Machine learning powered tooling for analysing climate alignment in the financial industry

越来越多的人期望资产所有者和管理者将大量资本配置到与气候相关的工具上。然而，与传统金融工具不同，传统金融工具的指标（如利率、期限、股息收益率）i)易于测量，ii)易于理解，气候融资指标由于以下几个原因（概述如下）更具挑战性。这意味着资产管理公司发现很难做出针对气候结果进行优化的决策，并且在一切照旧的基础上，是否会增加对此类投资的摩擦，以及在极限情况下，是否会导致气候融资领域内的资本错配置。在这个项目中，CAR与气候债券倡议组织（CBI）合作，提议开发一个应用程序，将气候和环境因素无缝地整合到金融决策中，彻底改变资产所有者和管理者与相关数据的交互方式。我们将构建一个应用程序，能够从披露文件中提取和汇总复杂的气候和环境数据，并根据评估气候变化指标的优先国家和国际协调框架（例如欧盟分类）分析输出。该应用程序将使金融市场参与者能够搜索和使用这些数据，从而使他们分析相关信息的能力发生重大变化，并使这些因素在常规决策过程中得到考虑。

Asset owners and managers are increasingly expected to allocate large proportions of capital towards climate-related instruments. However, unlike traditional financial instruments, which have metrics (e.g. interest rate, duration, dividend yield) that are i) easily measured, and ii) easily understood, climate finance metrics are significantly more challenging for several reasons (outlined below). This means asset managers find it difficult to make decisions which optimise for climate outcomes, and on a business-as-usual basis, can i) increase frictions towards such investments, and ii) at the limit, contribute to the misallocation of capital within the climate finance space.In this project, CAR, in partnership with the Climate Bonds Initiative (CBI), proposes to build an application to seamlessly integrate climate and environmental factors into financial decision making, overhauling how asset owners and managers interface with relevant data.We will build an application capable of extracting and summarising complex climate and environmental data from disclosure documentation, and analysing the output against prioritised national and international alignment frameworks for assessing climate change metrics (e.g. the EU Taxonomy). The application will make this data searchable and consumable by financial market participants, enabling a step change in their capacity to analyse relevant information and for these factors to be taken into account in regular decision making processes.