In the pre-sequence report of the industry rotation series, we mainly discuss how to effectively use all kinds of data and information to build an industry rotation model which can significantly predict the future industry cross-section income. Taking the landing of investment as the goal, this paper discusses the application and feasibility of the view of industry rotation in ETF investment. By constructing the subdivision industry system and optimizing the portfolio of ETF funds, this paper constructs the ETF investment model based on the view of industry rotation.

To build a subdivided industry system that is more suitable for the feasibility of investment.

In order to facilitate industry investment, with reference to the plate division of the GICS economic sector, a new industry segmentation system is constructed through split and splicing behavior on the basis of Wind3 and level 4 industries. In the process of construction, three principles of homogeneity, investment and completeness are emphatically considered: 1. According to the economic activity logic of the industry and the driving factors of the industry development, and fully consider the homogeneity and differences of the demand side; 2. The industry is required to have a certain degree of market attention and transaction activity. In addition to telecommunications services, insurance and other special industries with their own unique style, consider diversifying the risk of individual stocks and controlling the number of constituent shares in the industry. 3. The classification basically achieves full coverage of A-share listed companies, and there is no case that one company belongs to multiple industries at the same time. At the same time, take into account the scale of the industry as close as possible, so that the industry can be comparable.

Finally, the subdivided industry system is composed of 11 plates and 71 subdivided industries.

The combination of macro, meso and micro industry rotation model is effective in subdividing the industry system.

The industry rotation model with excellent effect on the Citic first-level industry system still shows a relatively good industry cross-section prediction ability and excellent grouping monotonicity when applied to the new industry segmentation system. Taking eight equal rights held industries as the industry portfolio, in the sample range of nearly 12 years from February 2010 to June 2021, the annualized income of the portfolio is 21.3%, the Sharp ratio is 0.89, and the maximum withdrawal is 47.8%. Compared with the full segment of the industry, the annualized excess return is 11.5%, the information ratio is 1.36, and the maximum relative withdrawal is 14.6%.

Through the combination optimization method, the view of industry segmentation is transformed into an investable ETF portfolio.

According to the complete component equity redistribution composed of all the industry indexes of the industry rotation portfolio as the fitting object, using all the ETF based on their respective constituent stock data, through the way of portfolio optimization, the composition distribution of the final ETF portfolio is as close as possible to the subdivision industry configuration that we want to achieve.

Starting from 2016, the ETF portfolio constructed by this method can basically realize the benefits brought by the view of industry rotation, and through the advantages of stock selection and allocation brought by the ETF tracking index, it can even further thicken the portfolio income on the basis of the original industry portfolio in the last three years. During the sample period from 2016 to June 2021, the annualized return of the portfolio optimization ETF portfolio is 18.1%, the Sharp ratio is 0.82, and the maximum withdrawal is 35%. All the data are slightly better than the corresponding industry portfolio performance.

The latest ETF portfolio position view in August

It is recommended to track the ETF portfolio of the following indices: CSI Chemical Industry theme Index, CSI Iron and Steel Index, CSI Household Appliances Index, CSI New Energy vehicle Index, CSI Coal Index, CSI 500 Information Technology Index, China Trading Service Semiconductor Chip Industry Index.