Predicting success of a new launch based on market sentiment in the Indian auto industry

Abstract

As per the efficient market hypothesis, newly available information which impact market perception gets immediately incorporated into the market price of the company’s stock. With this analysis we examine whether the share price movement around the launch or announcement of new car models can be used to predict the success or the failure of the car model in the long run. To examine this problem we explored two classification methods, Binary Logistic Regression and Linear Discriminant Analysis, and developed several models. The classification accuracy of the final Logistic Regression model is around 70%. The results indicate that higher the increase in share price adjusted with the index, higher is the chance of the car model being successful. Also, for two car models with same share price increase, the company with lower total asset and lower sales to fixed asset ratio has a higher chance of success.