杠杆率作用于企业创新的智能评价与鲁棒优化

Intelligent Evaluation and Robust Optimization of Leverage Ratio on Enterprise Innovation

摘要:

如何处理“促创新”与“稳杠杆”的关系,寻求既在目标创新度水平下又能规避债务风险的企业杠杆率,是推进中国高质量发展进程中亟需解决的一大问题。为此,本文首先结合PCA综合评价理论,构建杠杆率作用于企业创新的智能评价模型；然后设计基于群体智能优化的杠杆率鲁棒区间优化算法,得到目标创新度下的最优杠杆率区间。进一步采用中国上市公司数据进行检验,结果显示：所构建的基于RBF神经网络的智能评价模型,高效地评价出随机抽取的12家公司创新度水平；所设计的群优化算法在迭代过程中都能快速的收敛到最优值,且收敛的最小总账面杠杆率和最大总账面杠杆率都出现“中间高两头低”的现象,即目标创新度处于中间段(较大或不大)等级的企业,最小账面杠杆率和最大账面杠杆率都较高,目标创新度处于高段(非常大、很大)等级的企业和目标创新度处于低段(不大、较小)等级的企业,最小账面杠杆率和最大账面杠杆率都较低。本文的研究不仅实现了企业创新评价的智能化,还找到了不同创新度等级下的最优杠杆率区间,为帮助企业在风险可控之内实现持续性创新、金融机构负债规划及政府部门供给侧改革决策提供理论上的参考。

Abstract:

How to deal with the relationship between “promoting innovation” and “stabilizing leverage”, and seek the enterprise leverage ratio that can avoid debt risks at the same time under the target innovation degree is a major problem that needs to be solved in the process of promoting China"s high-quality development. Therefore, this paper firstly constructs an intelligent evaluation model of leverage ratio on enterprise innovation based on PCA comprehensive evaluation theory. Then, it designs a robust interval optimization algorithm of leverage ratio based on swarm intelligent optimization to obtain the optimal leverage ratio interval under the target innovation degree. Furthermore, it uses the data of Chinese listed companies to test. The results show that: the intelligent evaluation model based on RBF neural network can efficiently evaluate the innovation degree of 12 randomly selected companies; the designed swarm optimization algorithm can quickly converge to the optimal value in the iterative process, and the minimum and maximum total book leverage ratios of the convergence show the phenomenon of “high in the middle and low at the end”, that is, the minimum and maximum book leverage ratios of the enterprises with the target innovation degree in the middle section (large or not big) are higher, and the minimum and maximum book leverage ratios of the enterprises with the target innovation degree in the high section (very large, very big) and the low section (not big, small) are lower. The research of this paper not only realizes the intelligent evaluation of enterprise innovation, but also finds the optimal leverage ratio interval under different levels of innovation degree, which provides theoretical reference for helping enterprises to achieve sustainable innovation within controllable risks, financial institutions to plan debt, and government departments to make decisions on supply-side reform.