|
|
Data Envelopment Analysis – extension and application in automotive industry
Synková, Rut ; Jablonský, Josef (advisor) ; Cahlík, Tomáš (referee) ; Ivaničová, Zlatica (referee)
The Models of the Data Envelopment Analysis are the instruments for benchmarking of homogenous production units. The first models had been formulated by the end of 70th in the last century and since then have been the subject of interest in the theoretical area as well as in the analytical area. The thesis refers to the models of the Data Envelopment Analysis and its development in the theoretical area. The second goal of the thesis is an application of models of the Data Envelopment Analysis in industrial environment and illustration of its possible use by investment decision. The domestic literature on the area of the Data Envelopment Analysis has not been sufficient. The thesis is therefore mapping the present stage of knowledge in this area and further widens by the Allocation Models, Dynamic Analysis and the questions of non-controlled and imprecise variables. Models, which are formulated in the thesis, where afterwards applied on the data of the foreign companies. The numerical experiments were worked out by the software support for the models of the Data Envelopment Analysis, built in the environment of MS Excel. The main contributions of the thesis are in enlargement of the Dynamic Analysis by the partial continuous dynamic analysis and analysis of Efficiency Stability of the evaluated units. The other contribution is the wide application of the models of the Data Envelopment Analysis for benchmarking the companies, which mainly produce the motor vehicles. The thesis gives the overview of the often used models of the Data Envelopment Analysis as well as of the rarely used models. It discusses the possible solution to the special situations which may occur during the application. The thesis is divided into six chapters which, in typical case, contain next to the theory the illustrative application. The first chapter comprises of the subscription of the basic models of the Data Envelopment Analysis. In the second chapter, there are presented the Super-Efficiency Models and discussed the problems with zero inputs and outputs in these models. The overview of the Allocation Models is in following part. The Dynamic Analysis and analysis of Efficiency Stability are subjects of the fourth chapter. The fifth chapter is dedicated to non-controlled and imprecise variables. The last chapter focuses on application of the models in automotive industry and its possible use of results for estimated development of the evaluated units on Investment markets.
|
|
|
Multiple Marginalization and its Impact on Supply Chains' Efficiency
Zouhar, Jan ; Fiala, Petr (advisor) ; Cahlík, Tomáš (referee) ; Ivaničová, Zlatica (referee)
Double (or multiple) marginalization is often identified as the main source of a decentralized supply chain's (SC's) inefficiency. In its core lies the fact that if the agents constituting the SC choose their output prices according to the golden rule of profit maximization (that normally applies to a single firm that produces independently and sells directly to the end consumer), the prices in the SC tend to spiral up to an inefficient (equilibrium) level where both the consumer surplus and the SC's total profit are diminished. The aim of this paper is to analyze and quantify the impact of multiple marginalization on the behaviour of SC's that vary with respect to their structure (i.e. the number of agents and the links between them) and the shape of their cost and demand functions. The main gauge of this impact is the efficiency of a SC, defined as the ratio of the profit of a SC whose agents behave according to the model of multiple marginalization, and the potential profit of the SC (i.e. the maximum profit attainable under the conditions of complete coordination of prices within the chain). Besides efficiency, some other properties of a SC are studied, e.g. the distribution of the SC's profit among the individual agents or cost externalities within the SC. Three different models of multiple marginalization are studied in the paper. The first one is a linear model of multiple marginalization (i.e. a model with linear demand and cost functions); in this simplified setting we derived explicit formulae for values of the studied indicators. The second model is analogous to the first one only that it allows for non-linear demand and cost functions; in this case, the analysis is carried out using computer experiments with numeric algorithms. The last one is a dynamic model of multiple marginalization which studies the abovementioned price spiral through multi-agent simulation.
|
guest ::
