The Empirical Research Of The Multi-factor Stock Selection Mode Based On Industry Rotation

With vigorous development of the financial liberalization and intensive study on the modern financial theory,in the 1980 s,financial engineering emerged.It uses large and advanced financial information system to re-construct,disassemble and combinate existed financial data.Meanwhile financial engineering designs and develops financial products through establishing mathematical models and network diagrams,as well as using simulation technology and other analysis methods.Quantitative investment,an important branch of financial engineering,nowadays plays a significant role both in complex strategy combination and product development.Quantitative investment strategies could be classified to stock picking and choosing.Common stock-selection strategies are factors to pick stocks,chip stock selection,style,etc.Since the Merrill Lynch Investment Clock Theory was presented in 2004,industry rotation strategy has become a hot topic in asset-allocation research,and with constant improvement of the capital market,this strategy has wildly applied in financial products of fund companies.In fact,asset allocation is a definitely complex knowledge.Many financial institutions,especially mutual fund companies and asset management institutions,are actively exploring the allocation rules of China’s A-share market.These large types of assets can be allocated according to the industry,and the stock can be allocated according to the industry,so that the fund with the stock as the target can also be allocated according to the industry,which is the significance of studying the industry rotation.Because the economic cycle can reflect the liquidity of the whole economy,the first step is to judge the macro-economy before the industry stage.The purpose of this paper is to analysis by combining macro-economy and quantitative investment,so as to achieve the completion of the stock allocation strategy to verify the hypothesis.In this paper,portfolio theory,macro-economy and signal processing are s orted out to be combined with quantitative-stock-selection methods in order to construct the multi-factor-stock-selection strategy and the industry-rotation strate gy and conduct back test.First of all,based on investing logic,built on qualif ied investing experience and theory,and the Wind Quant platform,financial dat a,technical index data and valuation during 2016 and 2019 of all listed compa nies are selected to establish a factor set,which includes valuation factors,fina ncial quality factor,growth factors,scale factors,capital structure factor,the m omentum factor and liquidity factors.Secondly,perform pre-treatment on all da ta to screen out valid factors for building a multi-factor model and industry rot ation model.In construction of the industry rotation strategy,signal processing method should be conducted first to recognize the length of inventory cycle.After that,according to the CAPM theory,the 28 industries of SWS are divided into two types: cyclical and non-cyclical.Then,industry increasing value and production stock value are regarded as indicators of demand side and supply side respectively to obtain the latest stock cycle.Refer to the well-divided inventory cycle,properly configure industries.Allocate cyclical stock in passive de-stocking and negative re-stocking while deploy non-cyclical stock in passive re-stocking and negative de-stocking so as to establish the multi-factor-stock-selection model.Finally,based on comparison of before and after models with the CSI 800 index,the conclusion that the combination of the industry rotation strategy and the multi-factor model leads to the conclusion is as follows:(1)Based on the 39 candidate factors selected from valuation,quality,growth,structure and technical indicators,it is found that PE ratio,PB ratio,PS ratio,ROE and turnover ratio are more effective to construct the multi factor strategy with fixed industry style,and the cumulative yield is 45.98%,and the annualized return is 13.09%,which has achieved good returns;(2)Using the FFT to observe the obvious energy focus near the frequency point,it is verified that the short cycle of A-Share market is concentrated in 41.37 months,and based on the industrial added value and the finished product inventory of industrial enterprises as the index The multi factor model based on the industry rotation strategy is constructed,and the cumulative yield is 83.91%,and the annual yield is 21.91%.Both strategies beat the benchmark yield of 22.54% and the benchmark annualized yield of 6.62%;(3)During the whole model period,nonferrous metals,military industry,and communications were not selected for the three industries.The direct reason for the concentration was that the operation of different industry configurations in different periods was not carried out for the multi factor stock selection model with fixed industry style.