AFD_REGISTRATION_DOCUMENT_2017

CORPORATE SOCIAL RESPONSIBILITY

2

Report from one of the statutory auditors

Conclusion Based on this work and taking into consideration the limits referred to above – in particular the limits of the scope of quantitative environmental data which covers 56% of the workforce and the limits of the scope of some quantitative social data which covers 63% of the workforce – we certify that the management report includes the required CSR Information.

2.

Reasoned opinion on the reliability of the CSR Information

Nature and scope of work We held some ten meetings with the persons tasked with preparing the CSR Information in the departments responsible for the information collection process and, when necessary, with the persons responsible for internal control procedures and risk management, in order to: P ascertain whether the Standards are appropriate in terms of their relevance, completeness, reliability, neutrality and understandability, taking best practices for the sector into consideration when necessary; P verify that a collection, compilation, processing and review procedure has been established with the aim of ensuring that CSR Information is consistent and complete, and learn about internal control and risk management procedures relating to the preparation of CSR Information. We have determined the type and extent of our tests and controls based on the nature and importance of CSR Information with respect to the Company’s characteristics, the social and environmental issues linked to its activities, its priorities for sustainable development and best practices for the sectors. For the CSR Information that we deemed to be the most importantb (1) b: P at Group level, we examined source documents and conducted interviews to corroborate all qualitative information (organisation, policies, measures), we applied analytical procedures to the quantitative information and checked, using sampling techniques, the calculations as well as the consolidation of data and ensured that it was consistent with the other information in the management report; P we conducted interviews at the Agency’s head office to verify that procedures are properly applied and to identify any omissions. We also conducted detailed tests based on sampling, which consisted of checking the calculations made and reconciling data with supporting documentation. The selected sample represents 100% of the employees considered as parameters characteristic of the social data, 100% of the environmental data considered as parameters characteristicb (2) of the environmental issues and 100% of the quantitative corporate data presented. For other consolidated CSR Information, we assessed its consistency with our knowledge of the Company. Finally, we assessed the relevance of the explanations provided when certain information was totally or partially missing. We believe that the sampling methods and sample sizes that we used, based on our professional judgement, have allowed us to formulate an opinion with a moderate level of certainty. Greater certainty would require a broader audit. Because of the use of sampling methods and because of other limitations inherent in the operation of any information and internal control system, the risk of failing to detect a material misstatement in the CSR Information cannot be totally ruled out. Conclusion Based on this work, we found no material misstatement that would cause us to believe that the CSR Information is not, as a whole, presented in a truthful manner in compliance with the Standards. Paris-La Défense, 3 April 2018 KPMG S.A. Anne Garans Pascal Brouard Partner Partner Sustainability Services (1) Social indicators: Total headcount (breakdown by gender, age and region); External recruitment on permanent contracts; Number of dismissals (collective and individual); Percentage of supervisory positions held by women; Absenteeism rate; the total number of training hours; Average annual salary. Environmental indicators: Total paper consumption; Energy consumption; Emissions of greenhouse gases associated with energy consumption and employees’ business travel by air and by train. Societal indicators: Breakdown of AFD approvals by sector (loans outside France); Number and value of climate co-benefit projects; Estimated annual emissions of equivalent tonnes of CO2 avoided or reduced by mitigation projects; Share of sovereign financing >€100K under way which has been published on AFD’s website; Net total commitment to projects built on climate emissions as at 31bDecember 2017; Number of training hours dedicated to human rights. Qualitative information: Health and safety conditions in the workplace; Measures taken to promote equality between men and women; Training policies implemented; organizing the company to take into account environmental concerns and, if applicable, the procedures for environmental assessment or certification; Measures for preventing, recycling, reusing, other forms of recovery and elimination of waste; Adapting to the consequences of climate change; Conditions for dialogue with individuals or organisations interested in the company’s activity; Partnership or sponsorship initiatives; Anti-corruption initiatives. (2) See the list of environmental and social indicators set out in the footnote on page 3 of this report.

50

REGISTRATION DOCUMENT 2017

www.afd.fr