The Research On The Design And Implementation Of Multi-factor Quantitative Investment Management System

As the capital markets continuous improvement and institutional investorsgetting stronger, quantitative investment has been more widely recognized in the areaof asset management by virtue of its inherent advantages. With its practicality andstability, multi-factor quantitative investment strategies become an important branchof quantitative investment. Many public funds and securities companies launchedproducts based on the Multi-Factor quantitative investment strategies. However,because of the diversity of the factors, the asset managers have certain difficulties inunderstanding of the effectiveness of various types of factors. Not only the lack ofeffective criteria but a comprehensive evaluation of various factors to get an optimalweight is also a difficulty. After lots of research on the theory and practice, the paperseeks to develop research management system of quantitative investment strategybased on the information ratio principle, so as to enrich the quantitative managementteam with the knowledge of multi-factor strategy, enhance the efficiency ofquantitative investment management decision-making, and ultimately promote theperformance of investment management.This paper mainly explores the selection of multi-factor quantitative investmentstrategies based on the principle of information ratio, as well as the development ofquantitative investment management system. Firstly, it describes the theoretical basisof multi-factor quantitative investment strategies, the practical progress ofmulti-factor quantitative investment, and a detailed introduction of the core contentbased on the information model. In terms of factor selection, the author analyses theclassification of factor attributes, chooses information coefficient (IC) to capture theeffectiveness of factors quantitatively and uses the factor information ratio (IR) toevaluate the distinction and stability after effect. In terms of weight, the strategymodel derives the maximum IR value and then optimal weight using themean-variance principle and maximum value method. Secondly, it analyses the mainwork progress of normal investment analysis and problems of manual handling duringthe analysis. With these problems, it proposes some functional requirements to solvein the design of the system. After the data processing of ICIR model, it designsstructure modules of multi-factor quantitative investment management system. Thesystem mainly contains four parts, which are frontend user interaction layer、logic layer、ETL data acquisition module and database. Then, the thesis explains the dataprocessing in the system，the main function and key processes of each module. Finally,the system is used to do data analysis tests of ICIR model. The results show thatstrategy backtracking tests achieved good results and obvious distinction of differentevaluation packet. At the same time, this thesis points out the directions of furtherimprovement of the system.