Stock price index published by the BEI is intended to outline to investors which direction the market today. Forecasting price index needs to be done given the differences in opinion between technical groups with fundamentals. One method that can be used are exponential smoothing models. The purpose of this study was to determine the trend of sectoral stock price index as well as exponential smoothing models are best for forecasting stock price index. The data analyzed is the sectoral share price indices on the Stock Exchange in the form of weekly data for the last three years. The analysis was performed with software Eviews 8.1. The analysis showed the sectoral share price index of each sector for the long term is still fluctuating. Based on the analysis chart does not indicate there is a certain trend, either up or down tendency continually. Although since mid-2015 there were a downward trend, but in general for the October 2015 start a correction. Based on the analysis of these trends, the moving average forecasting method is not appropriate to do, given there was no sign of the trend along the period. Thus the exponential smoothing method is best applied in forecasting the stock price index. Of the several methods of forecasting is in exponential smoothing, Holt-Winters method multiplictive with three parameters is the best compared to the double method with one parameter and Holt-Winters no seasonal with two parameters. Selection is done by choosing the method of value Means Square Error (MSE), the smallest