of submitting new projects to the CDM has steadied at about 400 projects per quarter. That pace has not been really affected by uncertainties regarding the post Kyoto regime, even though most emission reductions are expected to occur well beyond 2012. Amid forecasts of plunging volumes of downstream carbon finance transactions and plummeting demand, the portfolio at entry point appears mainly supply driven. One explanation is that project developers are eager to bet on any external support that could help them overcome key barriers; convinced that the CDM will remain they believe it is worth developing and submitting applications, whose cost has fallen considerably. As a whole, the CDM has succeeded in unleashing a steady bottom up dynamic and has generated an unprecedented pipeline of low carbon development projects. It is strongly assumed that current growth of the CDM portfolio is constrained by procedural bottlenecks and that the steady project submission pace reflects the limited processing capacity of the overall chain.3 The bottlenecks most often cited are the limited number of accredited certifiers, known as Designated Operational Entities (DOEs); nonpermanence of the regulatory bodies; limited human resources for supporting the CDM in the UNFCCC Secretariat; and still imperfect procedures that must be improved on or reinvented while working at maximum capacity. High pressure at entry point, combined with development constraints on institutions, results in the steady expansion trends observed (for example, the pace of approving new methodologies, the growth of the portfolio, and increased GHG mitigation capacity, measured in tCO2e per year) (see Appendix B). Based on this solid linear trend (r2 > 0.99), the number of projects submitted to the CDM by the end of 2012 should total about 10,000 (figure 5.1). Extrapolating this trend to 2030 would yield more than 40,000 projects. As a result, the current annual GHG mitigation capacity of about 700 MtCO2e, which has been expanding steadily over the past three years at about 200 MtCO2e per year, would increase to about 1.5 GtCO2e per year in 2012 (figure 5.2). Extrapolating this trend to 2030 indicates a mitigation capacity of about 5 GtCO2e per year. Given the early stage of the CDM learning curve, these estimates may be considered conservative for three major reasons. First, as seen earlier, the untapped mitigation potential in non Annex I countries is far larger than 5 GtCO2e per year in 2030. Second, it can be expected that linking the proposed financing facility to the CDM will significantly increase the incentive for investment projects applying to the CDM. In particular, private banks' participation in the origination of new projects for the LCDF, in addition to the participation of public lenders including development banks, would allow a significant increase the CDM's capacity to reach out to additional potential projects. Finally, the above estimates are based on the current entry point capacity of the CDM, which is based on individual project application procedures.4 The combination of improving the CDM procedure, including by promoting a more programmatic or sectoral approach, and offering concessional financing through the LCDF can considerably enhance the capacity of the CDM as a portfolio aggregator for monitoring and certifying the mitigation performance of many projects. 20 World Bank Working Paper Figure 5.1. Historical and Projected Growth of the Number of Projects Submitted to the CDM by the End of 2012 10,000 9,000 8,000 Number of projects 7,000 6,000 5,000 4,000 3,000 2,000 y = 4.7223x ­ 183066 1,000 R 2 = 0.9981 0 Dec- May- Oct- Mar- Aug- Jan- Jun- Nov- May- Oct- Mar- Aug- Jan- Jun- Nov- Apr- Sep- Feb- Aug- Jan- 03 04 04 05 05 06 06 06 07 07 08 08 09 09 09 10 10 11 11 12 Year Source: Authors' calculations using UNFCCC data. Figure 5.2. Projected Mitigation Capacity of the CDM Portfolio by the End of 2012 1,600,000,000 1,500,000,000 1,400,000,000 1,300,000,000 1,200,000,000 1,100,000,000 tCO2/year 1,000,000,000 900,000,000 800,000,000 700,000,000 600,000,000 500,000,000 400,000,000 300,000,000 200,000,000 y = 567782x - 2E+10 100,000,000 R2 = 0.9959 0 Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- 03 04 04 05 05 06 06 07 07 08 08 09 09 10 10 11 11 12 12 Year Source: Authors' calculations using UNFCCC data. Therefore, it is proposed that: The CDM process, which has gained international legitimacy under the Kyoto Protocol and is expected to continue improving, should be used to validate the mitigation outcome of projects and also to identify potential candidate projects under the proposed LCDF. A Financing Facility for Low-Carbon Development 21 The official CDM validation should be set as an eligibility criterion and CDM certification of emissions reduction should be established as a binding agreement for every low carbon development project to be financed by the LDCF. This would ensure the full transparency and monitoring of the environmental performance of the LCDF. Notes 1 The Executive Board is a political body appointed by the COP to enact official decisions regarding practical modalities of the CDM. The MethPanel is a group of experts appointed by the Executive Board to provide technical recommendations on baseline and monitoring methodologies and CDM project cycle procedures, as well as address cross cutting policy issues, such as the definition of additionality, taking national policies into account in project host countries, and reducing transaction costs of small scale projects. The three principal authors of this study, all of whom are senior energy specialists, have served on MethPanel for three or more years. 2 Following slow initial growth (November 2001 to December 2003), the portfolio grew exponentially for several years before reaching a steady growth regime in 2006 (figure 5.1). More detailed information is available at http://cdmpipeline.org. 3 However, a fair criticism should compare transaction costs and timeline associated to the CDM validation and registration process to other national or international mechanisms like for instance MDB appraisal and financing processes or local environmental licensing. 4 The new programmatic approach, formally adopted at the 32nd meeting of the Executive Board in June 2007, has not yet been implemented in any significant way. Discussions about post 2012 modalities, which could, for example, rely on sectorwide benchmarking, are still at an early stage. Our calculations regarding the size of overall non Annex I potential are consistent with such assumptions. CHAPTER 6 Coordination with Carbon Markets: Complementarities of the LCDF and CDM T oday, all emissions reductions achieved by CDM projects can lead to the issuance of CERs that can be traded on the carbon market. To be registered under the CDM, candidate projects must demonstrate that the emissions reductions they would generate are additional; that is, without the CDM, the reductions would not have occurred. Demonstration of "additionality" is established during the validation stage of the official CDM process. Once validated, CDM projects must submit a request for registration; after successful registration, the projects can claim CERs for the emissions reductions they achieve over a limited time, known as the crediting period.1 Registration is usually requested when a project is about to initiate activities in order to avoid wasting part of the crediting period. Candidate projects can demonstrate additionality in many ways, such as difficulty in achieving financial closure or inability to compete financially with the baseline alternative.2 For example, many otherwise profitable energy efficiency projects, for which energy savings would ensure a short payback period, are stalled for lack of financing. With a large LCDF, such projects would no longer require carbon revenues for implementation.3 LCDF financing would suffice, especially if provided under concessional terms. Another significant portion of CDM projects would require both LCDF financing and carbon revenues to ensure competitiveness against baseline alternatives. Categorizing projects according to whether they require both LCDF financing and revenues from tradable CERs or LCDF financing alone would increase the efficient application of these instruments. Current CDM rules could easily be adapted to distinguish these two categories of mitigation projects via a two stage screening process: All projects applying for LCDF financing would need to demonstrate "environmental additionality"; that is, show that the baseline option would be more carbon intensive. The CDM methodology framework already provides proofed tools for testing the environmental additionality, namely the "baseline determination tool." Only projects that pass this test would be eligible for LCDF financing. 22 A Financing Facility for Low-Carbon Development 23 The CDM additionality test--a demonstration that, without the CDM, these emissions reductions would not have occurred--would be applied, taking into account the barrier removal effect of LCDF financing. For many projects, applying the CDM additionality tool(especially its financial modality, sometimes called "financial additionality") would be far easier than at present since it would be derived directly from the financial due diligence associated with applying for LCDF financing.4 LCDF projects that pass both additionality tests would be eligible for issuance of tradable CERs. Projects that pass the first, but not the second, test would be eligible for LCDF financing only and thus avoid inflating the carbon market.5 To ensure the cost effectiveness of the LCDF in delivering emission reductions and limit risk of lack of stringency at the early stage of implementation of the LCDF, it is proposed that a cap be established during the first years. This cap would limit to a maximum of 20 percent the volume of emission reductions achieved by projects financed by the LCDF that can become tradable CERs. The LCDF would give priority to projects that only need financing and would agree to finance projects requiring tradable CERs only up to that 20 percent limit. The limit would be checked quarterly. This cap would then be fine tuned when enough detailed knowledge regarding the financial additionality of the projects pipeline was accumulated. In short, using the existing CDM regulatory framework would permit sound and efficient development of both the lending facility and the carbon finance market. Notes 1 10 or 7 years, this can be renewed twice. 2 To date, many projects that have passed the validation stage are blocked from registration because of their inability to achieve financial closure. This common barrier, faced by any type of investment project in developing countries, was confirmed in the case of the above mentioned World Bank study in sub Saharan Africa. 3 Providing such projects tradable CERs, which they would sell on the carbon market, would eventually mean double financing of the same emissions reductions. 4 It is worth noting that this stage could benefit from any future improvement of the CDM modalities. 5 As a mandatory condition for benefiting from LCDF concessional financing, such projects would still be required to monitor their emissions reductions using an approved CDM methodology. CHAPTER 7 Conclusion T he global challenge of combating GHG emissions is of an order of magnitude that exceeds current institutional arrangements. Currently, the CDM based markets instruments of the Kyoto Protocol can only increase the internal rate of return of emission reduction projects in developing non Annex I countries. The existing carbon finance instruments cannot tackle the lack of lending capacity, which bars many developing countries from investment financing. The proposed LCDF would complement carbon finance by bringing the needed investment financing to mitigation projects in developing countries in the form of loans, at the required scale, on competitive and largely affordable terms. The LCDF would work in association with the international public institutions already involved in financing emission reduction efforts. The facility would leverage the financial monitoring and screening capacity of the private sector, as well as the technical expertise of the CDM, and would rely on an enhanced CDM to certify the mitigation performance of projects and to identify and channel projects. The LCDF could be set up quickly and function as a multilateral bank such as the World Bank or within it. The facility would have AAA status from its inception by receiving enough paid in capital to manage in steady state a loan portfolio of $1 trillion. Global diversification and conservative financial management would help ensure a positive annual net income on a sustainable basis. The LCDF would work in close cooperation with the CDM and use its existing infrastructure. It would therefore offer an incentive to radically enhance the productivity and the efficiency of existing processes at the CDM. The LCDF would also leverage the distribution and screening capacity of private banks, would work in collaboration with other international public institutions such as the regional development banks, the European Investment Bank, or the International Development Organization to further reduce operating costs. The LCDF could deliver in steady state, from 2030, 10 GtCO2e of emission reductions annually. This volume of abatement would constitute a significant step toward meeting the mitigation targets required to stabilize climate change, as estimated by the UNFCCC. 24 References Bakker, S. J. A., A. G. Arvanitakis, T. Bole, E. van de Brug, C. E. M. Doets, and A. Gilbert. 2007. Carbon Credit Supply Potential: A Bottom up Assessment of Mitigation Options Beyond 2012. ECN E 07 090. Petten: Energy Research Centre of the Netherlands. Barclays Capital. 2008. "So Long to the Longs." Monthly Carbon Standard, June. Barker, T., I. Bashmakov, A. Alharthi, M. Amann, L. Cifuentes, J. Drexhage, M. Duan, O. Edenhofer, B. Flannery, M. Grubb, M. Hoogwijk, F. I. Ibitoye, C. J. Jepma, W. A. Pizer, and K. Yamaji. 2007. "Mitigation from a Cross sectoral Perspective." In Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, eds. B. Metz, O. R. Davidson, P. R. Bosch, R. Dave, and L. A. Meyer. Cambridge and New York: Cambridge University Press. Haites, Erik. 2007. "Carbon Markets." Paper prepared for the United Stations Climate Change Secretariat, Bonn. Margaree Consultants, Inc., Toronto. Available at http://unfccc.int/files/cooperation_and_support/financial_mechanism/applicatio n/pdf/haites.pdf ------. 2008. Negotiations on Additional Investment and Financial Flows to Address Climate Change in Developing Countries. UNDP Environment and Energy Group, United Nations Development Program. IDEAcarbon. 2008. "The Long term Potential of the Carbon Market." Press Release, February 29. IEA (International Energy Agency). 2006. World Energy Outlook 2006. Paris: Organisation for Economic Co operation and Development (OECD)/IEA. IPCC (Intergovernmental Panel on Climate Change). 2007. Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II, and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, eds. R. K. Pachauri and A. Reisinger. Geneva: IPCC. McKinsey & Company. 2009. Pathways to a Low carbon Economy: Version 2 of the Global Greenhouse Gas Abatement Cost Curve. Available at www.mckinsey.com/ clientservice/ccsi/pathways_low_carbon_economy.asp New Carbon Finance. 2008. Press Release, January 28. http://www.newcarbon finance.com. Point Carbon. 2008. "Carbon Market Transactions: Dominated by Financials?" Carbon Market Analyst, May 21. Seres and Haites. 2008. "Analysis of Technology Transfer in CDM Projects." UNFCCC, Bonn. 25 26 World Bank Working Paper UNFCCC (United Nations Framework Convention on Climate Change). 2007. Investment and Financial Flows to Address Climate Change. Bonn: UNFCCC. Available at: at http://unfccc.int/files/cooperation_and_support/financial_ mechanism/application/pdf/background_paper.pdf USEPA (U.S. Environmental Protection Agency). 2006. Global Anthropogenic Non CO2 Greenhouse Gas Emissions: 1990­2020. Washington, DC: USEPA. Vasicek O. 2002. "The Distribution of Loan Portfolio Value." Risk, December: 160­162. WBCSD (World Business Council for Sustainable Development). 2002. World Bank. 2007. "State and Trends of the Carbon Market." World Bank, Washington, DC. ------. 2008a. "Low carbon Energy Projects for Development in Sub Saharan Africa-- Unveiling the Potential, Addressing the Barriers." World Bank, Washington, DC. ------. 2008b, "State and Trends of the Carbon Market." World Bank, Washington, DC. ------. 2009a; "Brazil Low Carbon Study." World Bank, Washington, DC. ------. 2009b; "Mexico Low Carbon Study." World Bank, Washington, DC. ------. 2009c "State and Trends of the Carbon Market." World Bank, Washington, DC. Appendixes 27 28 World Bank Working Paper Appendix A. Mitigation Potential in Non-Annex I Countries The IPCC indicates that for 2030 a global mitigation potential of 13.5 GtCO2e (9­17 GtCO2e)--half of which lies in developing countries--could be tapped at costs below $20 per tCO2e (Barker et al. 2007). For costs under $50 per tCO2e, this potential could reach 19.5 GtCO2e (13­26 GtCO2e), with slightly more than half located in developing countries. A credible, long term price signal of $20­$50 per tCO2e, along with actions targeting implementation barriers, could thus trigger multiple mitigation actions. Infrastructure, forestry, and agriculture sectors account for a significant share of mitigation opportunities. Buildings offer the cheapest form of mitigation, given that a significant portion of this potential has a "negative" cost. At costs below $20 per tCO2e, buildings represent 40 percent of all mitigation opportunities and 25­30 percent at costs of $50 per tCO2e. Forestry and agriculture sectors account for 30 percent of abatement opportunities in developing countries and 20­25 percent globally. Top down estimates are generally similar (with a predictably lower high end range), but sectoral breakdowns may vary (UNFCCC 2007). In reviewing estimates of mitigation costs in developing countries, Haites (2008) reported a maximum potential of 7.7 billion tCO2e annually by 2030. Energy efficiency, fuel switching, and reduced emissions from deforestation and degradation each accounted for roughly 25 percent, while reforestation and carbon storage and collection accounted for about 17 percent each and renewable energies 12 percent. Most recently, McKinsey & Company (2009) indicated a technology based, emission reduction potential below 60 per tCO2e at about 38 GtCO2e per year by 2030,1 on top of which behavioral change­based emission reductions could yield another 3.0­5.0 GtCO2e of abatement per year. According to this study, by 2030, 67 percent of the technology based potential or about 25 GtCO2e per year would be located in developing countries. Note 1 This study considered only technologies that provide emission reductions at an estimated cost less than 60 per tCO2e; the study estimated that another 3.0­6.0 GtCO2e per year of technology based abatement potential would be available at a cost of 60­100 per tCO2e. A Financing Facility for Low-Carbon Development 29 Appendix B. Historical and Projected Evolution of the CDM Pipeline: Rapid Widening of Scope and Quick Growth of the Project Portfolio and Mitigation Potential Although the CDM bureaucratic process could be improved--its key regulatory bodies (the Executive Board and MethPanel) are not yet permanent--the existing institutions have nonetheless succeeded in expanding the regulatory framework at a steady pace. As result, the range of mitigation options eligible to the CDM has grown rapidly. In addition, a growing flow of revisions and consolidations is widening the scope of already approved methodologies, allowing for even more projects to be developed. Since its official launching at Marrakesh on November 11, 2001, the CDM has unleashed an enormous bottom up dynamic, with participants from all over the world submitting proposals of methodologies to incorporate new mitigation activities into the CDM. For several years, every two to three months, the MethPanel receives rounds of 8­15 new proposals for review submitted by project participants (figure A2.1). In the first 30 rounds, some 300 methodologies, corresponding to as many mitigation activities, were proposed, illustrating the vitality of this supply driven process. Figure A2.1. Pace of Submitting New CDM Methodologies 350 25 Number of methodologies submitted per round 300 Total number of methodologies submitted 20 250 15 200 150 10 100 y = 0.1454x - 5474.7 R2 = 0.9962 5 50 No. new methodologies submitted New methodologies submitted per round Linear (No. new methodologies submitted) 0 0 Apr-03 Dec-03 Aug-04 Apr-05 Jan-06 Sep-06 May-07 Jan-08 Oct-08 Year Source: Authors' calculations using UNFCCC data. 30 World Bank Working Paper Figure A2.2. Pace of Approving CDM Methodologies 140 Total number of methodologies approved 120 100 80 60 40 y = 0.0649x - 2462.6 R2 = 0.98 20 No Approved Methodologies Linear (No Approved Methodologies) 0 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Dec-08 Year Source: Authors' calculations using UNFCCC data. Currently, the total number of approved methodologies exceeds 120, with 2 approved each month on average (figure A2.2). Once approved, a methodology can be used by any similar project worldwide, thereby unleashing a new segment of GHG mitigation projects. As a result, in only five years, more than 5,000 projects in 70 countries have applied for validation. Over the past two years, the pace of submitting new projects to the CDM has steadied at about 400 projects per quarter (figure A2.3). The validation portfolio, which now represents an annual GHG mitigation capacity of about 700 MtCO2e, has been expanding steadily over the past three years at about 200 MtCO2e per year (figure A2.4). It is strongly assumed that current growth of the CDM portfolio is constrained by procedural bottlenecks and that the steady pace of project submission reflects the limited processing capacity of the overall chain. The bottlenecks most often cited are the limited number of accredited certifiers, known as Designated Operational Entities (DOEs); nonpermanence of the regulatory bodies; limited human resources for supporting the CDM in the UNFCCC Secretariat; and still imperfect procedures that must be improved on or reinvented while working at maximum capacity. High pressure at entry point, combined with development constraints on institutions, results in the steady expansion trends observed (for example, the pace of approving new methodologies, the growth of the portfolio, and increased GHG mitigation capacity, measured in tCO2e per year). A Financing Facility for Low-Carbon Development 31 Figure A2.3. Number of CDM Projects Submitted for Validation 5,000 500 4,500 450 Numberof projects submitted per quarter 4,000 400 Total number of projects submitted 3,500 350 3,000 300 2,500 250 2,000 200 1,500 150 1,000 100 y = 4.8713x - 188918 500 50 R2 = 0.9975 0 0 Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- Apr- Jul- Oct- Jan- 04 04 04 04 05 05 05 05 06 06 06 06 07 07 07 07 08 08 08 08 09 Year Source: Authors' calculations using UNFCCC data. Figure A2.4. Mitigation Capacity of Projects That Have Applied to the CDM (Validation Stage) 800,000,000 700,000,000 600,000,000 500,000,000 tCO2/year 400,000,000 300,000,000 200,000,000 y = 567782x - 2E+10 100,000,000 R2 = 0.9959 0 Dec- Mar- Jun- Sep- Dec- Mar- Jun- Sep- Dec- Mar- Jun- Sep- Dec- Mar- Jun- Sep- Dec- Mar- Jun- Sep- Dec- 03 04 04 04 04 05 05 05 05 06 06 06 06 07 07 07 07 08 08 08 08 Year Source: Authors' calculations using UNFCCC data. 32 World Bank Working Paper Based on this solid linear trend (r2 > 0.99), the number of projects submitted to the CDM by the end of 2012 should total about 10,000 (figure A2.5). Extrapolating this trend to 2030 would yield more than 40,000 projects. A similar steady trend is also observed regarding the mitigation potential of the projects submitted for validation. Based on this solid trend, the mitigation capacity of projects submitted to the CDM by the end of 2012 should increase to about 1.5 GtCO2e per year (figure A2.6). Simply extrapolating this trend to 2030 indicates a mitigation capacity of about 5 GtCO2e per year by that date. To date, few if any CDM methodologies have been systematically implemented in any one country.1 Although China is rightly presented as a strong CDM leader, with more than 1,500 projects submitted for validation, the country has used so far only 37 percent of the approved CDM methodologies already used by non Annex I countries; among those it has used, nearly half (46 percent) have been applied to only one or two projects. In the case of India, more than 1,200 projects are in the validation pipeline; only 50 percent of the approved methodologies already in use in non Annex I countries have been used in India, more than one third of which have been applied to only one or two projects. Figure A2.5. Projected Number of Projects Submitted to the CDM by the End of 2012 10,000 9,000 8,000 Projected number of projects submitted 7,000 6,000 5,000 4,000 3,000 2,000 y = 4.7223x - 183066 1,000 R 2 = 0.9981 0 Dec- May- Oct- Mar- Aug- Jan- Jun- Nov- May- Oct- Mar- Aug- Jan- Jun- Nov- Apr- Sep- Feb- Aug- Jan- 03 04 04 05 05 06 06 06 07 07 08 08 09 09 09 10 10 11 11 12 Year Source: Authors' calculations using UNFCCC data. A Financing Facility for Low-Carbon Development 33 Figure A2.6. Projected Mitigation Capacity of the CDM Portfolio by the End of 2012 1,600,000,000 1,500,000,000 1,400,000,000 1,300,000,000 1,200,000,000 1,100,000,000 1,000,000,000 tCO2/year 900,000,000 800,000,000 700,000,000 600,000,000 500,000,000 400,000,000 300,000,000 200,000,000 y = 567782x - 2E+10 100,000,000 R 2 = 0.9959 0 Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- 03 04 04 05 05 06 06 07 07 08 08 09 09 10 10 11 11 12 12 Year Source: Authors' calculations using UNFCCC data. For other countries, the untapped potential is even greater. Brazil, for example, with 377 projects submitted for validation, has used just 36 percent of approved methodologies, 59 percent of which have been applied to only one or two projects. Indonesia, Malaysia, and Mexico have used 26, 24, and 18 percent of approved methodologies, respectively; in each case, the percentage applied to only one or two projects is about half. The percentage of unused methodologies would be even higher if those approved but not yet used, such as land use change, were taken into account. Results of a recent study of 44 countries in sub Saharan Africa, which counted facilities similar to those hosting CDM project proposals in other non Annex I countries, revealed more than 3,200 potential projects (more than 60 times the number of current projects in the Africa CDM portfolio); these correspond to GHG reductions of more than 700 MtCO2e per year in the energy sector alone. Considering that these 44 countries represent, at most, 5.4 percent of the energy based emissions of non Annex I countries, a simple calculation indicates, for the entire non Annex I CDM potential, an order of magnitude far greater than 60,000 projects and 13 GtCO2e per year. McKinsey & Company (2009) indicates a technology based potential for GHG reductions of 23 GtCO2e per year in non Annex I countries in 2030. Note 1 With the notable exception of methodology AM001 for destruction of HFC23 34 World Bank Working Paper Appendix C. LCDF Financial Model Initial Capital Estimate and Financial Viability Introduction--Objectives of the LCDF Financial Model The LCDF Financial Model takes as given the nature and the scope of the LCDF operations: channeling funds from capital markets to emission reduction projects on a massive scale, through the transformation of large amounts of AAA rated borrowings into a diversified portfolio of relatively small size, competitively priced loans. In terms of its size, the LCDF is expected to channel $100 billion a year to mitigation projects. Based on this value, which is consistent with the effort globally required to stabilize carbon in the atmosphere, the LCDF Financial Model derives the dollar values of some key financial parameters. Such parameters are essential to ensure the financial soundness of the LCDF and bring it into existence. The Financial Model starts from the LCDF balance sheet. Using both asset liability management and portfolio credit risk management techniques, it computes forecasts for the expected net income, the sensitivity to credit losses, and the size of the equity needed to ensure financial robustness (taken to be the AAA rating) on a sustainable basis. In particular the Financial Model produces the credit loss distribution of the loan portfolio. The LCDF Financial Model assumes the credit risk of emission reduction projects to be represented by a certain rating. The Model takes the cost structure of the LCDF facility to be primarily driven by the size of the loan portfolio and represents costs as a percentage of the outstanding loans. Solving the LCDF Model produces the value of two key parameters, which result from the assumptions taken and the financial viability objective of the facility (ensuring solvability at the AAA level on a sustainable basis). These parameters are: the spread charged on the loans by the LCDF, which will add to the borrowing costs to result in a low lending rate offered to emission reduction projects the initial level of paid in capital to meet the financial viability objective; in its design the LCDF assumes that this initial capital should be brought by Annex I countries. Hypotheses and Notations The balance sheet of the LCDF is comprised of two classes of assets: a portfolio of loans (L) and a portfolio of safe and liquid securities (P). Liabilities are also divided in two: borrowings (that is, bonds issued in international capital markets) and capital plus accumulated reserves (C). The financial structure of the LCDF is straightforward. Loans are matched dollar for dollar by borrowings (B) whose size is therefore equal to the size of the loan portfolio in amount. Capital and reserves are invested in the portfolio of liquid securities. Individual loans are supposed to be of the same individual amount, $100 million; of identical credit quality, equivalent to the BBB rating (Standard and Poor's); of the same maturity, 10 years; and denominated in U.S. dollars. For simplicity of exposition, loans are furthermore supposed to be bullet loans (the analysis would be unchanged A Financing Facility for Low-Carbon Development 35 for amortizing loans). The model is a discrete time model with 10 periods representing a year each. At the BBB rating, S&P provides, for time horizons going from one to ten years, the annual probabilities of default applicable to any cohort of new loans. In steady state, the portfolio will aggregate 10 cohorts of loans originated over the past 10 years. As a result, the default rate of the loan portfolio can be reasonably well approximated by the arithmetic average of the S&P default rates applicable to horizons one to ten years. We also assume that loans are correlated. The degree of correlation will be measured by the correlation parameter in the Vasicek (2002) model, which we will use to derive the loss distribution. Securities held by the LCDF are assumed to be liquid securities yielding a rate of return noted r. We assume this risk free rate to remain constant over time. In the simulation, for the purpose of computing the net income, r will be assigned a value, which is representative of the level of rates observed over the past 10 years. There is no discussion of whether the portfolio of securities should be benchmarked on U.S. Treasuries or on LIBOR as this is not the focus of the paper. The balance sheet structure separates the following: The loan portfolio financed by bonds, with both assets and liabilities indexed on LIBOR. Since all bonds issued would be swapped into U.S. dollar LIBOR, the LCDF liability facing the loan portfolio is equivalent to one large borrowing on which the LCDF has to pay the dollar LIBOR plus or minus a spread (representative of the AAA). All loans are also swapped into U.S. dollar LIBOR and the conditions (the interest rate charged) are based on the dollar LIBOR rate. This fraction of the balance sheet is insensitive to the level of interest rates and produces a net income that results from the lending margin, the operational costs, and the defaults occurring on the loans. The capital and reserves, assumed to be fully invested in securities returning r, which can be either the risk free rate (rate paid on U.S. Treasuries) or the LIBOR rate. The net income associated with this fraction of the balance sheet is directly dependent on the level of interest rates. Since this sensitivity is not the focus of the model we suppose that these rates remain constant. The LCDF will finance its loans by issuing bonds. At the end of a first year of operations, before allocation of the year's net income, and assuming there were no defaults in this first year, the balance sheet of the LCDF would be as below, expressed in millions of dollars. Assets Liabilities Securities--P Capital--C Loans--L Borrowings--B or 36 World Bank Working Paper Assets Liabilities Securities--68 Capital--68 Loans--100 Borrowings--100 We will detail below how capital can be estimated to be $68 million at the onset of the LCDF's operations. The net income is expressed as the net intermediation margin on loans plus the revenues from the portfolio minus operation costs minus losses on defaulting loans. Operation costs and expected losses are given as a percentage, respectively: e and of d of total loans outstanding. The notation e is estimated at around 15 basis points of total loans outstanding. The intermediation margin on loans, the percentage of the loan amount that comes in addition to the cost of borrowing to compute the interest charged to borrowers, is noted as s. With these assumptions and notations the expected net income is calculated as: rP L( s e (1 )d ) (1) where p is the recovery rate upon default and 1 ­ p is therefore the loss given default, as a percentage of the loan amount. We assume the recovery rate to be 30 percent, based on statistics and estimates made public by rating agencies. The notation d is the expected annual default rate on the loan portfolio. Modeling the Losses on the Loan Portfolio In steady state the expected average loss is the arithmetic average of expected annual losses for the 10 annual cohorts of 10 year loans, each one with its specific vintage year. We assume loans to have a BBB rating. We derive the data on expected annual losses from the statistical cumulative losses, calculated and published by S&P, for cohorts of loans with initial rating BBB, over the period 1981­2008. Losses are expressed in percentage of the initial principal. Years 1 2 3 4 5 6 7 8 9 10 Cumulative 0.24% 0.68% 1.17% 1.79% 2.43% 3.06% 3.59% 4.12% 4.63% 5.16% loss Annual 0.24% 0.44% 0.49% 0.62% 0.64% 0.63% 0.53% 0.53% 0.51% 0.53% loss Average annual loss: 0.52% We use the statistical average annual loss above as an estimate for the annual expected loss: d = 0.52 percent. The expected annual loss enables us to model the profitability of the LCDF for different values of the margin spread s and of the portfolio rate of return r. A Financing Facility for Low-Carbon Development 37 A full assessment of the financial viability and resilience of the LCDF, and of the appropriate level of capital to secure a AAA rating, requires that we know the distribution of losses on the loan portfolio so as to derive losses for various levels of occurrence probability, extreme and rare losses in particular. We start by detailing below the incidence of the assumption made on the degree of correlation between losses. Let us consider the portfolio in steady state, a portfolio comprised of 10,000 loans of $100 million each. Let us now suppose that their default characteristics are all independent. In such a case the number of loans n that can be in default is governed by the Poisson law. dn P( n) exp( d) n! (2) Based on the quantitative data published by S&P, the AAA rating would be obtained if the probability of default becomes less than 0.0004 percent.1 Therefore, we solve for the 99.9996 percent fractile, that is, we want to know the value for the number of loans in default in a given year, which would be an event with a probability of 0.0004 percent. The Poisson law establishes this number to be 86 loans. P( n 86 ) 0.0004 % This number of 86 loans in default corresponds, based on the assumptions above, to a loss of $6.02 billion (86 times $100 million times a 70 percent loss). P( loss $6.02 B) 0.0004 % The capital needed for a AAA rating would then be $6.02 billion. Figure A3.1: Number of Loans in Default, Cumulative Distribution and 99.9996% Level 120 100 80 Percent 60 40 20 0 1 9 17 25 33 41 49 57 65 73 81 89 97 105 Source: Authors' calculations using UNFCCC data. 38 World Bank Working Paper The hypothesis that all loans are independent is however a strong assumption. In order to model a possible correlation while keeping a tractable probability distribution, we use the expression developed by Vasicek in 2002 and published in Risk (Vasicek 2002) under the title Loan Portfolio Value. If we note: n nT the percentage of loans in default n on a total number of loans nt , the cumulative probability distribution is given by the probability that the percentage stays inferior to a certain number x, expressed as below: 1 P( x) N{ ( 1 N 1( x) N 1(d))} (3) where N and N ­1 are the cumulative normal distribution and its inverse, respectively. The correlation parameter is assigned the conservative value of 10 percent.2 In the Vasicek model, represents the correlation between the net values of any two borrowing projects taken from the portfolio of projects. We simulate the loan portfolio at the time when the LCDF has been in operation for 10 year and runs a $1 trillion balance sheet (with annual generations of loans of $100 billion each equally spanning the 10 previous years). We derive, from the cumulative distribution in equation (3), that the probability that the number of loans in default exceeds 1,126 is 0.0004 percent. P( n 1126 ) 0.0004 % This number corresponds to a loss of $78.8 billion. Compared to the $6 billion needed only when assuming independence, one can clearly see the significant impact of factoring in possible correlations between projects and loans. P( loss $78.8 B) 0.0004 % In steady state, after 10 years, the capital needed for the LCDF to sustain a AAA rating over time would have to be at least $78.8 billion. Financial Performance and Profitability of the LCDF over Time We can now simulate the financial performance of the LCDF. The facility is initially endowed with enough capital to generate capital and reserves of $78.8 billion in the tenth year of operation. Solving for this amount we find that the LCDF would need initial paid in capital of $68 billion to obtain, independently of any form of government guarantee, a AAA rating on a sustainable basis. More specifically, the evolution of the financial indicators in the first 10 years is retraced in table A3.1, where it is shown that capital plus reserves at the end of the tenth year of operation reach $80 billion, that is, slightly above the $78.8 billion needed. This level allows the LCDF to sustain its AAA thereafter on a steady state basis. We suppose that the level of interest rate r is 4 percent. Losses from defaults are calculated over the first 10 years by progressively adding up the cohorts of loans and A Financing Facility for Low-Carbon Development 39 using an average loan of default, which is the moving average of the annual default rates applicable to each cohort. With a 30 basis point spread charged on loans, the LCDF would show a very stable path in its first 10 years running up towards steady state regime. In particular, the net interest margin earned on the loan portfolio matches the operating costs plus the annual average losses on loans. At the end of the first 10 years, the LCDF can function with a balanced statement of account, yields a positive bet income that is fully distributed, and sustains operating costs of around $1.5 billion. Such a budget, comparable to the operating budget of an institution like the World Bank, reflects the presence of experts in infrastructure and development lending necessary to originate, select, and monitor loans, in partnership with institutions form the private and public sectors, at the best practice level. Table A3.1. Profitability of the LCDF over a 12-Year Period, Net Income and Return on Equity Year 1 2 3 4 5 6 7 8 9 10 11 12 Lending spread (bp) 30 30 30 30 30 30 30 30 30 30 30 30 Capital (in $B) 68.0 70.7 72.6 74.5 76.0 77.4 78.4 79.3 80.0 80.0 80.0 80.0 Loans (in $B) 100 200 300 400 500 600 700 800 900 1,000 1,000 1,000 Losses from defaults (168) (476) (819) (1,253) (1,701) (2,142) (2,513) (2,884) (3,241) (3,612) (3,612) (3,612) ($M) Portfolio income ($M) 2,720 2,121 2,179 2,234 2,281 2,321 2,353 2,380 2,401 2,401 2,401 2,401 Operating costs ($M) (150) (300) (450) (600) (750) (900) (1,050) (1,200) (1,350) (1,500) (1,500) (1,500) Loan margin ($M) 300 600 900 1,200 1,500 1,800 2,100 2,400 2,700 3,000 3,000 3,000 Net income ($M) 2,702 1,945 1,810 1,581 1,330 1,079 890 696 510 289 289 289 ROE 4% 3% 2% 2% 2% 1% 1% 1% 1% 0.4% 0.4% 0.4% Source: Authors' calculations using UNFCCC data. The 30 basis points lending margin is also comparable to the lending margin posted by multilateral development banks. At the time this paper was written, spreads on bonds of supranational institutions had moved upwards in the aftermath of the 2007­08 crisis. Using a precrisis cost of funds more typical of a AAA international financial institution (say LIBOR minus 20 basis points), we would come up with a lending rate in the vicinity of LIBOR plus 10 basis points, which can be considered as an extremely competitive lending rate, particularly compared to emerging market bond spreads (which have averaged 300 basis points over LIBOR over the past 10 years). Note that the calculations performed in this simulation were based on conservative assumptions. There would be room for improvement on the lending margin as the actual future net income stream of the LCDF allows it to build on reserves more rapidly than planned. Notes 1 See S&P Website (www.standardandpoors.com) and the CDO evaluator section. 2 This is a level considered as conservative by rating agencies, which further indicates that in case of a strong regional or sectoral concentration, the parameter value should be taken to be 20 percent. See the S&P Website for more details. Eco-Audit Environmental Benefits Statement The World Bank is committed to preserving Endangered Forests and natural resources. We print World Bank Working Papers and Country Studies on postconsumer recycled paper, processed chlorine free. The World Bank has formally agreed to follow the recommended standards for paper usage set by Green Press Initiative--a nonprofit program supporting publishers in using fiber that is not sourced from Endangered Forests. For more information, visit www.greenpressinitiative.org. In 2008, the printing of these books on recycled paper saved the following: Net Greenhouse Trees* Solid Waste Water Total Energy Gases 289 8,011 131,944 27,396 92 mil. *40 feet in height and Pounds CO2 6­8 inches in Pounds Gallons BTUs Equivalent diameter A Financing Facility for Low-Carbon Development is part of the World Bank Working Paper series. These papers are pub- lished to communicate the results of the Bank's ongoing research and to stimulate public discussion. This paper proposes an innovative financing mechanism, known as the Low Carbon Development Facility (LCDF), which would bring additional investment financing at con- cessional rates to low-carbon development projects in non- Annex I countries. The new financing would help to rapidly scale up project-based emissions avoidance in these coun- tries, up to 10 GtCO2e in 2030. The LCDF also could help the post-Copenhagen Green Climate Fund implement financial pledges made by Annex I countries to support projects, pro- grams, policies, and other activities in developing countries. The LCDF neither limits financing to incremental costs nor focuses on transformational projects. Therefore, the Facility does not substitute for the Global Environment Facility and the Clean Technology Fund, but rather helps scale up the innovative projects pioneered by these instruments. World Bank Working Papers are available individually or on standing order. This World Bank Working Paper series is also available online through the World Bank e-library (www.worldbank.org/newelibrary). ISBN 978-0-8213-8521-0 THE WORLD BANK SKU 18521