Clarity AI Response to the Japanese Financial Services Agency’s (JFSA) Code of Conduct for ESG Evaluation and Data Providers
As announced on January 26th, Clarity AI endorsed the Japanese Financial Services Agency’s Code of Conduct for ESG Evaluation and Data Providers. Here’s a comprehensive summary of our adherence to the Code.
1. Securing Quality
Clarity AI is a digital-native firm that offers its clients a comprehensive, customizable, fully-packaged Sustainability Tech Platform.
Clarity AI is neither an ESG rating provider nor just an ESG data products provider. In relation to the Code of Conduct, Clarity AI provides a sustainability technology platform that supports investors’ decision making in capital allocation, risk management and reporting. As such, Clarity AI does not provide ESG ratings but a software as a service platform and/or data, for clients to assess various areas of sustainability, including customizable and transparent ESG scores.
Clarity AI’s business model is based on subscriptions. The platform is available under a SaaS model by purchasing a license. Clients can choose which information and level of detail they want to integrate into their system (via API calls or widget integration), or consume directly through the Clarity AI webapp.
Clarity AI strives for excellence in terms of the quality of its processes, data and methodologies. Clarity AI offers scientific- and evidence-based methodologies powered by research and data science expertise, leveraging Machine Learning to collect and analyze data points, perform reliability checks and run estimation models at scale. All products are subject to rigorous standards and are regularly reviewed for improvements. Continuous learning is at the core of Clarity AI’s quality standards and we have multiple channels for users to share feedback in order to drive improvements.
1.1 For formulating and providing ESG evaluation and data, establishing necessary procedures to analyze in detail information that can be reasonably obtained.
1.2 Establishing cross-organizational and continuously applied methodologies to provide high-quality ESG evaluation and data, and disclosing it while paying attention to confidentiality, intellectual property, etc.
1.3 In order to ensure that the prescribed methodologies are applied consistently across the organization, disseminating them throughout the organization, as well as devising measures, such as horizontally reviewing under an appropriate system, or accumulating and sharing knowledge of evaluations to be provided.
1.4 Checking on a regular basis whether there would be any apparent discrepancy between the evaluation results and the service provision methodologies mentioned above, and updating methodologies as necessary (implementation of the PDCA cycle for evaluation).
1.5 Managing ESG evaluation methodologies and data on a continuous basis, checking or updating them regularly, and disclosing when the input data is usually obtained or updated by the providers (if evaluation and data items are diverse or of great numbers, doing this in a reasonable scope and manner, such as by consolidating or limiting the scope, taking into account their importance and usefulness based on user needs).
Clarity AI sustainability experts work together with data science experts to innovate, create, deploy and maintain reliable and transparent tools and scores by leveraging advanced technology (including artificial intelligence such as machine learning and natural language processing).
Data
In our data collection process, Clarity AI employs automated reliability checks leveraging machine learning algorithms, developed and continuously improved by data science and sustainability experts. When assessing historical data trends, company data, industry peers, and proprietary models, our algorithms are able to identify and eliminate inconsistencies. When a potentially incorrect data point is detected, our process involves human review, creating a virtuous cycle that leads to continuous learning and improvement. This approach allows us to provide our clients with high quality and reliable data and sets us apart from other methods of data accuracy assurance. We regularly communicate publicly regarding our approach, including via blogpost articles.
Methodology
Clarity AI has developed a number of different data and evaluation products that enable its users to assess the sustainability of different financial products and portfolios. As a software as a service provider, Clarity AI tools allow its clients to customize any products to their use case. Where customers choose to use products with pre-set methodologies and scores, Clarity AI’s assessments are rules-based and grounded in existing evidence and research. Detailed methodologies are made available to users in order to understand the data and methodologies behind Clarity AI products. Clarity AI staff also have access to these methodologies. Our modular technology platform connected to a single source of truth and integrating rule-based methodologies ensures their consistent application.
Clarity AI regularly updates its pre-set methodologies to ensure they follow the latest cutting edge research, analytics and regulatory changes. Frequency of updates depends on topics and clients needs. Clarity AI regularly engages with investment professionals, including through its Client Advisory Board, and with regulators through bilateral conversations and working groups to ensure our solutions are aligned with industry best practices. Any changes made as a result of this feedback loop are subject to a full assessment of their implications and communicated to clients.
1.6 In cases where ESG evaluation and data providing services are outsourced, taking necessary measures for the quality of ESG evaluation and data to be ensured including the outsourced party, such as, as necessary and depending on the nature and importance of the outsourced service, requesting the outsourced party to comply with 1. through 5. above.
Clarity AI enforces the strictest standards with all of its providers and regularly and comprehensively reviews the quality of the data and services it receives, by applying quality assurance processes. Clarity AI’s vendor contracts have specific clauses around quality control, including the application of penalties where quality thresholds are not met.
2. HR Development
Clarity AI recruits only the most talented staff and continuously trains and improves them through feedback loops. Clarity AI upholds the highest standards in terms of people development. Clarity AI’s values are excellence, passion and values, and the pillars of our culture are to be fact-based, diverse, meritocratic, transparent and flexible.
2.1 Collecting and analyzing information necessary to provide appropriate evaluation and data, and maintaining necessary professional resources and technologies to make relevant decisions.
2.2 In particular, taking necessary measures to ensure personnel engaged in ESG evaluation and data would have professional knowledge and carry out their duties in good faith.
2.3 Considering the nature of personnel evaluations that would appropriately evaluate personnel who engages in professional evaluations and working for providing high quality evaluations.
2.4 Recognizing, as top management of the institution, that securing and developing human resources is important element for continuously providing high quality evaluations, and taking actions as necessary.
Clarity AI is not a rating provider nor just a data provider. Clarity AI is a tech platform providing sustainability capabilities. Clarity AI does not provide ESG ratings, but allows users to build customized scores and provides pre-set scores and sustainability assessments.
Clarity AI’s research processes involve selecting the most reliable data and applying the most robust methodologies leveraging technology. Staff involved in these processes are highly trained specialists, ensuring Clarity AI continues to deliver a high quality product and service to its clients.
Data and information gathering
Data gathering process at Clarity AI is automated when possible, and further supported and overseen by our Data Research team that is focused on manual data quality reviews. The Data Science team supports the development of algorithms and tools for efficient and high quality data collection and data quality processes.
Recruitment & Continuous improvement and training
Clarity AI’s management recognises the importance of qualified personnel to deliver on our mission of bringing societal impact to market. Our recruitment processes ensure all new hires are subject to a rigorous selection process.Once hired, Clarity AI assesses its staff on a continuous basis and performance review cycles. Staff are assessed against key performance metrics.
Our staff continually exchange knowledge and have access to resources provided by Clarity AI and 3rd parties for continuous professional learning and development in their respective field of expertise. Many have undertaken and completed the Certificate in ESG Investing from the CFA Institute. In addition to performance review processes that lead to identification of growth opportunities that employees work with their manager to achieve, team members are developed continuously through continuous feedback and on the job learning, formal courses/training and attendance to global leading conferences and events which are subsidized by the company.
3. Conflict of Interest
Clarity AI’s business model minimizes its exposure to conflicts of interest and any conflict is addressed with proper processes and procedures. Clarity AI does not provide ESG ratings, but allows users to build customized scores and provides pre-set scores and sustainability assessments. The pre-set scores are quantitative outcomes of models and do not embed any qualitative assessment based on analysts’ views.
- In some instances the models underpinning the scores can be customized by our clients to reflect their own sustainability views, especially regarding the relative weights of the KPIs contributing to the aggregated score.
Clarity AI’s scores are based on three main types of information to qualify an entity’s sustainability performance: quantitative metrics, policies and controversies
- In the case that Clarity AI’s software displays pre-set scores, these are quantitative scores exclusively based on models which do not include analysts’ qualitative assessments. Models are applied in a consistent manner to the covered universe of companies. These cannot be influenced by the companies subject to the scoring.
- Clarity AI does not provide any ancillary services.
3.1 Identifying potential conflicts of interest that may affect the assessment and analysis conducted by the provider or its employees with respect to the services provided, and then establishing and disclosing effective policies to avoid, or appropriately manage and reduce the risk of, the conflict of interest.
3.2 Taking appropriate measures to ensure that other business relationship with a company subject to ESG evaluation or data does not affect the ESG evaluation or data, such as establishing a firewall between sales and evaluation divisions.
Clarity AI business is centered around providing ESG analytics based on data collected from publicly available sources and third party providers. Clarity AI does not provide ancillary activities that could lead to conflicts of interest. Furthermore, for pre-set methodologies and scores, our rules-based approach minimizes exposure to conflicts of interest arising as a result of analysts’ judgment. Nevertheless, conflicts may arise in the course of changing industry dynamics or further growth of the company. Clarity AI is developing future proof additional policies on the management of conflict of interest which would be applicable throughout the business.
All personnel working on ESG evaluation and data are separated from the sales division. Input from companies subject to our scores is only integrated where it relates to raw data values at the metric level. Clarity AI will correct errors in the underlying raw data that drive the scores only when these errors have been confirmed through publicly available evidence.
3.3 In cases evaluations are developed through questionnaire, paying attention to the contents and structure of service and questionnaire, so that there would principally be no such situation where the content of the questionnaire is unreasonably too complicated or difficult to understand and effectively respond without using the provider’s paid services.
Not applicable to Clarity AI as questionnaires are not used.
3.4 Taking appropriate steps to prevent their employees from engaging in securities or derivatives transactions that could create conflicts of interest with ESG evaluation and data provision services.
Clarity AI staff involved in research and analysis are not able to tamper with methodologies and data in a way to favor their own private financial transactions. Specific guidelines around staff securities or derivatives trading would be added if new potential conflicts of interest were to emerge.
3.5 Developing appropriate work and compensation structures for its own employees, and avoiding, or appropriately managing and reducing the risk of, potential conflicts of interest related to ESG evaluation and data provision services. For example, as necessary, assigning a staff member to conduct evaluation, separate from the staff member responsible for sales of ESG evaluation and data services.
Staff involved in the sale of ESG evaluation products are separated from the team members who are involved in research and building our models and methodologies. The research team is part of the digital product team, developing our software products. This is a separate organization to the sales team who sit under the go to market team, reporting to the CEO in a different organization and business structure.
3.6 Establishing measures to ensure that existing business relationship with companies subject to ESG evaluation and data provision does not affect the evaluation to the companies.
Clarity AI has not identified any conflicts of interest in this regard. It does not provide ancillary services to companies that could lead to a conflict of interest.
3.7 For the issuer pay model where compensation is received from the company subject to the evaluation, implementing detailed procedures to avoid conflicts of interests.
Not applicable, see above (3.6)
3.8 In cases where the same provider provides both the-subscriber-pay-model businesses and the-issuer-pay-model businesses, taking appropriate measures to prevent conflicts of interest in this regards.
Not applicable, see above (3.6)
4. Transparency
Transparency is one of the core values of Clarity AI. Clarity AI is a market leader in terms of offering full transparency to its clients on the methodology used (including customized methodologies by the user) and the source of each underlying data point. Clarity AI also provides clients with the ability to have a view at portfolio level, including both the entities and the investment funds that the client is invested in.
4.1 While giving necessary consideration to intellectual property, etc., ensuring the transparency of their services by recognizing that it is an essential and prioritized issue.
4.2 In order for users of ESG evaluation and data provision services to understand the basic content of the services, including what the evaluation aims to capture and how this is measured, disclosing the basic approach for providing services, including the purpose and basic methodology of evaluation.
4.3 In order to enable users and companies subject to evaluation to understand the basic structure of the evaluation, disclosing sufficient information on the methodologies and processes for formulating the evaluation, including any major updates on them, if any. When inquiries are received from companies subject to evaluation through a contact point, providing careful explanations to the extent practically possible.
4.4 Disclosing the sources of information that are used in the development of ESG evaluation and data. In particular, if estimated data is used, disclosing this fact and the basic methodology of estimation. If data sources and/or items are diverse or of great numbers, doing these in a reasonable scope and manner, such as by consolidating or limiting the scope, reflecting their importance and usefulness.
Transparency is one of the foundations upon which Clarity AI was founded and we provide transparency at every level possible:
- Methodological documents are published for all products and provide explanations of our methodologies. These explanations are also embedded into our software platforms.
- For calculated metrics (e.g., exposure to controversial activities) we include an explanation on how the metric has been constructed and the inputs that went into that calculation.
- Type and source of raw data specified: e.g. for quantitative metrics, we flag whether the data is reported or estimated, and whether the data is from Clarity AI or an external provider. For reported Clarity AI data we provide links to the company report where the data can be found and in some cases extracts from the report are included in our platform.
- For Clarity AI estimated data, we provide the confidence level of the estimate.
Our products allow the user to examine both the data source and methodology that underpins any assessment. We are known as a market leader on transparency and have been on the record to support transparency numerous times (see for example this).
Clarity AI investigates any queries on its data or outputs with the utmost diligence. Our outputs are based on data collected directly from reported entities or trusted third parties. Where estimates are used, the assumptions underpinning them are well documented and shared with users, along with the associated confidence interval of the estimate. Users of Clarity AI’s tools can, via an automated feedback loop, notify Clarity AI of any perceived errors in the data which are fully investigated.
4.5 Disclosing, in an easy-to-understand manner, the purpose, concept, and basic methodology of the evaluation; doing this in a reasonable scope and manner, such as by consolidating or limiting the scope, taking into consideration a provider’s situation and the importance and relevance of individual items. The items are for example the following:
- Purpose, approach, and intent of formulation of ESG evaluation and data
- Specific contents of evaluation methodologies (specific evaluation criteria, important indicators and weights in evaluation, businesses and companies
- Subject to evaluation, and other contents of methodologies that can lead to significant differences in evaluation results)
- Evaluation process (evaluation procedures and steps, checks and monitoring, etc.)
- Contact point where the evaluation results can be explained in detail
- Sources of information on which the evaluation is based, policy and status of estimated data usage, the update timings and estimation methodologies of data that is particularly important to the overall assessment
- With respect to the overall evaluation, the timing of evaluation and the timing of data creation, use, and update
- Changes made when the evaluation methodology is updated. Especially if any items are improved through the PDCA cycle, this fact and reasons for it.
All of this information is available to users of Clarity AI’s sustainability products. Through the detailed methodology documents, users can track each insight from the source data through the methodologies. Queries can be raised via an automated portal. Where clients have further, detailed queries, Clarity AI has a dedicated team of product specialists who can answer the granular queries. As mentioned above, any updates to methodologies or changes to underlying data are fully communicated in advance to impacted clients, along with an explanation of the change and its implications. Clarity AI also maintains a data change log that tracks any such changes to ensure they can be reviewed and audited at a later date as required.
5. Confidentiality
In the instances where we receive any confidential information, we have policies in place to ensure it is handled to the highest possible standards.
5.1 Establishing, disclosing and implementing the policies and procedures to protect information provided as confidential in the course of ESG evaluation and data services.
We process sensitive information in line with confidentiality obligations and ensure all of our providers do the same. In addition, we have a code of conduct and information security policies which also contain obligations to maintain confidentiality of the information that employees need to abide by.
5.2 Establishing, disclosing, and implementing the policies and procedures so that such confidential information will be used in accordance with the purpose of provision and not for the purposes other than ESG evaluation and data services, unless otherwise agreed.
An internal policy regarding data classification and data handling is already in place.
6. Communication with Companies
The majority of Clarity AI’s business is undertaken on a subscriber-pay basis. Our sustainability insights are derived from data gathered from public sources, proprietary models or third parties. In those instances where we communicate with companies, we ensure those communications are open and never influence the outcome of pre-defined, rules based assessments. We welcome any relevant, factual feedback on raw data from companies and offer (free) access to any inquiring companies to review all collected data and scores. Data errors or missing data points can be reported in an automated way directly to Clarity AI within its web-based tool.
6.1 When and if collecting information through surveys from a company subject to evaluation, notifying the company of the collection period sufficiently in advance. If available and where appropriate, entering, prior to the request, information that is already known to the providers, such as those publicly disclosed or submitted in the past, then seeking verification by the company in question.
Not applicable to Clarity AI as it does not use surveys.
6.2 Establishing a dedicated contact point where companies can send inquiries and raise issues regarding ESG evaluation and data provision, and informing the companies concerned or posting it in an easy-to-find manner.
Clarity AI publishes the relevant contact information on its website (including a contact form) and clients can contact Clarity AI directly in an automated way within its tools.
6.3 When disclosing ESG evaluation and data, subject to the institution’s evaluation methodologies and customer service policies, to the extent practically possible, expeditiously notifying or communicating to a company of the essential information sources of the evaluation and data, thereby allowing time for the company to check whether there are any significant deficiencies in the sources, such as factual errors.
We comply with this guideline and welcome challenges to our assessment and evaluations through well defined processes.
6.4 When a company subject to evaluation raises important or reasonable issues about the information source of evaluation and data, subject to its own evaluation methodologies and customer service policies, taking timely and appropriate measures such as allowing the company to at least assess the accuracy of the underlying important data and correcting errors if any.
Any reported incidents are reviewed by Clarity AI’s team in a timely manner.
6.5 As an ESG evaluation and data provider, disclosing a “procedures of engagement” regarding how it normally interacts with companies to be evaluated with respect to the evaluation and data it provides. The procedures would include elements such as when it requests information from companies, when and what companies could check with, how they could raise issues if any, and how the provider would be able to respond to such issues.
Not applicable to Clarity AI.
6.6 Subject to providers’ evaluation methods and customer service policies, considering the necessity of managing conflict of interest, and to the extent practically possible, conducting constructive dialogue with companies to be evaluated (for example, by providing feedback on evaluation results)
Not applicable to Clarity AI’s business model (as noted in Section 3).
