An empirical comparison of Exponential-Gamma with 'never triers' model, Weibull-Gamma with 'never triers' model and bass model for new automobiles trial calibration and forecasting in Chinese market

In fact, recent years China may surpass the United States as the world's largest auto market. At a time when Chinese automakers are trying to get more involved in the global industry, it becomes increasingly important to calibrate and forecast the new automobile penetration process successfully in opened and competitive market. In this study, we extends the Exponential model and Weibull- Gamma model to Exponential -Gamma with 'never triers' model (EG with NT) and Weibull-Gamma with 'never triers' model (WG with NT) by considering 'never triers' and difference between individuals. An empirical analysis was carried out subsequently to compare the accuracy of calibrating and forecasting ability between EG with NT model, WG with NT model and BASS model. Unlike some previous ones, it is concluded that when dealing with automobile trail data in Chinese market, WG with NT model provide significantly better calibrates and forecasts than other two models; EG with NT model is more accurate in forecasting than Bass model; Bass model is more accurate in calibrating than EG with NT model.