Abstract
Multivariate time series such as stock prices may be classified into some groups. The trend of group may have each tendency. A multivariate fuzzy trend model can decompose trends into a common trend and individual trends. However, the existing model does not take account of the trends of each group. Therefore, we propose a model which can decompose trends into the common trend, trend of each group and individual trends. The applicability of the model is shown by simulation studies. In addition the practical application is provided by using stock price data.
