Ping An Uses Artificial Intelligence to Drive New ESG Investment Strategies

Access to data from alternative sources can uncover hidden risks and opportunities

The Ping An Digital Economic Research Center (PADERC), a member of Ping An Insurance Company of China, has created four new investment strategies for environmental, social and corporate governance (ESG) investing using Ping An’s proprietary CN-ESG data for China A-shares, in light of surging demand in China for ESG ratings and data with wider coverage and a better fit for China’s market.

Ping An ESG framework aligns with international standards and Chinese regulations

The investment strategies detailed in the report, “Applications of Ping An CN-ESG Data and Framework in Quantitative Investment Strategy”, use the proprietary CN-ESG database and scoring framework developed by the Ping An Group. Ping An was the first asset owner in China to sign the United Nations Principles for Responsible Investment. The framework leverages Ping An’s expertise in finance and technology and aligns with international standards as well as guidelines from Chinese regulators to incorporate material topics for Chinese companies.

With technologies such as web crawlers, data mining, machine learning, knowledge graphs, natural language processing (NLP) and satellite remote sensing, the CN-ESG system can verify ESG disclosure-based data as well as mine non-disclosure-based data to provide investors with richer multi-dimensional information.

PADERC’s report provides an in-depth analysis on the data characteristics, effectiveness, and strategy back-testing results of the CN-ESG database and scoring framework, which covers more than 3,900 listed companies in the China A-share market with five years of historical data (2015-2019). The framework can provide quarterly results that are further adjusted based on news sentiment scores in real-time – compared to annual or semi-annual updates from most ESG rating providers.

Recommended AI News: Kasasa Partners with CU Strategic Planning to Help Credit Unions Better Serve the Underbanked

ESG factors independent of financial factors

PADERC found the Ping An’s CN-ESG scores among A-share companies is close to a normal distribution. The factor correlation test results show that scores have notable performance of quality factors. The overall correlation between CN-ESG factors and traditional financial factors is generally low, showing high levels of independence of ESG factors, which indicates these can provide new data and viewpoints for investment decisions.

The results of the factor layered test show that Ping An CN-ESG factors have a relatively strong positive screening effect on the Chinese Securities Index (CSI) 300 and CSI 800 stock pools. The financial window dressing factors constructed by evaluating the quality and authenticity of the company’s financial data yielded 11.61% of long-short gains since 2015.

ESG investment strategies that balance excess returns with ESG objectives

Based on CN-ESG data, PADERC constructed four types of ESG investment strategies that use artificial intelligence (AI) to balance excess investment returns and ESG investment targets:

1)     Ping An AI-ESG Selected 100 Strategy: This positive screening strategy selects companies with the highest ESG scores. Based on the broader CSI 800 stock pool, it can better leverage additional information from ESG scores. This strategy achieved an annualized excess return of 4.44%. The annual weighted average ESG score quantile of the portfolio is 94.2% among the benchmark stock pool.

2)     Ping An AI-ESG Enhancement Strategy: On the basis of ESG scores-based positive screening, PADERC added ESG factors to its Ping An Digital Economic Research Center 500+ No.1 AI Stock Selection Strategy and there was notable excess return. The AI stock selection strategy is based on linear and non-linear algorithms to capture complex market structures to predict the excess return of individual stocks. The Ping An AI-ESG Enhancement Strategy has an annualized excess return of 16.34%, and the annual weighted average ESG score quantile of the portfolio is 78.7% among the benchmark stock pool.

3)     CSI 300 ESG Style Index Series: The CSI 300 ESG Growth Index explores the growth value of the CSI 300 stocks, while controlling its tail risks. The CSI 300 ESG Low Volatility Index reinforces the stability features of ESG investment in both the short and long term. The ESG growth index achieved annualized excess returns of 5.67% and the low volatility index achieved 8.61% relative to the benchmark. The annual weighted average ESG score quantile of the portfolios are 75.1% (ESG growth index) and 73.1% (low volatility index) relative to the benchmark stock pool.

Further testing of excess returns shows that the above active management strategies have almost all achieved excess returns in adverse market conditions, including bond crises, annual bear market downturns, Sino-US trade war, and COVID-19, verifying the effectiveness of ESG factors in challenging environments.

4)     AI-ESG MAX Strategy: ESG enhancement of mainstream ETFs enables investors to gradually incorporate ESG concepts into their investing process without changing their traditional investing habits. Based on the CSI 300, controlling for sector deviation, this strategy sets tracking errors to 1%, 3% and 5%. Under different tracking error assumptions, the strategy maximizes ESG scores while achieving annualized excess returns of 3.61%, 3.40% and 3.43% respectively against the benchmark. The back-testing results of the strategy over the past five years show good performance, and excess returns were stable. This type of index enhancement strategy based on ESG factors could help drive an increase in the scale of ESG investing.

Recommended AI News: Mastercard SpendingPulse: U.S. Retail Sales Grew 3.0% This Holiday Season

Building a richer ESG strategy portfolio to meet investors’ diverse needs

Ping An’s CN-ESG framework will expand to include fixed income ESG data and climate risk-related AI-driven factors. It will enable more diverse investment options, such as ESG fixed income indices and climate risk-focused indices, to meet investors’ diverse needs. Ping An also developed a series of AI-ESG products focusing on corporate management, risk monitoring and analytics solutions for ESG and climate risk analysis, including portfolio sustainability footprint analysis, a portfolio adjustment tool, a sustainable funds screening tool, and climate risk asset pricing models to support ESG investment.

PADERC is a professional institution specializing in macroeconomics and policy research, using big data and artificial intelligence to provide insights on macroeconomic trends, including developments in ESG disclosures and ratings.

Recommended AI News: InMoment Data Reveals the Current Mindset of Affluent Investors, Links Their Customer Experience to Future Loyalty to Financial Firms