A Competitive Market Structure Model Using Asymmetric Switching Data of Consumer Durables

DOI Web Site 22 References

Bibliographic Information

Other Title
  • 耐久消費財の非対称スイッチングデータを用いた競争市場構造モデル
  • タイキュウ ショウヒザイ ノ ヒタイショウ スイッチングデータ オ モチイタ キョウソウ シジョウ コウゾウ モデル

Search this article

Abstract

There are several problems in competitive market-structure (CMS) models using brandswitching data. First, the previous researches of CMS models using brand switching data treat brand-switching behavior as zero order : the previous buying does not affect the present brand selection, but it is very hard to apply a zero order assumption especially to consumer durable goods. Second, most models treat switching data as symmetric, but the switching data of consumer durable goods is not symmetric in fact. Third, switching data is not dissimilarity but flow, so it is to be transformed to the appropriate proximity measure before being put into Cluster analysis or MDS. Fourth, the CMS using switching data has been being drawn by Cluster analysis or MDS, but there are not yet enough of them to portray the asymmetry of switching data of consumer durables. We focus on the identification of competitive market structure not only using asymmetric brand-switching data, but also considering first-order assumption of consumer durables. We propose a new approach that generates switching constants, which is based on the changing patterns of attributes between the newly purchased brand and the old brand. And we propose a transforming function that transforms asymmetric switching data into asymmetric proximity measure. Conditional probability is also used in transforming function to consider first-order trait of replacement behavior of consumer durables. Furthermore, a pictorial representation to identify competitive market structure is to be drawn by using asymmetric MDS (Hermitian Form Model). Our model is applied to American car trade-in data. We propose an interpreting algorithm for asymmetric MDS result. We also identify that the CMS of the American car market is composed of two attributes (cost of origin and price). The pictorial representation of asymmetry of switching data is well attained by asymmetric MDS and it shows the direction and the degree of difficulty of brand switching as well as the competitive relations of each brand.

Journal

References(22)*help

See more

Keywords

Details 詳細情報について

Report a problem

Back to top