模糊共享投入两阶段DEA及在物流企业绩效评价中的应用

发布时间：2018-08-23 12:30

【摘要】：物流业作为国民经济的重要组成部分,在整个社会发展中发挥着重要的作用,而物流企业作为物流业的主导力量之一,提升其绩效水平不仅有利于其自身实现长期稳定的发展,也能在一定程度上推动整个物流业的发展。然而,要提升物流企业绩效,其重要前提是对物流企业绩效做出科学、准确的评价。在众多分析方法中,数据包络分析(DEA)因其评价的客观性、不需要事先确定具体前沿函数形式等优势被广泛应用于包括物流企业在内的企业绩效评价中。但是,传统DEA将决策单元(物流系统)看成一个“黑箱”,忽视了生产系统内部的效率,传统网络DEA虽然可以评价子过程效率,但没有考虑共享投入情形。此外,由于数据波动、缺失等原因,实际得到的数据经常是非准确或模糊的。基于以上分析,本文首先根据Lertworasirikul等(2003)提出基于可能性测度的模糊DEA方法,对Chen等(2010)提出的共享投入两阶段DEA模型进行模糊拓展;然后,通过分析中国上市物流企业实际运营情况,构建包含总资产和应付职工薪酬两个共享投入在内的中国上市物流企业两阶段效率评价指标体系;再者,针对数据为准确值和模糊数两种不同情形,分别利用共享投入两阶段DEA模型和模糊共享投入两阶段DEA模型对我国23家上市物流企业的绩效进行评价,并比较不同情形下效率结果的异同,主要得到以下结论:1、以所选取的23家水上运输上市物流企业为样本,不考虑模糊情形,从其全局效率来看,考察期内,中国水上运输物流企业全局效率均值不足0.7,具有30%以上的提升空间;相较2014年,2015年样本上市物流企业的全局效率偏低,且不同物流企业全局效率之间的差距也在扩大。2、从样本物流企业子阶段效率情况来看,其盈利化阶段效率明显高于市场化阶段效率,样本物流企业市场化效率变动趋势与全局效率几乎同步,导致全局效率偏低的重要原因是样本物流企业市场化效率不足。3、从考察期内样本物流企业子阶段效率均值来看,只有6家企业的子阶段效率利用模式属于“高盈利化—高市场化”类型,大部分物流企业的利用模式偏离最优区域,整体效率的改进空间比较大。4、以样本物流企业2014~2015年年报数据为基础,构造其各项指标的三角模糊数,运用本文所提模糊共享投入两阶段DEA模型对其进行分析发现,全局以及阶段效率值随着可能性水平的提升而下降,这也在实证角度验证了本文所提出的几个定理。5、基于模糊共享投入两阶段DEA模型的上市物流企业全局以及子阶段可能性效率排名与基于准确值模型的全局以及子阶段效率均值排名差异不大,基本在后者排名基础上小幅度波动。但相较准确值模型,模糊共享投入两阶段DEA模型还具有可以处理实际生产过程中数据缺失或不确定性、能在一定程度反映决策者的风险偏好和经验判断等优势。
[Abstract]:As an important part of the national economy, the logistics industry plays an important role in the development of the whole society. As one of the leading forces of the logistics industry, the improvement of its performance level is not only conducive to its own long-term and stable development. Also can promote the development of the whole logistics industry to a certain extent. However, to improve the performance of logistics enterprises, its important premise is to make a scientific and accurate evaluation of the performance of logistics enterprises. Among many analysis methods, data Envelopment Analysis (DEA) is widely used in the performance evaluation of enterprises, including logistics enterprises, because of its objectivity, and does not need to determine the specific frontier function in advance. However, the traditional DEA regards the decision making unit (logistics system) as a "black box", neglecting the internal efficiency of the production system. Although the traditional network DEA can evaluate the efficiency of the sub-process, it does not consider the situation of shared input. In addition, due to data fluctuation, missing and other reasons, the actual data is often inaccurate or fuzzy. Based on the above analysis, this paper firstly proposes a fuzzy DEA method based on possibility measure according to Lertworasirikul et al. (2003), which extends the two-stage DEA model of shared input proposed by Chen et al. (2010). By analyzing the actual operation of listed logistics enterprises in China, this paper constructs a two-stage efficiency evaluation index system of Chinese listed logistics enterprises, which includes two shared inputs of total assets and employees' salary. In this paper, the performance of 23 listed logistics enterprises in China is evaluated by using the shared input two-stage DEA model and the fuzzy shared input two-stage DEA model for the two different cases of accurate value and fuzzy number. And comparing the similarities and differences of efficiency results in different situations, we can get the following conclusion: 1, taking 23 listed logistics enterprises of water transport as a sample, not considering the fuzzy situation, looking from its overall efficiency, during the inspection period, The global efficiency of Chinese water transportation logistics enterprises is less than 0.7, which has more than 30% improvement space. Compared with 2014, the global efficiency of the sample listed logistics enterprises in 2015 is lower. And the gap between the overall efficiency of different logistics enterprises is also expanding. From the sample logistics enterprise sub-stage efficiency, the profit stage efficiency is obviously higher than the market-oriented stage efficiency. The change trend of sample logistics enterprises' market-oriented efficiency is almost synchronized with the overall efficiency, which is the important reason that the sample logistics enterprises' market-oriented efficiency is insufficient. 3. From the mean of sample logistics enterprises' sub-stage efficiency during the investigation period, Only 6 enterprises belong to the type of "high profitability and high marketization", and most logistics enterprises deviate from the optimal region. The improvement space of overall efficiency is relatively large. Based on the data of sample logistics enterprises' 2014-2015 annual report, the triangular fuzzy number of each index is constructed, and the two-stage DEA model of fuzzy shared input proposed in this paper is used to analyze and find it. The global and phase efficiency values decrease with the increase of the possibility level. It also verifies several theorems. 5, the global and sub-stage possibility efficiency ranking of listed logistics enterprises based on fuzzy shared input two-stage DEA model and the global and sub-level based on accurate value model. There is little difference in the average ranking of stage efficiency. Basically in the latter rank on the basis of small fluctuations. However, compared with the accurate value model, the fuzzy shared input two-stage DEA model can deal with the lack of data or uncertainty in the actual production process, and can reflect the risk preference and empirical judgment of decision makers to a certain extent.
【学位授予单位】：贵州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F259.23;F272.92

【相似文献】