众筹项目筹资绩效的信息影响机制研究:基于Indiegogo平台证据

Information influence mechanism of crowdfunding financing performance:Based on evidence from the Indiegogo platform

众筹项目披露的信息是投资者了解项目及其创始人的主要途径,对项目筹资绩效具有重要作用;且投资者对待不同项目披露的信息的态度往往不同,而现有研究缺乏对项目营利属性与相关因素间内部作用的系统考虑。本文以详尽可能性模型(elaboration likelihood model,ELM)为理论基础,利用全球著名的Indiegogo众筹平台数据,将项目披露信息分为项目质量信息和社交网络信息,分别对应ELM理论的中心路径和边缘路径信息,进而探究其对筹资项目绩效的影响机制。结果表明:两类信息都对筹资绩效影响显著;其中营利类项目受中心路径信息影响更大,而非营利项目受边缘信息影响更大。其次,在上述结果基础上,本文根据项目的营利属性设计多任务学习模型,得出较优的预测实验结果,进一步证明了实证研究结论。本文研究完善并细化了众筹项目信息对不同项目筹资绩效的相关理论基础与实证依据,并探索以此设计机器学习模型的启发式思路;研究结论有助于指导项目创始人对披露信息进行优化管理,进而改进项目效益,提高众筹成功率。

Crowdfunding removes the limitation imposed by traditional financing methods,where investors tend to prefer enterprises with strong capital or good credit,and conveniently provides free and flexible financing for small,medium-sized and micro enterprises.In essence,crowdfunding involves attracting many small investments through online network to obtain the funds required for the project.Different from traditional financing,in which investors can obtain project information through networks and interviews,investors in crowdfunding projects can only understand the project through the limited information disclosed by funders.Because of this,the information disclosed by crowdfunding projects is the main source from which investors obtain information about projects and their founders;accordingly,such information plays an important role in project financing performance.Among the existing studies in this area,most focus on the projects′attributes as conveyed by the disclosure information,while ignoring the impact of the quality of the disclosure information itself on the projects′financing performance.In addition,research into founders′social characteristics focuses primarily on third-party social platforms,while there is minimal research into the social relationships on crowdfunding platforms themselves.Moreover,investors often have different attitudes towards the information disclosed by different projects;existing studies in this field,which mainly focus on one or several factors individually,lack a systematic consideration of the internal relationship between project profit attributes and related factors.Accordingly,this paper makes up for the shortcomings of existing studies,comprehensively considers the impact of projects′disclosure information and founders′social characteristics on the financing performance of crowdfunding projects,and introduces profit attributes as the adjustment variable used to explore the factors influencing the performance of different projects.The Elaboration Likelihood Model(ELM)is a social psychology framework,first proposed to help understand the effectiveness of persuasive communication.It explores how message characteristics affect consumers′attitude formation and decision-making behaviors by simulating consumers′exposure to product-related information.Therefore,with the help of ELM theory,this paper classifies the relevant factors affecting crowdfunding performance according to self-determination theory,and explores the mechanism on which these factors impact of crowdfunding project financing performance.Specifically,this paper uses data from the world-famous crowdfunding platform Indiegogo,which was first established in 2008 and has become one of the most popular incentive crowdfunding platforms in the world.The present paper then divides the information into two parts according to whether it is directly related to project quality,namely project quality information and social network information,which respectively represent central route and peripheral route information in ELM.Among them,central route information is directly related to the project in question:such information requires the receiver to critically think about and analyze it,consider the comparative advantages,and make a rational judgment regarding the goal.For its part,peripheral route information depends more on clues about the project′s targets and is indirectly related to the project.Subsequently,through empirical study,the paper explores the impact of different type of information on projects′financing performance,with the profit attributes as the adjustment variable.All variables passed the convergence validity test,and there was no multicollinearity problem.The results show that both central route and peripheral route information have an impact on financing performance.Investors in non-profitable projects are also found to be more vulnerable to the influence of peripheral route information and tend to spend less energy on thinking and analyzing.By contrast,investors in profitable projects tend to think more about the deeper meaning of the project′s information and invest more energy in analysis.Comprehensive robustness tests prove the stability and reliability of the obtained results.Furthermore,in the field of Internet finance,the prediction of investors′behaviors has also attracted the attention of academia and industry.Inspired by the empirical results,considering the regulatory role of projects′profit attributes,and noting that differences and relationships exist between the impact mechanisms of the two types of project information on projects′financing performance,this paper improves the design of machine learning models to achieve accurate financing performance predictions.Based on the central route and peripheral route in ELM,this paper designs a multi-task learning model to predict project performance.The designed model is then verified by conducting experiments on real datasets to demonstrate its effectiveness.Compared with single-task learning and random grouping learning,our model achieves better performance in terms of predicting both whether the projects would achieve their goals and the number of investors.We also design repeated experiments with different test sets to verify the universality of the conclusion.The results further prove that investors have different focuses on crowdfunding project information with different profit attributes.This paper improves and refines the theoretical and empirical basis of the impact mechanism of the information disclosed by crowdfunding projects on projects with different profit attributes.The use of ELM makes the research results more accurate and targeted,enabling the relevant theoretical basis to be perfected and refined.It also enriches the theoretical basis of crowdfunding performance.Furthermore,based on the empirical results,this paper explores the heuristic idea of designing machine learning models.The research can aid in understanding the impact mechanism of project information on financing for different crowdfunding projects,which will also help to guide project funders to optimize the management of information disclosure,thereby improving the projects′efficiency and the success rate of crowdfunding and making crowdfunding platforms more effective.