Market share is one of the most important measures in the valuation of prospering firm. It plays a role of composite indicator for the competitiveness of firm. So, the understanding of the underlying process of market share is inevitable factor for the econometricians and the business engager. Lately, the Korean Economy has been placed in the control of IMF. This shock will cause a lot of influence over the domestic economy. The idea that the information about the past shock-response experience will do us good for dealing with this kind of economic shocks is not new. Among numerous markets, we pay attention to the durable goods market, especially automobile market. The automobile market has large repercussion effect over the domestic economy on the issue of both national employment and technology integration. We divided the Korean automobile market into three segments: small, medium, and large-sized car, while each proportion of these segments has been changing slowly. We propose a Bayesian approach to detect and forecast structural changes in time series of the market shares in the domestic automobile market, especially for level shifts and drift changes, and compare the empirical results with other existing approaches.