In Li Renyu's view, FOF has a more systematic basic selection system, which can screen out fund managers with stable style and alpha ability according to quantitative and qualitative conditions such as medium and long-term performance, investment philosophy, investment style and investment and research support.
People-oriented is one of the evaluation criteria of Li Renyu Fund. In terms of investment concept, Li Renyu follows the investment concept of "AI+HI" and emphasizes the fund evaluation system combining qualitative and quantitative methods. She said that it is necessary to select fund managers with different styles, interpretable performance, replicable and sustainable in the medium and long term to establish a fund pool. Then, build an effective investment portfolio on the basis of high-quality fund pool, with the goal of obtaining higher risk-adjusted income, and provide each investor with a better investment plan.
AI model +HI judgment forms an asset allocation scheme
Li Renyu believes that the asset allocation scheme based on "AI+HI" has the characteristics of more scientific decision-making, more intelligent research and more flexible adjustment.
"First of all, artificial intelligence helps fund managers to regularly predict the future asset allocation ratio. At the same time, the improved risk budget model can help investment managers better control the exposure risk and expected return under an equity center. " Li Renyu said that the AI model can intelligently predict the location level through dynamic prediction, real-time tracking and location adjustment.
The asset allocation scheme based on AI model, supplemented by macro data, can help to judge whether it is necessary to adjust the internal structure. In the process of HI macro-perspective-assisted structural adjustment, the fund manager conducts cross-validation based on the marginal changes of different macro-indicators, comprehensively evaluates the overall macroeconomic situation, and gives suggestions on internal structural adjustment of the AI model scheme (including asset allocation suggestions, plate allocation suggestions, style allocation suggestions, etc.). ). AI model +HI judgment will eventually form a regular asset allocation plan, and will also make irregular dynamic adjustments according to the actual net worth performance and market style.
Excavate the effective factors of fund performance
At present, there are more than 10000 funds in the whole market with mixed performance. How to screen them?
Li Renyu believes that the whole market fund screening system can be based on "AI+HI". The screening system was designed and developed by Li Renyu, covering multi-dimensional fund characteristic indicators. At present, there are *** 177 factors in the system, and the specific factors are also increasing and expanding with the market.
Through multi-dimensional and three-dimensional fund classification system, quantitative research on effective factors of AI mining and qualitative analysis of fund managers' optimal adjustment and tracking, the system has established a scientific and standardized fund screening system and process, and strived to screen high-quality funds efficiently and accurately through this fund screening system.
In the fund classification system, Li Renyu pointed out that fund products are divided into three dimensions: position dimension, stock style dimension, position style dimension and net worth characteristic dimension. After three-dimensional classification of all-market funds based on multiple dimensions, the beta of each category can be consistent, which is convenient for investment managers to better compare alpha.
In terms of quantitative research, Li Renyu said: "We have independently established a huge fund factor library for different fund characteristic indicators. At the same time, based on the independent fund classification, we have excavated the fund factors that can effectively predict the future performance of such funds from many fund characteristic indicators by means of feature engineering in machine learning, and established a set of quantitative evaluation system through these factors, initially screening out a number of high-quality fund products. "