Building a closed loop of ESG assessment丨Decoding China's ESG ratings

The low quality of rating data and the large differences in rating methods of different institutions are prominent phenomena in the current ESG rating services, which not only affect the application of rating results by investment institutions, but also affect the trust of relevant parties in rating services.

Tao Linlin, head of the big data center of Menglang Sustainable Digital Technology (Shenzhen) Co., Ltd. (hereinafter referred to as "Menglang"), believes that the objectivity and scientificity of ESG assessment work are constrained by data quality, evaluation models, working models and other aspects. It is important that many institutions are constantly improving their data quality and evaluation models. At the same time, the internal working mode is also important.

In the ESG assessment, on the one hand, Menglang has established a three-level structure of "deliberation-decision-execution" to ensure the seriousness of the evaluation work; On the other hand, we will continue to strengthen investment in science and technology, improve the automation and intelligence of data collection, cross-validation, evaluation and scoring, and exception repair, so as to reduce the possible fluctuations in each link due to manual intervention.

Explore the consistency assessment of economic benefits and social values

Southern Weekly:

What is the background of the ESG assessment carried out by Monlang?

Tao Linlin:

The background of the ESG assessment is related to the international and domestic development of social impact investment, as well as the continuous promotion of the United Nations Sustainable Development Goals.

The predecessor of Menglang was the Social Value Investment Alliance (SIS). Since 2016, the Social Investment Alliance has been exploring and researching the relationship between social value and the capital market, carrying out comprehensive value evaluation of corporate sustainable development, and striving to comprehensively evaluate enterprises from the perspective of economic benefits and social value consistency.

In 2017, it developed a sustainable development value evaluation model for A-share listed companies. According to the evaluation results of the model, the "Sustainable Development Value of A-share Listed Companies Yili 99" ranking and related evaluation reports are released every year; In 2019, the "CSI Sustainable 100 Index" led by the Social Investment Alliance was launched; In 2020, the Social Investment Alliance and Bosera Fund jointly released the "Bosera CSI Sustainable 100 ETF" product, which was listed and traded. So far, a closed loop has been formed from value proposition - theoretical research - model development - scoring and rating - fund products.

In order to better promote the sustainable development of Chinese society, especially the business community, Mr. Ma Weihua, Chairman of the Social Investment Alliance, initiated the establishment of Menglang. Menglang extends and develops the concept of sustainable development value evaluation of the Social Investment Alliance, and at the same time solves the limitations of its own conditions in promoting business transformation and upgrading, so as to obtain more participation, support and synergy of social resources.

Southern Weekly:

What principles did Mon Wave follow in developing the ESG evaluation system?

Tao Linlin:

The ESG assessment system is based on the global consensus of the United Nations Sustainable Development Goals (SDGs) and the national strategy of the five development concepts, which are reflected in the evaluation model, indicator labels and evaluation criteria. Specifically, Menglang's ESG assessment follows the principles of substance, objectivity, inclusiveness, internationality, and locality.

The first is materiality, which comprehensively assesses the natural, political, economic, social, cultural and other environments in which the enterprise is located, as well as the main ESG risks and opportunities it faces, and reflects the special ESG issues and performance of the industry. The second is objectivity, which avoids human intervention in the rating model and evaluation process, and affects the credibility and authority of the assessment results. The third is inclusiveness, considering the diversity of political systems, market ecology, and cultural concepts in different countries, and constructing evaluation models according to local conditions. Fourth, ESG ratings should consider their applicability at home and abroad from the perspectives of financial liquidity and market openness and cooperation, and should not be "closed-door" and "self-talking". Fifth, it is local, fully considering the national conditions, enterprise conditions and people's feelings of the mainland, such as paying attention to the development needs of green and low-carbon, common prosperity and sustainable finance, as well as the concerns of various market players.

Under the guidance of these concepts and principles, Menglang has developed an ESG evaluation system framework composed of two parts: "screening sub-model" and "scoring sub-model". The "screening sub-model" is a negative rejection evaluation tool for sustainable development value assessment, and the "scoring sub-model" quantitatively scores the sustainable development value of listed companies.

Southern Weekly:

What are the considerations for dividing the assessment results into 9 basic levels and 19 enhancement levels in detail?

Tao Linlin:

The first is to benchmark against credit ratings. Most credit ratings are graded AAA-C with a "+/-" adjustment. Considering that the main users of ESG ratings are mainly capital markets, they have become accustomed to the meaning of credit ratings. Therefore, when classifying ESG rating results, we also try to use the same rating system to reduce the barriers to use.

The second is to consider the rationality of the distribution of scoring results. The sample size of listed companies' ESG assessment is large enough, and the range presented by the actual scoring results also supports the division of 9 basic levels. If it is only divided into four grades, A/B/C/D, it will lead to a concentration of ratings in a certain grade, and there is no differentiation. Judging from the rating results of A-share listed companies, the nine basic grades are basically normally distributed.

Finally, there are the results of the stock market backtest. At the beginning of the modeling, Menglang conducted a backtest of the stock market and found that the stock portfolio constructed from the rating results was well differentiated and monotonic.

"Working mechanism + technology empowerment" guarantees the scientificity of the assessment

Southern Weekly:

Does the evaluation take into account industry factors or regional differences?

Tao Linlin:

At the beginning of the modeling, we comprehensively considered the balance between the universality of ESG assessment and the differentiation of industries, and considered industry factors in multiple dimensions. According to Shenwan's industry standards, Menglang ESG Rating has set up 31 first-level industry models.

First of all, in the selection of indicators, we should follow the principle of substance, and comprehensively select general indicators and industry characteristic indicators. Secondly, in terms of weight allocation, different industry weights are set according to the relative materiality and importance of the index. Finally, in terms of index evaluation, the quantitative indicators are compared and scored within the industry, and the qualitative indicators also consider the actual situation of the industry to set different scoring standards.

Southern Weekly:

Data is the basis of ratings, how does Mon Wave ensure the quality of data?

Tao Linlin:

The data used in our rating mainly comes from public channels such as annual reports, social responsibility reports, sustainability reports, ESG reports, corporate official websites, temporary announcements, and regulatory information from regulatory authorities. In order to overcome the data dilemma, Menglang has taken the following measures:

First of all, a detailed work operation manual was formulated, which defined the requirements of data definition, type, nature, function, source, duration, quality and data samples and benchmarks, and built a prototype of the database.

Secondly, considering the factors such as database size, implementation efficiency and conflict of interest, a number of third-party institutions were selected as the partners of quantitative index data. A professional data production team was set up to conduct data retrieval of qualitative indicators.

Thirdly, a plan for the division of labor and coordination was formulated. The data team conducted back-to-back searches, cross-validation, a first review with the participation of the core group, a second review under the guidance of the Data Panel, and a third review under the "clear box" challenge (where the expert has the right to reconsider any ranking and access the company's full data, provided that the expert signs and follows a non-disclosure agreement).

Finally, it is verified by comparing the information organized by regulatory authorities, mainstream media and industry associations. After obtaining the data disclosed by the market, we corroborated the databases and blacklists published by the Ministry of Ecology and Environment, the State Administration of Taxation, the people's courts at all levels, and the competent ministries and commissions of various industries, and retrieved and analyzed the negative information disclosed by the mainstream media.

In order to continuously improve the efficiency and quality level of data collection. Since 2020, Menglang has set up a dedicated IT technical team to gradually promote the online, automated and intelligent evaluation of the production process.

Southern Weekly:

How to improve the objectivity and scientificity of evaluation work?

Tao Linlin:

The objectivity and scientificity of the evaluation work involve data quality, evaluation model, work mode, etc. Earlier we talked about evaluation indicators and data collection.

In terms of the mode of evaluation work, on the one hand, we ensure the seriousness of internal work by strengthening the internal decision-making mechanism; On the other hand, it is necessary to continuously strengthen technical means to minimize the possible volatility of each link due to manual intervention.

In terms of internal mechanisms, we have established and gradually improved the three-level structure of "deliberation-decision-execution". At the council level, the Evaluation Standards Advisory Committee (CSC) provides comments or suggestions on the elements of the model and the operation method; The Expert Committee on Evaluation Criteria (Social Committee) is responsible for the deliberation of major matters, such as the directional adjustment of the model, the establishment of projects for addition and deletion, etc.

At the decision-making level, the Evaluation Standards Working Committee (Social Standards Committee) is responsible for daily decision-making, such as the adjustment of model elements and the optimization of operating rules. The Social Standards Committee has four working groups: model, production monitoring, application products and data, which are responsible for the preliminary review of special topics. The working group is led by the Social Standards Committee and is open to qualified external experts.

At the executive level, the sustainable development value assessment project team is responsible for the implementation of the project, collecting and answering the suggestions and opinions of the Social Advisory Committee; Collate the information on major issues and submit them to the Social Special Committee for deliberation; Implement the resolutions of the Social Standards Committee; Cooperate with data suppliers and technical service providers to carry out evaluation and production work.

In terms of technology research and development, the data collection, data verification, initial scoring, abnormal data repair and other links will be automated and intelligent as much as possible to improve operation efficiency and reduce errors caused by manual work.

Supporting public research and investment applications

Southern Weekly:

What are the application scenarios of the ESG assessment of Menglang?

Tao Linlin:

There are two types of application scenarios: supporting public research and investment applications.

In terms of public research, Menlang not only publishes its own research reports based on ESG assessment results every year, but also provides support for universities, media, and research institutes that conduct research in the field of sustainable development.

In terms of investment application, on the one hand, Menglang has developed the CSI Sustainable 100 Index based on the ESG evaluation system, and applied the evaluation results to the Bosera CSI Sustainable 100 ETF. On the other hand, the ESG assessment results of Menglang have also become an important reference for securities firms and funds to carry out ESG investment research and investment decisions.

Southern Weekly:

In addition to ESG ratings, what other ESG data products has Menglang developed?

Tao Linlin:

At present, ESG ratings are mainly aimed at listed companies and bond issuers, and the scope of application is limited. Therefore, in addition to ESG ratings, for nearly 60 million surviving enterprises in the mainland's industrial and commercial database, Menglang has independently developed a carbon financing digital intelligence engine to carry out green asset identification and corporate carbon performance evaluation.

The "green identification" function of carbon financing provides financial institutions with green enterprises, green projects and green asset identification tools, helping financial institutions identify, identify and identify green customers and green assets, and build an efficient link bridge between green industries and financial institutions.

With the "carbon accounting and carbon rating model" as the core, the "carbon rating" function of carbon financing intelligently calculates the carbon emissions of enterprises through invoice data and automatically generates corporate carbon rating reports, and uses the credit products of financial institutions as the scenario to create a digital, intelligent and online "enterprise carbon rating linked loan", so as to realize the close linkage between corporate credit and corporate carbon performance, cover a large number of enterprise groups, and form an endogenous driving force for corporate carbon reduction.

Southern Weekly researcher Hou Minghui

Editor-in-charge: Sun Xiaowen